Vitrified Human Umbilical Arteries as Potential Grafts for Vascular Tissue Engineering

Abstract

Background

The development of a biological based small diameter vascular graft (d < 6 mm), that can be properly stored over a long time period at − 196 °C, in order to directly be used to the patients, still remains a challenge. In this study the decellularized umbilical arteries (UAs) where vitrified, evaluated their composition and implanted to a porcine model, thus serving as vascular graft.

Methods

Human UAs were decellularized using 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) and sodium dodecyl sulfate (SDS) detergents. Then, vitrified with vitrification solution 55 (VS55) solution, remained for 6 months in liquid nitrogen and their extracellular matrix composition was compared to conventionally cryopreserved UAs. Additionally, total hydroxyproline, sulphated glycosaminoglycan and DNA content were quantified in all samples. Finally, the vitrified umbilical arteries implanted as common carotid artery interposition graft to a porcine animal model.

Results

Decellularized and vitrified UAs characterized by proper preservation of extracellular matrix proteins and tissue architecture, whereas conventionally cryopreserved samples exhibited a disorganized structure. Total hydroxyproline content was preserved, although sulphated glycosaminoglycan and DNA contents presented significantly alterations in all samples. Implanted UAs successfully recellularized and remodeled as indicated by the histological analysis.

Conclusion

Decellularized and vitrified UAs retained their structure function properties and can be possible used as an alternative source for readily accessible small diameter vascular grafts.

Electronic supplementary material

The online version of this article (10.1007/s13770-020-00243-x) contains supplementary material, which is available to authorized users.

Introduction

Nowadays, several types of vascular grafts have been tested in vascular surgery. These grafts can be used as replacement of damaged vessels, occurred by a wide variety of human disorders. Among them, the cardiovascular disease (CVD) is one of the most lethal disorders globally [1]. Especially in Europe, the CVD is the primary cause of mortality in ages under 65 years, thus resulting in more than 667,000 deaths representing the 29% of all deaths, worldwide. It is estimated that 23.3 millions of deaths due to CVD will be occurred until 2030 [1]. As first line treatment is the use of small diameter vascular grafts (inner diameter < 6 mm) accompanied with anticoagulation treatment.

Saphenous vein (SV) is the most common autologous vascular graft, used in coronary artery bypass, although only 20% of patients with CVD have satisfactory vessels [2]. Alternative to SV, synthetic vascular grafts derived either from Dacron and expanded polytetrafluorethylene (ePTFE) have been applied in vascular surgery since 1960 [3, 4]. However, these vessel replacements are accompanied by severe adverse reactions including host immune reaction, thrombus formation and calcification. As a consequence, the patency rate of these grafts is decreased to 54% after 5 years of implantation, and new surgical procedure is needed [5]. Moreover, synthetic grafts do not possess any in vivo remodelling ability when compared to biological vascular grafts, thus cannot be easily applied in pediatric patients [6]. Until now, several approaches have been used in order to develop proper vascular grafts, including decellularization of natural matrices, glutaraldehyde fixation and 3D bio-printing [7, 8].

