Bc

IGMa

b

c

d

a

ARR2AA

KEUNG1C

1

mlmgislaolka

C4

(

h0

International Journal of Biological Macromolecules 107 (2018) 1765–1772

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

j ourna l h o mepa ge: www.elsev ier .com/ locate / i jb iomac

iosynthesis of dextran by Weissella confusa and its In vitro functionalharacteristics

rina Roscaa, Anca Roxana Petrovici a,∗, Dragos Peptanariua, Alina Nicolescua,ianina Dodib, Mihaela Avadaneia, Iuliu Cristian Ivanovc, Andra Cristina Bostanarud,ihai Maresd, Diana Ciolacua,∗

“Petru Poni” Institute of Macromolecular Chemistry, 700487, Iasi, RomaniaSCIENT - Research Center for Instrumental Analysis, 060114, Bucharest, RomaniaRegional Oncology Institute, 700483, Iasi, Romania“Ion Ionescu de la Brad” University, 700490, Iasi, Romania

r t i c l e i n f o

rticle history:eceived 21 July 2017eceived in revised form5 September 2017ccepted 9 October 2017vailable online 13 October 2017

eywords:

a b s t r a c t

The aim of this study was to monitor the influence of the fermentation conditions on the exopolysaccha-rides (EPS) biosynthesis. For this, different culture media compositions were tested on an isolated lacticacid bacteria (LAB) strain, identified by 16S rDNA sequence as being Weissella confusa. It was proved thatthis bacterial strain culture in MRS medium supplemented with 80 g/L sucrose and dissolved in UHTmilk produced up to 25.2 g/L of freeze-dried EPS, in static conditions, after 48 h of fermentative process.Using FTIR and NMR analysis, it was demonstrated that the obtained EPS is a dextran. The thermal analysisrevealed a dextran structure with high purity while GPC analysis depicted more fractions, which is normal

xopolysaccharideHT milkMRPC6S rDNA

for a biological obtained polymer. A concentration up to 3 mg/mL of dextran proved to have no cytotoxiceffect on normal human dermal fibroblasts (NHDF). Moreover, at this concentration, dextran breaks upto 70% of the biofilms formed by the Candida albicans SC5314 strain, and has no antimicrobial activityagainst standard bacterial strains. Due to their characteristics, these EPS are suitable as hydrophilic matrixfor controlled release of drugs in pharmaceutical industry.

ytotoxicity

. Introduction

Exopolysaccharides (EPS) are extracellular bio-acromolecules, who have received special attention in the

ast decade due to numerous applications, such as in the phar-aceutical, medical and food industries. These biopolymer are

enerally recognized as safe (GRAS) for human health and aredeal candidate for food industry, being used as gelling, emul-ifying, stabilizing and thickening agents [1]. EPS produced byactic acid bacteria (LAB) have immune-modulatory, antitumor,nti-inflammatory and immune-stimulator effects, and act asxidizing agents. Their structural properties determine the bio-

ogical activity and technological applications [2], therefore morenowledge about EPS isolation techniques, chemical compositionnd structure are of interest for the potential applications.

∗ Corresponding authors at: “Petru Poni” Institute of Macromolecular Chemistry,entre of Advanced Research in Bionanoconjugates and Biopolymers Department,1A Grigore Ghica-Voda Alley, 700487, Iasi, Romania.

E-mail addresses: petrovici.anca@icmpp.ro (A.R. Petrovici), dciolacu@icmpp.roD. Ciolacu).

ttps://doi.org/10.1016/j.ijbiomac.2017.10.048141-8130/© 2017 Elsevier B.V. All rights reserved.

© 2017 Elsevier B.V. All rights reserved.

Dextran is an EPS biosynthesized by several types of lacticacid bacteria, such as Leuconostoc mesenteroides, Lactobacillus bre-vis, Streptococcus mutants and Weissella confusa. Depending onthe strain and the composition of the culture medium, it can beobtained with a low or high molecular weight (10–150 kDa) [3].

This biopolymer is a very complex glycan composed of unitsof �-d-glucose with �-(1 → 6) linear bonds and different percent-ages of �-(1 → 4) �-(1 → 3) and �-(1 → 2) branches. The branchingdegree depends on the nature of dextransucrase biosynthesized bythe microbial strain. This enzyme hydrolyses the glycosidic bondin sucrose, releasing glucose which is further use in the biosynthe-sis of dextran and fructose. Both of them are involved in differentmetabolic processes [4]. EPS molecules are associated with oneanother, but can also interact with other molecules situated in theirproximity, such as proteins, lipids, inorganic ions or other macro-molecules found on the cell membrane surface [5].

The study of dextran obtained from the fermentation of Weis-sella spp. strains, especially Weissella confusa (W. confusa), has

recently entered into the attention of the scientific community. Itwas only in 2012 when the gene encoding (LBAE K39) of W. confusabiosynthesizes dextransucrase, which has a size of 180 kDa, wascompletely sequenced, thus becoming available for various appli-

1 logica

cnlawf[

aaDta[o

opouiomcgn

2

2

RACIp

2

aowemwt1gawfflooPUwqpoQp

766 I. Rosca et al. / International Journal of Bio

ations [6]. W. confusa is known to biosynthesis high amounts ofon-digestible oligosaccharides and mainly dextran as extracellu-

ar polysaccharides [7–9]. These polymers are receiving increasedttention because of both potential application as probiotics andide range of industrial uses, especially for bakeries [10–12] and

or the production of cereal-based fermented functional beverages13].

