This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 130.203.245.174

This content was downloaded on 11/11/2015 at 00:17

Please note that terms and conditions apply.

In vitro evaluation of carbon-nanotube-reinforced bioprintable vascular conduits

View the table of contents for this issue, or go to the journal homepage for more

2014 Nanotechnology 25 145101

(http://iopscience.iop.org/0957-4484/25/14/145101)

Home Search Collections Journals About Contact us My IOPscience

Nanotechnology

Nanotechnology 25 (2014) 145101 (10pp) doi:10.1088/0957-4484/25/14/145101

In vitro evaluation of carbon-nanotube-reinforced bioprintable vascular conduitsFarzaneh Dolati1,2, Yin Yu1,3, Yahui Zhang1,2, Aribet M De Jesus3,Edward A Sander3 and Ibrahim T Ozbolat1,2

1 Biomanufacturing Laboratory, Center for Computer-Aided Design, The University of Iowa, Iowa City,IA, USA2 Mechanical and Industrial Engineering Department, The University of Iowa, Iowa City, IA, USA3 Biomedical Engineering Department, The University of Iowa, Iowa City, IA, USA

E-mail: ibrahim-ozbolat@uiowa.edu

Received 31 October 2013, revised 21 January 2014Accepted for publication 17 February 2014Published 14 March 2014

AbstractVascularization of thick engineered tissue and organ constructs like the heart, liver, pancreasor kidney remains a major challenge in tissue engineering. Vascularization is needed to supplyoxygen and nutrients and remove waste in living tissues and organs through a network thatshould possess high perfusion ability and significant mechanical strength and elasticity. In thispaper, we introduce a fabrication process to print vascular conduits directly, where conduitswere reinforced with carbon nanotubes (CNTs) to enhance their mechanical properties andbioprintability. In vitro evaluation of printed conduits encapsulated in human coronary arterysmooth muscle cells was performed to characterize the effects of CNT reinforcement on themechanical, perfusion and biological performance of the conduits. Perfusion and permeability,cell viability, extracellular matrix formation and tissue histology were assessed and discussed,and it was concluded that CNT-reinforced vascular conduits provided a foundation formechanically appealing constructs where CNTs could be replaced with natural proteinnanofibers for further integration of these conduits in large-scale tissue fabrication.

Keywords: biofabrication, vascular tissue engineering, carbon nanotubes

(Some figures may appear in colour only in the online journal)

1. Introduction

Artificial blood vessels save the lives of many patients,especially in bypass applications such as shunts for dialysisor treatment of blood vessel failure; they also can be usedas supplement vessels for the fabrication of engineered thicktissues [1]. Based on the application of artificial vasculartissues, several features are required, including perfusability,mechanical elasticity and strength for pulsatile stress andsuture retention, diffusion properties, and the capabilityto transport nutrients, oxygen, and waste [2, 3]. In orderto develop vascular constructs biomimetically, one shouldthus consider incorporating these requirements during thefabrication process by choosing proper biomaterials andfabrication methods [2, 4].

Hydrogels are among the most commonly employedmatrices in tissue engineering because they are biocompatible,able to facilitate nutrient/oxygen transport, and are highlyhydrated three-dimensional (3D) networks that structurallyresemble the extracellular matrix (ECM) [2, 4, 5]. A suitablehydrogel for fabricating vascular conduits needs to satisfyseveral design parameters to promote the formation offunctional new tissues—for instance, physical parameters,biological properties, and biocompatibility [2, 5, 6]. Naturalpolymers, such as collagen, alginate, and chitosan, havebetter biocompatibility than synthetic polymer gels. However,they have limitations in mechanical properties (physicalparameters), which are very important design criteria in tissueengineering [6]. Therefore, modifying their properties bymeans of adding other polymers or fibers can be helpful.One of the most common natural hydrogels is alginate, which

0957-4484/14/145101+10$33.00 1 c© 2014 IOP Publishing Ltd Printed in the UK

Nanotechnology 25 (2014) 145101 F Dolati et al

is appropriate for fabricating tissue construct because it is anatural polymer, is abundant in nature, is highly biocompatible,and has low toxicity and macro-molecular properties similarto those of natural ECM [7]. It has a crosslinking-basedgelation property that enables printing through a coaxial nozzleunit. When biomaterial solution and crosslinker solution arefed simultaneously, the gelation of alginate vascular conduitsoccurs immediately, providing tubular conduits. The processof coaxial printing allows printing vascular conduits in anydesired shape in 3D [8]. In this work, the goal is to modifyalginate properties by adding carbon nanotubes to preparecomposite solutions for fabricating vascular conduits. Thecomposite solution consists of a biocompatible polymer asa matrix and is reinforced with fibers.

