ac

. GBC 1

Accepted 2 May 2008

Beta-amyloidAlzheimers diseasePhotopolymerizationOligosaccharide

e plica

e mo

membrane depolarization, changes in membrane capacitance [16],membrane destabilization [17], pore formation [18,19] or freeradical generation [20,21]. Consistent with these observations isthat the rst step in Ab toxicity is Ab binding to a membranecomponent.

tetraose (LNT), but varied in number and orientation of sialic acid inthe oligosaccharide. We tested the intrinsic toxicity of base oligo-saccharides and their relative abilities to prevent Ab toxicity in anN2A mouse neuroblastoma cell line. We developed a procedure tophotopolymerize the oligosaccharides, then puried and charac-terized the resulting multivalent sialic acid polymers. Finally weexamined the Ab binding and toxicity attenuation properties of thenovel photocrosslinked sialic acid containing polymers. This workcould contribute to the development of a new class of therapeutics

Contents lists availab

Biomat

ev

Biomaterials 29 (2008) 34083414* Corresponding author. Tel.: 1 410 455 3405; fax: 1 410 455 1049.dementia in the aging population [1,2], is characterized by thepresence of neurobrillary tangles and amyloid plaques [3]. Themain protein component of the plaques is beta-amyloid peptide(Ab) [46]. It has been hypothesized by many that Ab playsa causative role in the neurodegeneration associated with AD[3,4,710].

The mechanism by which Ab causes neurotoxicity is the subjectof much debate. Most believe that Ab toxicity is linked to the for-mation of aggregated species, with the most toxic species being anintermediate between monomer and brils [1114]. Some believethat Ab acts via association with the cell membrane [15], to cause

gangliosides containing multiple sialic acids, than it does for otherglycolipids [9,2830]. Based on these ndings, we hypothesizedthat we could develop a membrane mimic that was multivalent insialic acids, that would bind Ab, thus sequestering it, making the Abunavailable for cell binding, thereby reducing its neurotoxicity.

We chose to prepare the multivalent sialic acid materials via thephotopolymerization of a series of sialic acid containingoligosaccharides. The three oligosaccharides were 30-sialyl-N-acetyllactosamine (30SLN), sialyllacto-N-tetraose b (LST b), anddisialyllacto-N-tetraose (DSLNT). Each of the selected compoundscontained the same or similar carbohydrate backbone of lacto-N-Sialic acidMultivalent

1. Introduction

Alzheimers disease (AD), is thE-mail address: tgood@umbc.edu (T.A. Good).

0142-9612/$ see front matter 2008 Elsevier Ltd.doi:10.1016/j.biomaterials.2008.05.001use of a biomimic multivalent sialic acid compound that would compete with the cell surface for Abbinding. To explore this hypothesis, we synthesized a series of photocrosslinked sialic acid containingoligosaccharides and tested their ability to bind Ab and attenuate Ab toxicity in cell culture assays. Weshow that a polymer prepared via the photocrosslinking of disialyllacto-N-tetraose (DSLNT) was able toattenuate Ab toxicity at low micromolar concentrations without adversely affecting the cell viability.Polymers prepared from mono-sialyl-oligosaccharides were less effective at Ab toxicity attenuation.These results demonstrate the feasibility of using photocrosslinked sialyl-oligosaccharides for preventionof Ab toxicity in vitro and may provide insight into the design of new materials for use in attenuation ofAb toxicity associated with AD.

2008 Elsevier Ltd. All rights reserved.

st prevalent cause of

Increasing evidence indicates that sialic acid containing gan-gliosides are pivotal in the mechanism of Ab binding to cells [7,2227]. Ab has higher afnity for gangliosides which are clustered andKeywords:

dence, we have hypothesized that the Ab-membrane sialic acid interaction could be inhibited throughAvailable online 3 June 2008the Ab-neuronal membrane interaction plays a crucial role in Ab toxicity. More specically, it is thoughtthat Ab interacts with ganglioside rich and sialic acid rich regions of cell surfaces. In light of such evi-Development of photocrosslinked sialicfor use in Ab toxicity attenuation

Christopher B. Cowan a, Gerard L. Cote b, Theresa AaChemical and Biochemical Engineering, University of Maryland, Baltimore County, UMbBiomedical Engineering, Texas A&M University, College Station, TX 77843, USA

a r t i c l e i n f o

Article history:Received 21 February 2008

a b s t r a c t

b-Amyloid peptide (Ab), thdisease (AD), has been imp

journal homepage: www.elsAll rights reserved.id containing polymers

ood a,*

000 Hilltop Circle, Baltimore, MD 21250, USA

rimary protein component in senile plaques associated with Alzheimersted in neurotoxicity associated with AD. Previous studies have shown that

le at ScienceDirect

erials

ier .com/locate/biomater ia lsaimed at preventing Ab toxicity in AD.

teria2. Materials and methods

2.1. Materials

Sialic acid oligosaccharides were purchased from V-Labs, Inc (Covington, LA).The three oligosaccharides used all had either a lactosamine or lacto-N-tetraosebackbone with one or two sialic acids. DSLNT (Neu5Aca2-3Galb1-3(Neu5Aca2-6)-GlucNAcb1-3Galb1-4Glc), LST b (Galb1-3(Neu5Aca2-6)GlucNAcb1-3Galb1-4Glc),and 30SLN (Neu5Aca2-3Galb1-3GlucNAc) were used. Ab1-40 (>98% purity by HPLCand mass spectrometry) was purchased from Biosource International (Camarillo,CA). Mouse neuroblastoma N2A cells were purchased from ATCC (Manassas, VA).Cell culture reagents were purchased from Gibco-Invitrogen (Grand Island, NY).Propidium iodide (PI) was purchased from BD Biosciences (San Jose, CA). Iodo-Beads,G-5 desalting columns, Bolton hunter reagent (sulfo-SHPP) and CarboLink CouplingGel were purchased from Pierce Biotechnology (Rockford, IL). 125I was purchasedfrom Amersham Biosciences (Piscataway, NJ). Triethylamine, tetrabutylammoniumbromide, N-vinyl pyrrolidinone, and glycidyl methacrylate were purchased fromArcos Organics (Geel, Belgium). Irgacure 2959 was purchased from Ciba (Tarrytown,NJ). Periodic acid and resorcinol were purchased from Fisher (Fair Lawn, NJ). Allother chemicals were purchased from SigmaAldrich (St. Louis, MO).