Moreover, dextran has different applications in pharmaceuticalnd light industries [14]. It is used in medicine as antithromboticgents, reducing blood viscosity and increasing its volume [15].extran-based nanoparticles have applications in targeted drug

herapy, where dextran is used as coating, or it can be function-lized with other compounds in order to obtain specific properties16]. Dextran is also used as coating for protection against oxidationf metal nanoparticles [17].

The aim of this study was to find the optimum conditions forbtaining high EPS amounts by fermentative methods. For this pur-ose, the compositions of the culture medium were varied, and thebtained polymers were extracted, purified and characterized. Thesed bacterial strain was isolated from commercial yoghurt and

dentified by molecular biology techniques as Weissella confusa. Inrder to select the potential medical applications of the biopoly-er we conducted a series of biological assays: determination of

ytotoxicity on fibroblasts, antibacterial testing and the antifun-al susceptibility test against one of the most known pathogenowadays, Candida albicans.

. Materials and methods

.1. Microorganisms

The lactic acid bacteria strain coded PP29 was isolated fromomanian commercial yoghurt in the laboratories of Centre ofdvanced Research in Bionanoconjugates and Biopolymers (Intel-entru) of the “Petru Poni” Institute of Macromolecular Chemistry,

asi, and kept at −80 ◦C in Man Rogosa Sharpe medium (MRS) sup-lemented with 20% glycerol.

.2. Molecular identification of the bacterial strain

The PP29 LAB strain was identifying by 16S rRNA gene sequencenalysis. The bacterial DNA was extracted from 24 h cultures grownn MRS agar plates at 30 ◦C. DNA purification was made in duplicateith the Genomic DNA Purification Kit (Thermo Scientific) and the

lution was made in 100 �L nuclease free water. The spectrophoto-etric quantification was made in NanoDrop. A 10 ng/�L dilutionas made for both samples and 5 �L were used for the PCR reac-

ions. The primer pair: 27F 5′ AGAGTTTGATCMTGGCTCAG 3′ and492R 5′ TACGGYTACCTTGTTACGACTT 3′ were used for 16 S rDNAene amplification of [18]. The PCR reactions were performed in

total volume of 25 �L (12.5 �L GoTaq®

Master Mix, 2.5 �L for-ard primer 10 mM, 2.5 �L reverse primer 10 mM, 2.5 �L nuclease

ree water and 5 �L DNA). The reaction mixture was first incubatedor 10 min at 95 ◦C and then cycled for 35 times 30 s at 95 ◦C fol-owed by one cycle of 4 min at 60 ◦C. The electrophoresis migrationf the products was conducted in 2% agarose gel electrophoresis inrder to verify the reaction and the possible contaminations. TheCR products were purified with Wizard SV Gel and PCR Clean-p Kit (Promega) and a new electrophoresis migration was madeith the final products in order to verify the purification and to

uantify the quantity to be sequenced. Sequencing reactions were

repared using primers 27F/1492R. DNA sequencing was carriedut by using GenomeLabTM Dye Terminator Cycle Sequencing withuick Start Kit (Beckman Coulter). The sequencing products wereurified with glycogen, sodium acetate and Na2-EDTA, as indi-

l Macromolecules 107 (2018) 1765–1772

cated by the sequencing kit and were migrated in the sequencerGenomeLabTM GeXP Genetic Analysis System with the migrationprogram LFR-a. The sequences were interpreted, exported in Chro-mas Lite program (version 2.01) and examined with depositedsequences by using nucleotide BLAST program (NCBI, http://www.ncbi. nlm.nih.gov) and BLAST search tools [19].

2.3. Fermentations conditions

The strain isolation and purification was made in Petri dishesusing MRS agar supplemented with 1% CaCO3, incubated at 30 ◦Cfor 48 h [20].

For experimental fermentations were used three culture mediadenoted MDI, MDII and MDIII. The culture medium compositionswere the following: MDI: MRS (55.3 g/L), fructose (40 g/L), glu-cose (40 g/L), dissolve in distillate water; MDII: MRS (55.3 g/L) andsucrose (80 g/L) dissolved in distillate water; MDIII: MRS (55.3 g/L)and sucrose (80 g/L) dissolved in UHT milk (with the followingnutritional information/100 mL: energy value − 44 kcal, proteins− 3 g, lipids − 1.5 g (saturated fatty acids − 0.9 g), sugar − 4.5 g, cal-cium ions − 120 mg). All the fermentations were made in static (S)and dynamic conditions (D).

The culture medium was sterilized at 110 ◦C for 30 min and inoc-ulated with 30% of fresh inoculum (24 h) with A600nm of 0.5 [21].The samples were incubated at 33 ◦C for 48 h without pH correc-tion during fermentation, under static and dynamic conditions (at100 rpm in an orbital incubator) [22]. Before performing the EPSextraction and purification, the culture was heated at 100 ◦C for15 min in order to inactivate the enzymatic equipment capable ofdegrading the biopolymer [23].

2.4. EPS isolation and purification

The cells and proteins were removed by precipitation with 20%trichloroacetic acid (TCA) followed by centrifugation at 10,000 rpmfor 10 min at 4 ◦C. The EPS were separated by precipitation withthree volumes of cold ethanol for 24 h at 4 ◦C [22]. The EPS were col-lected by centrifugation at 12,000 rpm for 15 min at 4 ◦C, washedwith ethanol three times, resuspended in double distilled water(DDW) and subjected to dialysis through a membrane with a poros-ity of 14,000 Da against DDW for three days at room temperature.For analysis, the EPS samples were coded as it follows: I-PP-29 −the PP29 strain fermented in MDI in dynamic conditions, I-PP-29-S− the PP29 strain fermented in MDI in static conditions, II-PP-29 −the PP29 strain fermented in MDII in dynamic conditions, II-PP-29-S − the PP29 strain fermented in MDII in static conditions, III-PP-29− the PP29 strain fermented in MDIII in dynamic conditions, III-PP-29-S − the PP29 strain fermented in MDIII in static conditions andsubjected to freeze drying process. The amount of the polymer wasexpressed in g of dry biopolymer per liter culture medium [24].

