Description

The invention relates to a linear polyglycerol compound according to the preamble of claim 1 , to uses thereof according to the preambles of claims 10 and 1 1 , to a combined preparation according to the preamble of claim 12, to uses of this combined preparation according to the preamble of claim 13 and to a manufacturing method according to the preamble of claim 15.

Influenza infection causes millions of deaths worldwide every year despite of drugs and vaccines available. The influenza virus enters the host cell through multivalent interactions between viral hemagglutinin glycoproteins and sialic acid residues on the cell surface. Blocking the virus at this very first step of infection can be used as a preventive measure for influenza especially in high risk patients.

Presently, neuraminidase binding drugs like Tamiflu are being used prophylactically or at the early stage of an influenza infection in high dosage of 30 mg (in patients with body weight <15 kg) to 75 mg (body weight >40 kg) daily for at least two weeks in high risk patients by oral intake. The use of high dosage of this drug is also one of the reasons for development of Tamiflu-resistant viruses in patients infected with influenza. Tamiflu-resistant strains have already been found in 0.4 to 4 % of patients with higher resistant strains in children and this has been gradually increasing since year 2004 (cf. www.fda.gov/downloads/Drugs/DrugSafety/DrugShortages/UCM1838 50.pdf).

The emergence of drug-resistant influenza virus strains has made the discovery of new antiviral therapeutics indispensable. The idea of using multivalent inhibitors, which target hemagglutinin glycoproteins on influenza virus, as therapeutic or prophylactic drug has not been accepted well till now because the optimization of such inhibitors to achieve high efficiency without toxicity is a big challenging task.

The development of such a multivalent inhibitor demands a deep understanding of the influence of different factors like size, flexibility, ligand density, and linker length on multivalent interactions. In the past, different types of carrier backbones were adopted for the preparation of glycoconjugates bearing multiple sialic acid residues. These carrier backbones include sialoglycopeptides, polysaccharides, proteins, dendrimers, dendritic polyglycerol, liposomes, linear polymers, and nanostructures.

E.g., Papp et al. describes the inhibition of influenza viruses by hyperbranched polyglycerol derivatives [1 ]. Furthermore, Whitesides et al. and Matrosovich et al. reported on polyacrylamide sialosides, which were found to be efficient showing K, values in lower nanomolar range but serious concerns were raised about the toxicity of the polymers.

Erberich et al. describes, amongst others, a linear polyglycerol carrying 2-Acetylamino-2- deoxy-3,4,6-tri-0-acetyl-D-glucose residues as substituents [17].

WO 2009/032605 A2 describes bi-functional polymer-attached inhibitors of influenza virus.

It is an object of the present invention to provide novel nontoxic substances that can be used, amongst others, as antiviral drugs.

This object is achieved by linear polyglycerol compounds having the features according to the claim 1 . Such linear polyglycerol compounds comprise a backbone of linearly linked glycerol units that are optionally terminally substituted. According to the claimed invention, each linear polyglycerol compound carries a plurality of substituents in the nature of covalently bound residues chosen from the group consisting of sialic acid and sialic acid derivatives. Thereby, a degree of substitution (also referred to as degree of functionalization, DF) of the backbone is between 10 and 100 %. I.e., 10 % to 100 % of all hydroxide groups of the polyglycerol backbone are substituted by the before-mentioned substituents.

It is possible that all substituents are of the same chemical nature in a single linear polyglycerol compound molecule. Alternatively, it is possible that different substituents are present in one and the same linear polyglycerol compound molecule.

Surprisingly, these linear polyglycerol compounds showed a much better antiviral effect than the hyperbranched polyglycerol with an 1 1 carbon spacer between the polymer and the sialic acid ligand, which were described earlier by Papp et al. [1 ]. In a cell binding inhibition assay, IC ₅ o values of 200 μΜ sialic acid concentrations for both IPGSA 3 and 4 have been determined, while Papp et al. reported IC50 values between 1 .4 and 2.2 mM sialic acid concentrations. Furthermore, linear polyglycerol sialosides demonstrated to have much better antiviral effects than Tamiflu investigated by in vivo experiments.

Since all compounds that have been synthesized by the inventors so far have been non-toxic, the claimed compounds are, in an embodiment, compounds that are non-toxic for humans or animals, in particular non-toxic for humans, non-human mammals and rodents.

In an embodiment, the linear polyglycerol backbone additionally carries at least one further substituent in the nature of a covalently bound residue chosen from the group consisting of buffering agents, amines and sulfates. Thus, the backbone of the linear polyglycerol compound can be substituted by one or more buffering agents, one or more amines and/or by one or more sulfates along with the sialic acid residues and/or the sialic acid derivative residues.

In an embodiment, the substituents are at least one sialic acid derivative according to general formulae (I) or (II), wherein a covalent bond is formed between the backbone and residue Y of the sialic acid derivative:

wherein the residues have the following meaning in formula (I):

R ₂ = H, F, CI, Br, or NHC(NH)NH ₂ ,

R ₃ = H, OH, F, CI, or Br,

R ₄ = NHCH ₂ COOH, SCH ₂ COOH, NHCO(CH ₂ ) _n CH ₃ , or SCO(CH ₂ ) _n CH ₃ ,

R ₅ = H, OH, F, CI, Br, OCONH ₂ , or OCONH(CH ₂ ) _n CH ₃ ,

R ₆ = H, OH, F, CI, Br, OCONH ₂ , or OCONH(CH ₂ ) _n CH ₃ ,

R ₇ = H, OH, F, CI, Br, OCONH ₂ , or OCONH(CH ₂ ) _n CH ₃ , and

X = S and Y = CH ₂ , (CH ₂ ) ₃ OCH ₂ , or CH ₂ (CCHNNN)CH ₂ , or X = O and Y = NHCH ₂ , and

n = 0 to 10, independently from other variables n in the same molecule or in other molecules and wherein the residues have the following meaning in formula (II):

R ₂ = H, OH, or F,

R ₃ = H, F, or NHC(NH)NH ₂ ,

R ₄ = H, OH, F, CI, Br,

R ₅ = NHCH ₂ COOH, SCH ₂ COOH, NHCOCH ₂ OH, NHCO(CH ₂ ) _n CH ₃ , or SCO(CH ₂ ) _n CH ₃ ,

R ₆ = H, OH, F, CI, Br, OCONH ₂ , or OCONH(CH ₂ ) _n CH ₃ ,

R ₇ = H, OH, F, CI, Br, OCONH ₂ , or OCONH(CH ₂ ) _n CH ₃ , and

X = S and Y = CH ₂ , (CH ₂ ) ₃ OCH ₂ , or CH ₂ (CCHNNN)CH ₂ , or

X = NH and Y = COCH ₂ , and

n = 0 to 10, independently from other variables n in the same molecule or in other molecules, wherein the bond between the C2 atom and the C3 atom in formula (II) can be single bond or a double bond. Since the before-mentioned residues are based on sialic acid, the accordingly substituted linear polyglycerol compounds can also be denoted as linear polyglycerol sialosides.

According to formula (I), a linkage of the sialic acid (derivative) residue to the polyglycerol backbone is formed via the C-2 atom of the sialic acid (derivative) residue.

According to formula (II), a linkage of the sialic acid (derivative) residue to the polyglycerol backbone is formed via the -l atom of the sialic acid (derivative) residue.

In an embodiment, R ₂ and R ₃ in formula (I) both mean F so that the sialic acid derivative is a difluorosialic acid derivative.