2.2. Development of photocrosslinked sialic acid containing oligosaccharides

The development of the biocompatible photocrosslinked sialic acid containingpolymers was adapted from a procedure presented by Leach and coworkers [31] forthe generation of hyaluronic acid hydrogels. The rst step in the reaction was toderivatize the oligosaccharide with the photopolymerizable cross-linker, glycidylmethacrylate. The epoxide on the glycidylmethacrylatewas non-specically reactedto an alcohol on the oligosaccharide as follows. One milligram of oligosaccharide(DSLNT, 30SLN, or LST b) was added to 100 ml of dH2O. Triethylamine (2.2 ml) wasadded to the oligosaccharide to act as a strong nucleophile. Glycidyl methacrylate(2.2 ml) and tetrabutylammonium bromide (a surfactant, 2.2 mg) were added to themixture, which was then allowed to react overnight at room temperature. At thispoint, various methods were attempted to remove excess reagents including excessglycidylmethacrylate from themethacrylatemodied oligosaccharides, however, allresulted in extremely high losses of the methacrylate derivatized oligosaccharides,thus, the ensuing steps of the polymerization were carried out without purication.However, excess epoxide from the glycidyl methacrylate was deactivated by heatingthe reaction mixture at 60 C for 1 h. N-vinyl pyrrolidinone (3 ml) and Irgacure 2959(photoinitiator, 2 mg) were dissolved in the methacrylate derivatized oligosaccha-ride solution, at which time the reaction mixture was placed under a 365 nm lampfor 11 min, completing the polymerization procedure. The power delivered to theoligosaccharide during polymerization was approximately 70 mW/cm2.

2.3. Polymerized dot assay

To estimate the time of light exposure needed for crosslinking such thata slightly viscous solution was obtained, representing a photocrosslinked oligosac-charide of moderate molecular weight, a simple polymerized dot assay was de-veloped. Approximately 100 ml of glycidyl methacrylate derivatized oligosaccharidewas placed on a glass slide, towhich N-vinyl pyrrolidinone and Irgacurewere added.The mixture was irradiated for various lengths of time and the viscosity of theresulting polymer was qualitatively examined. At less than 10 min irradiation, theviscosity of the resulting mixture was indistinguishable from that of water. After14 min of irradiation a soft gel was formed, while after 21 min of irradiation a rmgel was formed. At between 11 and 12 min of irradiation, noticeable changes inviscosity of the reaction mixture were observed, however, no gelation was detected.

2.4. Purication of sialic acid polymer

Once the polymerizationwas completed, the polymer solution was puried andfractionated using ultraltration. Unreacted monomer, glycidyl methacrylate, andother reagents were removed from the polymer via separation using a 3000 Damolecular weight cutoff ultraltration membrane. Polymer that was retained afterultraltration was further fractionated using a 10,000 Da molecular weight cutofflter. The fraction of polymer that passed through the 10,000 Da membrane but wasretained by the 3000 Da membrane was used in the binding and toxicity studies.Using this fractionation strategy, we expected to collect crosslinked oligosaccharideswith a degree of polymerization between 2 and 8.

2.5. Proof of synthesis (FTIR)

Throughout the procedure of developing the photocrosslinked sialic acids, thechemical structures of the intermediates in reactions were checked using FourierTransform Infrared Spectroscopy (Perkin Elmer System 2000 FTIR, Boston, MA).

2.6. Periodateresorcinol assay for sialic acid determination

C.B. Cowan et al. / BiomaThe periodateresorcinol assay was utilized to colorimetrically determine theamount of sialic acid groups in a photocrosslinked polymer solution. In the assay,glycosidically bound sialic acids are oxidized with periodate, the product of which isreacted with the recorsinol reagent to give an easily detectable chromophore [32]. Abrief description of the microplate version of the assay is given. Forty microlitres ofa sialic acid containing sample were placed into a 96-well microplate. Ten micro-litres of 0.032 M periodic acid solutionwas added to each well and mixed by shakingthe plate for 5 min. The plate was then placed in an ice bath for 35 min to oxidize thesialic acids. Hundred microlitres of resorcinol reagent (0.6 g resorcinol in 60 ml of28% HCl, 40 ml of dH2O, and 25 mmol of CuSO4) was added to each well uponcompletion of the oxidation step and mixed on the shaker for another 5 min. Themicroplate was then heated at 80 C for 60 min in order to develop the chromo-phore. Upon completion, the microplate was removed from the oven, mixed for12 min, and absorbance at 650 nm was immediately read using a microplatereader. Alternatively, the chromophore could be stabilized by addition of 95% t-butylalcohol. Sialic acid concentration was linearly related to absorbance at 650 nm.Concentrations were estimated from a calibration using sialic acid standards.

2.7. Cell culture

N2A cells were grown in Minimal Essential Media (MEM) with the followingadded: 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 mg/ml streptomycin,2.5 mg/ml amphotericin B (fungizone), and 2.2 mg/ml NaHCO3. The cells were grownin a humidied 5% CO2/air incubator. All cells used during experimentation wereunder the 15th passage.

2.8. Ab solution

Lyophilized Ab was prepared by dissolving 1 mg of the peptide in 100 ml DMSO.The stock was then diluted to 20 mM in minimal essential media (MEM) and aggre-gated with gentle mixing at 34 rpm for 24 h. This method of aggregation consis-tently produced toxic aggregated species in our hands as determined by a number ofviability assays in different cell populations including the propidium iodide viabilityassay used with N2A cells.

2.9. Propidium iodide assay

Propidium iodide (PI) was used to measure the viability of the cells in all ex-periments. PI penetrates damaged membranes in dead and dying cells, and uo-resces when bound to double-stranded DNA [33]. Live cells are membraneimpermeant, resulting in low uorescence emission for this cell population. N2Acells were harvested and seeded in a 96-well at bottom plate at a density of 50,000cells/well. Cells were incubated for 24 h in MEM to allow cell adhesion, at whichtime mediumwas removed and replaced with either fresh medium, oligosaccharideor photocrosslined oligosaccharide, Ab, or Ab plus oligosaccharide or Ab plus pho-tocrosslined oligosaccharide. Cells were incubated for an additional 24 h in a CO2incubator, at which time cells were stained with PI for the viability assay. Super-natant from the wells was collected and placed in another 96-well conical bottomplate to make sure detached dead cells were counted in the viability assay. To theoriginal plate with attached live cells, 100 ml of 0.5 g/L trypsin was added to detachcells. Cells in both the supernatant containing plate and the trypsin treated platewere stained with 12 ml of 33 mM PI for 30 min. After staining, the cells from thesupernatant containing plate were then added back to the original plate, and thesuspension was gently mixed, such that the total cell population from the originalculture could be analyzed. Fluorescence staining of cells was assessed using owcytometry (FACSArray Bioanalyzer, BectonDickonson, Bedford, MA). Cells wereexcited with a 532 nm laser and uorescence was detected using a 564606 nmlter. Gating was done so as to obtain percentages of the total cell population thatwere viable. Normalized viability values were obtained by dividing the percentage ofviable cells in the sample by that in the control samples with no Ab or other agent.Signicance of the results was determined using a t test with p< 0.05, unless oth-erwise indicated. Normalized viability of cells with or without Ab varied betweenexperiments and batches of Ab used. Typical variations between experiments per-formed on different days were on the order of 10%. This variability is likely due toboth biological differences in cells along with differences in aggregation of Ab. Whenever possible, viability datawere always compared to data collected on the same dayto minimize these additional sources of variability.