2.5. Gel permeation chromatography analysis

To estimate the distribution of the EPS molar masses, gel per-meation chromatography (GPC) was used. Measurements (weightaverage of molecular weight number (Mw), the average molecularnumber (Mn) and the polydispersity index (PDI)) were recorded ona Polymer Laboratories System (PL-GPC 120, Varian) equipped withrefractive index detector and three PL-aquagel packed columnsfilled with beads of porous gel composed of vinyl copolymers(cross-linked) with polymeric hydroxyl functionality (8 �m par-ticle size and 20, 40 and 60 Å pore type), connected in series and

placed in the column oven at 30 ◦C. The samples concentration was0.8 mg/mL in H2O (filtered through a cellulose filter with 0.45 �mpore size) and 0.02 M NaNO3 solution was used as mobile phasewith a flow rate of 1.0 mL/min. The calibration curve was made

I. Rosca et al. / International Journal of Biological Macromolecules 107 (2018) 1765–1772 1767

EPS samples (from top to bottom).

o18Da

2

t2wbmmswi

2

arr0Iznsudp

2a

psGcrs

Fig. 2. The infrared spectra of I-PP-29 (blue trace) and III-PP-29 (red trace). (Forinterpretation of the references to colour in this figure legend, the reader is referred

Fig. 1. The infrared spectra of

n P-82 standard pullulan (Shodex Denko) with Mw 0.6 × 104, × 104, 2.17 × 104, 4.88 × 104 and 11.3 × 104, 104 × 21, 36, 6 × 104,0.5 × 104 g/mole in H2O, and 100 �L injection volume was used.ata recording and processing were made with Cirrus GPC onlinend offline software.

.6. Fourier-transform infrared spectroscopy (FTIR)

The FTIR spectra were recorded in transmission on a Bruker Ver-ex 70 spectrometer (Bruker Optics, Germany) at a resolution of

cm−1, on KBr pellets and for the data processing OPUS 6.5 soft-are (Bruker Optics, Germany) was used. The interest regions were

aseline-corrected by using the interactive concave rubber bandethod. In the curve fitting procedure, the position and the esti-ated number of the sub-bands were determined by using the

econd derivative spectrum. The mixed Gauss-Lorentz functionsere used, where the peak position was maintained fixed and the

ntensity, shape and width were considered as variable parameters.

.7. Nuclear magnetic resonance studies

For the NMR analysis, the EPS sample was dissolved in deuter-ted water with TSP as internal standard. The spectra have beenecorded at room temperature (approx. 24 ◦C). Chemical shifts areeported in ppm and referred to TSP (ref. 1H 0.00 ppm and 13C.00 ppm). The NMR spectra had been recorded on a Bruker AvanceII 400 MHz Spectrometer, equipped with a 5 mm inverse detection-gradient probe, operating at 400.1 and 100.6 MHz for 1H and 13Cuclei. Total correlation spectroscopy (TOCSY) and heteronuclearingle quantum coherence (HSQC) experiments have been recordedsing standard pulse sequences in the version with z-gradients, aselivered by Bruker with TopSpin 2.1 PL6 spectrometer control androcessing software.

.8. Thermal analysis measurements − thermogravimetry (TGA)nd differential scanning calorimetry (DSC)

The thermogravimetric (TG) analysis of EPS samples wereerformed on STA 449F1 Jupiter NETZSCH equipment. DSC mea-urements were performed on a Maia F3 200 DSC device (Netzsch,

ermany) using 10 mg of freeze-dried sample. Measurements werearried out in the 30–700 ◦C temperature range, applying a heatingate of 10 ◦C min−1. Nitrogen purge gas was used as inert atmo-phere at a flow rate of 50 mL/min. Samples were heated in open

to the web version of this article.)

Al2O3 crucibles. The device was calibrated for temperature andsensitivity with indium, according to standard procedure.

2.9. Biological tests

2.9.1. Cytotoxicity assayTo assess the cytotoxicity, the CellTiter 96sAQueous One Solu-

tion Cell Proliferation Assay, MTS (Promega) was performed onnormal human dermal fibroblasts, NHDF (PromoCell), by follow-ing the protocol recommended by the manufacturer [25]. Thecells were grown in DMEM: F12 medium (Lonza) supplementedwith 10% fetal bovine serum (Gibco), 1 mM sodium pyruvate(Lonza) and 1% penicillin–streptomycin–amphotericin B mixture(10 K/10 K/25 mg in 100 mL, Lonza). The same medium was usedto prepare serial dilutions of the EPS to be tested. To ensure celladhesion, NHDF were seeded at 5 × 103 cells per well in 96-welltissue culture plates and incubated for 24 h at 37 ◦C under a humid-ified atmosphere with 5% CO2. The medium was then replaced with100 mL per well of the previously prepared EPS dilutions and theplates were further incubated for 20 h. The MTS reagent was thenadded to each well and, after final 4 h incubation, the absorbance at

490 nm was recorded with an EnVisions plate reader (PerkinElmer).The experiment included 6 replicates and was repeated 3 times.

1768 I. Rosca et al. / International Journal of Biological Macromolecules 107 (2018) 1765–1772

F O. (A) 1H NMR spectra for the dextran produced by fermentation in comparison with 1HN ctra of III-PP-29 sample, (D) 1H, 13C HSQC spectra of III-PP-29 sample.

