Cite this: RSC Advances, 2013, 3, 15035

Selective nitrosation of modified dextran polymers3

Received 23rd April 2013,Accepted 12th June 2013

DOI: 10.1039/c3ra41988f

www.rsc.org/advances

Jessica M. Joslin,a Vinod B. Damodarana and Melissa M. Reynolds*ab

The formation of stable nitric oxide (NO) moieties onto polymer substrates is a critical part of NO materials

development. Depending upon the polymer functionality, nitrosation of the material can result in a variety

of different NO moieties. Toward the ability to selectively load NO in the presence of other reactive sites on

a polymer substrate, we investigate how the polymer functionality affects the ability to preferentially form

S-nitrosothiol groups over unwanted byproducts, such as N-nitrosamines. Using modified dextran as our

model polymer, we demonstrate that N-nitrosamines form on 2u amine sites under S-nitrosation

conditions, but are not a significant source of released NO under physiological pH. By tuning the synthetic

conditions associated with the polymer synthesis, we demonstrate that by replacing 2u amine sites on the

polymer with amide sites, N-nitrosation is effectively eliminated, resulting instead in predominant

S-nitrosation. The selectively S-nitrosated materials experience near-complete donor decomposition, giving

rise to NO under physiological pH. The ability to tune the availability and reactivity of different functional

groups is additionally helpful toward materials synthesis and application in general.

1. Introduction

The ability to load nitric oxide (NO) onto polymer backbonesin the form of an NO moiety is fundamental to thedevelopment of NO-releasing polymers, which demonstratepromising results as biomaterials.1–5 Since NO is a highlyreactive radical, NO functional groups enable stable storageuntil the therapeutic NO action is required from the material.The key feature associated with NO-releasing polymers is toexert fine control over the type, distribution and concentrationof NO donor in the system for highly tunable materials. Onepopular NO functional group is the S-nitrosothiol (RSNO)moiety, which is formed by nitrosation of a thiol site.S-nitrosothiols can be triggered to liberate NO through avariety of pathways, including heat-, light-, and metal ion-mediated,6 making this class of NO donor versatile dependingupon the material application. The incorporation of RSNOsdirectly onto polymer supports via nitrosation of thiol sites hasbeen previously noted.7–16 The primary method of investiga-tion for these materials has been based upon the quantifica-tion of NO generated by the nitrosated material. However, due

to the complexity of the macromolecular networks, there is thepotential for the formation of alternate NO donor groups,other than the intended RSNOs. Characterization of the NOmoieties formed during the nitrosation process is critical toelucidate the contribution of various functional groups on theresultant NO groups that form on the material and determin-ing which groups subsequently give rise to NO release.

Complete characterization of the NO loaded material mustbe two-fold: first, the NO moiety should be characterized todetermine the nature and efficiency of the nitrosation processand, secondarily, the NO release from the material should bemeasured in correspondence with the decomposition of theNO moiety. Depending upon the complexity of the polymerfunctionality, and due to the versatility of nitrosationchemistry,17 a variety of NO moieties can form during thenitrosation process (addition of NO+). In particular, this workwill emphasize the competitive formation of S-nitrosothiol andN-nitrosamine (RNNO) moieties on a mixed functional groupmodel polymer, dextran. As shown in Fig. 1, RSNO and RNNO

aDepartment of Chemistry, Colorado State University, Fort Collins, CO 80523, USA.

E-mail: melissa.reynolds@colostate.edu; Tel: +1 970 491 3775bSchool of Biomedical Engineering, Colorado State University, Fort Collins, CO

80523, USA

3 Electronic supplementary information (ESI) available: Dextran (1a, 1b)characterization of thiol incorporation: 1H NMR and IR; Spectroscopiccharacterization of nitrosated dextran derivatives: UV-vis (1b, 2b), IR (1a, 1b); NOrelease profiles (1a, 1b); Control experiments for 350 nm region interference:nitrosated dextran blank, nitrite analysis, baseline shift (1a, 1b); N-nitrosaminecontrol; Remaining RSNO control; Molar extinction coefficient determination(2a, 2b). See DOI: 10.1039/c3ra41988f

Fig. 1 Nitrosation of (a) 1u thiol or (b) 2u amine sites is accomplished wheneither functional group is exposed to a nitrosating agent (a source of NO+).

RSC Advances

PAPER

This journal is � The Royal Society of Chemistry 2013 RSC Adv., 2013, 3, 15035–15043 | 15035

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article OnlineView Journal | View Issue

moieties can form upon nitrosation of thiol and amine sites,respectively, that are present in the dextran system. It isimportant to characterize which NO moieties form and whichgive rise to NO, as well as consider which byproducts areundesirable. Specifically, N-nitrosamines have been demon-strated as toxic and carcinogenic compounds.18 We thereforeaim to prevent RNNO formation on polymer backbones,especially the biodegradable dextran polymers describedherein, where small molecule RNNO degradation productswould be made available in biological systems. To ourknowledge, no reports concerning the formation of RNNOmoieties during nitrosation of a thiol-containing material havebeen reported, which is a critical concern toward applying thematerials as clinically relevant devices.