Determining the orientation, de-aggregation, anddispersion of reinforcement by fibers can increase thestrength and resistance to deformation of the compositevasculature [3, 6, 7]. By homogeneously dispersing reinforcedfibers, the interfacial interaction with the matrix can increase,resulting in the enhancement of the mechanical propertiesof the composite [3, 7–9]. One approach to improvingthe mechanical properties of alginate is the integration ofcarbon nanotubes (CNTs), which are among the strongestmaterials known and possess a simple structure [10–12].In particular, single-walled carbon nanotubes (SWCNTs)and multi-walled carbon nanotubes (MWNTs) have uniquemechanical strength and stiffness, a high aspect ratio,are lightweight, and are flexible and low-density [10].Therefore, CNTs have outstanding potential as reinforcementsin composite materials. Several researchers used CNTs intheir studies to enhance mechanical properties and evaluatedcell viability and proliferation in tissue scaffolds. Yildirimet al investigated whether SWCNT-incorporated scaffolds hadbetter cell attachment and proliferation, thus having moreviable cells after seven days in comparison with non-reinforcedscaffolds [13]. Tissue scaffolds were fabricated by meansof a freeform fabrication technique through layer-by-layerdeposition of material. The mechanical testing showed thatthe composite reinforced with 1% SWCNT had higher tensilestrength. Mattson et al successfully used CNT as suitablesubstrates for nerve cell growth [14]. In that study, the flatsubstrate of MWCNTs was coated with a 4-hydroxynonenalbioactive molecule to create multiple neuritis. Then neuronswere grown on that substrate and generated a number ofbranches. However, neurons that were grown on the substrateswithout CNTs could generate only a few branches since theydid not resemble the cellular surfaces and ECM in the body.Hu et al used the same method as Mattson et al did [14]to incorporate CNTs as a substructure in order to culturethe hippocampal neuronal growth. The coated surface withMWCNTs was able to control the neuritis outgrowth, generatelonger neuritis length, and control the branching pattern [15].In addition, the study by Lobo et al showed that it was possibleto achieve higher cell viability and cell adhesion on MWCNTfilms [16].

In this paper, MWCNT-reinforced alginate vascularconduits were printed using a coaxial bioprinting process.Mechanical properties were evaluated and compared with

vascular conduits made of plain alginate with differentconcentrations of alginate and dosages of MWCNTs. Inaddition, the elongation and diffusion rate of the vascularconduit were investigated. Furthermore, cell viability tests andtissue histology studies were conducted to evaluate effect ofMWCNTs on cell viability, matrix deposition and the tissuegeneration process.

2. Materials and methods

2.1. Materials

Sodium alginate (purchased from Sigma Aldrich, UnitedKingdom) and calcium chloride (CaCl2) (purchased fromSigma Aldrich, Japan) were used to fabricate vascularconduits. Solutions of 3–4% (w/v) sodium alginate weredissolved in deionized water and placed in a shaker for 10 h at120 rpm. Similarly, 4% (w/v) calcium chloride solutions wereprepared using deionized water [8]. These concentrations werepreferred due to the manufacturability of functional conduitsbased on our earlier experiments.

Pristine MWCNTs were purchased from Nano Labs Inc.,USA. For preparation of the CNT solution a functionalizationmethod and an acid-washing treatment of nitric acid 70%(HNO3) were used to oxidize the pristine CNTs [17]. Carbonnanotubes (100 mg) were dispersed by sonication in 250 mlfor 1 h. Then, the mixture of MWCNT-HNO3 was refluxed at140 ◦C for 1.5 h while stirring. Next, the mixture was cooleddown to room temperature The concentration of the suspensionwas 70% (w/v).

For the first four composite samples, 1% (w/v) and 0.5%(w/v) MWCNT were dissolved in alginate solutions with 3%and 4% concentrations, and each sample was placed on a stirrerfor one day until the CNTs were dissolved homogeneously.Two samples of alginate only with concentrations of 3% and4% (w/v) were also produced.

2.2. Cell preparation

Human coronary artery smooth muscle cells (HCASMCs) (lifetechnologies, MA, USA) were used in this study to test cellviability when encapsulated in carbon-nanotube-reinforcedvascular conduits The HCASMCs were cultured at 37 ◦C in 5%CO2 in smooth muscle cell growth media (life technologies,MA, USA) supplemented with smooth muscle cell growthsupplement (life technologies, MA, USA), 100 µg µl−1

penicillin, 100 µg ml−1 streptomycin, and 2.5 µg µl−1

Fungizone. The culture media was changed every other day.Cells were harvested until we achieved a sufficient amount forbioprinting. After harvesting, cells were centrifuged down,re-suspended in a 4% sodium alginate and 1% MWCNTsolution, and gently mixed by a vortex mixer to get uniformdistribution. The cell-seeding density used in this study was10× 106 cells ml−1.

2.3. Fabrication of carbon-nanotube-reinforced vascularconduits

The experimental setup (see figures 1(a) and (b)) consisted of asingle-arm robotic printer (EFD Nordson, East Providence, RI)

2

Nanotechnology 25 (2014) 145101 F Dolati et al

Figure 1. The experimental setup for fabricating vascular conduits: (a) the coaxial nozzle system, (b) the bioprinter platform and (c)cross-sectional view of the coaxial nozzle assembly.