Toxicity inhibition constants, KI, were estimated from normalized viability ofcells, V, treated with Ab at different concentrations of photocrosslinked oligosac-charide, CP, by tting data from an inhibition curve (Eq. (1)) using a non-linear leastsquares regression. Vmax is the maximum viability measured at high concentrationsof photocrosslinked oligosaccharide.

V VmaxCPKI CP

(1)

2.10. Radiochemical binding assay

Binding of Ab to sialic acid polymers was determined using radiochemical

ls 29 (2008) 34083414 3409techniques as described previously [1]. In brief, Ab1-40 was radioiodinated viaa modied Bolton Hunter method. Sulfo-SHPP (100 nmol) was iodinated with200 mCi of 125I using the IodoBead catalyst. The reactionwas carried out for 15 min in

pH 8.0 borate buffer in a volume of 140 ml. The catalyst was removed and 10 nmol(100 ml) of freshly prepared Ab1-40 was added to the iodinated sulfo-SHPP. Thereactionwas allowed to proceed for 3 h at 4 C. The iodinated peptidewas separatedfrom free 125I using a G-5 desalting column. The peptidewas eluted from the columnwith phosphate buffer and stored at 4 C until use. The free activity associated withradiolabeled Ab was determined by precipitation of the peptide with 20 wt% tri-chloroacetic acid in the presence of 1 mg/ml bovine serum albumin. Precipitateactivity was greater than 80%.

The sialic acid polymer was immobilized to a CarboLink agarose coupling gelaccording to manufacturers directions. Sialic acid polymer (1 mg) was added to1 mL of 0.1 M sodium phosphate buffer at pH 7.0 and dissolved. Polymer containingbuffer (1 mL ) was added to NaIO4 giving a nal concentration of 23 mM. This so-lution was reacted for 30 min at room temperature in order to oxidize aldehydes onsugars. Oxidized sialic acid polymer was then added to 2 mL CarboLink gel andallowed to react overnight with mixing. The gel was then centrifuged, supernatantdecanted, then washed 3 times with phosphate buffered saline to remove polymerunbound to gel. Gels were stored at 4 C until use.

Binding assays were carried out by incubating 125I-labeled Ab of different con-centrations with 10 ml of the immobilized sialic acid polymer at room temperaturefor 2 h. Bovine serum albuminwas added at a nal concentration of 1 mg/ml to blocknon-specic binding. After 2 h, sufcient time for binding to come to equilibrium,free Ab was removed from that bound to the immobilized gel by centrifugation at12,000 g for 5 min. The activities of both bound and unbound Abwere counted usinga scintillation counter (Wallac microBeta, Perkin Elmer). To estimate non-specicbinding of Ab to the CarboLink gel without sialic acid polymer, different concen-

peaks from the FTIR, a peak located at 1722 cm1 associated withthe ester, a peak located at 1300 cm1 which corresponds to anepoxide, and a peak located at 1152 cm1 that is associatedwith thealiphatic ether. The DSLNT spectrum is characterized by a broadpeak (or peaks) in the region of 32002900 cm1 associated withamines and hydroxyl groups on the oligosaccharide. When glycidylmethacrylate was reacted with DSLNT, the epoxide peak in the FTIRspectrum diminished in intensity while the ester group remained.Prominent peaks in the DSLNT spectra in the region between 1700and 800 cm1 could also be seen in the spectrum taken of theglycidyl methacrylate derivatized DSLNT.

After photopolymerization, the absence of residual unreactedepoxide was conrmed by examining FTIR spectra for the epoxidepeak at 1152 cm1. The degree of polymerization was estimatedboth from fractionation results and MALDI-TOF MS. The presenceand concentration of sialic acid in the resulting polymer weredetermined via the resorcinol assay. Typical degrees of polymeri-zation were estimated to be between 2 and 8.

C.B. Cowan et al. / Biomaterials 29 (2008) 340834143410trations of Ab were added to the unmodied gel. Free and bound Ab activity wasdetermined as described for the sialic acid modied gel. At the highest concentra-tions tested, non-specic binding of Ab to the CarboLink gel accounts for approxi-mately 25% of the total binding.

The equilibrium dissociation constant for Ab to the sialic acid polymer wasdetermined from the t of a Langmuir isotherm (Eq. (2)) to the binding data usinga non-linear least squares regression algorithm.

B BmaxCAbKAb CAb

(2)

where B is the amount of bound Ab to the sialic acid polymer-gel corrected for theamount of Ab bound to the gel alone, CAb is the concentration of free Ab, Bmax is thetotal Ab binding sites on the surface of the gel, and KAb is the equilibrium dissoci-ation constant.

3. Results

3.1. Synthesis of polymer

FTIR was used to conrm addition of glycidyl methacrylate tooligosaccharides and provide an estimate of unreacted glycidylmethacrylate remaining in the reaction mixture. Fig. 1 showsrepresentative spectra of glycidyl methacrylate, DSLNT, and theproduct of the reaction. Glycidyl methacrylate has three notable

0

0.2

0.4

0.6

0.8

1

1.2

70011001500190023002700310035003900Wavenumber (cm

-1)

Ester Group Remains

Epoxide Group Deleted

CH2CH2

CH2C

CCH OH

OHR, R,O

O

OOCC

HR+

H2alkalineph > 10

Fig. 1. Representative FTIR spectra of reactants, DSLNT (grey line), glycidyl meth-acrylate (dotted line) and product of glycidyl methacrylate reaction with DSLNT (boldline). Difference spectrum of DSLNT reaction product minus DSLNT is shown offset

(ne line). Upon reaction, the peak corresponding to the epoxide functional group onthe glycidyl methacrylate is depleted, while the peak corresponding to the esterfunctionality is undiminished, indicating that the desired reaction had occurred.3.2. Intrinsic toxicity and Ab toxicity attenuation properties ofoligosaccharide monomers

One of the goals of this research was to develop a compoundthat was not only able to inhibit Ab toxicity but that was bio-compatible as well. To that end, the intrinsic toxicity of the sialicacid containing oligosaccharide monomers and polymers weretested before examining the Ab toxicity attenuation properties ofthe materials. The intrinsic toxicity of sialic acid containing oligo-saccharide monomers is shown in Fig. 2. The results show that atconcentrations of oligosaccharide 1 mM or below, minimal toxicityassociated with the compounds was observed. Above the 1 mMconcentration, DSLNT exhibited small but signicant toxicity toN2A cells after 24 h exposure.