The competitive formation of RNNOs has been consideredfor other small molecule and polymer systems containingamine sites for other classes of NO donors (i.e. N-diazenium-diolates).19–21 If RNNO formation has been considered for otherNO donor systems that do not directly involve nitrosation, thereis certainly a larger cause for concern regarding side reactionsduring the nitrosation of materials containing thiols amongother functional groups. More specifically, N-nitrosamineformation has been noted as a competitive process for smallmolecules undergoing NO exposure toward N-diazeniumdiolatemoiety formation.19 This concern was further realized forpolymers doped with N-diazeniumdiolate donors, where donorwas found to leach from the material and N-nitrosamines weresubsequently detected in the soaking solution.22 Even thoughnitrosation conditions aren’t employed for N-diazeniumdiolatesystems, any oxidized NO in the system could result innitrosation due to the formation of a nitrosating species. Assuch, RNNO formation has been considered for other NO donorsystems that do not directly involve nitrosation. For thiol-containing systems that rely on nitrosation processes for NOloading, there is certainly a larger cause for concern regardingside reactions during the nitrosation of materials containingthiols among other functional groups. Methods are thereforerequired to differentiate between the formation of thesedifferent NO donors in materials systems.

Small molecule RSNO and RNNO moieties have beenspectroscopically characterized, both with reported lmax valuesof 330–350 nm.23 Polymer systems containing RSNO or RNNOfunctionalities have been demonstrated to maintain similarabsorbance features as their respective small moleculeanalogues.9,20 Overall, because these moieties absorb in thesame wavelength region, it is difficult to characterize theformation of one over the other after initial nitrosation usingUV-visible spectroscopy. Minimal donor characterizationreported in the literature for NO loaded materials in additionto the overlapping UV-visible absorbance features associatedwith RSNO and RNNO functionalities have led to a further lackof consideration of nitrosation byproducts on these NO loadedpolymers.

As an alternative spectroscopic method, small moleculeRSNO and RNNO moieties have been characterized by IRspectroscopy.24,25 The NLO stretch of the nitroso group is

evident for both moieties. However, the frequency differencesbetween the N–S stretch of the RSNO (700–600 cm21) and theN–N stretch of the RNNO (1000–900 cm21) have been reportedin distinguishably different regions. As such, the characteristicIR absorbance properties could differentiate between eachmoiety to determine which form during nitrosation.

In addition to spectroscopic characterization, the RSNOand RNNO moieties can be distinguished from one anotherdue to the different stabilities of each functional group arisingfrom the different bond dissociation energies associated withthe homolytic cleavage of the X-NO bond. In general,theoretical methods have indicated that the N-NO bond isstronger than the S-NO bond, making the RNNO more stablethan the RSNO moiety.26 These differences in stability can beexploited to differentiate between the formation of eithermoiety in a materials system. The stability of some NOmoieties relative to others has consequences on the finalapplication of the material. If the ultimate goal is to result inthe maximum amount of NO release possible from thematerial, the formation of NO moieties that cannot contributeto NO release will not accomplish this, particularly if theformation of these stable moieties competes with the forma-tion of NO donor sites. It would instead be more beneficial totune the chemistry of the system to maximize formation of theNO products (i.e. S-nitrosothiols) that will decompose underphysiological conditions.

Toward the development of highly tunable NO materials(i.e. control over the types of NO moieties formed as well astheir concentrations), we report a system in which various NOfunctional groups were incorporated onto the backbone of adextran polymer. Dextran is a non-toxic polysaccharidepolymer that degrades naturally via enzymatic activity andhas been implicated as a therapeutic material.27 Dextran incombination with NO release appears promising for biomater-ials development due to the biodegradability and structuralproperties of dextran coupled to the therapeutic action ofNO.28–30 As such, functionalized dextran derivatives serve as agood model system for understanding the nitrosation pro-cesses that can give rise to different NO functionalities to exertcontrol over the NO loading process in a material system. Twodifferent polymer synthesis routes are presented that give riseto different sites available for nitrosation. Depending upon thepresence of different functional groups (thiol, amine, amide),the NO moiety formation (RSNO, RNNO) was impacted.Overall, we demonstrate that, depending upon the function-ality of the polymer system, selective S-nitrosation can beachieved. In addition to spectroscopically monitoring themajor nitrosation products, we track the behavior of thefunctional groups after NO release and are able to discernwhich groups give rise to NO.

This work demonstrates that spectrophotometric methodscan be used to identify the formation of unwanted nitrosationbyproducts in mixed NO moiety systems. Synthetic conditionscan then be tuned to allow for selective formation ofS-nitrosothiol donor moieties onto a model dextran backbone,which nearly completely release their NO payload. Overall, the

15036 | RSC Adv., 2013, 3, 15035–15043 This journal is � The Royal Society of Chemistry 2013

Paper RSC Advances

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

nitrosation processes described herein serve as a fundamentalbasis toward the investigation of S-nitrosothiol formation onpolymers with mixed functional groups and their resulting NOrelease behavior.

2. Experimental section

2.1 Materials

Dextran (from Leuconostoc spp. Mr y 40 000) was obtainedfrom Sigma, L-cysteine (98+%) from Alfa Aesar, cysteaminehydrochloride (>97%) from Fluka Analytical, and t-butyl nitrite(90%) from Aldrich. Unless otherwise stated, all phosphatebuffered saline (PBS) was prepared from tablet (OmniPur,EMD) and pH balanced to 7.4 using acid or base, if necessary.All solutions were prepared using Millipore filtered water (18.2MV) unless otherwise stated.