Figure 2. Biotense perfusion bioreactor: (a) tensile test system, (b) sample grips.

with a motion unit, a coaxial nozzle (see figures 1(a) and (c)),a syringe pump (New Era Pump System Inc., Farmingdale,RI) and a pressure regulator (EFD Nordson, East Providence,RI) for biomaterial and CaCl2 solution as the crosslinker. Thecoaxial nozzle was mounted on the bioprinter unit

The coaxial nozzle had three sections: a feed tube, anouter tube, and an inner tube (see figure 1(c)). The compositesolution was fed from the feed tube, and the inner tube wasused to feed the crosslinker solution. The composite solutionflowed into the space between the inner and outer tubes (seefigure 1(c)). A dispenser was connected to the syringe barrel,which contained the composite solution, and the syringe pumpwas connected to the barrel with the calcium chloride solution(see figure 1(d)). The sodium alginate was dispensed fromthe sheath section of the nozzle, while the crosslink solutionwas dispensed from the core section by compressed air. Whentwo solutions contacted, the vascular conduits were formedand printed. The coaxial assembly used in this study had a 25gauge (260 µm inner diameter (I.D.), 510 µm outer diameter(O.D.)) inner needle and a 14 gauge (1600 µm I.D., 2110 µmO.D.) outer needle. The dispensing rate for the CaCl2 solutionwas 16 ml min−1 and 12 ml min−1 for 4% and 3% plainalginate, respectively. The dispensing pressure for compositesolution was 20.7 kPa and 13.8 kPa for 4% and 3% alginate,respectively.

2.4. Mechanical testing

All printed vascular conduits were soaked in a calciumchloride solution for 24 h in order to minimize the effect

of residence time in the CaCl2 solution. Three differentrandom segments for each sample were fabricated to evaluatemechanical characterization using a Biotense PerfusionBioreactor (ADMET, Inc. Norwood, MA) (see figure 2).

The mechanical testing unit consisted of a linear actuator,sample grips, a bioreactor frame, and a 250 g load cell. The loadcell and closed-loop servo-controlled actuator can measure amaximum tensile load of 2 N and provide a stroke of 25 mm,respectively. The samples were mounted in the grips betweenpieces of sandpaper (to minimize slip), leaving a samplegauge length of 6–8 mm for mechanical loading. The vascularconduits were loaded to failure at a rate of 10 mm min−1.Samples that failed at the grips were discarded from analysis.Load–displacement data was recorded at 1 Hz through adata acquisition system (MTestQuattro System, ADMET, Inc.,Norwood, MA).

2.5. Perfusion test

The main function of vascular conduits is to provide nutrientsand oxygen to surrounding tissues as well as take away thewaste. In order to develop vascular conduits biomimetically,it is essential to evaluate the permeability capability of theengineered vascular conduits. Thus, a perfusion system, whichconsisted of a media reservoir, a peristaltic pump (ISMATEC,IDEX Corporation, Glattbrugg Switzerland), and a perfusionchamber (with a cover to prevent evaporation) (see figure 3(a)),was developed. A peristaltic pump was selected to providepulsatile media flow, and cell media was perfused from

3

Nanotechnology 25 (2014) 145101 F Dolati et al

Figure 3. (a) Perfusion system, (b) with media flow through the vascular conduit.

the media reservoir, through the pump and the vascularconduit, and pumped back to the media reservoir. Gauge25 (0.25 mm I.D., 0.52 mm O.D.) needles were insertedinto the fabricated vascular conduits. Surgery clips wereused to fix the vascular conduits during perfusion withoutleakage. Several combinations of fabrication parameters (i.e.,composite dispensing pressure and CaCI2 dispensing rate)were tested to obtain the ideal core diameter to obtain abest match for a gauge 25 needle. The criteria for thefabrication parameter selection was that the lumen diameterof the dispensed vascular conduit should be exactly sameas the size of a gauge 25 needle’s outer diameter, such thatthe needle could be inserted into the vascular conduit tightlyand no leakage should be allowed at the connection of theconduit and the needle. The original length of the perfusablevascular conduits was fixed at 8 cm. Perfusion experimentswere conducted with a 20 ml min−1 perfusion flow rate for 1 h.Elongation and diffusion rate measurements were conductedimmediately after 1 h of perfusion. The diffusion rate andelongation were measured for 4% alginate vascular conduits,4% alginate-0.5% MWCNT-reinforced vascular conduits, and4% alginate-1% MWCNT vascular conduits.

2.6. Cell viability

Immediately after printing, samples were washed with HBSSsupplemented with 100 U ml−1 penicillin, 100 µg ml−1

streptomycin, and 2.5 µg ml−1 fungizone for sterilizationbefore incubation. After washing, vascular conduits werecultured at 37 ◦C in 5% CO2. Cell viability assays wereperformed immediately after printing, as well as after threedays in vitro culture to evaluate cell survival. Plain vascularconduits with alginate served as a control group. Vascularconduits 5 cm in length were printed for each sample. For cellviability analysis, samples underwent fluorescent microscopicexamination. Vascular conduits were stained with calcium

acetoxymethyl ester (calcein AM) and ethidium homodimer-2(Invitrogen) at a concentration of 1.0 mM each. Calcein AMstains live cells for green, while ethidium homodimer-2 stainsdead cells red. After a 30-min incubation period, vascularconduits were imaged under a Leica fluorescent microscope(Leica Microsystems Inc., Buffalo Grove, IL, USA). Imageswere collected from three different locations randomly chosenfrom each sample. ImageJ (National Institutes of Health,Bethesda, MD, USA) was used for automated counting of red-and green-stained HCASMCs in each image, and percentagesof viable cells were calculated. Cellular droplets prior toprinting were used for initial control for all groups.

