Yin YuBiomanufacturing Laboratory,

Center for Computer-Aided Design,

The University of Iowa,

Iowa City, IA 52242;

Biomedical Engineering Department,

The University of Iowa,

Iowa City, IA 52242

Yahui ZhangBiomanufacturing Laboratory,

Center for Computer-Aided Design,

The University of Iowa,

Iowa City, IA 52242;

Mechanical and Industrial

Engineering Department,

The University of Iowa,

Iowa City, IA 52242

Ibrahim T. Ozbolat1Biomanufacturing Laboratory,

Center for Computer-Aided Design,

The University of Iowa,

Iowa City, IA 52242;

Mechanical and Industrial

Engineering Department,

The University of Iowa,

Iowa City, IA 52242

e-mail: ibrahim-ozbolat@uiowa.edu

A Hybrid Bioprinting Approachfor Scale-Up Tissue FabricationTissue engineering has been focused on the fabrication of vascularized 3D tissue for dec-ades. Most recently, bioprinting, especially tissue and organ printing, has shown greatpotential to enable automated robotic-based fabrication of 3D vascularized tissues andorgans that are readily available for in vitro studies or in vivo transplantation. Studieshave demonstrated the feasibility of the tissue printing process through bioprinting ofscaffold-free cellular constructs that are able to undergo self-assembly for tissue forma-tion; however, they are still limited in size and thickness due to the lack of a vascular net-work. In this paper, we present a framework concept for bioprinting 3D large-scaletissues with a perfusable vascular system in vitro to preserve cell viability and tissue mat-uration. With the help of a customized Multi-Arm Bioprinter (MABP), we lay out a hybridbioprinting system to fabricate scale-up tissues and organ models and demonstrated envi-sion its promising application for in vitro tissue engineering and its potential for thera-peutic purposes with our proof of concept study. [DOI: 10.1115/1.4028511]

Keywords: bioprinting, biomimetic design, organ fabrication

1 Introduction

Tissue engineering has shifted the paradigm of traditional medi-cine from therapeutically treating disease to replacing diseasedtissues and organ parts by regenerative medicine approaches. Thisremarkable leap in medical science has drawn great attention fordecades and has brought tremendous opportunities to not only dis-ease treatment, but also drug testing, disease modeling, and otherphysiological and pathological research [1].

Traditional tissue engineering has had great success in artificialtissue or organ fabrication by advancement of stem cell technol-ogy, materials science, and biomimetic system design. Ever sinceRobert Langer grew an ear on a mouse back by seeding cells ontoa biopolymer scaffold, this technology has enabled bioengineeredtissue such as cartilage, bone, skin, etc., to be grown in laborato-ries around the world [2]. These successes were largely limited totissues with avascular nature, which have lower oxygen consum-ing rates. However, a long-term obstacle for tissue engineeringhas been generation of vascularized tissue with blood perfusioncapability, which is critical for organogenesis, especially in vivo[3]. Traditional scaffolding methods often fail to provide an effi-cient media (oxygen, growth factor, water, etc.) transportationsystem for generation of thick tissues or organs. Several research-ers have attempted to control scaffold architecture to mediatemedia transport capabilities [4], but the majority have failed inorgan or tissue formation due to the lack of an integrated vascularsystem [5]. One possible solution to improve perfusion throughthick scaffolds is embedded microfluidic networks; however, mostof the microfluidic fabrication methods are multistep processesthat do not allow direct fabrication of a vessel-like structure fortissue integration [6].

Organ printing can be defined as layer-by-layer additive roboticfabrication of three-dimensional (3D) functional living tissues ororgans. It has been envisioned to push tissue engineering into anew era with the capability to do robotic large-scale living organfabrication [7]. It was inspired by the natural embryonic develop-mental process and used cell aggregates in a microspherical shapeas building blocks with a rapid-prototyping platform. These“mini-tissues” have certain measurable and controllable propertiesand can be assembled into macrotissues through the tissue fusionprocess or directed tissue self-assembly. With this method, scien-tists used multicellular spheroids made from smooth muscle cellsand fibroblasts to successfully generate vascular constructs [8]. Itwas predicted that this robotic-assisted biomimetic process wouldbe dramatically beneficial for tissue engineering in the near future[9]. In order to integrate a vasculature network within the tissueprinting process, the vasculature network should be practicallyprinted without requiring any support material; it also needs to beprinted in complex geometries on a larger scale to feed the oxy-genation need of larger tissues.

Cell aggregates possess accelerated fusion and folding capabil-ities upon contacting each other that is driven by a natural biologi-cal process [10]. High cell density within aggregates can be easilyobtained by the fusion process, in which traditional scaffoldingand cell-laden hydrogel usually depend on cell proliferation after-ward. Moreover, cell viability is enhanced due to the large cell-seeding density and reduced mechanical stress experienced com-pared with direct cell manipulation. Microscale organoids, inwhich heterocellular aggregates possess organ-like functions,have been generated in vitro for pancreatic, liver, and cartilage tis-sues [11–13]. Although these techniques seem to be impressivefor tissue engineering, especially as bio-ink for organ printing,their labor-intensive fabrication in limited scale makes theirapplicability for large-scale tissue/organ fabrication difficult.Moreover, printing them sequentially by ensuring contact betweeneach adjacent spheroid is another hurdle, given the extremely crit-ical handling and sterilization conditions. Without ensuring

1Corresponding author.Contributed by the Manufacturing Engineering Division of ASME for publication

in the JOURNAL OF MANUFACTURING SCIENCE AND ENGINEERING. Manuscript receivedApril 15, 2014; final manuscript received August 23, 2014; published online October24, 2014. Assoc. Editor: Darrell Wallace.