In an embodiment, R ₃ and R ₄ in formula (II) both mean F so that the sialic acid derivative is a difluorosialic acid derivative. In an embodiment, the substituents correspond to formula (I), wherein Ri = H, R ₂ = H, R ₃ = OH, R ₄ = NHCOCH ₃ , R ₅ = OH, R ₆ = OH, R ₇ = OH, X = S and Y = (CH ₂ ) ₃ OCH ₂ In an embodiment, the substituents correspond to formula (I), wherein Ri = H, R ₂ = H, R ₃ = OH, R ₄ = NHCOCH3, R ₅ = OH, R ₆ = OH, R ₇ = OH, X = S and Y = CH ₂ (CCHNNN)CH ₂ .

In an embodiment, the substituents correspond to formula (II), wherein Ri = H, R ₂ = H, R ₃ = H, R ₄ = OH, R ₅ = NHCOCH3, R ₆ = OH, R ₇ = OH, X = S and Y = (CH ₂ ) ₃ 0. In an embodiment, the substituents correspond to formula (II), wherein Ri = H, R ₂ = H, R ₃ = H, R ₄ = OH, R ₅ = NHCOCH3, R ₆ = OH, R ₇ = OH, X = S and Y = CH ₂ (CCHNNN)CH ₂ -.

The residue CH ₂ (CCHNNN)CH ₂ is a cyclic residue that can also be represented by the following formula (III):

Thereby, dashed lines indicate bonds that will be formed to neighboring residues or molecule parts.

In an embodiment, an accordingly substituted linear polyglycerol compound can correspond to the following formulae (IV), (V), (VI) or (VII):

Thereby, residues Ri to R ₇ can have the general meanings explained in connection to formulae (I) and (II) or the specific meanings indicated just above. In an embodiment, that the degree of substitution of the backbone is between 10 % and 80 %, in particular between 15 % and 95 %, in particular between 20 % and 90 %, in particular between 25 % and 85 %, in particular between 30 % and 80 %, in particular between 35 % and 75 %, in particular between 40 % and 70 % (in each case including the upper and lower limits). A very well suited range of substitution is between 40 % and 80 %. Another very well suited degree of substitution is between 42 % and 75 %.

In an embodiment, the median carbon atom of a first glycerol unit in the linear polyglycerol compound (C2 atom) is linked to one of the two terminal carbon atoms in a second glycerol unit (C1 atom) via an ether. Thereby, a 1 ,2-linkage between adjacent glycerol units in the linear polyglycerol compound is formed. It can also be denoted as 2,1 -linkage. In an embodiment, the terminal carbon atom of a first glycerol unit in the linear polyglycerol compound (C1 or C3 atom) is linked to one of the two terminal carbon atoms in a second glycerol unit (C1 or C3 atom) via an ether. Thereby, a 1 ,3-linkage between adjacent glycerol units in the linear polyglycerol compound is formed. It can also be denoted as 3,1 -linkage. In an embodiment, the glycerol units of the claimed compounds are either exclusively 1 ,2-linked or 1 ,3-linked to each other.

The general formulae of such 1 ,2-linked and 1 ,3-linked polyglycerol compounds are depicted below, wherein the meanings of the indicated residues in an embodiment are also indicated: 1 ,2-linked linear polyglycerol 1 ,3-linked linear polyglycerol

with

= 5 to 1000,

X, Y = independently from each other any organic residue with a functional group chosen from the group consisting of alcohol (hydroxyl), amine, thiol, azide, alkyne, alkene, carboxylic acid, aldehyde, ketone, halogen, isocyanate, isothiocyanate, Michael acceptor/donor group, Z = independently from other residues Z in the same polyglycerol molecule a residue chosen from the group consisting of OH, sialic acid, and sialic acid derivatives.

Thereby, 10 to 100 % of all residues Z are distinct from OH. In an embodiment, the compound (backbone plus substituents) has a number average molecular weight (M _n ) of 1 kDa to 100 kDa, in particular of 2 kDa to 90 kDa, in particular of 3 kDa to 80 kDa, in particular of 4 kDa to 70 kDa, in particular of 5 kDa to 60 kDa, in particular of 10 kDa to 60 kDa, in particular of 12 kDa to 55 kDa, in particular of 15 kDa to 50 kDa, in particular of 20 kDa to 45 kDa, in particular of 25 kDa to 50 kDa, in particular of 25 kDa to 45 kDa, in particular of 30 kDa to 55 kDa, in particular of 30 kDa to 52 kDa, in particular of 32 kDa to 50 kDa. Mn is thereby determined by gel permeation chromatography (GPC) as typically applied for this purpose by a person skilled in the art. Other generally suited methods for determining M _n are matrix-assisted laser desorption/ionization - time of flight mass spectroscopy (MALDI-TOF) and multi angle light scattering (MALS) leading to approximately the same results as GPC.

Thereby, a combination of the before mentioned molecular weight ranges with a degree of substitution of 30 % to 80 % is particularly well suited. In addition, a combination with a degree of substitution of 40 % to 80 % is also particularly well suited. Thus, in an embodiment, the compound might have a number average molecular weight (M _n ) of 30 kDa to 52 kDa and a degree of substitution of 30 % to 80 %. In another embodiment, the compound might have a number average molecular weight (M _n ) of 30 kDa to 52 kDa and a degree of substitution of 40 % to 80 %. These combinations are particularly suited if the substituent is a sialic acid or sialic acid derivative residue such as the substituents of formulae (IV) to (VII). In an embodiment, the linear polyglycerol compound is terminally functionalized with a functional group chosen from the group consisting of azide, bromide, chloride, iodide, fluoride, primary amine, secondary amine, tertiary amine, carboxylic acids, thiols, disulfides, ketals, acetals, aldehydes, and of hydroxyl, isocyanate, isothiocyanate, unsaturated carbon-carbon units having double or triple bonds, a Michael acceptor (such as an α,β-unsaturated aldehyde, ketone, ester, carboxylic acid amide, carboxylic acid imide, such as maleimide, or an α,β- unsaturated nitrile) and a Michael donor (such as organic copper compounds, amines, thiols, phenolate ions, cyanides and acrylates). I.e., at least one terminus of the polyglycerol compound carries an according reactive group in this embodiment. In an embodiment, the described linear polyglycerol compound is used as a medicament, namely as prophylactic or therapeutic agent. Suited areas of application are antiviral therapy (post infection) and prophylactic treatments of individuals to avoid a viral infection. Thus, the instant invention relates to method of treating a human or an animal (in particular a non-human mammal) in need thereof with a prophylactic or therapeutic agent comprising a linear polyglycerol compound according to the preceding explanations.

In an embodiment, the described linear polyglycerol compound is used as an antiviral agent, an antibacterial agent or an anti-inflammatory agent. Thus, the instant invention relates to method of treating a human or an animal (in particular a non-human mammal) in need thereof with an antiviral agent, an antibacterial agent or an anti-inflammatory agent comprising a linear polyglycerol compound according to the preceding explanations.

In an embodiment, the agent is used against Orthomyxoviridae, in particular influenza A virus, influenza B virus, and/or influenza C virus. Thereby, a use against influenza A virus is particularly suited.

It will become evident from the exemplary embodiments described subsequently that the linear polyglycerol compounds are particularly well suited to be used in combination with "classic" neuraminidase inhibitors like Oseltamivir (Tamiflu), Zanamivir (Relenza), Laninamivir (Inavir), and Peramivir since they show a synergistic effect regarding the virus replication inhibition. In an embodiment, the invention relates to a combined preparation comprising a linear polyglycerol compound according to the preceding explanations and a neuraminidase inhibitor. In an embodiment, these compounds are the only pharmaceutically active compounds in the combined preparation. Thereby, particularly well suited linear polyglycerol compounds are those carrying a sialic acids residue or sialic acid derivative residue as substituent. A particularly well suited neuraminidase inhibitor is Tamiflu.