2

EwtTg(ia

2

raammlatg

3

3b

iBb

t

Table 1The EPS amount extracted from the culture medium.

Fermentations conditions I-PP-29, g/L II-PP-29, g/L III-PP-29, g/L

ig. 3. NMR spectra of III-PP-29 EPS sample produced by W. confusa, recorded in D2

MR spectra of a commercial dextran, (B) 1H, 1H TOCSY spectrum, (C) 13C NMR spe

.9.2. The antifungal activityThe in vitro susceptibility testing was performed following the

UCAST EDef 7.2 guidelines using Candida albicans SC5314 strain,hich is known to form abundant biofilms with a complex struc-

ure. The stock solutions were prepared using water as a solvent.he active principle concentration range of the tested antifun-als was 0.0156–3 mg L−1. The minimum inhibitory concentrationsMICs) were determined using spectrophotometric technique on anMark Microplate Reader (Bio-Rad Laboratories, USA) at 405 nm,fter 24 h of incubation at 35 ◦C.

.9.3. Antimicrobial activityThe antimicrobial activity was considered on three different

eference strains, i.e. Escherichia coli ATCC 25922, Staphylococcusureus ATCC 6583 and Pseudomonas aeruginosa ATCC27853. Thentimicrobial activity was measured by the agar disk diffusionethod which supposes the addition of the EPS on the cultureedium pre-inoculated with the microbial suspension (0.5 McFar-

and standards optical turbidity in sterile saline solution, yielding suspension containing approximately 1 × 108 CFU mL−1 for allhree microorganisms), and measuring of the clear zone caused byrowth inhibition around the film disks after 24 h of incubation.

. Results and discussions

.1. The influence of the fermentative conditions on the EPSiosynthesis yield

The resulted sequences from PP29 LAB strain DNA sequenc-ng showed 98.4% identities with those available in the nucleotide

LAST data base and led to the identification of the Weissella confusaacterial species.

The fermentations of the W. confusa strain in three different cul-ure medium compositions, in static and dynamic conditions, lead

Static fermentations – – 25.2Dynamic fermentations 2.8 5.18 17.4

to different amounts of EPS biosynthesized after 48 h of fermenta-tive processes (Table 1). Our observations confirmed the literaturedata according to which the amount and the properties of the EPSdepend entirely on the composition of the culture medium and onthe fermentation conditions [26]. Dextran production by W. con-fusa is affected profoundly by nitrogen and carbon sources, and byinorganic salts present in the culture medium [27].

MDI and MDII in static conditions had inhibitory effect on EPSbiosynthesis due to the low oxygenation of the culture media. Inthe case of dynamic fermentations, after 48 h we obtained smallamounts of EPS. For MDI, the MRS carbon source was supplementedwith fructose and glucose, which are the components of sucrose,but there were obtained only 2.8 g/L EPS. This denoted that the LABenzymatic system does not recognized the components as sucrosebut as single sugars. For the MDII, the MRS carbon source was sup-plemented with sucrose, and after 48 h of fermentation there wereobtained almost double amounts of EPS (5.18 g/L EPS), which meansthat the proper sugar was sucrose. The MDIII culture medium wasvery productive, especially in static conditions there were obtained25.2 g/L EPS. This amount was obtained due to the presence inthe culture medium of milk, which supplements the MRS nitro-gen source, and also due to the calcium ions which enhances theviscosity of dextran [28].

3.2. Gel permeation chromatography analysis (GPC)

In order to characterize the EPS, gel permeation chromatog-raphy was used to obtain information on the molecular weight

I. Rosca et al. / International Journal of Biologica

Table 2Average molecular number (Mn), average molecular weight (Mw), and polydisper-sity index (PDI) of the EPS biosynthesis.

Sample fraction Mn, Da Mw, Da PDI

I-PP-29 1 2.4 * 105 4.5 * 105 1.862 4.5 * 104 6.6 * 104 1.46

II-PP-29 1 7.4 * 105 8.7 * 105 1.172 8.7 * 104 1.2 * 105 1.40

III-PP-29-S 1 9.9 * 105 1.2 * 106 1.24

evtiwaw

mmwoT

mpmflpiswmtw

3

awIaaecgcoafttot1(iv

obc

2 1.4 * 105 2.5 * 105 1.783 1.3 * 104 1.6 * 104 1.25

xpressed in fractions of different weight. Molecular weight is aery important variable that provides additional information onhe EPS physical properties, their molecule being dependent on theonic strength and therefore on their concentration in solution. The

eight average of the numerical molecular weight (Mw) and theverage molecular number (Mn) were estimated by comparisonith standard pullulan calibration curve.

Because the chromatographic method allows selectingolecules by size, an important parameter is the averageolecular weight that can reasonably approximate the moleculareight of the polymer fragments separated at the spine level. The

btained values for the EPS compounds extracted are presented inable 2.

Analyzing the results, it can be concluded that the numericolecular mass is influenced by the fermentation process, a

henomenon that determines the proportional increase of theolecular weight of the parent compound. For the EPS extracted

rom MDI culture media it was obtained a biopolymer with Mnower than the biopolymer extracted from MDII culture media sup-lemented with sucrose. This means that the sugar type is very

mportant for the obtained EPS with different Mn. In the MDIII andtatic conditions (III-PP-29-S) there were obtained three fractionsith high molecular weight and an uniform PDI. Also, by supple-enting the MRS nitrogen source, it was increased the amount and

he molecular weight of the biosynthesized polymer comparativelyith the fermentation made in MDII.