Journal of Manufacturing Science and Engineering DECEMBER 2014, Vol. 136 / 061013-1Copyright VC 2014 by ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

seamless contact in microscale resolution, spheroids can hardlyfuse with each other, easily leaving gaps and openings in the tis-sue as discussed in our recent review paper [3]. Besides, hydrogelsare required as a transferring medium to deposit spheroids; thismay not achieve complete scaffold-free fabrication and may faceside effects of biomaterial degradation. In addition, technologiesshould be developed to prevent spheroid fusion before printing;otherwise, nozzle clogging is inevitable. New technologies needto be developed to enable scalable, standardized spheroid fabrica-tion to allow large-scale automated tissue fabrication as well asminimize cell damage both transcriptionally and functionally. Interms of tissue maturation, the cellular fusion process needs to bewell characterized to understand cell interactions under variousmechanical and biological cues. Cellular reaction upon fusion alsoneeds to be addressed to elaborate on how extracellular matrix(ECM) is secreted by cells, and how to control ECM productionas well as remodeling to ensure the mechanical integrity andstrength of the regenerated tissues or organs. The most challeng-ing issue currently faced in organ printing technology is the gener-ation of a “built-in” vascular system within 3D thick tissues ororgans with perfusion capability [10]. An integrated bioreactor isalso in demand upon establishing a perfusable vascular networkfor continuously supplying oxygen and nutrients sufficient for cellgrowth and tissue maturation.

Our goal in this paper is to present a novel conceptual blueprintfor a large-scale vascularized tissue printing process via a hybridfabrication approach, which allows a vascular system and tissue tobe printed in tandem. We introduce a new biomimetic designapproach to develop perfusable tissues using a biologically drivenfusion, folding, and maturation process. Large-scale hybrid tissuescan be fabricated by the integration of a vasculature network inmultiscale with tissue-specific cell aggregate strands, which areconsidered tissue strands, followed by rapid fusion, remodeling,and maturation of the perfusable tissues. Geometric modeling forhybrid tissue construct fabrication is performed, and future direc-tions for tissue printing are also discussed.

2 Hybrid Fabrication of 3D Perfusable Tissues

In order to fabricate viable, functional tissues and organ coun-terparts in 3D, the integration of a vascular network is the fore-most component that needs to be taken care of. Withoutvascularization, engineered 3D thick tissues or organs cannot getenough nutrients, gas exchange, and waste removal, all of whichare needed for maturation during perfusion [14]. This results inlow cell viability and malfunction of artificial tissues. Systemsmust be developed to transport nutrients, growth factors, and oxy-gen to cells while extracting metabolic waste products such as lac-tic acid, carbon dioxide, and hydrogen ions so the cells can growand fuse together, forming the tissues. Cells in a large 3D tissueconstruct cannot maintain their metabolic functions without vas-cularization, which is traditionally provided by blood vessels [15].A vascular network is indeed a very complex hierarchy, wherelarge-scale arteries and veins are connected by a complex micro-scale capillary system, where media exchange takes placesbetween the blood and the tissue. Every single cell is supplied bycapillaries for its normal metabolism, in order to stay viable andbe able to proliferate for maintaining tissue integrity and function.Bioprinting technology, on the other hand, currently does notallow multiscale tissue fabrication, where bifurcated vessels arerequired to be manufactured with capillaries to mimic natural vas-cular anatomy. Although several researchers have investigateddeveloping vascular trees using computer models [16], only a fewattempts have been made toward fabricating bifurcated orbranched channels [17]. Successful maturation toward functionalmechanically integrated bifurcated blood vessel network is still achallenge. Thus, we introduce a novel concept to fabricate a vas-cular network within the hybrid tissue, which can be performed intwo steps: (1) bioprinting of a macroscale vasculature network intandem with vascularized tissue strands and (2) biologically

driven tissue fusion and assembly with envisioned capillarysprouting from macrovasculature to the tissue. The formerpresents more of a bioprinting approach than the latter, which ismainly biologically mediated.

2.1 Design for Hybrid Bioprinting Approach. In this paper,we introduce a new hybrid bioprinting approach to fabricate per-fusable tissues in 3D. It combines vasculature printing and tissue-specific cell aggregate strand printing in a single platform. Figure1 illustrates the concept for a scale-up tissue printing, where twounits of the bioprinter can run in tandem, printing two distinctmaterials. Strands are made of purely cells and their ECM, andthey will be used as building blocks to construct the scale-uporgan due to their quick fusion, folding, and maturation capabil-ities. In this concept, a printable vascular network in a continuoussingle luminal form in macroscale would be more practical to beintegrated with additive manufacturing-based bioprinting. In thisway, semipermeable vasculature can be printed continuously witha predetermined 0 deg/90 deg lay-down pattern similar to the tra-ditional additive manufacturing techniques. In order to eliminateblockage of vasculatures, arc fitting is used at U-turns betweentwo vasculatures during zigzag printing. Upon fabrication, oxy-genized media can be perfused through the macrovascular net-work and diffused out to the tissue strands to keep them viable. Inorder to print the scale-up hybrid model, the MABP developed inour recent work [18] can be used. The MABP (see Fig. 2(a)) facil-itates synchronization and co-ordination between multiple armsthrough a sensor-based system that enables crosstalk betweenmultiple arms so that two arms can communicate in an intelligentway, ensuring collision-free motion. One arm is equipped with acoaxial nozzle assembly for printing the vascular network, and theother arm is equipped with a custom unit to print tissue-specificcell aggregate strands (see Figs. 2(b)–2(d)). The nozzle configura-tion enables close motion of printer units. The introduced MABPhas the potential to reduce the fabrication time, which is crucialfor scale-up technologies, and would enable continuous deposi-tion, which yields more consistent results since material deposi-tion during starts and stops is not as uniform as it is during the restof the process. Continuous deposition also alleviates nozzle clog-ging since it generally occurs when the material is in static condi-tions. Most importantly, nozzle motions have more flexibility inz-axis, which is critical while printing vasculature with variationaldimensions that necessitate instantaneous changes between thenozzle tip and the printing stage to eliminate blockage of the