In Fig. 3 the viability of cells treated with 20 mM Ab and sialicacid containing oligomeric monomers is shown. The normalizedviability of N2A cells treated with aggregated Abwas 70%. The onlyoligosaccharide monomer that even marginally attenuated Abtoxicity was DSLNT, the oligosaccharide containing two sialic acidsper monomer.

3.3. Intrinsic toxicity and Ab toxicity attenuation properties ofphotocrosslinked oligosaccharides

The photocrosslinked oligosaccharide polymers were examinedfor both intrinsic toxicity and Ab toxicity attenuation properties

0.50.60.70.80.91.01.1

0 0.5 1 2[Oligosaccharide Monomer] (mM)

No

rm

alized

V

iab

ility

* *

Fig. 2. Viability of N2A neuroblastoma cells treated with sialic acid containing oligo-saccharide monomers for 24 h. DSLNT (solid bars), 30SLN (open bars), and LST b (stri-ped bars) were tested. Viability was assessed using a PI assay. Results were normalizedby viability of cells untreated with oligosaccharides. Positive controls, cells treatedwith 1.5% H2O2 for 4 h, were included in the assay. Viability of positive controls was

typically less than 20%. n 3 or more independent measurements. Error bars representstandard error of the mean. Asterisk (*) indicate viability was signicantly differentthan the negative control at p< 0.05.

using PI assays (Fig. 4). At concentrations below 1 mM, neitherphotocrosslinked DSLNT or photocrosslinked 30SLN exhibited anysignicant toxicity to cells. Data for photocrosslinked LST b are notshown, because at all concentrations tested, this polymer led to lossof viability in cells at levels comparable to positive controls (cellstreated with 1.5% H2O2). We noted that without careful purication

We examined the ability of photopolymerized DSLNT and 30SLNto attenuate the toxicity observed when cells were treated with20 mM Ab. As seen in Fig. 4, cells treated with Ab were only 78%viable, however, cells treated with high concentrations of eitherphotocrosslinked oligosaccharide and Ab, had viabilities signi-cantly higher than that cells treated with Ab alone (p< 0.05). At250 mM photocrosslinked sialic acid containing oligosaccharidescompletely attenuated Ab toxicity. Photopolymerized DSLNTperformed modestly better than photopolymerized 30SLN,p> 0.05, thus photopolymerized DSLNT was used in furtherstudies.

3.4. Relationship between Ab binding and toxicity attenuation ofphotopolymerized DSLNT

We hypothesized that multivalent sialic acid polymers such asphotopolymerized DSLNT would attenuate Ab toxicity by bindingto Ab, thus preventing Ab binding to the multivalent sialic acidregions on the cell surface. To test this hypothesis, we rst exam-ined if Ab bound to photopolymerized DSLNT, and then examined ifconcentrations of polymer necessary for Ab binding and toxicityinhibition were in the same range. As seen in Fig. 5, Ab binds tophotocrosslinked DSLNT. The equilibrium dissociation constant for

0.5

0.6

0.7

0.8

0.9

1.0

0.0 0.5 1.0 2.0[Oligosaccharide Monomer] (mM)

No

rm

alized

V

iab

ility

**

*

*

*

*

Fig. 3. Toxicity attenuation of 20 mM Ab using the sialic acid containing oligosaccharidemonomers. All cells were treated with 20 mM aggregated Ab for 24 h. At the same timeof Ab addition, different concentrations of oligosaccharides, DSLNT (solid bars), 30SLN(open bars), and LST b (striped bars), were added to cells. The dashed line representsviability of cells treated with Ab alone. The uncertainty in viability measurements ofcell treated with Ab alone for the experiment shown is with bars at 0 concentration ofoligosaccharides. Viability was assessed in N2A cells using the PI Assay. n 3 or moreindependent measurements. Error bars represent standard error. Asterisk (*) indicateviability was signicantly different than the Ab treated cells at p< 0.05.

C.B. Cowan et al. / Biomaterials 29 (2008) 34083414 3411of polymers by ultraltration with a 3000 Da molecular weightcutoff lter, all polymers were toxic at a level comparable to posi-tive controls (data not shown), indicating that unreacted reagentswere harmful to cells. We cannot conclude denitively if photo-polymerized LST b was toxic to cells or if we observed toxicity as-sociated with the photocrosslinked LST b as a result of incompletepurication of unreacted reagents.

0.9

1

iab

ility

*

*

**

**0.6

0.7

0.8

0 250 500

No

rm

alized

V

[Polymer] ( M)

Fig. 4. Viability of N2A cells treated with sialic acid containing photocrosslinkedoligosaccharides (polymers) in the presence or absence of 20 mM Ab. Viability of cellstreated with photopolymerized DSLNT without Ab (solid bars), photopolymerized30SLN without Ab (open bars), and photopolymerized oligosaccharides with 20 mM Ab(DSLNT: ne stripes; 30SLN: bold stripes). Photocrosslinked oligosaccharides wereadded at the same time as aggregated Ab. The dotted line represented viability of cellstreated with Ab alone. Cells were treated for 24 h prior to viability assessment usinga PI assay. Polymer concentrations are reported per sialic acid residue. n 3 or moreindependent measurements. Error bars represent standard error. All cells treated withphotopolymerized polymer and Ab had signicantly increased viability compared tocells treated with Ab alone. Asterisk (*) indicate that viability of cells treated with250 mM photopolymerized DSLNT and Abwas signicantly greater than viability of cellstreated with the same concentration of photopolymerized 30SLN and Ab. Asterisks (**)indicate that viability of cells treated with 500 mM photopolymerized DSLNT and Abwas signicantly greater than viability of cells treated with the same concentration ofphotopolymerized 30SLN and Ab. In all cases p< 0.05.the Ab binding to photopolymerized DSLNT, KAb, was estimated tobe 0.22 0.06 mM.