2.7. Tissue histology

Histochemistry was applied to sections of vasculature. After6 weeks of in vitro culturing in smooth muscle celldifferentiation media, frozen vascular conduits were fixedand sectioned at 5 µm for histological examination formarkers specific to smooth muscle cells. Verhoeff–Van Giesenmethod was used to visualize elastic fiber formation andcollagen deposition. Elastic fibers stain a blue black toblack color, while collagen stains a red color. Samples wereexamined under an Olympus BX-61 Brightfield-Fluorescentmicroscope (Olympus America, Melville, NY, USA) atdifferent magnifications.

2.8. Statistical analysis

The statistical significance of experimental data for themechanical and perfusion tests was determined by one-wayanalysis of variance (ANOVA) with a significance level ofp < 0.05 in Minitab 16. The paired-wise test was combinedwith the Tukey post-hoc test at a significance level of p< 0.05and used for the cell viability study in the Statistical Packagefor the Social Sciences (SPSS). Three samples were used foreach experimental group (n = 3).

4

Nanotechnology 25 (2014) 145101 F Dolati et al

Figure 4. Sample vascular conduits: (a) printed vascular conduits, (b) a zoomed image under light microscopy showing the wall and lumen,and (c) cell media successfully perfused through a meter-long printed conduit.

Figure 5. SEM images of vascular conduits: (a) structural integrity showing tubular shape, (b) 4% alginate reinforced with 1% MWCNTswith highlighted fibrous MWCNT at the fracture site of vascular conduits (see the dashed rectangle), (c) 4% plane alginate vasculature at thefracture site without appearance of fibrous shape in the spongy architecture, (d) inner wall of vascular conduit reinforced with 1% MWCNT,and (e) outer wall of vascular conduit reinforced with 1% MWCNT.

3. Results

3.1. Fabrication

In this study, vascular conduits were fabricated usingMWCNT-reinforced alginate and plain alginate (seefigure 4(a)). During the fabrication, the alginate dispensingpressure was set at 17.2 kPa, and the calcium chloride dispens-ing rate was set at 16 ml min−1. The printed vascular structureswere obtained in the form of conduits with well-definedtubular wall and lumen (see figure 4(b)). The average innerand outer diameters of the fabricated vascular conduits were461 ± 68 µm and 930 ± 43 µm, respectively. Vascularconduits were fabricated at more than a meter longwith successful perfusion capability, demonstrating theirfunctionality. Conduits could be manufactured at any desiredlength and pattern through printing without producing anyblockage or rupture that resulted in a leak or burst. Successfulperfusion through a meter-long vascular conduit was achieved

as shown in figure 4(c). In addition, they were highlypermeable, which enabled diffusion of media in a radialdirection similar to natural blood vessels. Moreover, theycould be printed layer by layer, and a 3D printed structurewith eight layers or more was achieved (data not shown). Avascular conduit with controllable dimensions in the micro-and sub-millimeter scales can be printed by altering theprocess parameters. For example, figure 4(b) shows the walland lumen under light microscopy. The wall dimensionscould be controlled by controlling the process parameters,instantaneously resulting in varying diameters across thelength.

Scanning electron microscope (SEM) figures of vascularconduits with 1% MWCNT and plain alginate are shown infigure 5. Figure 5(a) shows the lumen of vascular conduits,where an acceptable cylindricity was obtained, demonstratingthe structural integrity of the conduits. Figure 5(b) shows thestructure of a MWCNT-reinforced composite vascular conduitat the fracture, where fibrous MWCNTs are highlighted within

5

Nanotechnology 25 (2014) 145101 F Dolati et al

Figure 6. Comparison of (a) tensile strength of 3% alginate with 0.5% and 1% MWCNT-reinforced 3% alginate, (b) tensile strength of 4%plain alginate with 0.5% and 1% MWCNT-reinforced 4% alginate, (c) elastic modulus of 3% plain alginate with 0.5% and 1%MWCNT-reinforced 3% alginate, (d) elastic modulus of 4% plain alginate with 0.5% and 1% MWCNT-reinforced 4% alginate, (e) ultimatestrain of 3% plain alginate with 0.5% and 1% MWCNT-reinforced 3% alginate, (f) ultimate strain of 4% plain alginate with 0.5% and 1%MWCNT-reinforced 4% (* represents statistically significant difference p< 0.05).

the dashed rectangle. A similar structure did not appear in a4% plain alginate conduit, as presented in figure 5(c). In bothcases, sponge-like shapes were obtained at the fracture. Theinner and outer walls of the vascular conduit are illustrated infigures 5(d) and (e), respectively. Deformation was observedto some extend during the dehydration process.