Fig. 1 Concept of 3D tissue-printing technology: vascular net-work is printed in tandem with tissue-specific aggregatestrands to provide both mechanical support and mediaperfusion

061013-2 / Vol. 136, DECEMBER 2014 Transactions of the ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

lumen, which is quite difficult to achieve with a traditional single-arm bioprinter.

2.1.1 Direct Bioprinting of Macrovasculature Network. Thevasculature network in the proposed hybrid bioprinting conceptcan be directly printed using a pressure-assisted coaxial nozzleunit demonstrated in Fig. 2(c). Biomaterials such as sodium algi-nate, collagen, and chitosan or their composites encapsulating dif-ferent cell types existing in natural blood vessels such as smoothmuscle, fibroblasts, and endothelial cells can be printed withoutany support materials that make them practical for further applica-tions. With continuous flow chemical crosslinker through the coresection of the coaxial nozzle, vasculatures can be printed directly.Fabrication of vasculature with controllable dimensions in themicro- and sub-millimeter-scale is quite feasible as presented inour recent work [19]. Figure 3 shows examples of fabricated vas-culatures using 4% (w/v) sodium alginate crosslinked with 4%

calcium chloride (CaCl2) solution. The sodium alginate dispens-ing pressure was set at 21 kPa, and the CaCl2 dispensing rate wasset at 16 ml/min. The printing speed of the robot arm was set at14 mm/s. Second passage of human umbilical vein smooth musclecells was encapsulated with a density of 10� 106 cells/ml, wherefurther culture was provided under pulsatile flow mimicking ahemodynamically equivalent environment. Geometrically well-defined vasculatures can be manufactured at any geometry, length,and orientation, without occlusion or rupture that might result inleakage or burst (see Fig. 3(b)). In addition, they are highly per-meable, which enables diffusion of media in the radial directionupon perfusion with a pumping system, which is similar to naturalblood vessels. In our recent studies, we reached an 8.2 6 0.3 lL/min filtration rate within 3 h of perfusion [20]. Although theyhave acceptable mechanical properties to sustain both tensile andcompressive forces with reasonable matrix deposition in a shortperiod of time (see Fig. 3(c)), further enhancement will be helpfulto improve their strength through exogenous collagen and elastinfibers generated by methods such as electrospinning. Vascula-tures, in general, are able to support printed tissue strands, whichusually do not have sufficient mechanical strength and integrityalone to build 3D tissue. Vasculatures are printed into a highlyorganized 3D framework for tissue strands to attach on, providingan initial microenvironment in which they can develop and ulti-mately fuse to each other. In addition, previous studies have dem-onstrated the capability of the bioprinting system to fabricatevasculature with controllable dimensions in the micro- and sub-millimeter-scale by altering process parameters [20], which givessubstantial flexibility to the process. In other words, different-sized vasculatures can be printed to fit specific functional needs.

2.1.2 Bioprinting and Biofabrication of Vascularized TissueStrands. This section demonstrates tissue strands that can be usedas building blocks for the scale-up tissue printing process. In gen-eral, tissue strands can be generated by monoculturing or cocultur-ing cells including endothelial cells for capillary formation usingmicrofluidics- or molding-based fabrication approaches. Aggrega-tion of cells takes 1–2 weeks, depending on the applied fabricationtechnique as well as the type and size of cells used in the study.Harvested cells need to be mixed and aggregated at a predeter-mined optimal ratio for in vitro angiogenesis. Tissue strands canbe fabricated in any length; however, the diameter is limited tothe applied aggreation technique as well as diffusion limit of oxy-gen (see Fig. 4(a)). The greater the diameter, the lower the cell

Fig. 2 (a) The MABP [18], (b) printed cell aggregate strand, (c) working principle of the MABP with dual operating nozzles intandem, (d) printed vasculature showing perfusability in a perfusion culture chamber. The printer is located in a vertical flowhood in our clean room facility for the sterilization of the process.

Fig. 3 (a) A printed vasculature in perfusion chamber underpulsatile flow, (b) a printed vasculature in zigzag shape longerthan 80 cm in length demonstrating perfusion in a custom-made perfusion chamber, and (c) a histology image showingcollagen and smooth muscle deposition by human umbilicalsmooth muscle cells in 6 weeks.

Journal of Manufacturing Science and Engineering DECEMBER 2014, Vol. 136 / 061013-3

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

viabily in the inner sections of the tissue strands, although neovas-cularization is also achieved. The tissue strands are then loadedinto the bioprinter unit using custom tools and directly printedwith the vasculature network layer by layer predetermined by thetoolpath plan generated by a computer-aided design model.

