The FASEB Journal • Research Communication

Neoarteries grown in vivo using a tissue-engineeredhyaluronan-based scaffold

Barbara Zavan,* Vincenzo Vindigni,†,1 Sandro Lepidi,‡ Ilaria Iacopetti,�

Giampiero Avruscio,§ Giovanni Abatangelo,* and Roberta Cortivo**Department of Histology, Microbiology, and Medical Biotechnologies; †Clinic of Plastic Surgery,‡Clinic of Vascular Surgery, §Unit of Angiology, and �Department of Veterinary Sciences, Universityof Padova, Padova, Italy

ABSTRACT Vascular tissue engineering has emergedas a promising technology for the design of an ideal,responsive, living conduit with properties similar to thatof native tissue. The missing link in tissue-engineeredblood vessels is elastin biosynthesis. Several biomateri-als are currently used but few support elastin biosyn-thesis in a 3-D array. In previous studies, we demon-strated that a hyaluronan-based scaffold (HYAFF-11™)grafted in the infrarenal rat aorta successfully guidedthe complete regeneration of a well-functioning small-diameter (2 mm) neoartery. The aim of the presentstudy was to test the ability of HYAFF-11 biodegradablegrafts to develop into neovessels of larger size (4 mm)in a porcine model, focusing on extracellular matrix(ECM) remodeling and elastin biosynthesis. HYAFF-11tubes (diameter 4 mm, length 5 cm) were implanted inan end-to-end fashion in the common carotid artery.Grafts were analyzed for patency with a Duplex scanevery 15 days. ECM components were evaluated byhistological and molecular biological methods. All theanimals survived the observation period without com-plications. Intimal hyperplasia (initiating at the anasto-motic site) and graft thrombosis led to 3 cases of partialor complete occlusion, as demonstrated by histologicalexamination. There were no signs of stenoses or aneu-rysms in the remaining grafts. After 5 months, thebiomaterial was almost completely degraded and re-placed by a neoartery segment composed of maturesmooth muscle cells, collagen, and elastin fibers orga-nized in layers and was completely covered on theluminal surface by endothelial cells (vWF�). Whereas inprevious small animal studies, patency rates were notoptimal, those obtained in the present study usinghyaluronan-based grafts of larger size confirmed theability of these constructs to guide the development ofa well-functioning neoartery, with the remarkable addi-tional attribute of facilitating the formation of orga-nized layers of elastin fibers.—Zavan, B., Vindigni, V.,Lepidi, S., Iacopetti, I., Avruscio, G., Abatangelo, G.,Cortivo, R. Neoarteries grown in vivo using a tissue-engineered hyaluronan-based scaffold. FASEB J. 22,2853–2861 (2008)

Key Words: elastin � vascular prosthesis � remodeling � surgery

There is a considerable clinical need for alterna-tives to the autologous vein and artery tissues used forvascular reconstructive surgeries such as coronary by-pass, lower limb bypass, arteriovenous shunts, andrepair of congenital defects to the pulmonary outflowtract. Thus far, synthetic materials have not matchedthe efficacy of native tissues, particularly in small-diameter applications (1). The development of cardio-vascular tissue engineering introduced the possibility ofa living, biological graft that might mimic the func-tional elastic properties of native vessels (1, 2). A majorstructural element of arterial walls is elastic fiber, whichendows vessels with the critical property of elastic recoil(2). In arteries, elastin dictates tissue mechanics at lowstrains before stiffer collagen fibers are engaged. Elas-tin also confers elasticity, preventing dynamic tissuecreep by stretching under load and recoiling to itsoriginal configuration after the load is released. Inaddition to the mechanical responsiveness, elastin is apotent autocrine regulator to vascular smooth musclecell (SMC) activity, and this regulation is important forpreventing fibrocellular pathology. Finally, elastin deg-radation associated with matrix metalloproteinase(MMP) activity is a cell-mediated process, observed inalmost all types of vascular calcification. Elastin pep-tides associated with transforming growth factor-�1could induce osteogenic gene expression in SMCs,possibly via the elastin laminin receptor (ELR) signal-ing (1, 2). Their presence in a vascular graft wouldgreatly enhance design and patency (3–7). Elastic fibersalso influence vascular cell behavior through directinteraction and by regulation of growth factor activa-tion (3–7). Elastin knockout studies and clinical obser-vations have revealed the essential regulatory functionof elastin during artery development. In the absence ofextracellular elastin accumulation, smooth muscle pro-liferation leads to arterial stenosis (3–7). Thus, toensure appropriate mechanical function and preventserious complications, successful artery replacementsmust incorporate an elastic component. Approaches to