We measured Ab toxicity attenuation by photopolymerizedDSLNTat different concentrations of polymer (Fig. 6). We estimatedthe inhibition constants, KI, for photocrosslinked DSLNT at twodifferent concentrations of Ab, 20 and 40 mM, and found KI values of3.61.3 mM and 2.21.1 mM, respectively. The toxicity inhibitionconstants are not signicantly different from each other and onlyapproximately an order of magnitude greater than binding afnityof Ab to the photocrosslinked DSLNT.

While toxicity inhibition constants at the different concen-trations of Ab were the same, the maximum viability achievedor fractional increase in viability achieved in Ab treated cellsupon treatment with photopolymerized DSLNT, varied withdifferent concentrations of Ab. At 40 mM Ab, photopolymerizedDSLNT was less effective at attenuating Ab toxicity than at20 mM Ab.

0

0.5

1

1.5

2

2.5

3

3.5

4

0 50 100 150 200 250

Bo

un

d A

(p

mo

l)

[Free A ] (nM)

Fig. 5. Representative equilibrium binding isotherm of Ab bound to photocrosslinkedDSLNT. Filled circles represent experimental data. The line represents the t of themodel to the data. Isotherms were collected at room temperature. Ab was allowed to

come to equilibrium with 2.5 mg of immobilized polymer for 2 h before free and boundAb were measured. Bound Ab was estimated as Ab bound to polymer immobilized onthe CarboLink gel minus Ab bound to gel alone.

teria4. Discussion

In this work, we examined the feasibility of developing multi-valent sialic acid polymers that were both biocompatible and ableto attenuate Ab toxicity. Others have developed multivalent sialicacid polymers by conjugating sialic acid to a dendrimer backbonefor prevention of viral adhesion and infection that were effectiveboth in vitro and in vivo [34,35]. However, whenwe have previouslyused sialic acid modied dendrimers to attenuate Ab toxicity in cellculture models using neuroblastoma cells, dendrimer backbonetoxicity was evident at concentrations between 3 and 80 mMdendrimer, depending upon the dendrimer concentration [1,36].Therefore, we explored the development of multivalent sialic acidcontaining materials through different synthesis routes. Manyother investigators have used glycidyl methacrylate derivatizedpolysaccharides for preparing biocompatible hydrogels [3742],

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 20 40 60 80 100

Fra

ctio

na

l In

cre

as

e in

V

ia

bility

[DSLNT Polymer] ( M)

Fig. 6. Toxicity attenuation of Ab by photocrosslinked DSLNT as a function of polymerconcentration. Polymer concentrations are reported per sialic acid residue. N2A cellswere either treated with 20 mM (solid line) or 40 mM (dashed line) aggregated Ab.Different concentrations of polymer were added to cells at the same time as Ab. Cellswere treated for 24 h, after which viability was measured using a PI assay. Fractionalincrease in viability, dened as the viability of cells treated with Ab and photo-crosslinked DSLNT divided by the viability of cells treated with Ab alone, is plotted.n 3 or more independent measurements.

C.B. Cowan et al. / Bioma3412thus we explored the feasibility of using glycidyl methacrylate sialicacid oligosaccharides for preparing biocompatible soluble (lowmolecular weight) polymers for use in Ab toxicity attenuationapplications that might be useful in Alzheimers disease.

We rst tested the biocompatibility of both sialic acid containingoligosaccharide monomers and photocrosslinked polymers ina neuroblastoma cell line susceptible to Ab toxicity (Figs. 2 and 4).Below 1 mM, no signicant loss of cell viability was seen upontreatment of cells with sialic acid containing oligosaccharidemonomers, and for two of the sialic acid containing polymers madefrom photocrosslinked DSLNT and 30SLN, no signicant loss ofviability was seen at concentrations of 500 mM, the highestconcentrations tested. Photocrosslinked LST b was toxic to cells(leading to 20% or below viability in N2A cells, data not shown),however, we cannot rule out that toxicity of the polymer was due toresidual reagents from the polymerization that were not com-pletely puried from the polymer.

We examined the ability of sialic acid containing monomers andpolymers to attenuate Ab toxicity (Figs. 3 and 4). At concentrationsof 500 and 1 mM, monomer DSLNT and 30SLN modestly attenuatedAb toxicity, while LST b did not attenuate toxicity under the con-ditions tested (Fig. 3). In previous work, we have reported thatmonomeric sialic acid attenuated Ab toxicity at sialic acid concen-trations above 200 mM, but only at low (5 mM) concentrations of Ab[1]. Thus, modest toxicity attenuation seen with DSLNT and 30SLNare in line with results observed with the simpler saccharide, sialicacid. That LST b did not attenuate Ab toxicity under the conditionstested may be attributed to the different orientation of the sialicacid on the oligosaccharide (linked via a a2-6 linkage to N-acetyl-glucosamine compared to the a2-3 linkage to galactose on bothDSLNT and 30SLN). Most virions and siglecs (sialic acid bindingproteins) have specicity in the sialic acid linkage (a2-3 or a2-6)that they recognize [43,44]. During cancer metastasis, sialylation ofcell receptors is also orientation specic [4547]. Thus, it would notbe surprising if Ab was specic for sialic acid linkage, possibly ana2-3 linkage.

Photocrosslinked DSLNT and 30SLN were both able to almostcompletely attenuate Ab toxicity at high concentrations of polymer(Fig. 4). Photocrosslinked DSLNT and 30SLN had increased Ab tox-icity attenuation properties compared to monomeric DSLNT and30SLN. We attribute the enhanced toxicity attenuation properties ofthe sialic acid containing polymers to the increased sialic acidvalency in these materials. There are many examples of the role ofvalency in adhesion of ligands to multivalent materials [4853].More specic to Ab, others have shown for biological membranesthat Ab binding afnity to sialic acid containing gangliosides in-creased in clustered or multivalent regions of the membranecompared to membrane regions where sialic acids or gangliosideswere unclustered [25,30].

We hypothesized that multivalent sialic acid polymers madefrom photopolymerized DSLNT attenuated Ab toxicity by bindingAb and sequestering it, making it unable to bind to cell membranes.We demonstrate in Fig. 5 that Ab does bind to photopolymerizedDSLNTwith a binding afnity (equilibrium dissociation constant) of220 60 nM. Ab afnity to gangliosides in biological membranes,a potential binding site for Ab on the cell membrane, has beenreported to be on the order of 106 M[7,28]. The greater afnity ofAb for photopolymerized DSLNT than ganglioside rich regions ofthe cell surface might indicate that photopolymerized DSLNT couldeffectively compete with cell surface for Ab binding. Others haveshown that inhibitor afnity for Ab was correlated with the abilityof such inhibitors to attenuate Ab toxicity [54], supporting ourspeculation that Ab afnity for photopolymerized DSLNT would berelated to toxicity attenuation properties of the DSLNT polymer.