Our studies have shown that aggregated tissue strands presenthigh viability and rapid fusion capabilities. In this paper, we used2� 108 fibroblast cells and the average cell viability on day 1 post-fabrication was 75 6 0.5%, which gradually increased to 77 6 0.5%,and finally reached to 87 6 3% on day 7. When placed or printednext to each other, tissue strands underwent quick and seamlessfusion, forming a complete tissue patch in a week (see Figs. 4(b) and4(c)). Figure 4(d) shows F-actin expression, demonstrating thatfibroblasts in tissue strands are metabolically active as well. Other

studies have also shown great ability of angiogenesis within tissuespheroids made of endothelial cells and hepatocytes or pancreaticcells [11,21,22]. A similar phenomenon can be envisioned for tissuestrands as well if appropriated culture conditions are followed. Thisscaffold-free approach provides a practical approach in printinglarger constructs. It enables printing tissue aggregates continuously,which prevents gaps and breakage in the printed construct.

2.1.3 Geometric Modeling of Hybrid Constructs. In this sec-tion, the proposed design is mathematically represented in orderto fulfill fabrication and postfabrication requirements, which canbe used to determine design variables for bioprinting. As can beseen in Fig. 5, arc fitting is used at U-turns between two consecu-tive vasculatures during zigzag printing, which eliminates

Fig. 4 (a) Fabricated tissue strands, which are mechanically coherent and possess high cell viability. (b) and (c) Fusion of tis-sue strands under fluorescence microscope, where the gap between tissue strands closes in 7 days. (d) An immunostainingimage demonstrating F-actin expression (Note: 40,6-diamidino-2-phenylindole (DAPI) stained cell nucleus).

Fig. 5 A representative model of single layer hybrid bioprinting

061013-4 / Vol. 136, DECEMBER 2014 Transactions of the ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

occlusion of vasculatures. The first and the last vasculatures ineach layer should be printed longer than the tissue strands andother vasculatures in order to connect the construct to the perfu-sion system easily. During hybrid bioprinting, it is not practical ifthe tissue strand diameter is relatively smaller than that of vascu-latures, while tissue strands fuse to each other during in vitro cul-ture and contracts in some extent. Using smaller diameter tissuestrands thus bring several issues such as longer fabrication time,less structural integrity, and formation of gaps in the tissue whenfolding takes place. Tissue strands cannot be greater than the vas-culature in diameter as well, otherwise, clearance can be gener-ated between the tissue and vasculature, which results in mediaaccumulation in the clearance during perfusion. Thus, the diame-ter of the tissue strand should be same as that of the vasculature,which can be adjusted before or during bioprinting process. For asquare layer as shown in Fig. 5, the diameter of the vasculatureand tissue strands is equal and denoted by D (lm). In addition torelative size of diameters, the number of tissue strands betweentwo contiguous vasculatures is also important due to media diffu-sion requirement. Concentrations of solutes such as nutrients andoxygen around vasculatures change gradually. In order to providean efficient media transportation system, the variation of soluteconcentration gradients around vasculatures should be minimized,which can be achieved by designing distance between consecutivevasculatures. The distance between two contiguous vasculaturesat which the nutrients diffuse to based on cellular consumption ofnutrients is called the Krogh length kK �

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiDscO=RC

p(lm) [23],

where Ds (lm2/s), RC (mol/ls), and co are diffusion constant of the

concerning solute in the tissue strands, consumption rate of thesolute by cells, and concentration of the solute in the vasculature,respectively. The number of tissue strands between two contigu-ous vasculatures (n) can be thus calculated as

n � kK

D¼ 1

D

ffiffiffiffiffiffiffiffiffiffiDscO

RC

r(1)

where n has to be an integer. In order to maximize the consump-tion of the concerning solute, n should be the largest integer equalto or smaller than the right side of Eq. (1). In addition, the curva-ture of U-turn can be easily calculated based on the number of tis-sue strands and their diameter. The fabrication of tissue strand ishighly expensive and time consuming. It may take a month ormore to expand significantly high number of cells to be used intissue strand fabrication. On the other hand, preparation of materi-als for vasculature printing is negligibly cheap. In order to makethe tissue printing process practical and make the best use of thematerial, it is essential to calculate the volume of cell aggregatesbefore printing them. If the number of straight vasculatures is nc,the width W (lm) of the square construct can be calculated asW ¼ ncDþ 2Dþ ðnc � 1ÞnD, which can be reorganized as

W ¼ 3Dþ ðnc � 1Þðnþ 1ÞD (2)

The total volume of cell aggregates Vtotal (lm3) needed for tissuestrand printing for each layer can be then calculated as follows:

Vtotal ¼

3pD3

2þ pD3ðnc � 1Þ

84þ 9nþ nðnþ 1Þð2nc � 3Þ þ

ffiffiffiffiffiffiffiffiffiffiffiffiffin2 � 1p

þ 2xn

Xn� 1

2

i ¼ 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffin2 � 4iðiþ 1Þ � 1

p26664

37775;

for n ¼ 2mþ 1

3pD3

2þ pD3ðnc � 1Þ

84þ 9nþ nðnþ 1Þð2nc � 3Þ þ 2

Xn

2

i ¼ 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffin2 � 4ði2 � 1Þ � 1

p26664

37775;

for n ¼ 2m

8>>>>>>>>>>>>>>>>>><>>>>>>>>>>>>>>>>>>:

Subject to :

m 2N�;

xn ¼0 n ¼ 1

1 n 6¼ 1

�

(3)

The reader is referred to the Appendix for derivation of aboveequations. With mathematical formulations presented in this sec-tion, the operator can eliminate the risk of tissue death due to lackof oxygen or vital nutrients and make the tissue printing processpractical and maximize the utilization of cell aggregates, which isone of the challenges in this process.