1 Correspondence: Clinic of Plastic Surgery, University ofPadova, Via Giustiniani 2, 35100 Padova, Italy. E-mail:vincenzo.vindigni@unipd.it

doi: 10.1096/fj.08-107284

28530892-6638/08/0022-2853 © FASEB

date have included building tissue replacements on anelastin scaffold isolated from cadaveric tissues, supply-ing soluble tropoelastin to a cell culture, or designingbiocompatible synthetic elastic polymers (1). The mostpromising approach was described by L’Heureux et al.(8–10), who developed a completely autologous tech-nique called sheet-based tissue engineering. Dermalfibroblasts are obtained from a small skin biopsy andgrown in conditions that promote an engineered com-plete vessel, ready to be implanted in 6 months (8–10).This approach is time consuming, with total productiontime at �24 wk. In the present study, we investigated aresorbable biomaterial that can support artery remod-eling processes directly in vivo, without using chemicaland cellular in vitro preconditioning. This article pre-sents the results of the second part of a continuingstudy investigating the efficacy of a hyaluronan-basedvascular prosthesis, named ABAT (Artery Bioregenera-tion Assist Tube), as scaffolding material for the com-plete regeneration of a new artery tract (11, 12). Inparticular, we investigated the ability of the proposedbiomaterial to allow in vivo regeneration of a physio-logical arterial wall provided with an adequate elasticcomponent. Previously, we had demonstrated thatABAT induced complete artery regeneration in a ratexperimental model, orchestrating the vascular re-generation events needed for a very small arteryreconstruction (2 mm diameter) (11, 12). Thesepreliminary results, particularly the ability of ABATto allow for the complete regeneration of the elasticlayer, stimulated further investigations in a morerepresentative animal experimental model. We pro-duced 4-mm-diameter, 5-cm-length ABATs and testedtheir ability to substitute a common carotid arterytract in pigs. This work demonstrates and quantifieselastogenesis in a totally bioresorbable and biocom-patible graft that mimics the characteristic alignmentof native tissues in a pig experimental model. Thisalignment confers the mechanical strength of tissueand provides a template for organized extracellularmatrix (ECM) deposition.

MATERIALS AND METHODS

Vascular prosthesis fabrication

The biomaterial, derived from total esterification of hyaluro-nan (synthesized from 80–200 kDa sodium hyaluronate) withbenzyl alcohol and referred to as HYAFF-11, was dissolved indimethyl sulfoxide (DMSO). A core steel rotating cylindricalbar, A 4 mm, was coated with 1.5 ml of the HYAFF-11/DMSOsolution, which was subsequently coagulated in an ethanolbath. The shaped tubules were carefully removed from thecylinder, cut to 5.0 cm in length and 4 mm in diameter (50�m thick), washed in water, and air-flow dried. They werethen double-blister packed and sterilized by gamma radia-tion. These HYAFF-11 tubules were a gift of Fidia AdvancedBiopolymers (Abano-Terme, Italy).

Surgical implantation of the prostheses

The protocol was approved by the Institutional Animal CareCommittee of Padova University and conducted in compli-ance with U.S. National Research Council guidelines for theethical treatment of animals (13). Ten house swine aged 3 to4 months (29�2 kg) were used in this study. They weredeprived of food the night before surgery and receivednormal food throughout the postoperative period. Monolat-eral carotid artery graft interposition was performed througha cervical midline incision. The carotid arteries were pre-pared and freed on a length of �8 to 10 cm, taking care toprevent vascular spasm with externally applied papaverinesolution. After proximal and distal clamping of the carotidartery, a segment of �3 cm was excised in a beveled fashion.The prostheses were also beveled and trimmed to a length of5 cm. Proximal and distal anastomoses were performed with8-0 polypropylene continuous suture, and fibrin glue (Tissu-col; Baxter AG, Vienna, Austria) was used as a biologicalsealant to better seal the anastomotic site (Fig. 1). Through-out the survival period, animals were fed 0.5 g of aspirin daily.Duplex scan examinations were performed under inhalant-induced sedation (isoflurane 4%) after standard premedica-tion every 15 days.

Prostheses explants

Two animals were killed at each time point: 1, 2, 3, 4, and 5months. Before sacrifice, pigs underwent a general physicalexamination to evaluate their condition. Anesthesia was in-

Figure 1. Surgical implantation of the prosthesis. Proximal and distal anastomoses were performed with 8-0 polypropylenecontinuous suture (A; yellow arrow), and fibrin glue was used as a biological sealant to better seal the anastomotic site (B, C;yellow arrows). The biomaterial was transparent (A), allowing direct verification of surgical precision. When the clamps wereremoved (C), blood flow resumed, and the prosthesis appeared perfectly integrated with the original artery. There were nomacroscopic signs of abnormal tube dilation.