.3. Fourier-transform infrared spectroscopy (FTIR)

The FTIR spectra of samples I-PP-29 and II-PP-29, on one hand,nd III-PP-29 and III-PP-29-S, on the other hand, were very similarith each other (Fig. 1). Within the spectra of samples I-PP-29 and

I-PP-29, the overlapped absorption bands of hydroxyl stretchingnd NH amine stretching were noticed at 3394 cm−1. The amide Ind amide II bands (1668 cm−1 and 1545 cm−1) indicate the pres-nce of proteic residues. The region between 1490 and 1308 cm−1

overs signals from the bending vibrations of CH, CH2 and OHroups. The broad band with the maximum around 1058 cm−1 isharacteristics to glucopyranose fragments, being a superpositionf vibrations given by the C O C stretching and C OH stretchingnd bending. Identification of glycosidic linkages, of the bonded andree hydroxyl groups in various positions was made with the help ofhe second derivative spectra. The sub-band at 1054 cm−1, specifico �-d-glucoses, is mainly given by �(C2-O2H) [29,30]. A sub-bandbserved at 1022 cm−1 can be assigned to the � (C6-O6H) vibra-ions in H-bonded primary alcoholic groups. An intense peak at080 cm−1 arises from the overlapped �(C6-O6H), �(C4-C5) and �C1-H) vibrations [30,31]. The medium-to-intense sub-band peak-ng at 1129 cm−1 is confidently assigned to �(C4-O4H) stretchingibration [29].

The �(C O C) in glycosidic linkages of �-d-glucopyranoses isbserved at 1152 cm−1 and is further confirmed by the weaker sub-and at 832 cm−1 (�(C O)), which indicates the �-(1→3) anomericonfiguration [32,33]. The sub-band at 844 cm−1 is specific to �-

l Macromolecules 107 (2018) 1765–1772 1769

(1→6) linkages (Heyn, 1974). Therefore, the spectra of I-PP-29and II-PP-29 are very close to that of dextran, but with traces ofproteins and other polysaccharides. Instead, the FTIR spectra ofIII-PP-29 and III-PP-29-S do not show any signals from proteinresidues. The “sugar” band has a fine structure, where the vibra-tions corresponding to �(C6 O6H· · ·) (1015 cm−1), �(C3 O3H· · ·)(1040 cm−1), �(C6-O6H) + �(C4-C5) (1080 cm−1), or �as(C O C)and �(C O) in glycosidic linkages (1155 and 840 cm−1) are clearlyobserved (Fig. 2). Comparison with a sample of pure dextran fromcommercial source (M = 40.000) led to the almost perfect matchingof spectra.

3.4. NMR spectroscopy analysis

In order to confirm the presence of dextran in the analysedsample and to obtain more structural details there were recordedseveral one- and two-dimensional NMR experiments.

The 1H NMR spectrum of III-PP-29-S (Fig. 3A) was recordedwith suppression of the water signal, to reveal the less intense sig-nals from the region 4.5-5.5 ppm. The most intense signals wereassigned to the glucose units as it follows: � = 3.53 (1H, t, J = 9.4 Hz,H-4,), 3.59 (1H, dd, J = 9.7, 3.1 Hz, H-2), 3.73 (1H, t, J = 9.3 Hz, H-3),3.77 (1H, d, J = 9.8 Hz, H-6a or H-6b), 3.93 (1H, d, J = 9.1 Hz, H-5),4.00 (1H, d, J = 7.1 Hz, H-6a or H-6b), 4.99 (1H, d, J = 2.9 Hz, H-1).

The signal centered at 4.99 ppm has been previously assigned[34] to anomeric protons of �-(1 → 6)-linked glucose units, in themain chain. In the TOCSY spectrum presented in Fig. 3B it can benoticed the couplings of the anomeric proton from 4.99 ppm withH-2, H-3, H-4 and H-5 protons. These correlations confirmed thatthe investigated signals belong to the same spin system.

In the anomeric spectral region, three less intense signals wereobserved: a doublet centered at 4.67 ppm (J = 7.8 Hz), another dou-blet centered at 5.23 ppm (J = 3.2 Hz) and the third doublet centeredat 5.34 ppm (J = 3.7 Hz). The first two doublets belong to free glucoseremained from the culture medium, 4.67 ppm − anomeric protonfrom beta glucose and 5.23 ppm − anomeric proton from alphaglucose. The third doublet, from 5.34 ppm, has been previouslyassigned [34] to anomeric protons of �-(1 → 3)-linked glucoseunits, as branches of the main chain. The rest of the signals forthe branch glucose units could not be assigned because of their lowintensity and overlap with the signals from the main chain units.

Based on the two anomeric signals and the ratio of their inte-grals, the percentage of glycosidic linkages was establish as being96.2% for �-(1 → 6) and 3.8% for �-(1 → 3).

Information about carbon chemical shifts was obtained from the1D 13C NMR and 2D 1H, 13C HSQC spectra (Fig. 3C and D).

Based on the 1H–13C correlations from 1H, 13C HSQC spectrum,there were assigned the six intense signals as it follows: 68.5 (C-6),72.5 (C-4), 73.1 (C-5), 74.4 (C-2), 76.4 (C-3), 100.7 (C-1). All thesesignals belong to �-(1 → 6)-linked glucose units, in the main chain.From the same HSQC spectrum it was assigned the C-1 from thebranch �-(1 → 3)-linked glucose units at 102.6 ppm.

The obtained NMR data indicated the obtaining of anexopolysaccharide. Based on the reported data from literature [34]it was deduced that the obtained exopolysaccharide is dextran,results supported by the FTIR analysis as well.