In an embodiment, the invention relates to the simultaneous, separate sequential use of the pharmaceutically active compounds of the before-mentioned combined preparation in antiviral, antibacterial, or anti-inflammatory prophylaxis or therapy. The simultaneous use is particularly suited. However, the pharmaceutically active compounds (in particular exclusively a linear polyglycerol compound and a neuraminidase inhibitor) can be present in the combined preparation in a galenic form that enables a timely staggered release of the individual compounds so as to allow for sequential use of the individual pharmaceutically active substances of the combined preparation.

In an embodiment, the before-mentioned use is a use in antiviral prophylaxis or therapy. In a further embodiment, the antiviral prophylaxis or therapy is directed against influenza virus. Summarizing, the linear polyglycerol compounds described herein (IPG) are nontoxic and biocompatible and, therefore, can be explored for the development of multivalent glycoarchitectures bearing multiple substitutes, such as sialic acid residues or sialic acid derivative residues. In an aspect, the invention also relates to a method for manufacturing a linear polyglycerol compound according to the preceding explanations. This method comprises the following steps: a) providing a linear polyglycerol compound, comprising a backbone of linearly linked glycerol residues bearing hydroxyl groups or other functional groups chosen from the group consisting of allyl, azides, alkynes, alkenes, thiols, halogens, primary or secondary amines, carboxylic acids, aldehydes, ketons and any Michael donor or acceptor for conjugation of anionically charged entities, and b) causing a reaction of at least some of these hydroxyl groups or these other functional groups of the linear polyglycerol compound provided in step a) with a compound able to introduce a residue chosen from the group consisting of sialic acid and sialic acid derivatives into the linear polyglycerol compound.

In step a) allyl and azide a particularly suited functional groups. The reaction partners used in step b) might carry a thiol group or a thiopropargyl group to allow for formation of a covalent bond.

All embodiments of the described linear polyglycerol compound, the described combined preparation, their uses and the disclosed methods can be combined in any desired way and are transferable to any other or same category of subject-matter that is herein disclosed (i.e. from the compound to a method or from a method to a use or from a use to a method of from the compound to the combined preparation etc.).

Aspects of the invention will now be explained in more detail by examples and accompanying Figures. In the Figures:

FIG 1 shows a reaction scheme for 2 synthesis routes of different linear polyglycerol compounds;

FIG 2 shows volume/size distribution profiles for a linear polyglycerol and different linear polyglycerol compounds;

FIG 3A shows the results of hemagglutination inhibition with linear polyglycerol sialosides at different sialic acid densities;

FIG 3B shows the enhancement factor β of the tested linear polyglycerol sialoside based on their inhibition potential compared to monomeric a-2,6-sialyllactose;

FIG 3C shows the binding inhibition of R 18 labeled X 31 virus to human red blood cells with the most potent linear polyglycerol compounds;

FIG 4 shows the change in fluorescence upon binding of linear polyglycerol sialosides at different concentrations to fluorescently labeled X31 ;

FIG 5A shows cell viability after a treatment of MDCK II cells with inhibitors at different concentrations for 24 h; FIG 5B shows infection inhibition curves of the most potent linear polyglycerol compounds from an infection assay with X31 ;

FIG 5C shows infection inhibition with the most potent inhibitors 24 h after infection with X31 at 1 mM sialic acid equivalents;

FIG 5D shows normalization of infected cells to the untreated case;

FIG 6 shows the results of an experiment in which MDCK II cells were infected for 1 h with

X31 indicating the power of infection inhibition at a late state of infection;

FIG 7 A shows the results of an infection inhibition experiment with X31 in BALB/c mice using a linear polyglycerol compound as inhibitor; FIG 7B shows the results of an infection inhibition experiment with X31 in BALB/c mice using a linear polyglycerol compound and/or Tamiflu as a separate, secondary inhibitor;

FIG 8 shows the antiviral activity of LPG ₃ oSA ₀ .4o against two avian influenza strains post infection; and

FIG 9 shows the results of a neuraminidase activity inhibition assay.

The figures will be explained in more detail with respect to the following explanations of exemplary embodiments.

Exemplary Embodiments

A detailed analysis of random coiled particles of a sialic acid conjugate (Mw=32 kDa, 44% SA) of a linear polyglycerol (IPG or linPG or LPG) having a hydrodynamic diameter (dn) of 6.3 nm and a number average molecular weight (M _n ) of 10 kDa with dynamic light scattering, cryo- TEM as well as toxicity and efficacy assays was done. Subsequently, the results of this analysis together with the results of analyses of similar linear polyglycerol compounds will be presented, showing that low molecular weight linear polyglycerol sialosides are much more potent compounds than Tamiflu, in particular for a preventive treatment plan against influenza. Synthesis of linear sialic acid conjugated polyglycerols 2-6

For a detailed insight of the multivalent interactions between multivalent sialosides and hemagglutinin (HA) on influenza virus, a library of compounds was synthesized based on one polymeric backbone, namely, IPG 1 (M _n =10.1 kDa, D=1 .29). IPG 1 was synthesized by following previously reported procedures [2-6] and fully characterized by HNMR, ³ CNMR, and gel permeation chromatography (GPC). Thiopropargyl sialic acid was synthesized using methods previously published [7, 8].

Generally, the degree of sialic acid functionalization in the final compounds was controlled by using the limited equivalents of mesyl chloride in the first synthetic step. The consumption of all the azide groups in the click conjugation step was monitored by infrared spectroscopy (IR). Saturated EDTA solution was used during dialysis of the final product to remove Cu ions.

For the preparation of linear polyglycerol sialosides (IPGSAs or linPGSAs) 2-5, a three step procedure starting from IPG 1 was chosen: allylation, thiol-ene click reaction, and hydrolysis of the sialic acid conjugated product with 2M NaOH (see Figure 1 , upper line). The thiol-ene click reaction appeared to be much less complicated than the click reaction because of the less time required for reaction completion and also due to the absence of Cu. Here, the degree of sialic acid conjugation was controlled by degree of allylation which in turn controlled by using limited equivalents of NaH in the reaction mixture.

Consumption of all the allyl groups was easily monitored by HNMR. One compound in the IPGSA series, namely the compound IPGSA 6, was prepared by Cu-assisted click reaction also starting form IPG 1 also in three steps: three steps: mesylation, substitution by azide, and copper-catalyzed Sharpless/Huisgen click reaction IPGN ₃ with propargylated sialic acid (SA). This different synthesis approach was chosen to find out if there is any influence of the short thio-methylene-triazolyl or thio-propylene linker on the activity of the compounds (see Figure 1 , lower line). This was not found to be the case later on in the HA assay. With reference to Figure 1 , the synthesis of allylated IPG 1 a-1 d will now be explained. IPG 1 (0.1 g, 0.14 mmol OH to be functionalized) was dried at 60 °C overnight under high vacuum. Dried IPG was dissolved in dry dimethylformamide (DMF) (3 ml_) and cooled to 0 'C. To the stirred solution of IPG in dry DMF at 0 ^° C, NaH (4.59 mg, 0.18 mmol, 1 .3 eq., 95%) was added. After addition ice bath was removed and the temperature of the reaction mixture was allowed to reach room temperature. The reaction mixture was allowed to stir at room temperature for 2 h and cooled down again to 0 °C. The allyl bromide (24.12 μΙ, 0.28 mmol, 2.0 eq.) in dry DMF (1 ml_) was added drop wise to the reaction mixture using a syringe at 0° C. The ice bath was removed and after stirring for 16 hours at room temperature the reaction was quenched with water. The DMF was removed in vacuum and the resulting mixture was dialyzed in methanol (MeOH) to afford 90 mg of hPG-allyl 1 a. Degree of allylation was quantified as 10 % by Ή NMR. Yield = 84%. 1 H NMR (500 MHz, CD ₃ OD): 5(ppm) = 5.97-5.89 (m, 1 H, OCH ₂ CH=CH ₂ ), 5.23 (dd, 2H, J,=45Hz, J^I OHz, OCH ₂ CH=CH ₂ ), 4.01 (d, 1 H, J=5Hz, 3.71 -3.49 (m, IPG backbone). 13C{1 H} NMR (125 MHz, CD ₃ OD): 5(ppm) = 136.19 81 .55, 80.16, 71 .17, 70.95, 70.74, 70.58, 62.63 (IPG backbone). IR (CH ₃ OH): v(cm-1 ) = 3338, 2907, 2066, 1645, 1460, 1420, 1349, 1201 , 1260, 1 100, 997, 923. Table 1 shows the yields and different degree of allylation achieved as determined by 1 HNMR.