2854 Vol. 22 August 2008 ZAVAN ET AL.The FASEB Journal

duced as described for implantation. The prosthetic site wascarefully exposed at the point of previous surgical entry.Photographs and videos were taken for gross observation.After 10,000 U of intravenous heparin, grafts were excisedwith 2 cm of native proximal and distal carotid artery. Theharvested segments were rinsed carefully with heparinizedsaline. Prostheses were cut transversally at the midline into 2pieces and were fixed in formaldehyde and embeddingmedium for frozen tissue specimens (O.C.T. Tissue-Tek;Sakura Finetek, Torrance, CA, USA) for histological, immu-nohistochemical, and ultrastructural analyses. Some sampleswere entirely formalin fixed or cryopreserved, and full longi-tudinal sections were obtained for routine staining. Aftergraft harvesting, the animal was killed with an overdose ofbarbiturates (Nembutal sodium solution; Abbott Laborato-ries, Abbott Park, IL, USA).

Explant morphology

Tissue samples in fixative designated for histological orimmunohistochemical analysis were transferred to PBS with0.02% sodium azide after 4 h and stored at 4°C until the finalpreparation. The specimens were then dehydrated in gradedethanol (70–99.5%), cleared in xylene, and embedded inparaffin. Sections (5 �m) were cut on a standard microtome,and routine light microscopy examination was performedafter staining with hematoxylin and eosin (H&E) for cellinfiltration and with Weigert’s stain for detection of elastin.

Sections of the explanted graft were stained with H&E aswell as Weigert solution and evaluated for host cell infiltra-tion. Immunofluorescent studies of endothelial cells utilizedantibodies to von Willebrand factor (vWF; Dako, Carpenteria,CA, USA) and CD31 (Santa Cruz Biotechnology, Santa Cruz,CA, USA). To characterize vascular SMCs, cells were testedfor expression of myosin light chain kinase (MLCK) accord-ing to the method described by Vescovo et al. (14) and others.Negligible background was found in controls in which theprimary antibody was omitted. Negative controls were per-formed by replacing primary antibody with normal rabbitserum or phosphate-buffered salt solution. Immune reactionswere revealed by the fast red substrate (Sigma-Aldrich Corp.,St. Louis, MO, USA) and counterstained with hematoxylin(Sigma).

Real-time polymerase chain reaction (PCR)

One centimeter of explants was stored in RNAlater (Qiagen,Valencia, CA, USA) at 4°C for up to 3 wk. RNA was thenextracted from whole-hemisphere constructs (and for com-parison, from 50,000 cells cultured in flasks) using an RNeasykit (Qiagen) according to the manufacturer’s instructions.Briefly, constructs were incubated in 350 �l of RLT buffer(Qiagen) and homogenized via sonication for 20 s at roomtemperature. Samples were loaded onto RNeasy mini col-umns (Qiagen), and DNase treatment was performed accord-ing to the manufacturer. Total RNA was eluted with 50 �l ofH2O. Concentration of RNA was determined by measuringthe absorbance at 260 nm in a spectrophotometer (Nano-drop; Thermo Fisher, Waltham, MA, USA) and stored at�80°C.

For each target gene, primers and probes were selectedusing Primer3 software (Roche Molecular Diagnostics, Pleas-anton, CA, USA).

Tropoelastin: gcatttctccccgagatg/ggaactccacca-ggaatgg

Collagene type I: gggattccctggacctaaag/ggaacacc-tcgctctcca

MMP1: tgttcagctgtcccaagatg/ggttttcggaaggt-ccatagat

MMP2: ctggtgctgccacatttag/gggtgctgtaagccac-aga

CD31: aaggtggagtcgtgaaggtc/tgaaatgtactggag-gtttttcc

vWF: agtgcagacccaacttcacc/gtggggacactctttt-gcac

VEGF: tgcccgctgctgtctaat/tctccgctctgagcaaggMyo, MLC: gttgagggtctgcgtgtctt/aagttcagcac-

ccatgactgtaFU, prolyl 4-hydroxylase protein: ccaacaacgccagaga-

aga/tccctggagctcagattgatGene expression was measured using real-time quantitative

PCR on a Rotor-GeneTM 3500 (Corbett Research, Sydney,Australia). PCR reactions were carried out using the primersat 300 nm and the SYBR Green I (Invitrogen, Carlsbad, CA,USA; using 2 mm MgCl2) with 40 cycles of 15 s at 958C and 1min at 608C. All cDNA samples were analyzed in duplicate.Fluorescence thresholds (Ct) were determined automatically

Figure 2. Duplex scan images obtained at 15 days (A) and at 1 (B) and 4 months (C) after surgery. Yellow arrows (A, B) pointto the prosthesis. White arrows show the direction of blood flow. Green arrows (C) show the new artery tract and prosthesismacroscopical reabsorption 4 months after surgery.