3.5. Thermal analysis measurements − thermogravimetry (TGA)and differential scanning calorimetry (DSC)

By analyzing the thermal degradation curves, it can notice thatall four samples displayed comparable thermal degradation ten-

dency in two stages (Fig. 4, Table 3).

Since there is a direct relationship between carboxyl group con-tent and the material hydrophobicity, the initial material moisturecontent is given by the increased carboxyl group quantity. Regard-

1770 I. Rosca et al. / International Journal of Biological Macromolecules 107 (2018) 1765–1772

Fig. 4. TG and DTG curves of the studied samples.

Table 3Thermal characteristics of the studied samples.

Sample I-PP-29 II-PP-29 III-PP-29 III-PP-29-S

Ti, ◦C Tmax, ◦C Tf, ◦C �w, % Ti, ◦C Tmax, ◦C Tf, ◦C �w, % Ti, ◦C Tmax, ◦C Tf, ◦C �w, % Ti, ◦C Tmax, ◦C Tf, ◦C �w, %

Stage I 57 71 95 6.3 57 76 88 7.08 48 78 108 4.77 55 80 114 8.1Stage II 260 304 332 53.14 265 302 331 55.59 269 307 322 68.04 267 305 323 70.53Wrez, % 40.56 37.33 27.19 21.37

Ti − initial thermal degradation temperature; Tmax − temperature corresponding to the mcurves; Tf − final thermal degradation temperature; Wm − weight loss rate correspondindegradation (700 ◦C).

Fig. 5. Second heating curves of the studied samples.

im7sirr

Abrdmv

glucose and mannose, in a concentration of 4 mg/L, inhibited 65.82%of S. aureus biofilm formation, 43.58% P. aeruginosa and 33.41% ofE. coli [41]. In the same time, EPS based on glucose had antibacterialactivity against E. coli ATCC8739 and S. aureus ATCC6538 [42]. How-

ng to this, the samples structures suffer physical and chemicaloisture loss during the first thermal degradation stage, between

1 and 80 ◦C with a mass loss from 4.7-8.1% (Table 3) [35]. Theecond and principal degradation step is from 302 to 307 ◦C record-ng an important mass loss of 53.17–70.53% [36], corresponding toandom polymer backbone cleavages and depolymerisation, withemaining of inorganic material [37].

Fig. 5 depicts the second heating scans of the studied samples.ll samples showed a glass transition temperature domain (Tg)etween 217 and 223 ◦C. This aspect is in accordance with the dataeported by Scandola [38] reporting a Tg value of 223 ◦C for naturalextran. The producing of a more amorphous (i.e. enhanced seg-ental chain mobility) dextran could explain the lower Tg domain

alues, down to 217 ◦C.

aximum rate of decomposition for each stage, evaluated from the peaks of the DTGg to the Tmax values; Wrez − percentage of residue remained at the end of thermal

3.6. Biological tests

3.6.1. Cytotoxicity assayBy testing the biocompatibility it was proved that III-PP-29-S

sample, dextran-based, does not have cytotoxic effect on fibroblastsup to a maximum concentration of 3 mg/mL, when cell viabilitydrops below 70% (Fig. 6). Other results have been reported forEPS based on glucose and mannose, biosynthesized by Lactobacillusplantarum, where the fibroblast cell viability decreased to 70%, ata concentration of 100 mg/mL EPS [39]. These differences betweenresults may be related to the different type of lactic acid bacteriastrain and on the different type of EPS resulted from fermentation.For example, EPS based on galactose and glucose at a concentrationof 5–50 g/L did not have cytotoxic effect on the normal intesti-nal cells [40]. EPS based on galactose and glucose, at a dosage of5–50 �g/mL showed no cytotoxic effect on normal intestine 407cells [40].

3.6.2. The antifungal activityAntifungal susceptibility testing using the species Candida albi-

cans showed that III-PP-29-S can break by up to 70% of the biofilmsformed by the pathogenic fungus to a maximum concentration of3 mg/mL (Fig. 7), results that have not been reported yet in theliterature.

3.6.3. EPS antimicrobial activityEPS based on dextran showed no antimicrobial activity against

the reference strains. However, EPS based on small amounts ofsulphate (1.43%), uronic acids (2.98%), proteins (4.08%), galactose,

I. Rosca et al. / International Journal of Biological Macromolecules 107 (2018) 1765–1772 1771

Fig. 6. Relative cell viability at 48 h. The relative cell viability of NHDF cells was determined by MTS assay after 48 h of treatment with EPS. Data is represented as means ± SD(n = 3).

antifu

ee

4

empptiC

S

A

APaHa

R

[

[

[

[

[

Fig. 7. EPS

ver, EPS produced by this particular strain have no antimicrobialffect against the reference pathogens.

. Conclusions

FTIR and NMR analyses confirm the dextran structure of the EPSxtracted from experiment III-PP-29-S. According to TGA, the poly-er has a very stable structure with high purity and its amorphous

roperties were demonstrated by DSC analysis. Because the EPSroved to have antifungal properties, the next step will be to loadogether with this compound an antifungal drug in order to testts capacity to destroy the biofilm formation and the yeast cells ofandida albicans, which is nowadays an important health issue.

tatement of conflict of interest

The authors declare no competing financial interest.

cknowledgements

This work was supported by a grant of the Romanian Nationaluthority for Scientific Research, CNCS − UEFISCDI, project numberN-II-RU-TE-2014-4-0558. PhD I. Ros ca and MD PhD D. Peptanariure grateful also for the financial support from the European Union’sorizon 2020 research and innovation programme under grantgreement No 667387 WIDESPREAD 2-2014 SupraChem Lab.

eferences

[1] P.B. Devi, D. Kavitake, P.H. Shetty, Physico-chemical characterization of

galactan exopolysaccharide produced by Weissella confusa KR780676, Int. J.Biol. Macromol. 93 (2016) 822–828.