2.2 Envisioned Biologically Driven Capillary Sprouting:Bridging Capillaries and Main Vasculatures. Like all other tis-sue engineering techniques, generation of a functional tissue ororgan with an internal vascular network is an ultimate goal. Tosuccessfully create vascularization in multiscale within the intro-duced hybrid tissue concept, a fabricated hybrid constructs needto be placed in a custom-made perfusion chamber, which is thenconnected to a circulating tubing system to facilitate continuoussupply of the cell culture media in vitro (see Fig. 6). The perfusion

chamber serves as a bioreactor to not only ensure sufficient sup-port for hybrid tissue, but also to provide an environment similarto in vivo condition. The connecting tubing needs to be preciselyinserted into the inlet and outlet of vasculatures using a microposi-tioning system. Upon positioning the whole perfusion system intoan incubator, oxygenized cell culture media are delivered throughthe system using multichannel peristaltic pumps. The media in thereservoir need to be replenished at a regular base after culturing.Growth factors such as fibroblast growth factor and epidermalgrowth factor, which are essential for angiogenesis, are supple-mented within the circulating culture media. The printed tissuestrands fuse to each other in 3D, creating a larger tissue enclosingthe vasculature during in vitro incubation in a customized bioreac-tor with continuous media perfusion. More importantly, upon theformation of a 3D tissue construct, the abovementioned angiogen-esis growth factors are applied for further in vitro culturing to

Journal of Manufacturing Science and Engineering DECEMBER 2014, Vol. 136 / 061013-5

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

drive the natural process of vascularization between the main vas-culature and prevascularized tissue strands, creating sprouting.Ultimately, a thick 3D tissue construct can be fabricated with abiomimetic vasculature system and be readily available for trans-plantation, disease modeling, or drug screening. This will be amajor breakthrough toward fabrication of larger scale organs.

Studies have been done to construct perfusable blood vesselconnecting capillaries cardiac tissue by a 3D cell sheet fabricationtechnology, in which endothelial cells within cardiac cells sheetsprouted and connected to the main blood vessel upon perfusionof growth factor-rich culture media [23]. Other studies have alsoshown that a prevascularized hepatic bud, when transplantedin vivo, can successfully anastomose to the main blood vessel andsurvive for a long period of time, carrying out its function [11].All these highlights offer foreseeable potential for the hybrid bio-printing technique to have a similar nature-driven process uponperfusion. When the media are perfused through a continuous vas-cular network within the hybrid tissue, the biological signals aswell as the media gradient along the perfusion direction within themedia would guide endothelial cell reorganization, migration, andcarry out of angiogenesis within the tissue strands, more impor-tantly enable sprouting toward the media supply direction (Fig. 7).

Newly generated capillaries within tissue strands are expectedto create bridging with main vasculatures, so that media suppliedthrough these newly formed capillaries would guarantee the sur-vival of tissue strands for longer period of time. Prolonged mediacirculation within the newly generated vessel system would alsoaccelerate the tissue maturation process by supplying sufficientgrowth factors, which drive tissue-specific cells to secrete ECMand further facilitate tissue strands fusion from all directions in3D, producing a functional vascularized perfusable tissue. Later,the matured tissue could be used for drug testing by directly deliv-ering different drugs via the perfusing system to evaluate tissueresponse. Moreover, bioprinted tissue could be implanted in vivoby anastomosing the main vasculature to the host to replace dam-aged or diseased tissues or organs.

3 Proof-of-Concept Study

In this section, we demonstrate a proof of concept of the tissueself-assembly around a 1.5 cm long smooth muscle vasculature.The vasculature was fabricated using the parameters given in Sec.2.1.1, enabling media perfusion without any occlusion. It was lon-ger than the cell aggregate strands (5 mm long) so it could hook

Fig. 7 The concept of capillarization within maturated tissues: cut-away views demonstrate zoom-in view macrovasculaturesand cocultured tissue strands, where endothelial cells self-organize and create capillarization under hemodynamically equiva-lent media flow with required growth factors

Fig. 6 Custom-made perfusion system for fusion and maturation of tissues enclosing the vasculature network.Arrowheads show media flow direction. The whole system can be easily enclosed within an incubator.

061013-6 / Vol. 136, DECEMBER 2014 Transactions of the ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

up to the perfusion chamber. We kept the tissue construct in staticculture first upon fabrication to keep it mechanically and structur-ally integrated until sufficient maturation was obtained. After1 day in culture, fibroblast cell aggregate strands started fusing toeach other. Later, perfect maturation was obtained in 7 days. Ingeneral, the duration depends on the cell type and the culture con-ditions (i.e., static, perfusion, etc.) used. We tried to pull the vas-culature from the tissue and observed that the tissue attached tothe vasculature completely (see Figs. 8(a) and 8(b)). Cross-sectional view in Fig. 8(c) demonstrates that tissue strandsenclosed the vasculature and adhere to the smooth muscle matrixtightly. Therefore, tissue strands can be used for demonstratinglarge-scale hybrid tissue bioprinting concept.

4 Discussions

In this paper, scale-up hybrid tissue printing concept has beenpresented, which gives great promises with broad applications inbasic biomedical research as well as clinical scenarios. However,challenges and limitations are still inevitable in every aspect ofthis process.

First of all, tissue strand fabrication is a biomaterial-freeapproach, which needs a great number of cells that usuallyexceeds the capability of the traditional cell culture and expandingsetting in a short period of time. Thus, an advanced and acceler-ated cell culture techniques need to be designed to expedite cellexpanding to meet the demand for large numbers of cells in areducing culturing time. One way to fulfill this goal is to design3D culture substrates or a cell culture vessel [24]; also, special

chemicals or small molecules can be integrated into cell culturemedia to accelerate cell proliferation without losing their pheno-type in order to meet the required number of cells for tissue strandfabrication [25].