2855NEOARTERIES GROWN IN VIVO

by the software with efficiencies of amplification for thestudied genes ranging between 92 and 110%. For each cDNAsample, the Ct value of the reference gene R16 was subtractedfrom the Ct value of the target sequence to obtain the DCt.The level of expression was then calculated as 2_Ct andexpressed as the mean 6 sd of quadruplicate samples of 2separate experiments. Relative quantification of marker geneexpression is given as a percentage of the R16 product, andthe t test was applied. Experiments were performed with 2different explants and repeated at least 3 times. Values wereexpressed as the mean � sd.

Proliferation assays

The 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazoliumbromide (MTT) -based (thiazolyl blue) cytotoxicity test ana-lyzes mitochondria viability; that is, only functional mitochon-dria can oxidize an MTT solution, giving a typical blue-violetend product. This assay is an indirect method to assess cellgrowth and proliferation because the optical density valuecan be correlated to the number of cells. At each time point(1, 2, 3, 4, 5 months), a 1-cm length of the central portion ofthe explanted tube was evaluated with the MTT test. Briefly,supernatant was gently harvested from the multiwell tissueculture plate, and 1 ml of MTT solution (0.8 �g/ml in PBS)was added. Cultures were returned to the incubator and, after3 h, supernatant was again harvested. Each scaffold (1 cm ofexplant) was then transferred to an Eppendorf microtube(Eppendorf, Hamburg, Germany) and 1 ml of extractionsolution (0.01 N HCl in isopropanol) added. Eppendorfmicrotubes were vortexed vigorously for 5 min to allow totalcolor release from the scaffolds, then centrifuged at 14,000rpm for 5 min, and supernatants were read at 534 nm.

Cell numbers

Cell numbers were determined using a DNeasy kit (Qiagen)to isolate total DNA from 1 cm of explant. At each time point(1, 2, 3, 4, and 5 months), a 1-cm length of explanted tube wasevaluated for cell numbers. Briefly, constructs were digestedfollowing the manufacturer’s protocol for tissue isolation,using overnight incubation in proteinase K (Qiagen). Theconcentration of DNA was detected by measuring the absor-bance at 260 nm in a spectrophotometer. Cell numbers werethen determined from a standard curve (micrograms DNA vs.cell number) generated by DNA extraction from countedcells. The standard curve was linear over the tested range of5–80 �g DNA (r�0, 99).

RESULTS

Gross observations at implantation and duplexscan analyses

This prosthetic material was shown previously to be easyto handle from a surgical point of view, being soft and

Figure 3. A) Functional duplex scan of normal controlateralcarotid artery. B, C) Functional duplex scan studies after 1and 5 months, confirming regular blood flow throughout theprosthesis (B) and the new reconstituted artery tract (C).D) Yellow arrow points to progressive occlusion of the pros-thesis because of steady thickening of the arterial wall at theanastomotic site after 3 months. A total of 3 prostheses werecompletely occluded at duplex scan examination after 2, 3,and 4 months.

2856 Vol. 22 August 2008 ZAVAN ET AL.The FASEB Journal

elastic, as well as easy to cut and suture. Regarding thepresent experiments, it had the ideal texture to allowsuturing the anastomoses with 8-0 polypropylene con-tinuous suture (Fig. 1A, B). Fibrin glue was needed toseal small pores in the prosthetic material at theanastomotic sites (Fig. 1C). When blood flow resumedthrough the prosthesis, pulsatile dilation of the tubewas noticeable with the naked eye. Blood loss from thetube ceased after fibrin glue application (Fig. 1C).Figure 1B, C presents the tube before and after therelease of vascular clamps at implantation. The bioma-terial was transparent (Fig. 1A), allowing direct verifi-cation of surgical precision. When the clamps wereremoved (Fig. 1C), blood flow resumed, and the pros-thesis appeared perfectly integrated with the originalartery. There were no macroscopic signs of abnormaltube dilation. The anastomoses were solid and wellintegrated with the original artery (Fig. 1C).