[2] D. Kavitake, P.B. Devi, S.P. Singh, P.H. Shetty, Characterization of a novelgalactan produced by Weissella confusa KR780676 from an acidic fermentedfood, Int. J. Biol. Macromol. 86 (2016) 681–689.

[

ngal assay.

[3] B. Srinivas, P. Naga Padma, Green synthesis of silver nanoparticles usingdextran from Weissella confusa, Int. J. Sci. Environ. Technol. 5 (2016) 827–838.

[4] S. Shukla, A. Goyal, Medium optimization of fermentation for enhanceddextrans production from Weissella confusa Cab3 by statistical methods, Curr.Biotechnol. 2 (2013) 39–46.

[5] A. Mishra, B. Jha, Microbial exopolysacchrides, in: E. Rosenberg, E.F. DeLong, F.Thompson, S. Lory, E. Stackebrandt (Eds.), The Prokaryotes: AppliedBacteriology and Biotechnology, 4th ed., Springer, Berlin Heidelberg, 2013, pp.179–192.

[6] S. Shukla, Q. Shi, N.H. Maina, M. Juvonen, M. Tenkanen, A. Goyal, Weissellaconfusa Cab3 dextransucrase: properties and in vitro synthesis of dextran andglucooligosaccharides, Carbohyd. Polym. 101 (2014) 554–564.

[7] N.H. Maina, L. Virkki, H. Pynnönen, H. Maaheimo, M. Tenkanen, Structuralanalysis of enzyme-resistant isomaltooligosaccharides reveals the elongationof �-(1 → 3)-linked branches in Weissella confusa dextran,Biomacromolecules 12 (2011) 409–418.

[8] M.S. Bounaix, H. Robert, V. Gabriel, S. Morel, M. Remaud-Siméon, B. Gabriel, C.Fontagné-Faucher, Characterization of dextran producing Weissella strainsisolated from sourdough and evidence of consitutive dextransucraseexpression, FEMS Microbiol. Lett. 311 (2010) 18–26.

[9] M. Amari, L.F. Arango, V. Gabriel, H. Robert, S. Morel, C. Moulis, B. Gabriel, M.Remaud-Siméon, C. Fontagné-Faucher, Characterization of a novel sucrosefrom Weissella confusa isolated from sourdough, Appl. Microbiol. Biotechnol.97 (2013) 5413–5422.

10] K. Katina, N.H. Maina, R. Juvonen, L. Flander, L. Johansson, L. Virkki, M.Tenkanen, A. Laitila, In situ production and analysis of Weissella confusadextran in wheat sourdough, Food Microbiol. 26 (2009) 734–743.

11] R. Coda, R. DiCagno, M. Gobbetti, C.G. Rizzello, Sourdough lactic acid bacteria:exploration of non-wheat cereal-based fermentation, Food Microbiol. 37(2014) 51–58.

12] I. Kajala, Q. Shi, A. Nyyssölä, N.H. Maina, Y. Hou, K. Katina, M. Tenkanen, R.Juvonen, Cloning and characterization of a Weissella confusa dextransucraseand its application in high fibre baking, PLoS One 10 (2015) e0116418, http://dx.doi.org/10.1371/jour-nal.pone.0116418.

13] E. Zannini, A. Mauch, S. Galle, M. Gänzle, A. Coffey, E.K. Arendt, J.P. Taylor,D.M. Waters, Barley malt wort fermentation by exopolysaccharide-formingWeissella cibaria MG1 for the production of a novel beverage, J. Appl.Microbiol. 115 (2013) 1379–1387.

14] H. Gloria Hernández, S. Livings, J.M. Aguilera, A. Chiralt, Phase transitions of

dairy proteins, dextrans and their mixtures as a function of waterinteractions, Food Hydrocolloid. 25 (5) (2011) 1311–1318.

15] M. Naessens, A.N. Cerdobbel, W. Soetaert, E.J. Vandamme, Leuconostocdextransucrase and dextran: production, properties and applications, J. Chem.Technol. Biotechnol. 80 (2005) 845–860.

1 logica

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[exopolysaccharides by Lactobacillus helveticus MB2-1 and its functional

772 I. Rosca et al. / International Journal of Bio

16] S.C. McBain, H.H. Yiu, J. Dobson, Magnetic nanoparticles for gene and drugdelivery, Int. J. Nanomed. 3 (2008) 169–180.

17] S.L. Easo, P.V. Mohanan, Hepatotoxicity evaluation of dextran stabilized ironoxide nanoparticles in Wistar rats, Int. J. Pharm. 509 (2016) 28–34.

18] J.A. Frank, C.J. Reich, S. Sharma, J.S. Weisbaum, B.A. Wilson, G.J. Olsen, Criticalevaluation of two primers commonly used for the amplification of bacterial16rRNA genes, Appl. Environ. Microb. 74 (8) (2008) 2461–2470.

19] S.F. Altschul, T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, D.J.Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein databasesearch programs, Nucleic Acids Res. 25 (1997) 3389–3402.

20] C. Schiraldi, V. Valli, A. Molinaro, M. Carteni, M. De Rosa, Exopolysaccharidesproduction in Lactobacillus bulgaricus and Lactobacillus casei exploitingmicrofiltration, J. Ind. Microbiol. Biotechnol. 33 (5) (2006) 384–390.

21] C.T. Liu, I.T. Hsu, C.C. Chou, P.R. Lo, R.C. Yu, Exopolysaccharide production ofLactobacillus salivarius BCRC 14759 and Bifidobacterium bifidum BCRC 14615,World J. Microb. Biotechnol. 25 (2009) 883–890.