Under this scope, the human umbilical cord with its containing arteries might be a good candidate for the development of small diameter vascular grafts. The human umbilical cord is consisted of two arteries and one vein and its length is 60–120 cm [9, 10]. The umbilical vein is responsible for the transportation of oxygen and nutrients to the fetus, while the arteries are transporting non oxygenic blood back to placenta. According to the literature the use of human umbilical vein as potential vascular grafts has been described since 1950 [11]. However, severe complications have been reported when UV was used as a vascular graft, such as handling difficulties, development of intima hyperplasia due to incompliance and impaired patency rate [11]. On the other hand, the human umbilical arteries (UAs) didn’t get the full attention by the research society. The human (UAs) have a lumen diameter of 1–3 mm and can be easily isolated without any invasive procedure. In addition, UAs unlike to UVs are characterized by defined vessel layers (tunica intima, media and adventitia), are easier to be handled by the surgeons, and may represent similar compliance as the coronary arteries. The above data in combination with the fact that UAs are characterized by in vivo remodeling properties, make them ideal candidates for small diameter vascular graft engineering. The study of human umbilical arteries (UAs) as potential vascular scaffolds have been described by few groups earlier [12, 13]. Most research groups applied the decellularization method to UAs in order to serve as vascular grafts [12, 13]. However, the decellularization process demands 3–5 days following by a 5 days sterilization cycle and the scaffolds could be applicable for implantation even after 2 weeks [12, 13]. In order the vascular grafts to be fully functional, long culture period with specialized cellular populations (endothelial cells and vascular smooth muscle cells) in vessel bioreactors, is needed. This prolonging manufacturing time could be life threating for high risk patients with CVD who demand immediate vascular transplantation. A possible solution to address this issue might be the vitrification of the decellularized scaffolds, while the grafts could ideally be ready for transplantation upon demanding. In contrast to unsatisfactory conventional cryopreservation methods that were applied in vascular grafts in the past, the vitrification approach might be a more feasible method [14].

The vitrification method initially applied to the proper preservation of human oocytes [15]. Vitrification solutions are consisting of different cryoprotective agents, which can prevent the ice crystal formation and properly preserve the tissue extracellular matrix (ECM) [14]. Moreover, the vitrification approaches have been applied to vessels such as carotid artery and saphenous vein with successful results [16].

In this study, the human UAs were decellularized according to a previously published decellularization protocol [13], followed by vitrification and storage at liquid nitrogen for a time period of 6 months. As control groups, stored UAs with conventional cryopreservation method, were used. Then, histological, biochemical and biomechanical analysis performed in UAs in order to evaluate properly the composition of the produced scaffolds. Finally, the functional assessment of vitrified UAs as vascular grafts were tested by vascular implantation experiments in a porcine model. More specifically, replacement of porcine common carotid artery with the vitrified UAs was performed and remained for 30 days. Then, the grafts were explanted, analyzed histologically and compared with the initial harvested porcine common carotid artery.

Materials and methods

Reception of human umbilical cords

Human umbilical cords (n = 10, l = 30 cm) were collected from end term normal and caesarian deliveries (gestational ages 38–40 weeks) by experienced midwives after signed informed consent by the mothers. The informed consent has been signed by the mothers before the delivery, is in accordance with the ethical standards of Greek National Ethical Committee and accepted by the ethical board of Biomedical Research Foundation Academy of Athens (Ref 2578/32). Detailed description of the umbilical cord samples used in this study, is provided in Table S1–S3. Then, the human umbilical cords were transported properly to the Hellenic Cord Blood Bank (HCBB) and processed immediately or stored maximum for 24 h at 4 °C.

Isolation of human umbilical arteries

The human umbilical cords briefly rinsed in Phosphate Buffer Saline 1 × (PBS 1×, Gibco, Life Technologies, Grand Island, NY, USA) for removal of excess blood and blood clots. Then, the Wharton’s Jelly tissue was removed with the use of sterile surgical instruments and the human UAs were isolated in a mean length of 20 cm. Each UA were cut into 7 segments of 2 cm and kept into PBS 1× (Gibco) supplemented with 10 U/ml Penicillin (Gibco) and 10 μg/ml Streptomycin (Gibco) until further use. Specifically, from each UA, the 6 UA segments were initially submitted to decellularization and from those, 4 segments were vitrified and stored for a time period of 1, 2, 4 and 6 months. The remaining decellularized UA segment was conventionally cryopreserved and used as positive control and the remaining non decellularized (native) UA segment was used as negative control (Fig. S1).

Decellularization of human umbilical arteries

The UA segments (n = 60, l = 2 cm) were initially decellularized, according to a previous described protocol [13]. Briefly, the UA segments were submitted in CHAPS buffer, pH 7 (8 mM CHAPS, 1 M NaCl and 25 mM EDTA in PBS 1×, Gibco) for 22 h, followed by incubation in SDS buffer pH 7(1.8 mM SDS, 1 M NaCl and 25 mM EDTA in PBS 1×, Gibco) for another 22 h. Finally, the UA segments were incubated in α-MEM supplemented with 40% v/v FBS for 48 h at 37 °C under continuous agitation.