Despite partial or complete occlusions in 3 prosthe-ses, documented by histological analysis, animal sur-vival rate was 100%, and no signs of vascular insuffi-ciency in peripheral regions were observed. Duplexscan studies showed no aneurismatic dilation or col-lapse of the prostheses at any time point. As explainedin detail later, the 3 occlusions were found to be due tointimal hyperplasia. Figure 2 presents duplex scanimages obtained at 15 days (Fig. 2A) and at 1 (Fig. 2B)and 4 months (Fig. 2C) after surgery. The figure showsthe integrity of the prosthesis at different time points

and its macroscopical reabsorption after 4 months.Figure 3 presents functional duplex scan studies after 1and 5 months, confirming regular blood flow through-out the prosthesis (Fig. 3B) and the new reconstitutedartery tract (Fig. 3C). Figure 3D shows progressiveocclusion of the prosthesis due to steady thickening ofthe arterial wall at the anastomotic site after 3 months.Progressive occlusion was clearly due to a technicalproblem at the anastomotic site; surgical anastomosiscreated a mismated diameter between original arteryand prostheses. In conclusion, a total of 2 prostheseswere completely occluded at duplex scan examinationafter 2 and 3 months, 1 prosthesis was partially oc-cluded after 4 months, and 7 prostheses did not showsigns of occlusion or intimal hyperplasia.

Histological analyses

Figure 4 presents a transversal section of the prosthesisafter 4 months. The specimen was harvested at themidline, and a nearly complete occlusion of the lumenwas evident (Fig. 4A, B). The occlusion was probablydue to progressive intimal hyperplasia. No biomaterialwas evident, as it had been substituted by neoadventitialtissue that was completely fused with medium (Fig. 4A,B). Although patency rates seemed inferior in previoussmall-animal studies, hyaluronan-based grafts of largersize confirmed the ability of hyaluronan to guide thedevelopment of a well-functioning neoartery, including

Figure 4. Transversal section of the prosthesis after 4 months. The specimen was harvested at the midline, and a nearly completeocclusion of the lumen is evident, indicated by H&E stain (A) and Weigert’s stain (B). No biomaterial is evident. Black arrowspoint to the endothelial layer lining the luminal surface (D; H&E stain) and the well-defined elastic layers (E; Weigert’s stain).Panels C and F (Weigert’s stain) show normal contralateral carotid artery; elastic layers are diffuse in the arterial wall context.Organization of the intima in the new artery tract (D, E) is identical to the native vessel (F). Scale bars � 2 mm.

2857NEOARTERIES GROWN IN VIVO

the remarkable attribute of facilitating the organizationof layers of elastin fibers (Fig. 4B, C). Figure 5A, B showsa completely occluded artery 5 months after surgery.The occlusion was probably due to late thrombusformation related to intimal hyperplasia at the anasto-motic site, because artery wall regeneration was com-plete and clearly included elastic fiber layers (Fig. 5B,white arrows).

Figure 6 presents a transversal section of the prosthe-sis 5 months after surgery. The specimen was harvestedat the midline. Flattened cells were observed on theluminal surface, surrounded by other cell components,which collectively resembled a natural artery wall withno signs of intimal hyperplasia (Fig. 6A, C). Weigertstain demonstrated a linear, continuous, and very wellrepresented elastic component (Fig. 6B, D). The vascu-lar lumen was lined with cells that stained positively forCD31 and vWF (Fig. 7A, B). Immunohistochemical andimmunofluorescence analysis confirmed the presenceof a continuous, thin layer of vascular SMCs (Fig. 7C,D), as indicated by positive staining with an antibodyspecific for MLCK.

Extracellular analyses

The ECM was evaluated with both traditional histolog-ical staining (Weigert for elastic fibers; Fig. 6D) andwith molecular biology techniques (real-time PCR forgene expression; Fig. 8). RNA was extracted everymonth in an attempt to characterize gene expressionduring the tissue growth and development periods.Semiquantitative real-time PCR (Fig. 8) was then usedto quantify the relative expression of genes, with a focuson the genes involved in ECM synthesis and remodel-ing. Collagen type I synthesis was increased over time.The levels of collagen type I doubled from the first tothe fifth month. In contrast, tropoelastin expressiondecreased from month 1 to month 5, following the timecourse of elastic fiber organization.

Cellular analyses

Figure 9 reports cell numbers (Fig. 9A) and metabolicactivity (MTT test; Fig. 9B). Large numbers of cells werepresent in the implanted scaffold. Autologous cellscolonized the scaffolds in a time-dependent manner,from a baseline of �7 � 106 cells (for 1 cm of scaffoldexplanted) 1 month after implantation to 12 � 106 cells4 months after implantation (Fig. 9A). Figure 9B re-veals that the cells were active; the increases in MTTvalues were correlated with increasing mitochondrialactivity.

Real-time PCR methods were applied to study thephenotype of the cells present within the implant. After1 month, the scaffolds were colonized principally withfibroblasts, the number of which was constant over time(Fig. 10, FU). SMCs (Fig. 10, myo) were also presentfrom the first month, and the number slowly increasedover time.

Endothelial cells were analyzed for CD31, vWF, andvascular endothelial growth factor (VEGF) expression.As reported in Fig. 10, 1 month after implantation,endothelial cells were few, but the number increasednotably over time.