2.2 Methods

2.2.1 Polymer synthesis. Dextran was modified to containpendant cysteine or cysteamine groups using two differentsynthetic approaches. Syntheses and their respective charac-terization were repeated in at least triplicate to ensurereproducibility of the procedures. All experiments wereperformed in triplicate with the average and standarddeviation reported.

a. Scheme 1: Reductive amination. The first syntheticapproach used to modify dextran with a thiol pendant involvedthe use of reductive amination to attach a cysteamine (1a) orcysteine (1b) group onto the polymer backbone, as shown inScheme 1.31 To accomplish this, dextran (1 g) was prepared inwater (30 mL), followed by treatment with sodium periodate(0.8 g, 3.73 mmol, Sigma Aldrich) and concentrated sulfuricacid (0.3 g, 3 mmol, 0.8 eq. periodate, BDH). The mixture wasshielded from light under stirring for 1.5 h to allow forperiodate oxidation of the dextran. The reaction was termi-

nated via treatment with ethylene glycol (0.18 g, 2.84 mmol,0.76 eq., 95%, Acros) for a half an hour followed byneutralization with sodium acetate (0.2 M, Mallinckrodt)solution. Dextran aldehyde derivative was isolated by dialysisusing a Spectra/Por1 dialysis membrane with a molecularweight cut-off of 8000 Da using Millipore water with multiplechanges. The dialyzed dextran aldehyde solution was cooledto 0 uC in an ice bath and subsequently treated withcysteamine hydrochloride or cysteine (1.05 eq. relative toperiodate quantity). The pH was adjusted to 8.5 via sodiumhydroxide (1 M, Fisher Scientific) addition and stirred at 0 uCfor 1 h. To accomplish the reductive amination, sodiumcyanoborohydride (0.25 g, 1 eq., Acros) was added to thesolution and stirred at 0 uC for 2 h, after which time theproduct was neutralized by adding acetic acid (10%, BDH).Following an overnight dialysis, the product was treated withdithiothreitol (0.05 g, Fluka) for one hour at room tempera-ture to reduce any disulfide bonds that may have formedduring the reaction period. The final thiolated product wasfurther extensively dialyzed and finally isolated after freezedrying. The 1H NMR and ATR-IR spectra corresponding to thereductive amination products are included in the ESI3 (Fig. S1and S2). 1H NMR d (ppm) (400 MHz, D2O) Dextran–cysteamine (1a): 2.87 (m, –CH2–CH2–SH), 3.35–3.85 (m,dextran H C2–C6) and 4.83 (m, dextran H C1). Dextran-cysteine (1b): 2.74–3.25 (m, –CH2–SH), 3.35–3.85 (m, –NH–CH– overlapped with dextran H C2–C6) and 4.83 (m, dextranH C1). IR nmax/cm21 Dextran–cysteamine (1a): 3600–3000(–OH), 2890 (–CH), 1247 (–C–N–), and 984 (–CH2 of a-1,6-linkage). Dextran-cysteine (1b): 3600–3000 (–OH), 2890 (–CH),1577 (–COOH), 1256 (–C–N–), and 1003 (–CH2 of a-1,6-linkage).

b. Scheme 2: Carboxymethyl intermediate. The secondsynthetic approach to achieve thiol-modification of dextraninvolved the formation of a carboxymethyl intermediate thatwas further functionalized with cysteamine (2a) or cysteine(2b), as shown in Scheme 2. The dextran derivatives were

Scheme 1 Dextran modification via reductive amination results in the linkage of cysteamine (1a) or cysteine (1b) to the polymer backbone via a 2u amine linkage.

This journal is � The Royal Society of Chemistry 2013 RSC Adv., 2013, 3, 15035–15043 | 15037

RSC Advances Paper

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

synthesized and characterized by 1H NMR and ATR-IR asreported in a previous publication from our group.28 In brief,dextran hydroxyl groups were first modified with a carbox-ymethyl tether by treating with monochloroacetic acid (MCA)under alkaline conditions followed by acidification with aceticacid. Carboxyl groups of the modified dextran derivatives werethen covalently modified with aminothiol derivatives throughstable amide linkages using carbodiimide chemistry.

2.2.2 Ellman’s assay for thiol quantification. To quantify theamount of thiol that was incorporated onto the dextranbackbone for all derivatives, the Ellman’s assay was per-formed. All standard, sample and Ellman’s reagent solutionswere prepared in a phosphate buffer (100 mM), which wasprepared by adding NaH2PO4 (100 mM, sodium phosphate,monobasic, anhydrous, molecular biology grade, BDH) toNa2HPO4 (100 mM, sodium phosphate, dibasic, anhydrous,ACS grade, Mallinckrodt Chemicals) until pH 8 was reached.Thiol standards were prepared from a stock solution ofcysteine (0.012 M) via serial dilution. A solution of DTNB (10mM, 5,59-dithiobis(2-nitrobenzoic acid), Sigma), the Ellman’sreagent, was prepared. Each modified dextran sample wasdissolved in buffer at 5 mg mL21 concentration. Aliquots (100mL) of each standard/sample solution were added to DTNBsolution (100 mL) and brought to a final volume of 4 mL. Thesolutions were agitated for 1 h at room temperature, and theresulting solutions were pipetted into a 96-well plate in 200 mLaliquots. The absorbance values of the DTNB-treated sampleswere read at 414 nm on a Synergy 2 microtiter plate reader(BioTek, Winooski, VT, USA).