Another critical issue is the usage of chemically crosslinkedhydrogels for vasculature bioprinting. In order to generate adirectly printed, mechanically strong vasculature, a higher con-centration of hydrogel needs to be used. In that case, cells arelargely entrapped individually within the hydrogel matrix with arelatively lower infiltration rate of cells. It also limits migration orproliferation of cells, which results in less cell–cell interaction,limits normal cell function, and results in less ECM deposition. Inthe future, other materials like collagen and elastin may also beintegrated within a low concentration of chemically crosslinkedhydrogel to supplement for mechanical strength as well as provid-ing a more optimal condition for cell growth and function. Lastbut not least, the mechanical property of the printed vasculatureneeds further optimization, since tissue strands would generategreat forces from all directions on the vasculature. Nanoscalematerials such as electrospun protein fibers can be introducedwithin the vasculature wall to enhance its mechanical properties.Capillary sprouting from the macrovascular network to tissuestrands also depends on the biological performance of the vascula-tures as well as the applied perfusion setting. Therefore, a highlyporous vasculature network is essential; it will enable endothelialcells to migrate and organize themselves based on the flow direc-tions and applied growth factors to trigger them. Endothelial cellscan be seeded during the bioprinting process or can be perfusedand glued to the luminal surface of the vasculature during the per-fusion process. Gluing endothelial cells is well established in theliterature and currently used for decellularized organs [26].Although this has been demonstrated in a few recent Naturepapers [23,27] using both artificial and natural vessels, more stud-ies should be conducted to understand this phenomenon further todevelop a well-established protocol.

One of the important issues after the hybrid bioprinting processis that the printed tissue construct cannot be directly connected tothe perfusion unit while the tissue and the vasculature are notstructurally integrated and cannot withstand perfusion. The mediaflow, under certain pressures ranging from 20 to 55 mmHg andflow rates from 30 to 50 ml/min for a biomimetic hemodynami-cally equivalent environment, can easily blow the whole con-struct. Therefore, the construct should be kept in static mediaculture and treated very gently during handling and media addi-tion. Once sufficient mechanical coherency is obtained, the con-struct can be transferred to the perfusion unit for further fusionand maturation processes. In order to avoid any potential issuesafter the bioprinting process, the tissue construct can be directlyprinted into the perfusion chamber. The designed perfusion cham-ber should hold the construct precisely as a support structure untiltissue fusion is achieved. Upon successful tissue fusion and matu-ration, the free-standing tissue can be easily used for different pur-poses, such as in vitro drug testing or transplantation to an animalmodel, particularly into a highly vascularized place in vivo.

Polymer-free bioprinting is one of the most exciting directionsin tissue printing. It enables rapid fabrication of tissues and over-comes all the drawbacks associated with polymers, such as degra-dation and related toxic products, limited cell infiltration andencapsulation capability, poor cell migration and proliferationinside the polymer matrix, and less chance for vascularization.Despite the great advantages, mechanical properties are the majordrawback, and careful investigation should be conducted toachieve acceptable mechanical rigidity before and after the bio-printing process. In general, culturing tissue strands for a longerperiod of time generates better mechanical properties becausecells deposit more ECM, particularly elastin and collagen pro-teins; however, their fusion and adhesion capabilities decreasewhile maturation is completed during the tissue strand fabricationprocess. Thus, the cell needs to be guided biologically to depositsatisfactory collagen and elastin in a shorter period of time to

Fig. 8 (a) Demonstration of self-assembly of fibroblast tissuestrands around a vasculature supporting pulsatile perfusion,which demonstrates the feasibility of the concept. (b) Zoom-inand (c) cross-sectional view.

Journal of Manufacturing Science and Engineering DECEMBER 2014, Vol. 136 / 061013-7

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

provide mechanical strength; these are the major proteins in theconnective tissue stromal of parenchymal organs. Better mechani-cal coherency also helps the operator to load the bio-ink easilywithout any challenges. In this regard, novel nozzle configurationsshould be developed, which enable the loading and printing ofthese strands with minimum structural damage, preserving theirintegrity.

5 Conclusions

This paper demonstrates a new approach in fabrication of livingtissues and organs on a larger scale using 3D printing technology,where scale-up tissue printing can be performed in two majorsteps: (i) printing a continuous vasculature network in tandemwith the rest of the multicellular vascularized tissue strands and(ii) biological maturation where tissue fusion and maturationoccurs along with envisioned biologically driven capillary sprout-ing from the main vasculature steam. We have demonstrated theproof-of-concept of tissue self-assembly around a smaller-scalemacrovasculature. Although the MABP can be used in developinglarger-scale tissues with the proposed design strategy along with anassociated toolpath plan for both arms, further research in both bio-printing technology and developmental biology needs to be con-ducted to biomimetically develop tissues for transplant and in vitrotesting for various tissue types. For future work, we will expand ourfabricated hybrid tissue models in 3D with more cell types cocul-tured into tissue strands and work on creating neovascularizationand capillary sprouting within free-standing tissue systems.

Acknowledgment

This study was supported by the National Institutes of Health(NIH) and the Institute for Clinical and Translational Sciences(ICTS) Grant No. ULIRR024979, NSF CMMI CAREER award #1349716, and a grant from the Diabetes Action Research and

Education Foundation. We thank Dr. Adil Akkouch for the immu-nostaining image.