DISCUSSION

Vascular tissue engineering has emerged as one of themost promising approaches to producing mechanicallycompetent vascular substitutes. Despite the great prom-ise engineered blood vessels have to offer, elastin, acritical structural component, is notably absent frommost end products of vessel tissue engineering (1).Elastin is a critical structural and regulatory matrixprotein that plays a dominant role in conferring elas-ticity to the vessel wall (2–7). Although cardiovasculartissue engineering is being actively investigated, verylittle attention has been given to elastin, which, inaddition to mechanical responsiveness, is a potentautocrine mediator of vascular SMC activity, whoseregulation is important for preventing fibrocellularpathology.

Figure 5. A completely occluded artery 5 months after sur-gery. A) Complete occlusion due to late thrombus formation(H&E stain). B) White arrows point to elastic fibers (H&Estain). Scale bars � 2 mm.

2858 Vol. 22 August 2008 ZAVAN ET AL.The FASEB Journal

In vascular tissue engineering, scaffolds play an im-portant role in elastin biosynthesis by SMCs. It isconceivable that in previous techniques (15–19), thechange from 2- to 3-D culture conditions utilizing

porous scaffolds may have contributed to a loss inelastin biosynthesis by SMCs. It is now known that thescaffold must provide the appropriate chemistry forSMCs to secrete elastin.

Figure 6. Transversal section ofthe prosthesis 5 months after sur-gery. The specimen was harvestedat the midline. Flattened cells areseen on the luminal surface (D;black arrows), surrounded byother cell components, which col-lectively resemble a natural arterywall with no signs of intimal hyper-plasia (A, B). Weigert’s stain (B, E)demonstrates a linear, continuous,and very well represented elastic

component. A duplex scan study confirmed regular blood flow throughout the prosthesis 5 months after surgery (C). Scalebars � 2 mm.

Figure 7. Immunohistochemical and immunoflu-orescence analysis. The vascular lumen is linedwith cells that stained positively for CD31 (A) andvWF (B). A continuous, thin layer of vascular SMCs(C, D), indicated by positive staining with anantibody specific for MLCK, is very well repre-sented in the arterial wall context. Scale bars �2 mm.

2859NEOARTERIES GROWN IN VIVO

Few scaffolds are able to promote elastin biosynthe-sis. One of the most promising ECM-derived polymersis hyaluronan. In vitro studies with gel preparations ofhyaluronan (20, 21) confirm elangiogenesis in neona-tal SMCs. Recently, human vascular cells, such as endo-thelial cells (22) and SMCs (23), have been grown invitro on HYAFF constructs to develop new tissue-engi-neered vascular substitutes. Previous in vivo studies (11,12) have confirmed that tubes of benzylic esters ofhyaluronic acid (namely HYAFF-11 biomaterials) cansequentially orchestrate the vascular regenerationevents needed for very small artery reconstruction.

In light of these results, a tubular scaffold was used as

an in vivo temporary absorbable guide in 10 pigs tosubstitute a common carotid artery tract. We focusedour attention on elastic layer regeneration, the missinglink in engineered vascular prostheses. The most prom-ising and novel findings of the present study involvedcomplete regeneration of the elastic layers and thecomplete reabsorption of the biomaterial, 4 and 5months after implantation, respectively. Ultrasoundimaging demonstrated a constant diameter of theABAT throughout the 5 months of the study, indicatingthat the prostheses were mechanically stable duringincorporation into the surrounding tissue. These data,coupled with positive histological observations and theabsence of aneurysm formation in the long-term, indi-cate that this ABAT has promise as a clinical prosthesis,although further mechanical tests must be performed.

The absence of lymphocyte and neutrophil infiltra-tion attested to the biocompatibility of the material,which subsequently was entirely degraded 4 monthsafter implantation. HYAFF-11 has already been shownto be very well tolerated and to elicit no adversereactions in clinical practice (22, 24). Macroscopic andhistological results confirmed the ability of the prosthe-sis to create a “biomimetic” environment during tissueformation.

Data regarding cell numbers and metabolism con-firmed the colonization of the scaffolds by autologouscells such as fibroblasts and SMCs, particularly duringthe first 2 months, when these cells helped to secreteECM components. Molecular analysis of myo expres-sion by SMCs was high during the first month of theimplant. Probably the majority of elastic fibers were alsoproduced during the first month, because tropoelastinexpression was then at maximum levels. After post-translational modifications, the tropoelastin monomer,which is soluble, nonglycosylated, and highly hydro-phobic, is organized into elastic polymers that formconcentric rings of elastic lamellae around the mediallayer of arteries, as shown in Fig. 6. Weigert’s staininghighlighted elastin fibers, revealing the presence ofthese well-organized elastin rings. This component isnotably absent from most engineered vessels generated

Figure 8. Gene expression of extracellular components. Tro-poelastin, collagen (coll I), MMP1, and MMP2 throughoutthe 5-month observation period after implantation. Real-timePCR results were normalized to r16 expression at each timepoint. Results are expressed as mean and range.