2.2.3 NO loading. Nitrosation of each dextran derivative wasachieved by adding anhydrous methanol (4 mL, ACS grade,Macron Chemicals, stored over 4 Å molecular sieves) to therespective dextran derivative (50 mg) in an amber, EPA-certified vial (EnviroWare, Fisher Scientific) with a septum-containing lid. The t-butyl nitrite reagent was pretreated with

10 w/v% EDTA ((ethylenedinitrilo)tetraacetic acid, disodiumsalt, dihydrate, ACS grade, EMD) and subsequently injected(0.4 mL) into the dextran suspension. The sample remainedunder stirring overnight, protected from light. After nitrosa-tion, the sample was pumped under vacuum for 2 h to removethe methanol and excess t-butyl nitrite.

To serve as an N-nitrosamine control for characterizationcomparison, N-nitrosoproline was prepared. Proline (50 mg,99%, Alfa Aesar) was suspended in anhydrous methanol (4mL), followed by injection of EDTA-treated t-butyl nitrite (0.4mL). The solution was stirred overnight, and the nitrosatedproduct was isolated by vacuum as described above.

2.2.4 NO moiety, NO, and byproduct characterization. Allnitrosated products were analyzed before and after NOanalysis by solution phase UV-visible spectroscopy on aNicolet Evolution 300 spectrophotometer (Thermo ElectronCorporation, Madison, WI, USA). The polymer solutions wereprepared at 0.75 mg mL21 concentration, where the sampleswere sonicated for the time required to solubilize the polymer.The derivatives prepared via reductive amination (1a, 1b) wereanalyzed before and after nitrosation using a Nicolet 6700 FT-IR fitted with an ATR sample stage for solid-phase IRmeasurements.

NO release was directly detected via chemiluminescenceusing Sievers 280i Nitric Oxide Analyzers (NOA, GE Analytical,Boulder, CO, USA). All samples were dissolved in PBS (0.75 mgmL21) and agitated for the amount of time required for thesample to completely dissolve, followed by injection into theNOA cell. NOA measurements were performed at a 5 s timeinterval, over the duration of time necessary for the measure-ments to reach baseline. The derivatives prepared via reductiveamination (1a, 1b) required about 10 h to reach baseline. Thecysteamine derivative prepared via carboxymethyl intermedi-ate (2a) required about 10 h to reach baseline, while thecysteine derivative (2b) required about 20 h. It is important to

Scheme 2 Dextran modification via a carboxymethyl intermediate results in the linkage of cysteamine (2a) or cysteine (2b) to the polymer backbone via an amidelinkage.

15038 | RSC Adv., 2013, 3, 15035–15043 This journal is � The Royal Society of Chemistry 2013

Paper RSC Advances

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

note that all NOA experiments were performed under solutionphase (PBS, pH 7.4) at room temperature, exposed to ambientlight. The mode of RSNO decomposition employed in theseexperiments, therefore, is due to heat and light initiated NOrelease.

To assess nitrite in the polymer solution recovered after NOanalysis, the Griess assay was performed. Sodium nitrite(99.999%, Alfa Aesar) standards were prepared in theconcentration range 0–32 mM via serial dilution from a stock10 mM solution. Aliquots (212 mL) of recovered polymer PBSsolutions (0.75 mg mL21) and nitrite standards were pipettedinto a 96-well plate and further treated with sulfanilic acidsolution (21 mL of 12.5 mM, prepared from 99%, Aldrich) andhydrochloric acid solution (21 mL of 6 M, prepared from 36.5–38.0%, BDH Aristar) after refrigeration. Further treatment withNEDA solution (22 mL of 12.5 mM, N-(1-naphthyl)ethylenedia-mine dihydrochloride, Acros) resulted in pink-colored solu-tions (lmax of 540 nm). After 15 min of agitation, theabsorbance values at 540 nm were measured using a BioTekSynergy 2 microtiter plate reader.

3. Results and discussion

3.1 Dextran derivatives prepared by reductive amination

Thiol incorporation using Scheme 1 resulted in 0.459 ¡ 0.004and 0.296 ¡ 0.007 mmol thiol g21 for the cysteamine (1a) andcysteine (1b) derivatives, respectively, (Table 1) with the thiolpendant group attached through an amine linkage.Nitrosation of each derivative was accomplished by addingt-butyl nitrite, a well-established nitrosating agent that servesto transfer a nitrosonium ion (NO+) to a nucleophilic site.Fig. 2 indicates the spectrum of the cysteamine derivative (1a)after nitrosation where an absorbance feature at 325–350 nmindicates the presence of nitroso products. Each nitrosateddextran derivative underwent NO analysis of the dextransolutions (0.75 mg mL21 in PBS) while exposed to ambientlight for the time required for the NO measurements to reachbaseline. Table 1 outlines the NO recovery over the duration ofthe NO analysis period, where 0.037 ¡ 0.003 and 0.057 ¡

0.009 mmol NO g21 were recovered for derivatives 1a and 1b,respectively. The persistence of an absorbance feature between325–350 nm for the cysteamine derivative (1a) after completeNO recovery (Fig. 1) indicates a stable byproduct that hasformed either during nitrosation or the NO release period.