Cryopreservation of human umbilical arteries

The decellularized UA segments (n = 40, l = 2 cm) were immersed for 6 times of 15 min into precooled vitrification solution (VS55) consisting of 3.10 M DMSO, 3.10 M formamide and 2.21 M 1,2-propanediol in Euro-Collins solution (Gibco). Then, the UA segments were placed into cryotubes containing the precooled vitrification solution and rapidly cooled (43 °C/min) to − 100 °C, followed by slow cooling (3 °C/min) to − 135 °C. The next day, the cryotubes contained the UA segments were stored in liquid nitrogen at − 196 °C for a time period of 1, 2, 4 and 6 months.

Regarding the conventional cryopreservation method, the UA segments (n = 10, l = 2 cm) were initially placed in cryotubes with precooled cryopreservation solution for 20 min at 4 °C and then placed in Mr. Frosty containers and frozen overnight at slow cooling rate of 1 °C/min until reaching − 80 °C. The cryopreservation solution consisted of 40% v/v α-MEM, 50% v/v FBS and 10% v/v DMSO, 2.5% w/v chondroitin sulfate (Gibco). The cryotubes finally stored into liquid nitrogen at − 196 °C for 1 month.

Thawing procedure of umbilical arteries segments

In order to evaluate the efficacy of vitrification method in storage of UAs, the thawing procedure of vitrified UA segments performed after 1, 2, 4 and 6 months. The cryotubes contained the vitrified UA segments were removed from liquid nitrogen at the desired time points and immediately immersed in waterbath (Memmert, Schwabach, Germany) at 37 °C. Then, the UA segments removed carefully from the cryotubes, placed in 50 ml polypropylene falcon tube in order to remove the vitrification solution in a step wise manner. Briefly, the vitrified UA segments were placed in solution contained 1 M Dextran in PBS 1 × (Gibco) and centrifuged. The supernatant was discarded and a new solution contained 0.5 M Dextran in PBS 1 × (Gibco) was added followed by centrifugation. Finally, the UA segments were placed in PBS 1 × (Gibco) and centrifuged for a last time. The centrifugation was performed at 500 g for 6 min. The cryotubes contained the conventional cryopreserved UA segments were removed from the liquid nitrogen and immediately immersed in water bath at 37 °C until completely thawed. Then, UA segments placed in 50 ml polypropylene falcon tube with PBS 1× and centrifuged at 500 g for 6 min.

Histological analysis of umbilical arteries segments

UA segments (n = 70, l = 2 cm) from all experimental conditions were submitted to histological analysis. For this purpose, the UA segments were fixed in 10% v/v neutral formalin (Sigma-Aldrich, Merck, Darmstadt, Germany), emended in paraffin and cut into 5 μm sections. Then, Hematoxylin and Eosin (H&E, Sigma-Aldrich), Sirius Red (Sigma-Aldrich) and Orcein stain (Sigma-Aldrich) were performed in order to determine the presence of cell nuclei, collagen and elastin respectively. Images were acquired with Leica DM L2 light microscope (Leica Microsystems, Weltzar, Germany) and processed with ImageJ 1.46r (Wane Rasband, National Institutes of Health, Bethesda, MD, USA).

Scanning electron microscopy

UA segments (n = 12, l = 5 mm) were processed for scanning electron microscopy (SEM). Specifically, SEM analysis was performed in non decellularized (n = 3), decellularized (n = 3), vitrified and stored for 6 months (n = 3) and conventionally cryopreserved (n = 3) UA segments for 1 month. Briefly, the UA segments were fixed with 1% v/v glutaraldehyde solution (Sigma-Aldrich) in 0.1 M sodium cacodylate buffer (Sigma-Aldrich) for 30 min. Then, the samples were washed with distilled water for 5 min and dehydrated with 20 min exchanges in 70% v/v, 80% v/v, 95% v/v aqueous ethanol and absolute ethanol. Finally, the UA segments were placed in hexamethyldisilazane solution (Sigma-Aldrich) for 10 min, air dried and sputter-coated with gold (Cressington Sputter, Coater 108 auto, Watford, United Kingdom). The samples were examined with Phillips XL-30 scanning electron microscope (Phillips, FEI, Hillsboro, OR, USA).