Table 1 : Yield and degree of functionalization of IPG-allyl

IPG-allyl 1 a 1 b 1 c 1 d

DF(%) 10 44 73 100

Yield % 84 80 86 87

Synthesis of IPGSA _pro tected 2-5

IPG-allyl 1 a (0.08g, 0.102 mmol of allyl group) was dissolved in dry methanol (3ml). Added freshly prepared thio-sialic acid (77.46 mg, 0.153 mmol, 1 .5 eq.) and 2,2-dimethoxy-2- phenylacetophenone (DMPA) as radical initiator (26.14 mg, 0.102 mmol, 1 .0 eq.). The catalytic amount of f/7s(2-carboxyethyl)phosphine hydrochloride (TCEP-HCI) was added to the reaction to avoid oxidation of thio-sialic acid. The solution was degassed by flushing argon through the reaction mixture for 10 minutes. The reaction mixture was stirred and irradiated with UV light using high pressure UV lamp at room temperature for 2 hours. The reaction completion was monitored by 1 HNMR by looking for disappearance of resonance peak corresponding to one of the allyl protons at δ 5.85. The reaction mixture was dialyzed to remove DMPA and excess of sialic acid derivative against methanol to afford the 107.8 mg IPGSA _pro tec 2. Conversion: quantitative (determined by H NMR). Yield= 82%. 1 H NMR (500 MHz, CD ₃ OD): 5(ppm): 5.35 (brs, 3H, H-8, NH, H-7, SA), 4.84 (m, 1 H, H-4, SA), 4.57-3.49 [m, SA (H-9, H-5, H-6, COOCH ₃ , -SCH ₂ CH ₂ CH20-), and IPG backbone], 2.85-2.61 (m, 3H, SA, H-3e, - SCH ₂ CH2CH ₂ 0-), 2.17-1 .83 (m, 18H, 8SA, 4xOCOCH ₃ , H-3a, -NHCOCH ₃ , -SCH2CH ₂ CH ₂ 0-). 13C{1 H} NMR (125 MHz, CD ₃ OD): 5(ppm)=173.34, 172.36, 171 .77, 171 .62, 171 .43, 170.08 (- COOCH ₃ , 4x-COCH ₃ , -NHCOCH ₃ ), 84.41 (SA C-2), 81 .56-62.56 (IPG backbone, SA: C-6, C- 4, C-8, C-7, C-9), 53.60 (-COOCH ₃ ), 49.97, 49.85 (C-5), 39.15 (C-3), 31 .08 (-SCH ₂ CH2CH ₂ 0- ), 26.75-20.88 (-NHCOCH3, -SC/- 2CH2CH2O-, 4x-COCH ₃ ). IR (CH ₃ OH): 3363, 2932, 2874, 1735, 1662, 1548, 1369, 1216, 1 1 19, 1076, 1030, 951 cm .

Table 2 shows the yields and degree of functionalization of IPGSAprotected 2-5 achieved. Table 2: Yield and degree of functionalization of IPGSA _pro tected 2-5

IPGSAprotected 2 3 4 5

DF(%) 10 44 73 100

Yield % 82 77 70 80

Synthesis of IPGSAs 2-5

2M NaOH (5ml) was added to the IPGSA _pro tected 2 (0.09g, 0.07mmol of ligands) and stirred at room temperature for 2 h. The reaction mixture was neutralized by adding 2M HCI solution and dialyzed against water for 3 days. The solvent of the dialysis was changed thrice a day. The aqueous solution obtained after dialysis was lyophilized to afford the pure 72.5 mg IPGSA 2. Conversion: quantitative (determined by H NMR). Yield= 92%. Ή NMR (500 MHz, D ₂ 0, δ): 3.88-3.56 [m, SA (H-8, NH, H-7, H-4, H-9, H-5, H-6, -SCH ₂ CH ₂ CH20-) ; IPG backbone], 2.81 - 2.70 (m, 3H, SA, H-3e, -SCH ₂ CH2CH ₂ 0-), 1 .75-2.08 (m, 3H, SA, H-3a, -NHCOCH3, - SCH2CH ₂ CH ₂ 0-). 13C{1 H} NMR (125 MHz, D ₂ 0): 5(ppm) = 175.06, 174.20 (-COOH NHCOCH3), 86.21 (SA C-2), 79.69-60.78 (IPG backbone, SA: C-6, C-4, C-8, C-7, C-9), 51 .78, 51 .62 (C-5), 41 .17 (C-3), 29.19 (-SCH ₂ CH2CH ₂ 0-), 26.25 (-NHCOCH ₃ ), 22.07 (- SCH2CH ₂ CH ₂ 0-). IR (solid): 3358, 3292, 2926, 2876, 1740, 1660, 1604, 1371 , 1223, 1 120, 1037 cm ^"1 . .

Table 3 shows the yields, DF, and average molecular weight as determined by 1 HNMR/GPC of the final compounds IPGSA 2-5.

Table 3: Yield of IPGSA 2-5

IPG-SA 2 3 4 5

DF(%) 10 44 73 100

Yield % 92 87 44 79 Mw (kDa) 15 31 .7 46 59

(NMR)

*GPC (H ₂ 0)

Mn (kDa) 12.4 39.25 33.70 49.25

Mw (kDa) 16.93 67.62 54.02 63.56

PDI 1 .35 1 .72 1 .60 1 .29

Synthesis of IPG-OMs 1e

IPG 1 (0.1 g, 0.27 mmol OH to be functionalized) was dried at 60 'Ό overnight under high vacuum. Dried IPG was dissolved in dry DMF (3 ml_) followed by addition of triethylamine (57.18 μΙ_, 0.41 mmol, 1 .5 eq with respect to OH group) and cooled to 0 'Ό. Then methane sulphonyl chloride (MsCI) (27.16μΙ, 0.35mmol, 1 .3 eq. with respect to OH group) in dry DMF (1 ml_) was added drop wise using a syringe at 0 °C. After stirring the solution for 1 h at 0 °C, the mixture was allowed to reach room temperature and was stirred for 16 h. The dry DMF was removed in vacuo and the resulting mixture was dialyzed in MeOH to afford 96.4 mg of IPG-OMs as a honey-like product. Degree of mesylation was quantified as 20% by H NMR.

Yield = 80%. ¹ H NMR (500 MHz, D ₂ 0, δ): 4.43-3.52 (m, CH ₂ and CH, PG scaffold); 3.16 (brs, CH ₃ mesyl), IR (film): v = 3398, 2913, 2877, 1653, 1459, 1349, 1 168, 1 109, 1081 , 931 , 838 cm ^-1 .