The IR spectrum shown in Fig. 3a for the cysteaminederivative (1a) indicates the appearance of a frequency band aty1365 cm21, which is indicative of the NLO stretch of thenitroso moiety. Fig. 3b further shows the appearance of S–Nand N–N stretching bands at y750 cm21 and y830 cm21,respectively, that are indicative of the formation of bothS-nitrosothiol and N-nitrosamine moieties after the nitrosa-tion period. The initial UV-vis spectrum of either derivativeafter nitrosation does not allow for the distinction between theRSNO and RNNO moieties; however, the distinct IR bandspresent after nitrosation indicate that mixed nitroso productsare forming. Overall, spectroscopic evidence qualitativelysuggests that both the 2u amine linkages and the 1u thiolsites present in the system experience nitrosation when thedextran system is exposed to t-butyl nitrite nitrosationconditions (see Fig. 4a). For all spectroscopic analysis of thedextran derivatives prepared via reductive amination, similarabsorbance features are exhibited for either the cysteamine orcysteine derivative. For simplicity, the cysteamine derivative(1a) data is presented here, while the cysteine derivative (1b) isincluded in the ESI3 (Fig. S3–S6).

To ensure that the persistence of the 325–350 nmabsorbance feature is due predominantly to stable RNNOs inthe system, controls were run to account for any absorbance

Table 1 The thiol content associated with each thiolated dextran derivative as determined by the Ellman’s assay and the amount of NO recovered from thecorresponding nitrosated materials. All measurements are reported as an average and standard deviation of n ¢ 3 trials

Pendant group Sample ID Thiol content [mmol g21] NO recovery [mmol g21]a % NO recovery (vs. thiol content)

cysteamine 1a b 0.459 ¡ 0.004 0.037 ¡ 0.003 8%2a c 0.485 ¡ 0.007 0.161 ¡ 0.007 33%

cysteine 1b b 0.296 ¡ 0.007 0.057 ¡ 0.009 19%2b c 0.299 ¡ 0.005 0.158 ¡ 0.007 53%

a NO recovery conditions at room temperature, exposed to ambient light under pH 7.4 phosphate buffered saline at 0.75 mg mL21 concentration.b Thiol attachment accomplished through Scheme 1: reductive amination. c Thiol attachment accomplished through Scheme 2: carboxymethylintermediate.

Fig. 2 UV-vis spectra for the nitrosated cysteamine dextran derivative (1a)prepared via reductive amination before and after NO analysis (0.75 mg mL21).Each spectrum represents the average of n = 3 trials with a standard deviation ,

8%.

This journal is � The Royal Society of Chemistry 2013 RSC Adv., 2013, 3, 15035–15043 | 15039

RSC Advances Paper

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

contribution due to interferences. There are various factorsother than RNNO formation that could contribute to anabsorbance feature in the y350 nm region: a) residual t-butylnitrite nitrosating agent, b) nitrite formed in the system due toprematurely released NO reacting under oxygenated condi-tions (lmax = 354 nm for nitrite), c) any O- or C- nitrosationoccurring on the dextran backbone, or d) any shift in thepolymer solution baseline during the NO analysis period.Overall, it was found that nitrite was present in the system atlow enough concentrations to not interfere in the UV-visspectrum. Additionally, non-thiolated dextran was subjected tonitrosation conditions and the resulting spectrum indicatedno absorbance features due to detectable residual t-butylnitrite or O- and C- nitrosation. The absorbance spectra beforeand after NO analysis did not experience any shift in thebaseline that would contribute to the appearance of anabsorbance feature. An in-depth analysis of each of thesecontrol experiments can be found in the ESI3 (Section S4)where, overall, the only contribution to the remaining 325–350nm absorbance feature was due to the presence of stableRNNOs in the system. A control N-nitrosoproline experiment

was also run where the absorbance feature due to the N-NOmoiety matches the absorbance feature remaining for thenitrosated dextran derivatives after NO analysis. Additionally,to confirm that the remaining absorbance feature was not dueto remaining RSNOs in the system, the solution of cysteaminedextran (1a) recovered after NO analysis was subjected toexposure to UV light (Blak-Ray B-100AP High Intensity; 100Watt, 365 nm) for 30 min, where no significant change in theabsorbance feature was observed. Since RSNOs are known todecompose via a light-initiated pathway,6 any remainingRSNOs would have experienced decomposition upon intenseUV exposure. The spectra corresponding to theN-nitrosoproline and UV-exposed derivatives can be found inthe ESI3 (Fig. S11–S13).

For these dextran derivatives with mixed functional groups,competitive nitrosation is occurring on the amine and thiolsites. Competitive RNNO and RSNO formation has beenreported for small molecules under acidic nitrite conditionswhere, depending upon the reaction conditions (i.e. pH),RNNO formation was enhanced.32 We find that RNNOformation must also be a consideration for polymers exposedto t-butyl nitrite conditions. Previous literature suggests thatfor small molecule mixed systems, during the nitrosationprocess, RSNO formation could dominate followed by transni-trosation to result in RNNO formation.33–35 However, theappearance of both the S–N and N–N stretching modes in theIR after dextran nitrosation suggests that the RNNO formsduring the nitrosation period.

3.2 Dextran derivatives prepared by carboxymethylintermediate

As mentioned previously, N-nitrosamines have been found tobe carcinogenic and toxic and their formation on thesedextran systems is a critical drawback for use in biomedicalapplications. To eliminate RNNO formation on the modifieddextran backbone, a different synthetic approach was taken.By using chemistries involving a carboxymethyl intermediate(Scheme 2), the cysteamine (2a) and cysteine (2b) residueswere attached to the dextran by an amide linkage instead of an

Fig. 3 Highlighted IR regions for dextran–cysteamine derivative (1a) prepared via reductive amination before and after nitrosation where (a) the 1370–1350 cm21

region indicates an absorbance feature (1) corresponding to a NLO stretch and (b) the 850–720 cm21 region indicates absorbance features (2) corresponding to a N–N stretch and (3) corresponding to a S–N stretch. Each spectrum represents the average of n = 3 trials with a standard deviation ¡ 2%.