DNA quantification

For the DNA quantification analysis, the UA segments (n = 40) were digested with a lysis buffer contained 0.1 M Tris pH 8, 0.2 M NaCl, 5 mM EDTA and 25 mg/ml Proteinase K (Sigma-Aldrich), incubated for 12 h at 55 °C. After the tissue lysis, inactivation Proteinase K (Sigma-Aldrich) at 60 °C for 5 min of was occurred. The DNA was isolated, diluted in 50 μl of RNA-se free water (Sigma-Aldrich) and measured photometrically at 260 to 280 nm. Furthermore, the DNA content of all samples was analyzed by electrophoresis with 1% w/v agarose gel. Images were acquired and processed with UVITEC Imaging System (Cleaverscientific, Warwickshire, United Kingdom).

Quantification of collagen and sulphated glysoaminoglycan content

UA segments including non decellularized, decellularized, vitrified (stored for 6 months) and conventionally cryopresereved samples, (n = 70) were digested in 125 μg/ml papain buffer (Sigma-Aldrich) at 60 °C for 12 h and the total collagen amount in each sample was quantified with the Hydroxyproline Assay Kit (Sigma-Aldrich) according to manufacturer’s instructions. Briefly, the samples were hydrolyzed with 12 M HCl, dried and incubated with Chloramine T/oxidation buffer and DMAB reagent. Finally, the hydroxyproline content was determined photmetrically at 560 nm by interpolation to hydroxyproline standard curve To quantify the sulfated glycosaminoglycan (sGAG) content, samples (n = 70) were digested in 125 μg/ml papain buffer, followed by addition of dimethyleneblue (Sigma-Aldrich) and photometrically quantification at 525 nm. The sGAG content was calculated by interpolation to standard curve. Dilutions of 3, 6, 12, 25, 50, 100 and 150 μg/ml chondroitin sulfate were used for the development of the standard curve.

Biomechanical testing

Biomechanical analysis in non decellularized (n = 5, l = 2 cm), decellularized (n = 5, l = 2 cm), vitrified (n = 5,, l = 2 cm) and conventionally cryopreserved (n = 5, l = 2 cm) UAs was performed in order to evaluate their mechanical properties. Uniaxial testing on longitudinal direction strips was performed in all samples. Mechanical analysis was conducted in a Zwick/Roell tensile tester (model Z 0.5, Zwick GmbH & Co. KG, Ulm, Germany) equipped with a 200 N load cell. The longitudinal strips of all samples were tested at room temperature, and continuously sprayed with PBS 1× (Gibco), for rehydration. Then, the strips were clamped at their ends, using sandpaper, under zero strain on the tensile tester, which produced sample preloading of 0.005 N before the operating program started to collect the data. During mechanical analysis, the samples were preconditioned for 10 cycles at a rate of 10 mm/min. Sample extension (Δl, in mm) and corresponded generated load (F, in Newtons) were converted to engineering strain (ε), and engineering stress (σ, in MPa). Elastin (El-E) and collagen (Col-E) phase slope, transition stress (σTrans) and strain (εTrans), ultimate tensile strength (σUTS) and failure strain (εUTS) were used in order to analyze the stress–strain behavior of each sample.

Cytotoxicity assay

The cytotoxicity effects of decellularization, vitrification or conventional cryopreservation solutions were evaluated by contact and tissue extracts cytotoxicity assay. Contact cytotoxicity test was performed in six-well plates (Orange Scientific, Braine-l’Alleud, Belgium), where decellularized, vitrified and conventionally cryopreserved UA segments (n = 15) were placed as rings. Human mesenchymal stromal cells, passage 3 (MSCs P3) derived from Wharton’s Jelly tissue were seeded in each well at a density of 1 × 10⁵ cells until reaching 80% confluency. Plates were incubated at 37 °C with 5% CO₂ for 48 h. Native UA segments seeded with MSCs P3 and MSCs P3 alone, were used as negative control. As positive control MSCs P3 with 1.5 mM SDS (Gibco) was used.