3.2. Mechanical characterization of vascular conduits

The results of the tensile stress test showing tensile strength,elastic modulus, and ultimate strain of vascular conduits forall types of samples are presented in figure 6. From theplots, one can compare the mechanical properties of sampleswith and without MWCNT and different concentrations ofbiomaterial solution. Figure 6(a) shows that the tensile strength

increased significantly with each addition of MWCNT from110 ± 5.8 kPa in 3% alginate only gels to 161 ± 24 kPain gels with 0.5% MWCNT and 238 ± 14 kPa in gelswith 1% MWCNT. The tensile strength of 4% alginate alsoincreased with the inclusion of MWCNT from 382 ± 19 kPain alginate alone gels to 420 ± 22 kPa in gels 0.5% MWCNTto 422 ± 22 kPa in gels with 1% MWCNT (see figure 6(b)).As presented in figure 6(c), elastic modulus for 3% alginategels also increased significantly with the addition of MWCNTfrom 105 ± 7.5 kPa to 174 ± 13 kPa and 305 ± 17.5 kPa, foralginate alone, alginate with 0.5% MWCNT, and alginate with1.0% MWCNT, respectively. Using the same reinforcementpercentages, elastic modulus also increased from 341 ± 23 to593 ± 28 and 667 ± 35 kPa for 4% alginate (figure 6(d)).For groups in figures 6(c) and (d), there was a significant

6

Nanotechnology 25 (2014) 145101 F Dolati et al

difference in the elastic modulus between plain alginate andthe reinforced ones; however, no significant difference wasobserved in the elastic modulus between the reinforced oneswith different MWCNT concentrations for 4%

In contrast to tensile strength and elastic modulus, ultimatestrain decreased (though not significantly) for both 3% and4% alginate gels as MWCNT concentration was increased(see figures 6(e) and (f)). In 3% alginate gels, ultimate straindecreased from 0.82± 0.18 to 0.79± 0.03 and 0.68± 0.18 forgel alone, gel with 0.5% MWCNT and gel with 1% MWCNT,respectively. In 4% alginate gels, ultimate strain decreasedfrom 0.69 ± 0.09 to 0.68 ± 0.13 and 0.64 ± 0.22 for gelalone, gel with 0.5% MWCNT and gel with 1% MWCNT,respectively.

Samples with higher concentration of alginate andMWCNT had higher tensile strength and elastic modulusand lower ultimate strain compared to samples with lowerconcentrations of these components. Differences in thesemechanical properties can be explained in part by theproperties of the alginate itself. The higher the concentrationand viscosity of the alginate solution, the stronger are theintermolecular interactions between polymer chains and themore entanglements form [18]. Therefore, one would expectthat the 4% alginate gel’s increased tensile strength and elasticmodulus and decreased ultimate strain can be attributed togreater cohesion between the polymer chains due to morephysical entanglements. These could be due to distribution,dispersion and alignment of the MWCNT in alginate, whichcould affect the physical and mechanical properties and theorientation of nanomaterial along the vascular conduits. Theaddition of reinforcing MWCNT also affected the mechanicalproperties, although much more so in 3% alginate gels thanin 4% alginate gels. The distribution and dispersion of theMWCNT in alginate appears to affect the microstructure ofthe gel, which in turn can control the physical propertiesof the vascular conduits. The incorporation of MWCNT ina lower alginate solution likely improved the mechanicalproperties of the conduits more substantially than in the higherconcentration alginate solutions due to the presence of fewerpolymer chain entanglements. If the ratio of MWCNT toalginate concentration had been maintained for both 3% and4% alginate gels it is possible that the inclusion of MWCNTwould have produced a stronger affect in the 4% gels.

3.3. Perfusion study

The results of the perfusion study show that all vascularconduits elongated from 8 to 8.4 ± 0.05 cm after one hour ofmedia perfusion. This was primarily caused by the weight ofvascular conduits. Due to gravity, the weight of the perfusedmedia as well as the weight of the vascular conduit itselfstretched the conduit toward the bottom of the perfusionchamber. Elongation terminated as long as the vascular conduitreached the bottom of perfusion chamber. The diffusion ratesfor 3 h of perfusion are shown in figure 7. The diffusion ratefor 4% plain alginate, 4% plain alginate-0.5% MWCNT, and4% alginate-1% MWCNT were 7.2± 0.48 µl min−1, 8.4±0.68 µl min−1, and 7± 0.71 µl min−1, respectively. Although

Figure 7. Effect of MWCNT concentration on diffusion rate ofvascular conduits.

the 4% alginate to 1% MWCNT group had the highestdiffusion rate, no statistical difference was observed amongthose three groups, which might be due to the negligible effectof MWCNT reinforcement on the permeability of the alginatepolymer structure. In other words, the pore size of alginatepolymer, which is approximately 30–450 nm for 4% [20], ismuch greater than the diameter of MWCNT (approximately30 nm, according to the manufacturer’s specifications), andhence the diffusion of media was not affected by thereinforcement of MWCNTs at low concentration. However,one can speculate that the diffusion rate could decrease if thereinforcement concentration increased significantly.