Fig. 4 (a) For the dextran derivatives (1a, 1b) containing both 2u amine and 1uthiol sites, the introduction of t-butyl nitrite results in competitive nitrosation toform N-nitrosamine and S-nitrosothiol moieties, respectively. (b) Dextranderivatives (2a, 2b) containing amide and 1u thiol sites will selectively yieldS-nitrosothiol moieties, which decompose completely to give rise to NO anddisulfide formation.

15040 | RSC Adv., 2013, 3, 15035–15043 This journal is � The Royal Society of Chemistry 2013

Paper RSC Advances

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

amine linkage. The incorporation of the thiol group using thissynthetic approach resulted in 0.485 ¡ 0.007 and 0.299 ¡

0.005 mmol thiol g21 for the cysteamine (2a) and cysteine (2b)derivatives, respectively. Solutions of each nitrosated deriva-tive (0.75 mg mL21 in PBS) underwent NO analysis at roomtemperature with the samples exposed to ambient light. Thecysteamine (2a) and cysteine (2b) derivatives released 0.161 ¡

0.007 and 0.158 ¡ 0.007 mmol NO g21, respectively (Table 1).Fig. 5 shows the representative UV-vis absorbance spectra

for the cysteamine derivative (2a) before and after NO analysis.The initial absorbance feature occurs near 335 nm, indicatingnitroso product formation. The absorbance spectrum for thecysteine derivative (2b) is nearly identical (see ESI3, Fig. S14).The initial 335 nm peak decays nearly completely during NOanalysis (89% decomposition for 2a, 91% for 2b) leaving nosignificant absorbance feature. The lack of an absorbancefeature after NO release indicates that no RNNOs are detectedin the system and that the initial absorbance feature afternitrosation is due predominantly to RSNOs formed duringnitrosation (Fig. 4b). RNNO formation has been effectivelyeliminated by tuning the polymer functionality and the RSNOmoieties formed during nitrosation nearly completely decom-pose to result in NO release. The electron-withdrawing natureof the carbonyl group of the amide functional group serves tomake the amine less negative when compared to the 2u aminecase, resulting in a lower affinity for the NO+ nitrosatingspecies. Additionally, the steric crowding due to the carbonylgroup could block significant interaction of the NO+ with thenitrogen of the amide. Thus, nitrosation of the amide nitrogenis eliminated, resulting in predominant and selectiveS-nitrosation for the derivatives prepared via carboxymethylintermediate approach (2a, 2b).

Molar extinction coefficient values were determined bycorrelating the absorbance value at 335 nm to the concentra-tion of RSNO moiety. Solutions of each nitrosated dextranderivative were prepared at concentrations ranging from 0.05–1.0 mg polymer mL21. The NO recovery (mol per g of material)was corrected by the % decomposition of the moiety to yield a

conversion factor to express the g polymer mL21 concentra-tions in mol S-nitrosothiol L21 units. The molar extinctioncoefficient values were determined to be 1115 ¡ 42 M21 cm21

for the cysteamine derivative (2a) and 869 ¡ 12 M21 cm21 forthe cysteine derivative (2b) from the slope of the correspond-ing Beer’s law plot (see ESI3, Fig. S15). Both molar extinctioncoefficient values are within the reported order of magnitude(y103 M21 cm21) as those reported for small moleculeS-nitrosothiols in aqueous solution,6 indicating once againthat RSNOs are the predominant nitrosation product.

3.3 Impact of NO moiety formation on NO loading and release

The absorbance values at 335 nm associated with thenitrosated dextran derivatives containing an amide linkage(baseline corrected with a 0.75 mg mL21 solution of thecorresponding non-nitrosated derivative) were 0.143 ¡ 0.002for the cysteamine derivative (2a) and 0.116 ¡ 0.004 for thecysteine derivative (2b). By converting the absorbance valuesusing the corresponding molar extinction coefficients, theinitial RSNO concentrations were found to be 0.171 ¡ 0.007mmol g21 for 2a and 0.177 ¡ 0.007 mmol g21 for 2b, whichcorresponds to 35 ¡ 2 and 59 ¡ 3% nitrosation based uponthe initial thiol content for the cysteamine and cysteinederivatives, respectively.