Figure 9. A) Cell proliferation over the 5-month observationperiod after implantation. Cell numbers were determined viaDNA assay. Results are expressed as the mean and range.B) Cell proliferation over the 5-month observation periodafter implantation. Cell proliferation was determined viaMTT assay. Results are expressed as mean and range.

Figure 10. Gene expression of cellular markers during the5-month observation period after implantation. For endothe-lial cells, the markers used were CD31, vWF (vw), vascularendothelial growth factor (vegf), SMCs (myo), and fibroblast(FU). Real-time PCR results were normalized to r16 expres-sion at each time point. Results are expressed as mean andrange.

2860 Vol. 22 August 2008 ZAVAN ET AL.The FASEB Journal

with other techniques. At the same time, collagen typeI expression increased over time, entrapping SMCs andthus provoking a reduction in tropoelastin synthesis.Expression of the collagenase MMP1 was stronglydown-regulated from its low initial level, indicating thestability of the collagen produced. The expressionpattern of gelatinase MMP2 reflected an increase inECM degradation in the second month as the tissuedeveloped. Given the numerous substrates of MMP2and the role of MMP2 in basement membrane remod-eling in particular, its increased expression indicatedan increasing degree of remodeling, consistent with thedegree of development observed histologically (25).

The regenerated vessel grew inside the tubule with-out signs of infiltration into the biomaterial wall. Thepresence of endothelial cells revealed by CD31, vWF,and VEGF expression increased after the first month ofimplantation, thus ensuring the thromboresistance ofthe hyaluronan tube.

The feasibility of creating a completely biodegrad-able vascular regenerating guide directly in vivo, with-out chemical and cellular preconditioning in vitro, is ofgreat interest. This approach may overcome the mainproblems related to present techniques. Tissue-engi-neered grafts currently being investigated require anextended period of preparation, and thus they cannotbe used in emergency situations (26–28). Furthermore,the prolonged duration of culture increases the risk ofinfection and raises costs in terms of personnel, equip-ment, and materials needed.

In conclusion, this hyaluronan-based tube seemed tofulfill the major requirements of a microvascular scaf-fold in that it allowed for vascular tissue growth, facili-tated endothelialization of the luminal surface, andpromoted elastin biosynthesis.

REFERENCES

1. L’Heureux, N., Dusserre, N., Marini, A., Garrido, S., de laFuente, L., and McAllister, T. (2007) Technology insight: theevolution of tissue-engineered vascular grafts–from research toclinical practice. Nat. Clin. Pract. Cardiovasc. Med. 4, 389–395

2. Patel, A., Fine, B., Sandig, M., and Mequanint, K. (2006) Elastinbiosynthesis: The missing link in tissue-engineered blood ves-sels. Cardiovasc. Res. 71, 40–49

3. Karnik, S. K., Brooke, B. S., Antonio, B. G., Sorensen, L., Wythe,J. D., and Schwartz, R. S. (2003) A critical role for elastinsignaling in vascular morphogenesis and disease. Development130, 411–423

4. Long, J. L., and Tranquillo, R. T. (2003) Elastic fiber productionin cardiovascular tissue-equivalents. Matrix Biol. 22, 339–350

5. Li, D. Y., Brooke, B., Davis, E. C., Mecham, R. P., Sorensenk,L. K., and Boakk, B. B. (1998) Elastin is an essential determinantof arterial morphogenesis. Nature 393, 276–280

6. Kielty, C. M., Sherratt, M. J., and Shuttleworth, C. A. (2002)Elastic fibres. J. Cell Sci. 115, 2817–2828

7. Ratcliffe, A. (2000) Tissue engineering of vascular grafts. MatrixBiol. 19, 353–357

8. L’Heureux, N., Germain, L., Labbe, R., and Auger, F. A. (1993)In vitro construction of a human blood vessel from culturedvascular cells: a morphologic study. J. Vasc. Surg. 17, 499–509

9. L’Heureux, N., McAllister, T. N., and de la Fuente, L.M. (2007)Tissue-engineered blood vessel for adult arterial revasculariza-tion. N. Engl. J. Med. 357, 1451–1453

10. L’Heureux, N., Dusserre, N., Konig, G., Victor, B., Keire, P.,Wight, T. N., Chronos, N. A., Kyles, A. E., Gregory, C. R., Hoyt,G., Robbins, R. C., and McAllister, T. N. (2006) Humantissue-engineered blood vessels for adult arterial revasculariza-tion. Nat. Med. 12, 361–365