Appendix

For a square network with a width W, the distance between thecenters of two adjacent vasculatures l (lm) can be calculated as

l ¼ ðnþ 1ÞD (A1)

There are two types of tissue strands: tissue strands between vas-culatures and tissue strands confining the entire layer. There existtwo tissue strands confining the entire layer with a length of W.For those tissue strands between vasculatures, the situation isquite complex. The total number of tissue strands between vascu-latures in one layer of printing can be calculated as (nc� 1)n. Thelength of tissue strands between vasculatures depends on theradius of curvature of the U-turns (which equals l) as well as thenumber of tissue strands between two adjacent vasculatures(which is n). As shown in Fig. 5, the length of tissue strandsbetween vasculatures comprises two parts, ls (lm) and lc (lm),where lc is the length of the tissue strands inside the U-turn curva-ture, and ls is the part of the tissue strands printed in parallel to thestraight vasculatures. For all tissue strands, ls remains same andcan be calculated using the following equation:

ls ¼ W � lþ D

2(A2)

Figure 9(a) shows the U-turn section of a vasculature. R (lm) isthe inner radius of the curvature, which can also be represented asR ¼ ðl� D=2Þ. lc can be calculated using the Pythagorean theo-rem (see Figs. 9(b) and 9(c)). Thus, the total length of tissuestrands inside the curvature can be calculated as

lc ¼

1

2

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffilðl� 2DÞ

pþ xn

Xn� 1

2

i ¼ 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffil2 � 2lD� 4iðiþ 1ÞD2

pfor n ¼ 2m� 1

Xn

2

i ¼ 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffil2 � 2lD� ð4i2 � 1ÞD2

pfor n ¼ 2m

8>>>>>>>>>><>>>>>>>>>>:

Subject to :

m 2N�;

xn ¼0 n ¼ 1

1 n 6¼ 1

�

(A3)

Fig. 9 (a) U-turn section of a vasculature, (b) two tissue strands are printed between two consecutive vasculatures (n 5 2), and(c) three tissue strands are printed between two consecutive vasculatures (n 5 3)

061013-8 / Vol. 136, DECEMBER 2014 Transactions of the ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

In summary, the total length of tissue strands Lt (lm) in one layerof the hybrid construct can be represented as

Lt ¼ 2W þ nðnc � 1Þls þ ðnc � 1Þlc (A4)

The total volume of cell aggregate (Vt) needed for tissue strand print-ing in each layer of the hybrid structure can be then calculated as

Vt ¼pD2Lt

4(A5)

By combining Eqs. (3), (A1), (A3), (A4), and (A5), Eq. (A6) canbe obtained as

Vt ¼

3pD3

2þ pD3ðnc � 1Þ

84þ 9nþ nðnþ 1Þð2nc � 3Þ þ

ffiffiffiffiffiffiffiffiffiffiffiffiffin2 � 1p

þ 2xn

Xn� 1

2

i ¼ 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffin2 � 4iðiþ 1Þ � 1

p26664

37775;

for n ¼ 2mþ 1

3pD3

2þ pD3ðnc � 1Þ

84þ 9nþ nðnþ 1Þð2nc � 3Þ þ 2

Xn

2

i ¼ 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffin2 � 4ði2 � 1Þ � 1

p26664

37775;

for n ¼ 2m

8>>>>>>>>>>>>>>>>>><>>>>>>>>>>>>>>>>>>:

Subject to :

m 2N�;

xn ¼0 n ¼ 1

1 n 6¼ 1

�

(A6)

References

[1] Mason, C., 2005, “Tissue Engineering Skin: A Paradigm Shift in Wound Care,”Med. Device Technol., 16(10), pp. 32–33.

[2] Langer, R., 2000, “Tissue Engineering,” Mol. Ther., 1(1), pp. 12–15.[3] Ozbolat, I. T., and Yu, Y., 2013, “Bioprinting Toward Organ Fabrication: Chal-

lenges and Future Trends,” IEEE Trans. Biomed. Eng., 60(3), pp. 691–699.[4] Shor, L., G€uceri, S., Gandhi, M., Wen, X., and Sun, W., 2008, “Solid Freeform

Fabrication of Polycaprolactone/Hydroxyapatite Tissue Scaffolds,” ASME J.Manuf. Sci. Eng., 130(2), p. 021018.

[5] Griffith, L. G., and Naughton, G., 2002, “Tissue Engineering: Current Chal-lenges and Expanding Opportunities,” Science, 295(5557), pp. 1009–1014.

[6] Andersson, H., and Van Den Berg, A., 2004, “Microfabrication and Microflui-dics for Tissue Engineering: State of the Art and Future Opportunities,” LabChip, 4(2), pp. 98–103.

[7] Mironov, V., Kasyanov, V., and Markwald, R. R., 2011, “Organ Printing: FromBioprinter to Organ Biofabrication Line,” Curr. Opin. Biotechnol., 22(5), pp.667–673.

[8] Visconti, R. P., Kasyanov, V., Gentile, C., Zhang, J., Markwald, R. R., and Mir-onov, V., 2010, “Towards Organ Printing: Engineering an Intra-OrganBranched Vascular Tree,” Expert. Opin. Biol. Ther., 10(3), pp. 409–420.

[9] Marga, F., Jakab, K., Khatiwala, C., Shepherd, B., Dorfman, S., Hubbard, B.,Colbert, S., and Forgacs, G., 2012, “Toward Engineering Functional OrganModules by Additive Manufacturing,” Biofabrication, 4(2), p. 022001.