Tissue extract cytotoxicity assay was performed by digesting the non-decellularized, decellularized, vitrified and conventionally cryopreserved UA segments (n = 20) with 25 μg/ml Proteinase K in PBS 1× (Gibco) overnight at 56 °C. Then, the Proteinase K was heat inactivated at 60 °C and the occurred tissue extracts were placed with 5 × 10⁴ MSCs P3 in each well with 1 ml of regular culture medium. The regular culture medium consisted of α-MEM (Gibco) supplemented with 15% v/v FBS (Gibco), 10 U/ml Penicillin (Gibco), 10 μg/ml Streptomycin (Gibco) 2 mM l-glutamine (Gibco). The plates were incubated at 37 °C with 5% CO2 for 48 h.

After the incubation period, the samples washed with PBS 1 × and fixed in 10% v/v neutral buffered formalin for 10 min, followed by Giemsa staining (Sigma-Aldrich) for 5 min. Finally, the six-well plated washed with distilled water, air-dried and examined under light microscope DM L2 (Leica Microsystems). The images were processed with ImageJ 1.46r (Wane Rasband, National Institutes of Health).

Implantation of vitrified umbilical arteries to porcine model

The functional assessment of vitrified UAs (stored for 6 months, n = 3, l = 5 cm) as vascular graft was conducted by implantation experiments in porcine model (n = 3), 3–4 months old, weighting 40–50 kg. All care and handling of the animals were provided according to the Guide for the Care and Use of Laboratory Animals of BRFAA, conformed to the Directive 2010/63/EU of the European Parliament, were in accordance with the ethical standards of the Greek National Ethical Committee and approved by our Institution’s ethical board (No 5959). Specifically, replacement of porcine common carotid artery with the vitrified UA was performed. The animals anesthetized and given clopidogrel (30 mg/kg) as anti-platelet agent 30 min before the surgery. Then, the animals were opened and the common carotid artery was exposed under standard sterile conditions. The common carotid artery was cross-clamped and divided between proximal (close to the heart) and distal (close to the head) areas. The vitrified umbilical artery was inserted, replaced the common carotid artery, and finally sutured. The excluded segment of porcine common carotid artery was served as negative control. After confirmation of normal blood flow through the graft, the clamps removed and the wound was closed. The animals were recovered from the surgical procedure and maintained without any anti-coagulation or anti-platelet treatment.

The implants were remained for a time period of 30 days, then explanted and submitted for histological analysis. Vitrified UAs prior to implantation, porcine common carotid artery and midpoints of harvested grafts were fixed in 10% v/v neutral-formalin (Sigma-Aldrich), paraffin embedded and sectioned at 5 μm. H&E (Sigma-Aldrich) and Orcein stain (Sigma-Aldrich) were performed to the above vessels. In addition, indirect immunofluorescence for elastin was performed. Briefly, slides were deparaffinized, rehydrated and blocked. Sections incubated with monoclonal antibody against elastin (1:2000, Sigma-Aldrich) followed by FITC-conjugated mouse IgG (1:80, Sigma-Aldrich), dehydrated and mounted. The images were obtained with LEICA SP5 II confocal microscope with LAS Suite v2 software (Leica Microsystems). 3D reconstructions of the obtained images were performed with Volocity 6.1.2 (PerkinElmer Inc., Waltham, MA, USA).