Synthesis of IPG-N ₃ 1f

IPG-OMs 1e, DF=20% (0.09 g, 0.20 mmol OMs) was dissolved in dry DMF (5 ml_) in a one- neck flask equipped with a refluxing condenser. NaN ₃ (65.52 mg, 1 .00 mmol, 5 eq. with respect to OMs group) was added to the reaction mixture and stirred at 60 °C for 3 days. The resulting salts were filtered off. DMF from the reaction mixture was evaporated using rotor vapor and the residue was dialyzed in MeOH to obtain 0.081 g of IPG-N ₃ 1f. Conversion: quantitative (determined by Ή NMR). Yield= 98%. Ή NMR (500 MHz, MeOD, δ): 3.72-3.38 (m, CH ₂ and CH, IPG scaffold), IR (CH ₃ OH): v = 3349, 2922, 2878, 2098 (N ₃ ), 2034, 1633, 1540, 1 160, 1041 cm ^"1 .

Synthesis of IPGSA 6 IPG-N ₃ 1 f, DF=20% (0.07 g, 0.1 7 mmol N ₃ groups) was dissolved in 4ml of THF-H ₂ 0 (1 :1 ) solvent mixture followed by addition of solution of propargylated SA derivative (1 20.6 mg, 0.22 mmol, 1 .3 eq. with respect to N ₃ groups) in THF (1 ml). CuS0 ₄ -5H ₂ 0 (8.4 mg, 0.034mmol, 0.2 eq with respect to N ₃ groups) was dissolved in 0.2 ml of H ₂ 0 and added to the solution of Na.ascorbate (67.3mg, 0.34mmol, 2.0 eq with respect to N ₃ groups) in 0.4 ml H ₂ 0. The resultant solution of salts in water was added drop wise to the solution containing IPG-N ₃ and propargylated SA. The reaction mixture was degassed thoroughly with argon for 5 minutes and then allowed to stir for 3 days at room temperature under argon atmosphere. The progress of the reaction was monitored by IR. When the peak at -2100cm ¹ corresponding to azide stretching frequency completely disappeared, solvent of the reaction mixture was evaporated at rota vapor. 2M NaOH (8ml) was added to the residue, and stirred at room temperature for 2hrs. The reaction mixture was neutralized by adding 2M HCI solution and dialyzed against water and saturated EDTA solution for 2 days and again using water for 4 days. The solvent of the dialysis was changed thrice a day. The aqueous solution obtained after dialysis was lyophilized to afford the pure 98.5 mg IPGSA 6.

Conversion : quantitative (determined by Ή NMR). Yield= 78%. Ή NMR (500 MHz, D ₂ 0, δ): 8.01 (s, 1 H, C=CH), 4.69-3.45 [m, SA (H-8, NH, H-7, H-4, H-9, H-5, H-6); SCH ₂ C=C; and IPG core], 2.81 -2.80 (m, 1 H, SA, H-3e), 2.01 (s, 1 H, SA, NHAc), 1 .88-1 .84 (m, 1 H, SA, H-3a). 13C{1 H} NMR (125 MHz, D20) : 1 74.93 (0=C, NHAc), 172.1 9 (C-1 , SA), 144.21 (C=CH), 125.35 (C=CH), 83.72 (C-2, SA); 79.86, 79.75, 79.65, 7954, 79.43, 77.61 , 74.89, 71 .12, 68.93, 68.67, 68.34, 68.20, 68.09, 67.88, 62.80, 62.57, 60.74, 60.66 (IPG backbone, C-6, C-8, C-7, C-9, C-4, SA), 51 .66, 51 .20 (C-5, SA), 40.01 (C-3), 23.13 (NHAc), 22.06 (-CH ₂ -, SA). IR (film) : v = 3355, 2929, 2879, 1698, 1 640, 1 558, 1457, 1430, 1 374, 1346, 1 31 5, 1 276, 1223, 1 120, 1060, 955 cm ^"1 . PDI=1 .28.

All the final sialic acid conjugates were well characterized by HNMR, ³ CNMR, IR, and GPC. The sialic acid ligands were statistically distributed on the polymer backbone of IPG. The number of sialic acid ligands per polymer in the final compounds were determined by degree of functionalization (DF) (also referred to as degree of substitution) as obtained by HNMR and the molecular weight of the polyglycerol core (cf. Table 4 below). In case of IPGSA 5 with 1 00% SA functionalization, there is no free hydroxyl group on the IPG chain.

Detailed description of synthesis and characterization of intermediates and final compounds All reagents and solvents were purchased from commercial suppliers and used without further purification. Reactions requiring dry or oxygen-free conditions were carried out under argon in Schlenck glassware. H and ³ C NMR spectra were recorded on Bruker AMX 500 (500 and 125 MHz for 1 H and 13CNMR, respectively) and Delta Joel Eclipse 700 (700 and 175 MHz for H and ³ CNMR, respectively) spectrometer at 25 ^° C and calibrated by using the deuterated solvent peak. Infrared (IR) spectra were recorded with a Nicolet AVATAR 320 FT-IR 5 SXC (Thermo Fisher Scientific, Waltham, MA, USA) with a DTGS detector from 4000 to 650 cm ^"1 . For ESI measurements a TSQ 7000 (Finnigan Mat) instrument and for high-resolution mass spectra a JEOL JMS-SX-102A spectrometer was used. Molecular weight distributions of IPG 1 , and the post functionalized polymers determined by means of GPC coupled to a refractive index detector (Rl) obtaining the complete distribution (Mn, Mp, Mw, dispersity). Measurements were carried out under highly diluted conditions (5 mg/mL) from a GPC consisting of an Agilent 1 100 solvent delivery system with pump, manual injector, and an Agilent differential refractometer. Three 30 cm columns (PPS: Polymer Standards Service GmbH, Germany; Suprema 100 A, 1000 A, 3000 A with 5 and 10 μηι particle size) were used to separate aqueous polymer samples using water with 0.1 N NaN0 ₃ as the mobile phase at a flow rate of 1 mL/min. The columns were operated at room temperature with the Rl detector at 50 'C. The calibration was performed by using certified standards pullulan (linear) and dextran (branched) from PSS. WinGPC Unity from PSS was used for data acquirement and interpretation. DLS measurements of the various polymers were conducted by using a NanoDLS particle sizer (Brookhaven Instruments Corp.) at 25 ^° C. Aqueous samples were filtered through 0.2 mm filters prior to analysis. Water of Millipore quality was used in all experiments. An USHIO super high mercury lamp (USH 102d, 100 W) was used for UV assisted thiol-ene click reaction.

Naturally occurring sialic acids constitute a family of more than 50 structurally distinct nine- carbon 3-deoxy-ulosonic acids, the most widespread derivative being 5-N-acetyl-neuraminic acid (Neu5Ac). In the context of the instant disclosure, the abbreviation "sialic acid" (SA) is used for for Neu5Ac. The polymers were characterized by NMR, and SEC for the determination of absolute molecular weights and dispersity (Mw/Mn). Number of ligands per particle in the final compounds were determined by degree of functionalization as obtained by HNMR and the molecular weight of the polyglycerol core. Postfunctionalized average molecular weight of the final compounds was calculated by using the Mn of the polyglycerol core and the experimental degree of functionalization as determined by HNMR.

Characterization of IPGSAs 2-6 Size analysis in aqueous solution

Dynamic light scattering was used for the determination of hydrodynamic diameter (dn) of final sialic acid conjugates IPGSAs 2-6 in PBS (pH 7.4, 10 mM). All polymer conjugates were freely soluble in aqueous solution as all sample solution appeared clear. Measurements were carried out at a concentration of 1 g/L for all the polymers. The results have been summarized in Table 4, and the corresponding volume size distribution profiles are displayed in Figure 2.

Table 4: Characterization of IPG 1 and IPGSAs 2-6.