Ab binding to DSLNT polymers does not appear to completelysaturate at the concentrations tested. Given the propensity of Ab toaggregate as a function of concentration [5,55], the shape of the Abbinding isotherms may indicate non-specic binding to DSLNTpolymers or may be reective of different aggregation states of Abinteracting with the DSLNT polymer. The method used to evaluateAb binding the DSLNT polymers cannot be used to discriminatebetween these possibilities.

When we measured the Ab toxicity attenuation as a function ofphotocrosslinked DSLNT, we found that photocrosslinked DSLNTinhibited toxicity with an inhibition constant of 23 mM, regardlessof the Ab concentration. This inhibition constant is similar to thatfound when using sialic acid dendrimers of similar valency to at-tenuate Ab toxicity [1]. DSLNT polymers attenuated Ab toxicity aswell as sialic acid dendrimers without the associated cytotoxicity ofthe dendrimeric materials. The DSLNT polymers attenuated Abtoxicity with an inhibition constant that was not a function of Abconcentration, but with a maximum increase in viability that wasa function of Ab concentration suggests that the mechanism oftoxicity attenuation of Ab might not be direct competition for Abbinding. Analogous to what one might see in the presence of aninhibitor when examining the rate of enzymatic reactions, wewould have expected that if inhibition was competitive, the ap-parent inhibition constant would change as a function of Ab, butmaximum viability (just like the maximum rate of reaction) wouldremain unchanged. One explanation of our results is that Ab binds

ls 29 (2008) 34083414to polymeric DSLNT, and the polymer-DSLNT-Ab complex kills cells,but at a much lower rate and/or with much less efcacy that

teriaunbound Ab. Regardless of the mechanism by which photo-crosslinked DSLNT attenuates Ab toxicity, we have shown thatphotopolymerized DSLNT is able to effectively attenuate Ab toxicitywithout the associated toxicity we had previously seen withmultivalent sialic acid dendrimers [1,36].

We observed Ab toxicity attenuation at concentrations of pho-tocrosslinked DSLNT in the 23 mM concentration range, which islikely too high a concentration for an in vivo application as a ther-apeutic for Alzheimers disease treatment. In previous work, weshowed that multivalent sialic acid dendrimers were able to at-tenuate Ab toxicity at lower concentrations (in the high nanomolarrange) and that toxicity attenuation was a function of sialic acidvalency, along with other parameters [1,36]. We believe Ab afnityfor DSLNT polymers should increase with valency. Increasing thedegree of polymerization of the DSLNT photopolymers should yieldpolymers that are able to attenuate Ab toxicity at lower concen-trations. However, delivery of soluble DSLNT polymers in vivo eitherin the blood, in the CSF, or across the blood brain barrier, will be-come increasingly more difcult as polymer size increases. Anotherapproach to improve toxicity attenuation properties of sialic acidpolymers without signicantly increasing the size of the polymerwould be to alter the size of the crosslinking molecules such thatthe spacing of sialic acid residues in the polymer more closelymatched suspected binding sites of sialic acid on the Ab oligomer orbril. Delivery of multivalent sialic acid polymers in the bloodshould be feasible, however, intracerebral delivery may be con-siderably more problematic and highly dependent upon the lip-ophilicity of the polymer backbone. Recent studies suggest whilesome Ab can be cleared from the cerebral cortex by agentsdelivered in the blood [56], more effective clearance is seen whenantibodies or other agents which sequester Ab are delivered to thebrain [57].

5. Conclusions

We demonstrated the feasibility of synthesizing photo-crosslinked multivalent sialic acid materials from sialic acidcontaining oligosaccharides and glycidyl methacrylate. The pho-tocrosslinked sialic acid containing polymers were able to attenu-ate Ab toxicity in a cell culture model without any notablecytotoxicity at concentrations up to 500 mM. We show that photo-polymerized DSLNT binds to Ab with an equilibrium dissociationconstant on the order of 200 nM, suggesting that these materialsbind Ab with greater afnity than cell surface gangliosides. Thephotopolymerized DSLNT attenuated Ab toxicity at concentrationsas low as 2 mM. We believe with further optimization, both bindingafnity and toxicity attenuation properties of photocrosslinkedsialic acid polymers could be improved. These materials providea useful tool in the examination of the mechanism of Ab mediatedcell toxicity and the development of new strategies to prevent Abtoxicity associated with Alzheimers disease.

Acknowledgments

Financial support for this workwas provided by a grant from theNational Institutes of Health (R21 NS050346) to T.A.G. and G.L.C.

References

[1] Patel D, Henry J, Good T. Attenuation of beta-amyloid induced toxicity by sialicacid-conjugated dendrimeric polymers. Biochim Biophys Acta 2006;1760(12):18029.

[2] Brookmeyer R, Gray S, Kawas C. Projections of Alzheimers disease in theUnited States and the public health impact of delaying disease onset. Am J

C.B. Cowan et al. / BiomaPublic Health 1998;88(9):133742.[3] Gong Y, Chang L, Viola KL, Lacor PN, Lambert MP, Finch CE, et al. Alzheimers

disease-affected brain: presence of oligomeric A beta ligands (ADDLs) suggestsa molecular basis for reversible memory loss. Proc Natl Acad Sci U S A 2003;100(18):1041722.

[4] Hensley K, Buttereld DA, Mattson M, Aksenova M, Harris M, Wu JF, et al. Amodel for beta-amyloid aggregation and neurotoxicity based on the freeradical generating capacity of the peptide: implications of molecular shrap-nel for Alzheimers disease. Proc West Pharmacol Soc 1995;38:11320.

[5] Ward RV, Jennings KH, Jepras R, Neville W, Owen DE, Hawkins J, et al. Frac-tionation and characterization of oligomeric, protobrillar and brillar formsof beta-amyloid peptide. Biochem J 2000;348(Pt 1):13744.

[6] Doig AJ, Hughes E, Burke RM, Su TJ, Heenan RK, Lu J. Inhibition of toxicityand protobril formation in the amyloid-beta peptide beta(2535) usingN-methylated derivatives. Biochem Soc Trans 2002;30(4):53742.

[7] Choo-Smith LP, Surewicz WK. The interaction between Alzheimer amyloidbeta(140) peptide and ganglioside GM1-containing membranes. FEBS Lett1997;402(23):958.