Morphometric analysis of vitrified UA before and after implantation

Inner (lumen) diameter and thickness of the vessel wall were measured as the average value of proximal, middle and distal points of vitrified UAs before and after implantation to the animals and compared to porcine common carotid artery. The site of UA close to placenta was defined as the proximal point, the area close to fetus was defined as the distal point and the intermediate point between them was defined as the middle point. As proximal point of implanted UA and carotid artery was defined the site of the vessel close to the heart, as distal point the site of the vessel close to head and the intermediate space was the middle point. The inner diameter was measured based on the histological sections stained with H&E and Orcein stain. In addition, the number of elastin fibers were measured manually in triplicate in vitrified, implanted UA and common carotid artery by two independent observers, using the with ImageJ 1.46r (Wane Rasband, National Institutes of Health) and presented as total number, number of elastic fibers at the tunica media (media layer) and tunica adventitia (outer layer) of the vessel grafts.

Statistical analysis

Statistical analysis was performed by using Graph Pad Prism v 6.01 (GraphPad Software, San Diego, CA, USA). Comparisons of total collagen, sGAG contents and morphometric data between all samples were performed with Welch’s t test. Comparison of DNA content and biomechanical results between all samples was performed with unpaired non-parametric Kruskal–Wallis test. Statistical significant difference between group values was considered when p value was less than 0.05. Indicated values are mean ± standard deviation.

Results

Histological analysis of the human umbilical arteries

The evaluation of the decellularization and vitrification approaches was performed by histological stainings (H&E, Sirius Red, Orcein stain) and SEM. More specifically, the decellularized UA segments characterized by the absence of cell and nuclei materials as indicated by the H&E. Furthermore, decellularized and vitrified UA segments presented good preservation of the extracellular matrix (ECM) components such as collagens and elastin, as indicated by Sirius Red and Orcein stains (Fig. 1). Elastin fibers were observed only at tunica intima (inner layer close to the lumen), accompanying by a few at tunica media of UAs. No elastin fibers were observed at tunica adventitia of UA segments.

Fig. 1.

A–P Histological analysis of non decellularized, decellularized, vitrified and conventionally cryopreserved human umbilical arteries A, E Non decellularized (negative control) UA stained with H&E, I Sirius Red and M Orcein stain. B, F Decellularized UA stained with H&E, J Sirius Red and N Orcein stain. C, G Vitrified UA and stored at − 196 °C for 6 months stained with H&E, K Sirius Red and O Orcein stain. D, H Conventionally cryopreserved UA (positive control) stained with H&E, L Sirius Red and P Orcein stain. Images A–D are presented with original magnification 2.5×, scale bars 500 μm. Images E-P are presented with original magnification 10×, scale bars 100 μm. L: Lumen, TM: Tunica Media, TA: Tunica Adventitia

Vitrified UA segments that were stored for 1,2,4 and 6 months, characterized by similar ECM formation, indicating the successful preservation at − 196 °C (Fig. S2). On the other hand, the conventionally cryopreserved UA segments presented a totally destructed ECM and a quite altered vessel structure. In addition, SEM images revealed the preservation of the ECM of the decellularized and vitrified UA segments, while the ECM of the conventionally cryopreserved UA segments characterized by total destruction of collagen fibers, due to ice crystal formation (Figs. 1, 2).

Fig. 2.

SEM images of human umbilical arteries. A Non decellularized (negative control) UA. B Conventionally cryopreserved (positive control) UA. C Decellularized UA. D Vitrified UA and stored at − 196 °C for 6 months. Original magnification 1000×, scale bars 10 μm. L: Lumen, TM: Tunica Media

DNA quantification

In order to evaluate the success of decellularization method in human umbilical arteries, total DNA quantification was performed. In addition, DNA content was quantified in vitrified (stored for 6 months) and conventionally cryopreserved UA segments. The DNA amount (ng DNA/mg Tissue) of UA before the decellularization was 1637 ± 119 ng DNA/mg dry tissue, while after the decellularization was 82 ± 10 ng DNA/mg dry tissue. Additionally, vitrified and conventionally cryopreserved UAs characterized by similar amounts of DNA with the decellularized samples. Specifically, the DNA content of vitrified and conventionally cryopreserved UA segments was 79 ± 6 and 72 ± 4 ng DNA/mg dry tissue, respectively (Fig. 3). Statistically significant differences were observed in DNA content between native and decellularized, vitrified and conventionally cryopreserved UAs (p < 0.005). Furthermore, the DNA of decellularized, vitrified and conventionally cryopreserved UA segments was analyzed by DNA electrophoresis in 1% w/v agarose gel. No bands were observed of the aforementioned samples whereas the native sample characterized by a dense DNA band (Fig. S3).