3.4. Cell viability

As demonstrated in figure 8(a), the printing process causedsome cell damage; quantitative red-fluorescent-labeled deadcells were observed from the fluorescence image along theconduit channel wall. Compared with the initial control, cellviability was decreased from 75.6 ± 0.04% to 53.3 ± 0.01%for plain alginate vascular conduits, and from 72.2 ± 0.02%to 53.9± 0.01% for MWCNT-reinforced upon printing (seefigure 8(a)). Nevertheless, cells were able to recover fromdamage during in vitro culture after the printing process, ascell viability in day 3 reached 71.3± 0.02% and 68.8± 0.01%for plain alginate and MWCNT-reinforced vascular conduits,respectively. These percentages were close to the initial controlgroup. This demonstrated that cells were not completely dead(see figure 8(b)); instead they might be damaged to some extentand able to recover and grow in culture.

No significant difference was observed between thecontrol group and the MWCNT-reinforced ones. Thus, thelow concentration of MWCNT reinforcement in this researchdid not affect the viability of HCASMC in short-term invitro culture; however, biodegradable nanofibers are preferredfor long-term culture or in vivo studies while the livingenvironment cannot absorb CNTs by natural means.

3.5. Tissue histology

The tissue histology study showed different characteristicsbetween CNT-free (positive control group) and CNT-reinforced conduits. In CNT-reinforced conduits, damagedcells within the conduit wall can be easily identified by broken

7

Nanotechnology 25 (2014) 145101 F Dolati et al

Figure 8. (a) Cell viability over time in plain alginate vascular conduits and MWCNT-reinforced ones, (b) fluorescent microscopic imagefrom three-day cultured MWCNT-reinforced vascular conduit showed most of the cells are viable (green), and a minimal amount of deadcells (red) were also observed.

Figure 9. CNT-reinforced conduits: (a) transverse section showing damaged cells (shown with red arrows) within conduit wall, (b) radiallongitudinal section showing damaged cells (shown with red arrows) within conduit wall, (c) positive control group without CNTreinforcement, where cells were healthy (shown with blue arrows) and produced substantial matrix in the outer section of the wall in 6weeks (highlighted in dashes).

cell nuclei, as indicated by the red arrowheads in figures 9(a)and (b). In CNT-free conduits, the majority of the cells wereintact, with rounded nuclei uniformly distributed within theconduit wall (see figure 9(c), blue arrowheads). Some livecells were still observed, as shown in figure 9(b). In addition,a number of cells migrated to the outermost section of theconduit wall and proliferated, forming a well-aligned cellularlayer with substantial ECM formation, as shown in the dashedbox in figure 9(c) In CNT-reinforced conduits, almost no ECMformation was observed, and elastin and collagen were stainednegative. On the other hand, plain conduits stained positive forcollagen as shown in figure 9(c) and stained positive for elastinin some other histological sections (data not shown here).

4. Discussions

Our experiments with printing CNT-reinforced vascularconduits were successfully demonstrated using sodiumalginate because of its appealing gelation properties, whichwere highly suitable for the coaxial printing process. Usingprinting, complex shapes in 3D can be fabricated throughlayer-by-layer stacking of vascular conduits. In general,vascular conduits with an external diameter ranging from500µm to 2 mm and a lumen diameter ranging from 150µm to1 mm can be printed using the demonstrated system. One of the

major advantages of the introduced bioprinting process is thatit enables direct printing of the vascular conduits without theneed for any temporary support material that is to be removedthereafter, such as thermo-sensitive hydrogels, i.e., agaroseand collagen.

In order to fabricate functional vascular conduits, oneshould ensure that the mechanical, structural, biological andperfusion capabilities of vascular conduits are acceptable.In our experiments, we reached an 8.2 ± 0.3 µl min−1

diffusion rate within three hours of perfusion, which enabledhigh viability of encapsulated cells due to the super-diffusiveproperties of alginate. As alginate concentration increases,mesh network size diminishes, and a compact structureforms, resulting in slower diffusion of the media [19]. Inthe meantime, mechanical and structural characteristics ofthe vascular conduits decrease as the material concentrationincreases. In general, viability and cell migration capabilitydecrease as the concentration of alginate increases. Thus, thereis a tradeoff between functionality and the cell viability andbiological performance of the conduits. Our initial trials withlow concentrations of alginate, such as 2.5%, showed that thebioprinting system did not produce structurally well-defined,mechanically acceptable perfusable conduits. Most of thetime, the vascular conduits collapsed when low-concentrationranges were used. Therefore, nanofiber reinforcement (such as

8

Nanotechnology 25 (2014) 145101 F Dolati et al

natural polymers and proteins) might be ideal for enhancing themechanical, structural and perfusion characteristics of vascularconduits with biologically reasonable properties.