[8] Arispe N, Rojas E, Pollard HB. Alzheimer disease amyloid beta protein formscalcium channels in bilayer membranes: blockade by tromethamine andaluminum. Proc Natl Acad Sci U S A 1993;90(2):56771.

[9] Kakio A, Yano Y, Takai D, Kuroda Y, Matsumoto O, Kozutsumi Y, et al. In-teraction between amyloid beta-protein aggregates and membranes. J Pept Sci2004;10(10):61221.

[10] Mott RT, Hulette CM. Neuropathology of Alzheimers disease. NeuroimagingClin N Am 2005;15(4):75565. ix.

[11] Hoshi M, Sato M, Matsumoto S, Noguchi A, Yasutake K, Yoshida N, et al.Spherical aggregates of beta-amyloid (amylospheroid) show high neurotox-icity and activate tau protein kinase I/glycogen synthase kinase-3beta. ProcNatl Acad Sci U S A 2003;100(11):63705.

[12] Wang SS, Becerra-Arteaga A, Good TA. Development of a novel diffusion-basedmethod to estimate the size of the aggregated abeta species responsible forneurotoxicity. Biotechnol Bioeng 2002;80(1):509.

[13] Kayed R, Sokolov Y, Edmonds B, McIntire TM, Milton SC, Hall JE, et al. Per-meabilization of lipid bilayers is a common conformation-dependent activityof soluble amyloid oligomers in protein misfolding diseases. J Biol Chem 2004;279(45):463636.

[14] Dahlgren KN, Manelli AM, Stine Jr WB, Baker LK, Krafft GA, LaDu MJ. Oligo-meric and brillar species of amyloid-beta peptides differentially affectneuronal viability. J Biol Chem 2002;277(35):3204653.

[15] Whitson JS, Mims MP, Strittmatter WJ, Yamaki T, Morrisett JD, Appel SH. At-tenuation of the neurotoxic effect of A beta amyloid peptide by apolipoproteinE. Biochem Biophys Res Commun 1994;199(1):16370.

[16] Sokolov Y, Kozak JA, Kayed R, Chanturiya A, Glabe C, Hall JE. Soluble amyloidoligomers increase bilayer conductance by altering dielectric structure. J GenPhysiol 2006;128(6):63747.

[17] Kilsdonk EP, Yancey PG, Stoudt GW, Bangerter FW, Johnson WJ, Phillips MC,et al. Cellular cholesterol efux mediated by cyclodextrins. J Biol Chem 1995;270(29):172506.

[18] Arispe N, Pollard HB, Rojas E. Giant multilevel cation channels formed byAlzheimer disease amyloid beta-protein [A beta P-(140)] in bilayer mem-branes. Proc Natl Acad Sci U S A 1993;90(22):105737.

[19] Diaz JC, Linnehan J, Pollard H, Arispe N. Histidines 13 and 14 in the Abetasequence are targets for inhibition of Alzheimers disease Abeta ion channeland cytotoxicity. Biol Res 2006;39(3):44760.

[20] Buttereld DA, Grifn S, Munch G, Pasinetti GM. Amyloid beta-peptide andamyloid pathology are central to the oxidative stress and inammatorycascades under which Alzheimers disease brain exists. J Alzheimers Dis 2002;4(3):193201.

[21] Varadarajan S, Yatin S, Aksenova M, Buttereld DA. Review: Alzheimers am-yloid beta-peptide-associated free radical oxidative stress and neurotoxicity. JStruct Biol 2000;130(23):184208.

[22] Ariga T, Kobayashi K, Hasegawa A, Kiso M, Ishida H, Miyatake T. Character-ization of high-afnity binding between gangliosides and amyloid beta-protein. Arch Biochem Biophys 2001;388(2):22530.

[23] Ariga T, Yu RK. GM1 inhibits amyloid beta-protein-induced cytokine release.Neurochem Res 1999;24(2):21926.

[24] Matsuzaki K, Horikiri C. Interactions of amyloid beta-peptide (140) withganglioside-containing membranes. Biochemistry 1999;38(13):413742.

[25] Wakabayashi M, Okada T, Kozutsumi Y, Matsuzaki K. GM1 ganglioside-mediated accumulation of amyloid beta-protein on cell membranes. BiochemBiophys Res Commun 2005;328(4):101923.

[26] Yanagisawa K, Odaka A, Suzuki N, Ihara Y. GM1 ganglioside-bound amyloidbeta-protein (A beta): a possible form of preamyloid in Alzheimers disease.Nat Med 1995;1(10):10626.

[27] Zha Q, Ruan Y, Hartmann T, Beyreuther K, Zhang D. GM1 ganglioside regulatesthe proteolysis of amyloid precursor protein. Mol Psychiatry 2004;9(10):94652.

[28] Ariga T, Kiso M, Hasegawa A, Miyatake T. Gangliosides inhibit the release ofinterleukin-1beta in amyloid beta-protein-treated human monocytic cells. JMol Neurosci 2001;17(3):3717.

[29] Kakio A, Nishimoto S, Yanagisawa K, Kozutsumi Y, Matsuzaki K. Interactions ofamyloid beta-protein with various gangliosides in raft-like membranes:importance of GM1 ganglioside-bound form as an endogenous seed forAlzheimer amyloid. Biochemistry 2002;41(23):738590.

[30] Kakio A, Nishimoto SI, Yanagisawa K, Kozutsumi Y, Matsuzaki K. Cholesterol-

ls 29 (2008) 34083414 3413dependent formation of GM1 ganglioside-bound amyloid beta-protein, anendogenous seed for Alzheimer amyloid. J Biol Chem 2001;276(27):2498590.

[31] Baier Leach J, Bivens KA, Patrick Jr CW, Schmidt CE. Photocrosslinked hyalur-onic acid hydrogels: natural, biodegradable tissue engineering scaffolds. Bio-technol Bioeng 2003;82(5):57889.

[32] Bhavanandan VP, Sheykhnazari M. Adaptation of the periodateresorcinolmethod for determination of sialic acids to a microassay using microtiter platereader. Anal Biochem 1993;213(2):43840.

[33] van Engeland M, Ramaekers CS, Schutte B, Reutelingsperger CPM. A novelassay to measure loss of plasma membrane asymmetry during apoptosis ofadherent cells in culture. Cytometry 1996;24:1319.

[34] Reuter JD, Myc A, Hayes MM, Gan Z, Roy R, Qin D, et al. Inhibition of viraladhesion and infection by sialic-acid-conjugated dendritic polymers. Bio-conjug Chem 1999;10(2):2718.