PGOH/ M _n ^a (kDa) DF ^b % Vol. Z _av . ^c (d. nm) PDI ^C SA polymer ^d PGSA

11 10 - 6.38±0.16 0.71 ±0.04

12 15 10 9.81 ±2.13, (95.36)% 0.45±0.06 14

13 31 .7 44 10.29±1 .10, (98.8%) 0.59±0.008 60

14 46 73 10.24±1 .05, (87.6%) 0.32±0.04 99

15 59 100 1 1 .26±0.77, (96.1 )% 0.42±0.07 135

16 21 20 7.17±0.35 0.68±0.01 27 ^a IPG 11 : average molecular weight 10 kDa was used; Mw/Mn = 1 .29; postfunctionalized Mn calculated by using the Mn of the polyglycerol core and the experimental degree of functionalization.

b determined by ¹ HNMR analysis

c determined by DLS (PBS, 10 mM, pH 7.4) ; the values represent the means of at least three experiments ± standard deviation

d the number of sialic acid (SA) units per polymer was calculated from DF as determined by ¹ HNMR

Functionalized IPGSAs 2-6 were found to be bigger than the non-functionalized IPG 1 although the increase in hydrodynamic diameter with increasing SA conjugation was a bit random (Figure 2). A representative compound, namely IPGSA 4 was analyzed by cryo-TEM at two different concentrations, namely, 1 μΜ and 5 μΜ. Most of the IPGSA 4 conjugates were observed as small spherical clusters with a diameter of less than 7 nm along with some aggregates in the form of large spherical morphologies with a diameter of greater than 30 nm. The morphology of the lin PGSA particles were not perfectly spherical. This can be due to the presence of coiled structure of linear polymer in aqueous medium. Generally, a tendency to form aggregates was found to be higher at higher concentration of IPGSA 4 in cryo-TEM micrographs at different concentrations, although the concentrations used in the different assays were much lower than 1 μΜ nanoparticles. The dhi of IPGSA 4 was observed to be 10.24 nm along with some aggregates in the size range of 40-50 nm in DLS. This is in line with the size observed in cryo-TEM micrographs.

Hemagglutination inhibition (HAI) assay

The sialic acid conjugated polyglycerols were investigated for hemagglutination inhibition (HAI) using 4 HA units of influenza A virus (X-31 , H3N2) strain as a preliminary screening test, as shown in Figures 3A and 3B. To express the effect of multivalency, 2,6-sialyllactose was used as a control.

Figure 3A shows the results of hemagglutination inhibition with linear polyglycerol sialosides at different sialic acid densities. Error bars indicate the standard error of the mean (n≥3). Figure 3B shows the enhancement factor of each tested sialoside based on its inhibition potency compared to monomeric 2,6-sialyllactose. Figure 3C shows the results of binding inhibition of r18 labeled X31 virus to human red blood cells with the most potent linear polyglycerol compounds as inhibitor. Error bars indicate the standard error of the mean (n≥3). Data were fitted with a sigmoidal dose response to obtain IC ₅ o values. The results of figures 3A to 3C are summarized in Table 5.

To get an idea how differently functionalized linear polyglycerol sialosides (IPGSAs) behave in the HAI assay, five IPGSAs 2-6 with increasing degree of SA functionalization with hydrodynamic diameter ranging from 7-1 1 nm were synthesized and screened. The two compounds IPGSA 3 with 44% and IPGSA 4 with 73% SA functionalization appeared to have much lower K, values 19.53 μΜ and 8.46 μΜ, respectively, than the other IPGSA compounds.

In the following section, a simple statistical mechanical model is provided that can qualitatively explain the observed trends, and possibly shed some light on the origins of the observed behavior.

Table 5: Inhibitor constants from the hemagglutination and erythrocyte binding inhibition assay of the most potent candidates and monomeric a-2,6-sialyllactose as control. All values are expressed in terms of sialic acid concentration or nanoparticle concentration with its standard error of the mean (n≥3).

Compound Ki(HAI) Binding

Ki(HAI) Enhancement Binding

(SA%, Particle inhibition

SA [μΜ] factor (β) inhibition scaffold) [nM] ICSOSA [mM] IC50 particle

[μΜ] α-2,6-

481 15±1413 - 1

sialyllactose

IPGSA 1

- 2000000 ^* - - 4000 ^* (0%,10kDa)

IPGSA 2 200.0 ±

2000 ^* 139860 ^* - 2.8±0.3

(10%,-I OkDa) 21 .4

IPGSA 3

19.5±5.9 328.0±98.4 2463 0.2±0.0 3.3±0 (44%,10kDa)

IPGSA 4

5.9±2.7 59.2±27.2 8212 0.2±0.1 2.0±1 .0 (73%,10kDa)

IPGSA 5

1000.0±0.0 7407.4±0.0 48 1 ^* 7.4 ^* (100%,10kDa)

IPGSA 6 1860.0±620. 68888.9 ±

26 4 ^* 148.1 ^* (20%,-I OkDa) 0 22963.0

^* Maximum tested concentration

Determination of binding affinity of polyglycerol sialosides to whole virus particles with microscale thermophoresis

Figure 4 shows a change in fluorescence upon binding of polyglycerol sialosides at different concentrations to fluorescently labeled X31 . The examined IPGSAs show one binding step. Binding constants are described as apparent dissociation constants (Kdapp) derived from a one site binding fit of each binding step. Error bars indicate the standard error of the mean (n≥3). The determined values are summarized in the following Table 6.

Table 6: Apparent dissociation constants (Kdapp) of the most potent inhibitors.

Compound Kdapp

(SA%, scaffold) Particle [μΜ]

a-2,6-sialyllactose 2500

IPGSA 1 (0%,10kDa) -

IPGSA 3 (44%,10kDa) 10.1 ±2.9

IPGSA 4 (73%,10kDa) 0.6±0.3 Theoretical model for the binding of IPGSA with influenza virus

An important question arising from the experimental data of Figure 3A is under which conditions should one expect an optimum value for the degree of functionalization DF, and what are the parameters controlling the position of this minimum.

In this regard, the inventors propose a simple model in qualitative agreement with the obtained data, based on the statistical mechanics of multivalent binding applied to the specific system used. In the following, only the main assumptions and results of this model will be summarized. Typically, hemagglutination inhibition due to multivalent constructs is proportional to the construct dissociation constant [9], which in turn is proportional to the exponential of the binding free-energy AGbmd. Naively, one could think that a larger DF would always correspond to a lower AGbind since more binding configurations can form, but this is not always the case. For the system disclosed here, the distance between possible binding sites and the length of the ligands makes it impossible for a ligand to be within binding distance of two different receptors at the same time. In this case, binding occurs always with an "indifferent topology" [10], and the number of binding configurations W is just equal to the number of ligands N within binding distance from the viral receptor. In this case, AGbind decreases only logarithmically with N, and hence with DF, since the two are directly proportional to each other. However, the steric repulsion between a construct and the binding surface grows linearly with N, since each of the unbound ligands will feel some repulsion from the surface due to excluded volume effects. The full calculation of AGbmd ives:

_ AGsteric AG _hond where k _B is Boltzmann's constant and T the system temperature. In equation 1 , AGsteric measures the repulsive free-energy due to steric interactions between a single ligand and the surface of the virus when the multivalent construct is at binding distance from the surface. AGbond is the energy of formation of a single bond. AS _aV idit _y > 0 is the further attractive contribution that arises in multivalent binding coming from the various distinct possible binding configurations achievable [10].

Given equation 1 , AGbind has a minimum at a particular value of N (and hence DF), given by Nopt = 1 /AGsteric/kEsT, as also observed in the experimental data of Figure 3. The repulsive contribution AGsteric can be calculated by modelling the ligands as rigid rods, and two further experimental observations can then be explained. First, the leading term in AGstenc does not depend on the nanoparticle curvature, in accordance with the experimental data. Furthermore, within the same description, AGsteric kEsT is lower for a ligand grafted to a linear polymer compared to when it is grafted on the impenetrable surface of a nanoparticle, which also explain why the minimum for IPGSA is shifted to higher values.