Carbon-nanotube reinforcement increased the ultimatetensile stress by 11% and the modulus of elasticity by 94%. Itdecreased the ultimate strain by 18%. We achieved 70% cellviability three days post-printing that increased thereafter inshort-term culture. In general, cells proliferate and reach over90% viability in a week [16]. Particularly, the reinforcementenhanced the mechanical properties of low-concentrationalginate. The improvement of ultimate tensile stress and elasticmodulus demonstrated that the stress was transferred betweenthe hydrogel matrix and the carbon nanotubes reinforced inthe vascular conduit’s composites. In addition, the externalstress was effectively transferred from the alginate matrix tothe reinforced MWCNTs through the better bonded interfacesin alginate-reinforced MWCNT composites [22, 23]. Theinterfacial adhesion between CNTs and the matrix is oneof the crucial factors affecting the mechanical properties ofthe composites. In fact, MWCNTs have a greater affinity tothe polymer matrix, resulting in a significant improvement inmechanical properties [22, 23].

As demonstrated in the histology images, cells withinCNT-reinforced conduits were damaged and underwent celldeath. They were not able to repopulate or to produce ECMin the long term. Nevertheless, in plain vascular conduits (thepositive control group), cells were largely intact and were ableto migrate and proliferate towards the outer section of thechannel wall. We can speculate that it was probably guidedby the media gradient during culture, since the outer sectioninteracted with cell culture media extensively, especiallysmooth muscle cell growth factors. Also, repopulated cellswere able to produce intercellular ECM and form a sheetcovering the outer section of the channel wall. Some matrixformation was observed in the inner surface of the wall as alining, and that was separated from the wall during the cuttingprocess of the histology study. However, matrix formation inthe inner surface could be improved by keeping the pulsatilemedia flow through the lumen. In this way, HCASMCs couldalign their proliferation with respect to the pulsatile flow asin the natural blood vessels. Although a cell viability testshowed acceptable cell viability in CNT-reinforced conduitsin short-term culture, only a limited number of cells were ableto survive and carry out their functions properly in long-termculture.

Despite their benefit in improving the mechanicalproperties of conduits, CNTs have side effects such astoxicity in long-term culture or in vivo testing [21]. Thus,in order to exploit the mechanical benefits of printed fibercomposites, other non-toxic materials should be investigated.Biodegradable materials, such as polylactide, polyglycolide,and its derivatives, can be added to alginate in the form ofnanofibers to increase the mechanical strength of conduitswhile also increasing biocompatibility for cell growth andfunction [24]. Alternatively, short extruded collagen fibersthat exist on the mm-micron scale could be added fora similar effect [25]. Collagen type-1 and elastin are theessential components of natural blood vessels that give

ultimate mechanical properties (strength and elasticity).Matrix deposition including these proteins was observed inthe positive control group, and reinforcing them further duringthe printing process could improve the mechanobiologicalcharacteristics of the vascular conduits considerably. In thisregard, the fabricated vascular conduits can be tested in vivosafely for long-term monitoring of the performance of theconduits.

5. Conclusion

In this paper, we demonstrated a new practical techniquefor vasculature fabrication, where micro vascular conduitswere directly printed using a coaxial nozzle configuration.Vascular conduits in this work were reinforced with CNTsto improve the mechanical properties that are essential forfurther applications such as generating blood vessels orproducing supplement vasculatures for engineered tissues.Mechanical and perfusion testing was conducted for CNT-reinforced vascular conduits, and biological characterizationwas performed in the short and long term to understandcellular viability and extracellular matrix formation. Althoughshort-term results were acceptable, CNT-induced toxicityreduced the biological performance of the conduits.

In sum, this work provides a foundation for direct printingof mechanically reinforced vascular conduits, where CNTscould be replaced with natural polymer nanofibers for furtherapplications such as scale-up tissue fabrication or blood vesselgeneration. Effectiveness of natural polymer fibers has alreadybeen demonstrated in the literature [25]. For future studies,we will thus reinforce electro-spun collagen type-1 nanofibersalong with lower percentages of polymer solution and generatean endothelial lining inside the lumen to biomimeticallyfabricate a vasculature network. The network will be integratedwith tissue printing process using the Multi-Arm Bioprinterpublished in our recent work [26]. It will promote oxygenationand media transport, be a part of the hybrid tissue in shortperiod of time and degrade eventually.

Acknowledgments

This study was supported by the National Institutes of Health(NIH) and the Institute for Clinical and Translational Science(ICTS) grant number ULIRR024979, and by a US Departmentof Education Graduate Assistance in Areas of National NeedFellowship (GAANN P200A120071). The authors would liketo thank Professor David Cwiertny from The University ofIowa’s Civil and Environmental Engineering Department forproviding MWCNTs.