It is interesting that, despite significantly different thiolcontent for the carboxymethyl intermediate polymers (0.459 ¡

0.004 mmol g21 for 2a, 0.296 ¡ 0.007 mmol g21 for 2b),similar RSNO concentrations resulted after nitrosation. In fact,the RSNO concentrations after nitrosation (0.171 ¡ 0.007mmol g21 for 2a, 0.177 ¡ 0.007 mmol g21 for 2b) arestatistically of the same population at the 90% confidence level(CL). Despite different amounts of thiol residues present ineach system, this would suggest a fundamental limit to thenitrosation under these conditions. Additionally, each deriva-tive with the same RSNO concentration decomposed to thesame extent (90%), resulting in total NO payloads that arestatistically of the same population (90% CL). Despite thesame total NO recoveries, the cysteamine (2a) and cysteine (2b)derivatives yield different release kinetic profiles (Fig. 6). Eachprofile exhibits a relatively flat, steady state NO release,followed by a decay to baseline. However, the cysteaminederivative (2a) releases a steady state value of y0.8 nmol NO,which is roughly twice the steady state NO of the cysteinederivative (2b), which releases y0.4 nmol NO at steady state.Additionally, the cysteamine derivative (2a) has completed itsNO release within y10 h, while the cysteine derivative (2b)requires y20 h to release the same NO payload. Overall, wefind that the functionality associated with the thiol pendantgroup (cysteamine, cysteine) results in different NO releasekinetics. The ability to tune the NO release kinetics basedupon differences in the polymer functionality is a criticalrequirement for polymer systems intended for differenttherapeutic applications.

Since the UV-vis absorbance features of the RNNO andRSNO moieties for the reductive amination dextran derivatives(1a, 1b) overlap, we cannot quantify the extent of nitrosationassociated with these derivatives. However, the total NOrecovery of each derivative appears to be impacted dependingupon the functionality of the nitrosated polymer. The ratio of

Fig. 5 UV-vis spectra for the nitrosated cysteamine dextran derivative (2a)prepared via carboxymethyl intermediate approach before and after NO analysis(0.75 mg mL21). All spectra represent n ¢ 3 with a standard deviation ¡ 9%.

This journal is � The Royal Society of Chemistry 2013 RSC Adv., 2013, 3, 15035–15043 | 15041

RSC Advances Paper

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

released NO relative to the amount of available thiol sites (seeTable 1) is much lower for the amine containing materials(y10–20%) compared to the amide containing materials(y30–50%). The polymers containing the amine linkageresulted in lower overall NO recovery when compared to thosewith the amide linkage. Since the amine-containing polymerscan experience nitrosation of both the thiol and the aminesites, and because the RNNO moieties are relatively stable,only the RSNO is giving rise to the detected NO. In general, theformation of stable RNNO moieties during the nitrosationprocess will not contribute to the NO release, thus lowering theNO yield associated with the amine containing derivatives. Bymaking available for nitrosation only the thiol sites using thecarboxymethyl intermediate approach, this alternate aminesink for the NO is eliminated and most of the NO loaded asRSNO is recovered as indicated by near-complete disappear-ance of the 335 nm RSNO absorbance band. It can also beconsidered that the formation of disulfide in the presence ofoxygen due to thiol oxidation36 could be resulting in lowerthiol availability, which would also impact the amount ofRSNO formed. However, because similar thiol contents areexhibited for a given thiol pendant regardless of the syntheticscheme employed (Table 1), the rate of disulfide formation willbe similar, not impacting the thiol availability significantlybetween synthetic approaches. Therefore, the increase in NOrecovery when switching from the reductive amination to thecarboxymethyl intermediate approach can be attributed to theelimination of competitive RNNO formation. The eliminationof harmful RNNO species on the polymer backbone isadditionally helpful toward the implementation of thesematerials for bioapplications.

4. Conclusions

Herein, we have reported the synthesis and characterization ofnitrosated biodegradable dextran derivatives containingcysteine and cysteamine pendant groups. Depending uponthe synthetic approach employed to prepare the derivative,amine or amide functional groups were available in additionto the thiol sites. After spectroscopic analysis of the nitrosatedamine-containing derivatives, it was found that N-nitrosaminefunctional groups formed competitively with S-nitrosothiols.The S-nitrosothiol moiety gave rise to NO release, while theN-nitrosamine remained stable under physiological pH.N-nitrosamines are undesirable side products due to theirtoxic and carcinogenic nature. As such, tuning the polymerchemistry such that the material contained amide instead ofamine linkages resulted in primarily S-nitrosothiol formation.S-nitrosothiol groups completely decomposed, giving rise tohigher amounts of NO when compared to the nitrosatedamine-containing derivatives. Overall, we present that thefunctional groups surrounding the thiol site on a polymer canbe tuned to directly impact the NO moieties that are formedduring nitrosation and the corresponding NO release proper-ties. Special care must be taken for polymer systems contain-ing multiple functional groups available for nitrosation.Toward the ultimate goal of creating NO-releasing polymerswith increased NO capacities, it is critical to identify andquantify how the NO is stored on the polymer and which of theNO moieties in the system are capable of releasing NO forbioactivity.

Acknowledgements

We would like to acknowledge financial support for thisresearch from Colorado State University and the Departmentof Defense Congressionally Directed Medical ResearchProgram (DOD-CDMRP). This research was supported byfunds from the Boettcher Foundation’s Webb-WaringBiomedical Research Program. V.B.D was supported from theDOD-CDMRP and J.M.J. was supported by the BoettcherFoundation’s Webb-Warning Biomedical Research Program.

References

1 M. C. Frost, M. M. Reynolds and M. E. Meyerhoff,Biomaterials, 2005, 26, 1685–1693.

2 V. N. Varu, N. D. Tsihlis and M. R. Kibbe, Vasc. Endovasc.Surg., 2009, 43, 121–131.

3 A. B. Seabra and N. Duran, J. Mater. Chem., 2010, 20,1624–1637.

4 D. A. Riccio and M. H. Schoenfisch, Chem. Soc. Rev., 2012,41, 3731–3741.

5 A. W. Carpenter and M. H. Schoenfisch, Chem. Soc. Rev.,2012, 41, 3742–3752.

6 D. L. H. Williams, Acc. Chem. Res., 1999, 32, 869–876.7 K. S. Bohl and J. L. West, Biomaterials, 2000, 21, 2273–2278.

Fig. 6 Representative real-time NO release profiles for the cysteamine (2a) andcysteine (2b) dextran derivatives prepared via carboxymethyl intermediateapproach (0.75 mg mL21 in PBS, room temperature). The cysteamine derivative(2a) exhibits a steady state NO release of y0.8 nmol, while the cysteinederivative (2b) exhibits y0.4 nmol NO.