Whereas in general the exact value of N _op t depends on the specific model used to describe the ligand, and an accurate calculation might require use of detailed atomistic simulations, the possible presence of a minimum arises from a very general mechanism of competition between binding entropy and steric repulsion, and is thus a general aspect of the system disclosed here.

Infection inhibition at the early stage of infection

Figure 5A shows cell viability after a treatment of MDCK II cells with inhibitors at different concentrations for 24 h. Cell viability was monitored by using a tetrazolium salt (promega). Error bars indicate the standard error of the mean (n≥3).

Figure 5B shows infection inhibition curves of the most potent inhibitors from an infection assay with X31 at a multiplicity of infection (MOI) of 0.05. X31 was pretreated with inhibitor for 30 min, then added to cells for 1 h, followed by a washing step to remove unbound virus. 24h after infection the cell viability was monitored by using a tetrazolium salt (promega), which gets reduced upon contact to mitochondria. Error bars indicate the standard error of the mean (n≥3). The dose response was fitted with a sigmoidal dose response to obtain IC ₅ o values (summarized in Table 7).

Figure 5C shows infection inhibition with 1 mM sialic acid of the most potent inhibitors 24 h after infection with X31 at an MOI of 0.1 . X31 was pretreated with inhibitor for 30 min, then added to cells for 1 h, followed by a washing step to remove unbound virus. Viral nucleoprotein expression was monitored by immunostaining against X31 nucleoprotein. Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI) to assess the ratio of infected cells.

Figure 5D shows normalization of infected cells to the untreated case. Inhibitor treated cells showed significantly lower amount of viral protein expression 24 h after infection, while a control inhibitor did not. Error bars indicate the standard error of the mean (n≥3). A variance analysis (ANOVA) was implied to determine significance ( ^*** p<0.001 ). The experimental results depicted in Figures 5A to 5D are summarized in the following Table 7.

Table 7: Infection inhibition constants (IC50) of all inhibitors after infection of MDCK II cells at an MOI of 0.05 for 24 h.

The IC50 value has been calculated for all IPGSAs. Within the investigated concentration, no toxicity could be detected. Therefore, there is no need to calculate an SI value (therapeutic index).

Infection inhibition at the late state of infection

Figure 6 shows the results of an experiment in which MDCK II cells were infected for 1 h with X31 at an MOI of 0.1 . Then inhibitor was added at different concentrations. 24 h after infection the viral titer was determined by a plaque assay. Error bars indicate the standard deviation (n≥3).

Study in mouse

Figure 7A and 7B show the results of an infection inhibition experiment with X31 in BALB/c mice. X31 was pretreated with inhibitor at the indicated concentrations, followed by intranasal application of the mixture to sedated mice.

In the experiments, the results of which are depicted in Figure 7A, the infection-related loss in body weight was daily controlled for 1 1 days. IPGSA 3 at 5 mM and 1 mM sialic acid concentrations was compared to infected/untreated, uninfected/untreated or infected/control treated (linPGOH 1 ) cases. The effect of a cytokine storm related loss in body weight was controled by using UV inactivated X31 , which was not infectious, but still able to hemagglutinate as the untreated virus. Error bars indicate the standard error of the mean (n≥5).

Figure 7B shows the results of a comparison of IPGSA 3 to equimolar nanoparticle concentrations of Oseltamivir (tamiflu) or the combiantion of both inhibitors. Error bars indicate the standard error of the mean (n≥5).

Summarizing, IPGSA 3 with 44% SA was best in its inhibition potency compared to its molecular mass. Therefore, it was further investigated at the late stage of infection and in vivo. under both experimental settings the inhibitor was able to reduce infectivity significantly. In vivo 5 mM sialic acid of IPGSA 13 (83μΜ nanoparticle) was able to almost fully block infection. The addition of Tamiflu further boosted the inhibition, while the Tamiflu alone was not able to cause potent infection inhibition. Therefore, a combined usage of both IPGSA, in particular IPGSA 13, and Tamiflu for a therapeutic application is herewith proposed.

Post infection inhibition of viral propagation

Linear polyglycerols of the type IPGi ₀ SA ₀ .4o (degree of functionalization with sialic acid: 40 %, 10 kDa scaffold) are able to inhibit viral propagation post infection of the human seasonal influenza strain X31/H3N2 and two avian influenza strains Mallard/H3N2 and ltaly/H7N1 . The results of an according experiment are depicted in Figure 8.

MDCK II cells were infected for 1 h with A/X31 (H3N2), A/Mallard/439/2004 (H3N2) or A/turkey/ltaly/472/1999 (H7N1 ) virus at an MOI 0.01 . Subsequently, cells were washed and incubated in the infection medium supplemented with control substance or inhibitor at indicated concentrations (corresponding to 167 nM nanoparticle concentrations for IPG 1 or 10 μΜ SA equivalents for IPGi ₀ SA ₀ .4o, respectively). 24 h post infection (p.i.) the viral titer was determined by a plaque assay. The error bars indicate the standard error of the mean (SEM) (n≥3). The obtained data was statistically analyzed using a two-tailed Mann-Whitney t-test. The observed inhibition of viral propagation was found to be significant ( ^** p<0.01 , ^*** p<0.001 , ^**** p<0.0001 ).

These findings demonstrate the high efficacy of linear polyglycerol sialosides as antiviral compounds in a therapeutic experiment setup.

Neuraminidase activity inhibition assay

Influenza neuraminidase is a tetrameric glycoprotein on the viral envelope. Its role for viral entry is still under debate, however, it is indispensable for viral release from host cells. The enzyme has sialidase activity, hydrolyzing terminal Neu5Ac residues of glycobranches. Tamiflu (Oseltamivir) or Relenza (Zanamvir) acting as competitive inhibitors demonstrated potent enzyme inhibition (low nM IC ₅ o values), and as a consequence antiviral activity of viral propagation in vivo and in vitro. However, a fast development of resistant strains against these inhibitors showed the limitation of such drugs for upcoming influenza mutants. The inventors' studies on viral neuraminidase activity in the presence of inhibitors suggest the highly potent ability of linear polyglycerol sialosides to interact and inhibit the catalytic neuraminidase activity with an IC ₅ o of 62 μΜ. The results are shown in Figure 9.

To obtain these results, X31 /H3N2 virus (64 HAU) was maintained in the presence of a dilution series with IPGi ₀ SA ₀ .4o for 30 minutes at room temperature, before the substrate MUNANA (20-(4-methylumbelliferyl)-a-D-N-acetylneuraminic acid) was added to the virus-inhibitor mix. Upon enzymatic conversion of MUNANA to the fluorescent methylumbelliferone, fluorescence (excitation: 360 nm; emission: 450 nm) was recorded using a platereader.

Since IPGi ₀ SA ₀ .4o 1 b is functionalized with the natural sialoside Neu5Ac, mutation of the structurally conserved and essential binding pocket of neuraminidase is unlikely. Taken together, linear polyglycerol sialosides are able to bind two essential target sites (SA binding site of HA and NA) at the same time, and due to its bio mimicking properties this compound might less likely provoke a development of resistance upon mutation of the binding site. Additionally, linear polyglycerol sialosides have a much lower IC ₅ o value than hyperbranched polymers carrying a sialic acid residue previously described. Detailed description of materials and methods used for biological assays

Virus material : X31 viruses (influenza strain A/Aichi/2/68 H3N2, reassorted with A/PuertoRico/8/1934 H1 N1 ) were harvested from allantoic fluid of hen eggs. Virus isolates were clarified upon low speed centrifugation (300 g, 10 min) and concentrated by ultracentrifugation (100,000 g, 1 h). Protein concentrations were determined using a standard BCA assay (Kit from Thermo Fisher Scientific, Belgium). Viral titer was determined by TCID50 and hemagglutination. Hemagglutination units (HAU) have been assessed with 1 % human red blood cells (DRK Berlin (AB+)) in PBS [1 ].