References

[1] L’Heureux N, Dusserre N, Konig G, Victor B, Keire P,Wight T N and McAllister T N 2006 Humantissue-engineered blood vessels for adult arterialrevascularization Nature Med. 12 361–5

[2] Yow K-H, Ingram J, Korossis S A, Ingham E andHomer-Vanniasinkam S 2006 Tissue engineering ofvascular conduits Br. J. Surg. 93 652–61

9

Nanotechnology 25 (2014) 145101 F Dolati et al

[3] Yildirim E, Yin X, Guceri S and Sun W A preliminary studyon using multi-nozzle polymer deposition system tofabricate composite alginate/carbon nanotube tissuescaffolds. Proc. 17th Solid Freeform Fabrication Symp.pp 84–95

[4] Kai D, Prabhakaran M P, Stahl B, Eblenkamp M,Wintermantel E and Ramakrishna S 2012 Mechanicalproperties and in vitro behavior of nanofiber–hydrogelcomposites for tissue engineering applicationsNanotechnology 23 095705

[5] Bu H, Kjøniksen A L, Knudsen K D and Nystrom B 2004Rheological and structural properties of aqueous alginateduring gelation via the Ugi multicomponent condensationreaction Biomacromolecules 5 1470–9

[6] Coleman J N, Khan U, Blau W J and Gun’ko Y K 2006 Smallbut strong: a review of the mechanical properties of carbonnanotube–polymer composites Carbon 44 1624–52

[7] Augst A D, Kong H J and Mooney D J 2006 Alginatehydrogels as biomaterials Macromol. Biosci. 6 623–33

[8] Spitalsky Z, Tasis D, Papagelis K and Galiotis C 2010 Carbonnanotube–polymer composites: chemistry, processing,mechanical and electrical properties Prog. Polym. Sci.35 357–401

[9] Yakobson B I and Avouris P 2001 Mechanical properties ofcarbon nanotubes Carbon Nanotubes (Berlin: Springer)pp 287–327

[10] Salvetat J P, Bonard J M, Thomson N H, Kulik A J, Forro L,Benoit W and Zuppiroli L 1999 Mechanical properties ofcarbon nanotubes Appl. Phys. A 69 255–60

[11] Andrews R and Weisenberger M C 2004 Carbon nanotubepolymer composites Curr. Opin. Solid State Mater. Sci.8 31–7

[12] Yildirim E D, Yin X, Nair K and Sun W 2008 Fabrication,characterization, and biocompatibility of single-walledcarbon nanotube-reinforced alginate composite scaffoldsmanufactured using freeform fabrication techniqueJ. Biomed. Mater. Res. B 87 406–14

[13] Mattson M P, Haddon R C and Rao A M 2000 Molecularfunctionalization of carbon nanotubes and use as substratesfor neuronal growth J. Mol. Neurosci. 14 175–82

[14] Hu H, Ni Y, Montana V, Haddon R C and Parpura V 2004Chemically functionalized carbon nanotubes as substratesfor neuronal growth Nano Lett. 4 507–11

[15] Lobo A O, Antunes E F, Machado A H A, Pacheco-Soares C,Trava-Airoldi V J and Corat E J 2008 Cell viability andadhesion on as grown multi-wall carbon nanotube filmsMater. Sci. Eng. C 28 264–9

[16] Zhang Y, Yu Y, Chen H and Ozbolat I T 2013Characterization of printable cellular micro-fluidic channelsfor tissue engineering Biofabrication 5 025004

[17] Smith B, Wepasnick K, Schrote K E, Cho H H, Ball W P andFairbrother D H 2009 Influence of surface oxides on thecolloidal stability of multi-walled carbon nanotubes: astructure–property relationship Langmuir 25 9767–76

[18] Khalil S E D 2005 Deposition and structural formation of 3Dalginate tissue scaffolds Doctoral Dissertation DrexelUniversity

[19] Ozbolat I T and Koc B 2012 3D hybrid wound devices forspatiotemporally controlled release kinetics Comput.Methods Programs Biomed. 108 922–31

[20] Chan A W and Neufeld R J 2009 Modeling the controllablepH-responsive swelling and pore size of networked alginatebased biomaterials Biomaterials 30 6119–29

[21] Magrez A, Kasas S, Salicio V, Pasquier N, Seo J W, Celio Mand Forro L 2006 Cellular toxicity of carbon-basednanomaterials Nano Lett. 6 1121–5

[22] Thostenson E T and Chou T W 2002 Aligned multi-walledcarbon nanotube-reinforced composites: processing andmechanical characterization J. Phys. D: Appl. Phys.35 L77

[23] Boccaccini A R and Gerhardt L C 2010 Carbon nanotubecomposite scaffolds and coatings for tissue engineeringapplications Key Eng. Mater. 441 31–52

[24] Rim N G, Shin C S and Shin H 2013 Current approaches toelectrospun nanofibers for tissue engineering Biomed.Mater. 8 014102

[25] Gentleman E, Nauman E A, Dee K C and Livesay G A 2004Short collagen fibers provide control of contraction andpermeability in fibroblast-seeded collagen gels Tissue Eng.10 421–7

[26] Ozbolat I T, Chen H and Ozbolat I T 2014 Development of‘Multi-arm Bioprinter’ for hybrid biofabrication of tissueengineering constructs Robot. Comput.-Integr. Manuf.30 295–304

10