15042 | RSC Adv., 2013, 3, 15035–15043 This journal is � The Royal Society of Chemistry 2013

Paper RSC Advances

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online

8 M. C. Frost and M. E. Meyerhoff, J. Biomed. Mater. Res., PartA, 2005, 72A, 409–419.

9 A. B. Seabra, R. da Silva and M. G. de Oliveira,Biomacromolecules, 2005, 6, 2512–2520.

10 D. A. Riccio, K. P. Dobmeier, E. M. Hetrick, B. J. Privett, H.S. Paul and M. H. Schoenfisch, Biomaterials, 2009, 30,4494–4502.

11 Y. Li and P. I. Lee, Mol. Pharmaceutics, 2010, 7, 254–266.12 A. B. Seabra, D. Martins, M. Simoes, R. da Silva, M. Brocchi

and M. G. de Oliveira, Artif. Organs, 2010, 34, E204–E214.13 P. N. Coneski, K. S. Rao and M. H. Schoenfisch,

Biomacromolecules, 2010, 11, 3208–3215.14 P. N. Coneski and M. H. Schoenfisch, Polym. Chem., 2011,

2, 906–913.15 V. B. Damodaran, J. M. Joslin, K. A. Wold, S. M. Lantvit and

M. M. Reynolds, J. Mater. Chem., 2012, 22, 5990–6001.16 D. A. Riccio, P. N. Coneski, S. P. Nichols, A. D. Broadnax

and M. H. Schoenfisch, ACS Appl. Mater. Interfaces, 2012, 4,796–804.

17 D. L. H. Williams, Nitrosation Reactions and the Chemistry ofNitric Oxide, Elsevier, 2004.

18 R. C. Shank, Toxicol. Appl. Pharmacol., 1975, 31, 361–368.19 P. N. Coneski and M. H. Schoenfisch, Org. Lett., 2009, 11,

5462–5465.20 Z. R. Zhou and M. E. Meyerhoff, Biomacromolecules, 2005,

6, 780–789.21 H. Zhang, The University of Michigan, 2002.22 K. A. Mowery, M. H. Schoenfisch, J. E. Saavedra, L.

K. Keefer and M. E. Meyerhoff, Biomaterials, 2000, 21, 9–21.23 P. G. Wang, M. Xian, X. P. Tang, X. J. Wu, Z. Wen, T. W. Cai

and A. J. Janczuk, Chem. Rev., 2002, 102, 1091–1134.

24 K. Szacilowski and Z. Stasicka, Prog. React. Kinet. Mech.,2001, 26, 1–58.

25 R. L. Williams, R. J. Pace and G. J. Jeacocke, Spectrochim.Acta, 1964, 20, 225–236.

26 Y. Fu, Y. Mou, B. L. Lin, L. Liu and Q. X. Guo, J. Phys. Chem.A, 2002, 106, 12386–12392.

27 M. Naessens, A. Cerdobbel, W. Soetaert and E.J. Vandamme, J. Chem. Technol. Biotechnol., 2005, 80,845–860.

28 V. B. Damodaran, L. W. Place, M. J. Kipper and M.M. Reynolds, J. Mater. Chem., 2012, 22, 23038–23048.

29 M. M. Reynolds, S. D. Witzeling, V. B. Damodaran, T.N. Medeiros, R. D. Knodle, M. A. Edwards, P. P. Lookianand M. A. Brown, Biochem. Biophys. Res. Commun., 2013,431, 647–651.

30 M. M. Reynolds, S. D. Witzeling, V. B. Damodaran, D.M. Jarigese, M. A. Edwards, R. D. Knodle, P. P. Lookian andM. A. Brown, J. Vet. Sci. Med., 2013, in press.

31 V. B. Damodaran, J. M. Joslin and M. M. Reynolds, PMSEPrepr., 2011, 105, 814–815.

32 R. Davies, R. C. Massey and D. J. McWeeny, J. Sci. FoodAgric., 1978, 29, 62–70.

33 M. J. Dennis, R. Davies and D. J. McWeeny, J. Sci. FoodAgric., 1979, 30, 639–645.

34 N. R. Shenoy and A. S. U. Choughuley, Cancer Lett., 1992,65, 227–232.

35 M. Keshive, S. Singh, J. S. Wishnok, S. R. Tannenbaum andW. M. Deen, Chem. Res. Toxicol., 1996, 9, 988–993.

36 N. Hisano, H. Iwata, Y. Teramura, H. Chen and Y. Ikada, J.Polym. Sci., Part A: Polym. Chem., 2011, 49, 671–679.

This journal is � The Royal Society of Chemistry 2013 RSC Adv., 2013, 3, 15035–15043 | 15043

RSC Advances Paper

Publ

ishe

d on

13

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f A

berd

een

on 1

6/08

/201

3 15

:17:

31.

View Article Online