Cell culture: MDCK-II (Madin-Darby Canine Kidney Epithelial) cells have been maintained in DMEM (supplemented with 10% fetal calf serum and streptomycin/penicillin) at 37°C and 5% C0 ₂ . Hemagglutination inhibition assay (HAI): Inhibitors were two-fold serially diluted in PBS. Then, 4 HAU containing 4-10 ⁷ virus particles [1 1 ] were added to all wells. After 30 min incubation at room temperature, 50 μΙ of a 1 % erythrocyte solution (~2-10 ^s cells/μΙ) was added, gently mixed and incubated for 60 min at room temperature. Humanerythrocytes predominately carry oc-2,6 ^' -sialosugars. The inhibitor constant ( ί(ΗΑΙ)) reflects the lowest inhibitor concentration, which is necessary to achieve 100% inhibition. To check for full hemagglutination inhibition, the microtiter plate was tilted by 60 ° to cause droplet formation from the red blood cell pellet [12]. Single cell binding inhibition assay: Experiments were performed as previously described [1 ]. In short, influenza A virus (X31 ) was labeled with 20 μΜ octadecylrhodamine B chloride (R18) in PBS and free dye was removed via virus pelleting. 5 μg labeled virus was incubated with two-fold serial dilutions of binding inhibitors, and incubated for 30 min at room temperature (0.5 μ9/μΙ X31 ). The virus-inhibitor complex was mixed with 40 μΙ of a 1 % erythrocyte solution and incubated for 30 min at room temperature. Unbound virus was removed upon centrifugation for 10 min at 1500 g. The resuspended cell pellet was analyzed for R18 fluorescence by flow cytometry (Beckton Dickinson FACScan flow cytometer). Data were acquired by using BD CellQuest. From each measurement, a single cell population was gated and median fluorescence was normalized to an untreated virus control.

Infection inhibition assay: Viral cytotoxicity inhibition assays have been performed using an MTS reagent (Promega). 3- 10 ⁴ MDCK II cells were seeded the day before infection. X31 have been pretreated with inhibitors in a two-fold dilution series for 30 min at room temperature under slight agitation. Cells were washed once with PBS (Mg ²⁺ , Ca ²⁺ ) and pretreated virus (MOI 0.05) was added for 1 h at room temperature to allow binding. Unbound virus was removed by washing once with 37 ^< Ό infection medium (DMEM, 2 mM Glutamine, 0.1 % FCS, 0.1 % BSA, 2.5 μg TPCK, Pen/Strep), and subsequently incubated for 24 h at 37 °C. Two hours before the readout, 20 μΙ MTS solution was added to each well and put back to 37 °C for the remaining time. Finally, absorbance at 490 nm was measured and data were normalized to maximum viral cytotoxicity.

(Inf ected _treated — Infected _untreated )

Infection inhibition (%) =

(Uninf ected _untreated Infected _untreated )

Viral NP expression analysis: X31 virus has been pretreated in an inhibitor titration experiment with polyglycerols for and maintained for 30 min at room temperature. The day before MDCK-II, 5x10 ⁴ have been seeded in an 8-well plate (ibidi). On the next day, cells have been washed once with PBS ⁺⁺ and infected with treated or untreated X31 (MOI 0.1 ) for 1 hour on ice to synchronize the infection. Unbound virus was remove by discarding the supernatant and washing with infection media (DMEM, 0.1 % FCS, 0.1 % BSA). Following, cells were incubated for 24 h at 37 ^< Ό and stained for viral nucleoprotein and cell nuclei with DAPI.

Here, cells have been washed, fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.5% (w/v) triton X100. Further, cells have been blocked with blocking buffer (PBS ⁺⁺ , supplemented with 3% BSA and 100 mM glycine) for 2 hours at room temperature (rt). Immunostaining was performed with fluoresceinisothiocyanat (FITC) coupled mouse aNP monoclonal antibody (1 :500 in PBS ⁺⁺ with 1 % BSA) by incubating over night at 4 ^< €. Further, cells were stained for their nuclei with DAPI. Stained cells have been washed with PBS++ and then used for microscopy. Next, images were recorded by using an inverted confocal microscope (Olympus 1X81 ). Subsequently, images were analysed for viral NP expression and DAPI signal by using an imageJ software. The NP signal was expressed in ratio to infected, untreated cells.

Additionally, references [13] to [16] might be considered interesting. Thereby, references 13- 15 describe L- and P- selectin inhibition and anti-inflammatory activity of sulfated and non- sulfated polymers based on hyperbranched polyglycerols. Reference [16] regards the anti- pathogenic activity of multivalent glycoconjugates.

References cited in the preceding sections

[1 ] Papp, I. ; Sieben, C ; Sisson, A. L ; Kostka, J.; Bottcher, C; Ludwig, K. ; Herrmann, A. ; Haag, R. ChemBioChem 2011 , 12, 887-895

[2] Sunder, A.; Hanselmann, R. ; Frey, H. ; Mulhaupt, R. Macromolecules 1999, 32, 4240-4246;

[3] Haag, R.; Sunder, A; Stumbe, J. F. J. Am. Chem.Soc. 2000, 122, 2954-2955.

[4] Sunder, A.; Mulhaupt, R.; Haag, R.; Frey, H. Adv. Mater. 2000, 12, 235-239.

[5] Ul-haq, M. I. ; Shenoi, R. A. ; Brooks, D. E.; Kizhakkedathu, J. N. J. Polym. Sc. Part A:

Polym. Chem. 2013, 51, 2614-262.

[6] Wilms, D. ; Wurm, F. ; Nieberle, J.; Bohm, P. ; Kemmer-Jonas, U.; Frey, H. Macromolecules

2009, 42, 3230-3236.

[7] Gan, Z.; Roy, R. Can. J. Chem. 2002, 80, 908-916.

[8] Ogura, H. ; Furuhata, K.; Itoh, M.; Shitori, Y. Carbohydr. Res. 1986, 158, 37-51

[9] Vonneman et al, "Size Dependence of Steric Shielding and Multivalency Effects for Globular Binding Inhibitors", J/ACS 2015, 137.

[10] Kitov, P. I. and Bundle, D.R. "On the Nature of the Multivalency Effect: A Thermodynamic Model", JACS 2003, 125.

[1 1 ] Cross, G. Semin. Avian Exot. Pet Med 2002, 11, 1 5-18. The preparation and chemical characteristics of hemoglobin-free ghosts of human erythrocytes. J.T. Dodge 1962, Archives of biochemistry and biophysics.

Dernedde, J ; Papp, I; Enders, S; Wedepohl, S; Paulus, F; Haag, R. Journal of Carbohydrate Chemistry 2011 , 30, 347-360.

Papp, I; Dernedde, J ; Enders, S; Haag, R. Chem. Comm. 2008, 44, 5851 -5853.

Dernedde, J ; Rausch, A; Weinhart, M ; S. Enders, S; et al, Proc. Nat. Aca. Sci. USA 2010, 107, 19679-19684.

Bernardi, A; Jimenez-Barbero, J ; Casnati, A. ; et al. Chem. Soc. Rev. 2013, 42, 4709-4727. Erberich, M. ; Keul, H.; Moller, M. Macromolecules 2007, 40, 3070-3079.