11. Lepidi, S., Grego, F., Vindigni, V., Zavan, B., Tonello, C., Deriu,G. P., Abatangelo, G., and Cortivo, R. (2006) Hyaluronan biode-gradable scaffold for small-caliber artery grafting: preliminaryresults in an animal model. Eur. J. Vasc. Endovasc. Surg. 32, 411–417

12. Lepidi, S., Abatangelo, G., Vindigni, V., Deriu, G. P., Zavan, B.,Tonello, C., and Cortivo, R. (2006) In vivo regeneration ofsmall-diameter (2 mm) arteries using a polymer scaffold. FASEBJ. 20, 103–105

13. National Research Council (1996) Guide for the Care and Use ofLaboratory Animals, National Academy Press, Washington, DC, USA

14. Vescovo, G., Scannapieco, G., Spano, E., Calliari, I., Leprotti, C.,Serafini, F., Ambrosio, G. B., and Dalla Libera, L. (1996) Myosinlight chain kinase in vascular smooth muscle. An immunohisto-chemical study in normotensive and spontaneously hypertensiverats. BAM 6, 183–187

15. Alsberg, E., Anderson, K. W., Albeiruti, A., Rowley, J. A., andMooney, D. J. (2002) Engineering growing tissues. Proc. Natl.Acad. Sci. U. S. A. 99, 12025–12030

16. Mann, B. K., and West, J. L. (2001) Tissue engineering in thecardiovascular system: progress towards a tissue engineeredheart. Anat. Rec. 263, 367–371

17. Matsumura, G., Hibino, N., Ikada, Y., Kurosawa, H., andShin’oka, T. (2003) Successful application of tissue engineeredvascular autografts: clinical experience. Biomaterials 24, 2303–2308

18. Shin’oka, T., Matsumura, G., Hibino, N., Naito, Y., Watanabe,M., Konuma, T., Sakamoto, T., Nagatsu, M., and Kurosawa, H.(2005) Midterm clinical result of tissue-engineered vascularautografts seeded with autologous bone marrow cells. J. Thorac.Cardiovasc. Surg. 129, 1330–1338

19. Weinberg, C. B., and Bell, E. (1986) A blood vessel constructedfrom collagen and cultured vascular cells. Science 231, 397–400

20. Joddar, B., Ibrahim, S., and Ramamurthi, A. (2007) Impact ofdelivery mode of hyaluronan oligomers on elastogenic re-sponses of adult vascular smooth muscle cells. Biomaterials 28,3918–3927

21. Joddar, B., and Ramamurthi, A. (2006) Elastogenic effects ofexogenous hyaluronan oligosaccharides on vascular smoothmuscle cells. Biomaterials 27, 5698–5707

22. Turner, J., Kielty, C. M., Walker, M. G., and Canfield, A. E.(2004) A novel hyaluronan-based biomaterial (Hyaff-11) as ascaffold for endothelial cells in tissue engineered vascular grafts.Biomaterials 25, 5955–5964

23. Remuzzi, A., Mantero, S., Colombo, M., Morigi, M., Binda, E.,Camozzi, D., and Imberti, B. (2004) Vascular smooth musclecells on hyaluronic acid: culture and mechanical characterizationof an engineered vascular construct. Tissue Eng. 10, 699–710

24. Tonello, C., Zavan, B., Cortivo, R., Brun, P., Panfilo, S., andAbatangelo, G. (2003) In vitro reconstruction of human dermalequivalent enriched with endothelial cells. Biomaterials 24, 1205–1211

25. Ross, J.J., and Tranquillo, R. T. (2003) ECM gene expressioncorrelates with in vitro tissue growth and development in fibringel remodeled by neonatal smooth muscle cells. Matrix Biol. 22,477–90

26. Ye, Q., Zund, G., Jockenhoevel, S., Hoerstrup, S. P., Schoeber-lein, A., Grunenfelder, J., and Turina, M. (2000) Tissue engi-neering in cardiovascular surgery: new approach to developcompletely human autologous tissue. Eur. J. Cardiothorac. Sur.17, 449–454

27. Zhang, W. J., Liu, W., Cui, L., and Cao, Y. (2007) Tissueengineering of blood vessel. J. Cell. Mol. Med. 11, 945–957

28. Kerdjoudj, H., Moby, V., Berthelemy, N., Gentils, M., Boura, C.,Bordenave, L., Stoltz, J. F., and Menu, P. (2007) The ideal smallarterial substitute: Role of cell seeding and tissue engineering.Clin. Hemorheol. Microcirc. 37, 89–98

Received for publication January 31, 2008.Accepted for publication March 13, 2008.

2861NEOARTERIES GROWN IN VIVO