[35] Landers JJ, Cao Z, Lee I, Piehler LT, Myc PP, Myc A, et al. Prevention of inuenzapneumonitis by sialic acid-conjugated dendritic polymers. J Infect Dis 2002;186(9):122230.

[36] Patel DA, Henry JE, Good TA. Attenuation of beta-amyloid-induced toxicity bysialic-acid-conjugated dendrimers: role of sialic acid attachment. Brain Res2007;1161:95105.

[37] Kim IS, Jeong YI, Kim DH, Lee YH, Kim SH. Albumin release from biodegradablehydrogels composed of dextran and poly(ethylene glycol) macromer. ArchPharm Res 2001;24(1):6973.

[38] Trudel J, Massia SP. Assessment of the cytotoxicity of photocrosslinked dextranand hyaluronan-based hydrogels to vascular smooth muscle cells. Bio-materials 2002;23(16):3299307.

[39] Chen FM, Zhao YM, Zhang R, Jin T, Sun HH, Wu ZF, et al. Periodontal re-generation using novel glycidyl methacrylated dextran (Dex-GMA)/gelatinscaffolds containing microspheres loaded with bone morphogenetic proteins.J Control Release 2007;121(12):8190.

[40] Flores-Ramirez N, Elizalde-Pena EA, Vasquez-Garcia SR, Gonzalez-Hernandez J, Martinez-Ruvalcaba A, Sanchez IC, et al. Characterization anddegradation of functionalized chitosan with glycidyl methacrylate. J BiomaterSci Polym Ed 2005;16(4):47388.

[41] Leach JB, Bivens KA, Collins CN, Schmidt CE. Development of photo-crosslinkable hyaluronic acid-polyethylene glycol-peptide composite hydro-gels for soft tissue engineering. J Biomed Mater Res A 2004;70(1):7482.

[42] Reis AV, Cavalcanti OA, Rubira AF, Muniz EC. Synthesis and characterization ofhydrogels formed from a glycidyl methacrylate derivative of galactomannan.Int J Pharm 2003;267(12):1325.

[43] Laidler P, Litynska A, Hoja-Lukowicz D, Labedz M, Przybylo M, Ciolczyk-Wierzbicka D, et al. Characterization of glycosylation and adherent prop-erties of melanoma cell lines. Cancer Immunol Immunother 2006;55(1):

[44] van Rossenberg SM, Sliedregt LA, Autar R, Piperi C, Van Der Merwe AP, vanBerkel TJ, et al. A structurefunction study of ligand recognition by CD22beta. JBiol Chem 2001;276(16):1296773.

[45] Nakata D, Troy 2nd FA. Degree of polymerization (DP) of polysialic acid(polySia) on neural cell adhesion molecules (N-CAMS): development andapplication of a new strategy to accurately determine the DP of polySia chainson N-CAMS. J Biol Chem 2005;280(46):3830516.

[46] Chang WW, Yu CY, Lin TW, Wang PH, Tsai YC. Soyasaponin I decreases theexpression of alpha2,3-linked sialic acid on the cell surface and suppresses themetastatic potential of B16F10 melanoma cells. Biochem Biophys Res Com-mun 2006;341(2):6149.

[47] Aubert M, Panicot L, Crotte C, Gibier P, Lombardo D, Sadoulet MO, et al. Res-toration of alpha(1,2) fucosyltransferase activity decreases adhesive andmetastatic properties of human pancreatic cancer cells. Cancer Res 2000;60(5):144956.

[48] Huskens J. Multivalent interactions at interfaces. Curr Opin Chem Biol 2006;10(6):53743.

[49] Kumar PV, Asthana A, Dutta T, Jain NK. Intracellular macrophage uptake ofrifampicin loadedmannosylated dendrimers. J Drug Target 2006;14(8):54656.

[50] Lees WJ, Spaltenstein A, Kingery-Wood JE, Whitesides GM. Polyacrylamidesbearing pendant alpha-sialoside groups strongly inhibit agglutination oferythrocytes by inuenza A virus: multivalency and steric stabilization ofparticulate biological systems. J Med Chem 1994;37(20):341933.

[51] Polizzotti BD, Kiick KL. Effects of polymer structure on the inhibition of choleratoxin by linear polypeptide-based glycopolymers. Biomacromolecules 2006;7(2):48390.

[52] Prieto MJ, Bacigalupe D, Pardini O, Amalvy JI, Venturini C, Morilla MJ, et al.Nanomolar cationic dendrimeric sulfadiazine as potential antitoxoplasmicagent. Int J Pharm 2006;326(12):1608.

[53] Woller EK, Cloninger MJ. The lectin-binding properties of six generations ofmannose-functionalized dendrimers. Org Lett 2002;4(1):710.

[54] Cairo CW, Strzelec A, Murphy RM, Kiessling LL. Afnity-based inhibition ofbeta-amyloid toxicity. Biochemistry 2002;41(27):86209.

[55] Pallitto MM, Murphy RM. A mathematical model of the kinetics of beta-amyloid bril growth from the denatured state. Biophys J 2001;81(3):180522.

[56] Bergamaschini L, Rossi E, Storini C, Pizzimenti S, Distaso M, Perego C, et al.Peripheral treatment with Enoxaparin, a low molecular weight heparin, re-duces plaques and b-Amyloid accumulation in a mouse model of Alzheimersdisease. J Neurosci 2004;24(17):41816.

[57] Vasilevko V, Xu F, Previti ML, Van Nostrand WE, Cribbs DH. Experimentalinvestigation of antibody-mediated clearance mechanisms of amyloid-beta in

C.B. Cowan et al. / Biomaterials 29 (2008) 3408341434141128. CNS of Tg-SwDI transgenic mice. J Neurosci 2007;27:1337683.

Development of photocrosslinked sialic acid containing polymers for use in Abeta toxicity attenuationIntroductionMaterials and methodsMaterialsDevelopment of photocrosslinked sialic acid containing oligosaccharidesPolymerized dot assayPurification of sialic acid polymerProof of synthesis (FTIR)Periodate-resorcinol assay for sialic acid determinationCell cultureAbeta solutionPropidium iodide assayRadiochemical binding assay

ResultsSynthesis of polymerIntrinsic toxicity and Abeta toxicity attenuation properties of oligosaccharide monomersIntrinsic toxicity and Abeta toxicity attenuation properties of photocrosslinked oligosaccharidesRelationship between Abeta binding and toxicity attenuation of photopolymerized DSLNT

DiscussionConclusionsAcknowledgmentsReferences