[10] Mironov, V., Visconti, R. P., Kasyanov, V., Forgacs, G., Drake, C. J., andMarkwald, R. R., 2009, “Organ Printing: Tissue Spheroids as Building Blocks,”Biomaterials, 30(12), pp. 2164–2174.

[11] Takebe, T., Sekine, K., Enomura, M., Koike, H., Kimura, M., Ogaeri, T.,Zhang, R.-R., Ueno, Y., Zheng, Y.-W., and Koike, N., 2013, “Vascularized andFunctional Human Liver From an iPSC-Derived Organ Bud Transplant,”Nature, 499(7459), pp. 481–484.

[12] Greggio, C., De Franceschi, F., Figueiredo-Larsen, M., Gobaa, S., Ranga, A.,Semb, H., Lutolf, M., and Grapin-Botton, A., 2013, “Artificial Three-dimensional Niches Deconstruct Pancreas Development in Vitro,” Develop-ment, 140(21), pp. 4452–4462.

[13] Schubert, T., Anders, S., Neumann, E., Sch€olmerich, J., Hofst€adter, F., Grifka,J., M€uller-Ladner, U., Libera, J., and Schedel, J., 2009, “Long-Term Effects ofChondrospheres on Cartilage Lesions in an Autologous Chondrocyte Implanta-tion Model as Investigated in the SCID Mouse Model,” Int. J. Mol. Med.,23(4), pp. 455–460.

[14] Novosel, E. C., Kleinhans, C., and Kluger, P. J., 2011, “Vascularization Is theKey Challenge in Tissue Engineering,” Adv. Drug Deliv. Rev., 63(4), pp.300–311.

[15] Yu, Y., Zhang, Y., Martin, J. A., and Ozbolat, I. T., 2013, “Evaluation of CellViability and Functionality in Vessel-Like Bioprintable Cell-laden TubularChannels,” ASME J. Biomech. Eng., 135(9), p. 91011.

[16] Mondy, W. L., Cameron, D., Timmermans, J.-P., Clerk, N. D., Sasov, A., Cas-teleyn, C., and Piegl, L., 2009, “Computer-Aided Design of MicrovasculatureSystems for Use in Vascular Scaffold Production,” Biofabrication, 1(3),p. 035002.

[17] Norotte, C., Marga, F. S., Niklason, L. E., and Forgacs, G., 2009, “Scaffold-Free Vascular Tissue Engineering Using Bioprinting,” Biomaterials, 30(30),pp. 5910–5917.

[18] Ozbolat, I. T., Chen, H., and Yu, Y., 2014, “Development of ‘Multi-Arm Bio-printer’ for Hybrid Biofabrication of Tissue Engineering Constructs,” RobotComput. Integr. Manuf., 30(3), pp. 295–304.

[19] Zhang, Y., Yu, Y., Chen, H., and Ozbolat, I. T., 2013, “Characterization ofPrintable Cellular Micro-Fluidic Channels for Tissue Engineering,” Biofabrica-tion, 5(2), p. 025004.

[20] Zhang, Y., Yu, Y., and Ozbolat, I. T., 2013, “Direct Bioprinting of Vessel-LikeTubular Microfluidic Channels,” ASME J. Nanotechnol. Eng. Med., 4(2),p. 021001.

[21] Fujimori, H., Asahina, K., Shimizu-Saito, K., Ikeda, R., Tanaka, Y., Teramoto,K., Morita, I., and Teraoka, H., 2008, “Vascular Endothelial Growth Factor Pro-motes Proliferation and Function of Hepatocyte-Like Cells in Embryoid BodiesFormed From Mouse Embryonic Stem Cells,” J. Hepatol., 48(6), pp. 962–973.

[22] Brissova, M., Fowler, M., Wiebe, P., Shostak, A., Shiota, M., Radhika, A., Lin,P. C., Gannon, M., and Powers, A. C., 2004, “Intraislet Endothelial Cells Con-tribute to Revascularization of Transplanted Pancreatic Islets,” Diabetes, 53(5),pp. 1318–1325.

[23] Sekine, H., Shimizu, T., Sakaguchi, K., Dobashi, I., Wada, M., Yamato, M.,Kobayashi, E., Umezu, M., and Okano, T., 2013, “In Vitro Fabrication of Func-tional Three-Dimensional Tissues With Perfusable Blood Vessels,” Nat.Commun., 4(1399), pp. 1–10.

[24] Liu, H., Collins, S. F., and Suggs, L. J., 2006, “Three-dimensional Culture forExpansion and Differentiation of Mouse Embryonic Stem Cells,” Biomaterials,27(36), pp. 6004–6014.

[25] Vander Heiden, M. G., Plas, D. R., Rathmell, J. C., Fox, C. J., Harris, M. H.,and Thompson, C. B., 2001, “Growth Factors Can Influence Cell Growth andSurvival Through Effects on Glucose Metabolism,” Mol. Cell Biol., 21(17), pp.5899–5912.

[26] Song, J. J., Guyette, J. P., Gilpin, S. E., Gonzalez, G., Vacanti, J. P., and Ott, H.C., 2013, “Regeneration and Experimental Orthotopic Transplantation of a Bio-engineered Kidney,” Nat. Med., 19(5), pp. 646–651.

[27] Sakaguchi, K., Shimizu, T., Horaguchi, S., Sekine, H., Yamato, M., Umezu,M., and Okano, T., 2013, “In Vitro Engineering of Vascularized TissueSurrogates,” Sci. Rep., 3(1316), pp. 1–7.

Journal of Manufacturing Science and Engineering DECEMBER 2014, Vol. 136 / 061013-9

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 11/10/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use