22] A. Tayuan, G.W. Tannock, S. Rodtong, Growth and exopolysaccharideproduction by Weissella sp. from low-cost substitutes for sucrose, Afr. J.Microbiol. Res. 5 (22) (2011) 3693–3701.

23] N. Sengul, S. Isik, B. Aslim, G. Ucar, A.E. Demirbag, The effect ofexopolysaccharide-producing probiotic strains on gut oxidative damage inexperimental colitis, Digest. Dis. Sci. 56 (3) (2011) 707–714.

24] S. Palomba, S. Cavella, E. Torrieri, A. Piccolo, P. Mazzei, G. Blaiotta, V.Ventorino, O. Pepe, Wheat sourdough from Leuconostoc lactis andLactobacillus curvatus exopolysaccharide-producing starter culture:polyphasic screening, homopolysaccharide composition and viscoelasticbehavior, Appl. Environ. Microb. 78 (8) (2012) 2737–2747.

25] CellTiter 96® Aqueous One Solution Cell Proliferation Assay, technicalbulletin, Promega Corporation, 2012.

26] Y. Wang, C. Li, P. Liu, Z. Ahmed, P. Xiao, X. Bai, Physical characterization ofexopolysaccharide produced by Lactobacillus plantarum KF5 isolated fromTibet Kefir, Carbohyd. Polym. 82 (2010) 895–903.

27] S. Shukla, A. Goyal, 16S rRNA based identification of a glucan hyper-producingWeissella confusa, Enzym. Res. (2011) 250842, http://dx.doi.org/10.4061/2011/250842.

28] A.N. de Belder, Dextran, edited by Handbooks from Amersham Biosciences UK

Limited Amersham Place, Little Chalfont, Buckinghamshire HP7 9NA, England,18-1166-12, 2003.

29] Y. Maréchal, M. Milas, M. Rinaudo, Hydration of hyaluronan polysaccharideobserved by IR spectrometry. III. Structure and mechanism of hydration,Biopolymers (Biospectroscopy) 72 (2003) 162–173.

[

l Macromolecules 107 (2018) 1765–1772

30] M. Kanou, K. Nakanishi, A. Hashimoto, T. Kameoka, Influences ofmonosaccharides and its glycosidic linkage on infrared spectral characteristicsof disaccharides in aqueous solutions, Appl. Spectrosc. 59 (7) (2005) 885–892.

31] K.I. Shingel, Determination of structural peculiarities of dextran, pullulan andgamma-irradiated pullulan by Fourier-transform IR spectroscopy, Carbohyd.Res. 337 (2002) 1445–1451.

32] J.J. Cael, J.L. Koenig, J. Blackwell, Infrared and raman spectroscopy ofcarbohydrates. Part VI: Normal coordinate analysis of V-amylose,Biopolymers 14 (1975) 1885–1903.

33] N.N. Siddiqui, A. Aman, A. Silipo, S.A.U. Qadera, A. Molinaro, Structuralanalysis and characterization of dextran produced by wild and mutant strainsof Leuconostoc mesenteroides, Carbohyd. Polym. 99 (2014) 331–338.

34] N.H. Maina, M. Tenkanen, H. Maaheimo, R. Juvonen, L. Virkki, NMRspectroscopic analysis of exopolysaccharides produced by Leuconostoccitreum and Weissella confuse, Carbohydr. Res. 343 (2008) 1446–1455.

35] Z. Ahmed, Y. Wang, N. Anjum, H. Ahmad, A. Ahmad, M. Raza, Characterizationof new exopolysaccharides produced by coculturing of L. kefiranofaciens withyoghurt strains, Int. J. Biol. Macromol. 59 (2013) 377–383.

36] D. Kothari, J.M. Rao Tingirikari, A. Goyal, In vitro analysis of dextran fromLeuconostoc mesenteroides NRRL B-1426 for functional food application,Bioact. Carbohydr. Dietary Fibre 6 (2015) 55–61.

37] J. Maia, R.A. Carvalho, J.F.J. Coelho, P.N. Simões, M.H. Gil, Insight on theperiodate oxidation of dextran and its structural vicissitudes, Polymer 52(2011) 258–265.

38] M. Scandola, G. Ceccorulli, M. Pizzoli, Molecular motions of polysaccharides inthe solid state: dextran, pullulan and amylose, Int. J. Biol. Macromol. 13 (4)(1991) 254–260.

39] S.V. Dilna, H. Surya, R.G. Aswathy, K.K. Varsha, D.N. Sakthikumar, A. Paudey,K.M. Nampoothiri, Characterization of an exopolysaccharide with potentialhealth-benefit properties from a probiotic Lactobacillus plantaroum RJF4,LWT-Food Sci. Technol. 64 (2015) 1179–1186.

40] C.T. Liu, F.J. Chu, C.C. Chou, R.C. Yu, Antiproliferative and anticytotoxic effectsof cells fractions and exopolysaccharides from Lactobacillus casei 01, Mutat.Res. Genet. Toxicol. Environ. Mutagen. 721 (2011) 157–162.

41] W. Li, J. Ji, X. Rui, J. Yu, W. Tang, X. Chen, M. Jiang, Production of

characteristics in vitro, LWT-Food Sci. Technol. 59 (2014) 732–739.42] I. Trablesi, S.B. Shima, H. Chaabane, B.S. Riadh, Purification and

characterization of a novel exopolysaccharide produced by Lactobacillus sp.Ca6, Int. J. Biol. Macromol. 74 (2015) 541–546.