Fig. 3.

Biochemical and DNA analysis in UA segments. A Total Hydroxyproline content of native, decellularized, vitrified (and stored for 6 months), and conventionally cryopreserved UA segments. Statistical significant difference in total hydroxyproline content was observed between native, decellularized, vitrified and conventionally cryopreserved UA segments, p < 0.05. B Sulphated GAG content of native, decellularized, vitrified (and stored for 6 months), and conventionally cryopreserved UA segments. Statistical significant difference in sGAG content was observed between native, decellularized, vitrified and conventionally cryopreserved UA segments, p < 0.001. C DNA content of native, decellularized, vitrified (and stored for 6 months), and conventionally cryopreserved UA segments. Statistical significant difference in DNA content was observed between native, decellularized, vitrified and conventionally cryopreserved UA segments, p < 0.05. NATIVE: Native-Non decellularized, DECEL: Decellularized, VITRIF 6 TH M: Vitrified (and stored for 6 months), CRYOPR: Conventionally cryopreserved UA segments

Biochemical analysis

Total hydroxyproline content of decellularized, vitrified and conventionally cryopreserved UA segments was higher when compared to native samples. The amount of hydroxyproline in native, decellularized, vitrified and conventionally cryopreserved was 62 ± 3, 100 ± 10, 84 ± 13, 62 ± 14 μg hydroxyproline/mg dry tissue, respectively (Fig. 3 and S4). Statistically significant decrease in total hydroxyproline content was observed between decellularized, vitrified and conventionally cryopreserved UA segments (p < 0.05). Regarding to sulphated GAG content, native, decellularized, vitrified and conventionally cryopreserved UA segments characterized by 2.1 ± 0.1, 1.5 ± 0.1, 1.4 ± 0.1 and 0.7 ± 0.2 μg sGAG/mg dry tissue, respectively (Fig. 3 and S4). Statistical significant decrease in sGAG content was observed between native and all samples (p < 0.001). Additionally, decrease in sGAG content, with statistically significant value, was observed between decellularized, vitrified and conventionally cryopreserved UA segments (p < 0.05).

Biomechanical analysis of UAs

Uniaxial testing of non decellularized, decellularized, vitrified and conventionally cryopreserved arteries was performed in order to determine their stress–strain behavior. As shown in Table 1, decellularized, vitrified and conventionally cryopreserved UAs presented to be stiffer than non decellularized samples. Specifically, statistically significant differences were observed in transition stress, ultimate tensile strain (p < 0.05), whereas no statistically differences were observed in transition strain, ultimate tensile stress, elastin and collagen phase slope.

Table 1.

Biomechanical analysis of non decellularized, decellularized, vitrified and conventionally cryopreserved UAs

Biomechanical Testing Parameters	Non decellularized-native UAs	Decellularized UAs	Vitrified UAs—stored for 6 months	Convent. Cryopr. UAs	p value
Transition stress − σΤrans (MPa)	0.11 ± 0.02	0.19 ± 0.03	0.20 ± 0.03	0.19 ± 0.03	0.011
Transition strain − εΤrans (−)	0.33 ± 0.08	0.36 ± 0.08	0.36 ± 0.07	0.40 ± 0.08	0.394
Ultimate tensile stress σUTS (MPa)	0.96 ± 0.04	1.23 ± 0.06	1.26 ± 0.04	1.27 ± 0.04	0.362
Ultimate tensile strain − εUTS	0.76 ± 0.13	0.96 ± 0.14	0.81 ± 0.14	1.07 ± 0.15	0.024
Elastin phase slope − El-E (MPa)	0.09 ± 0.02	0.13 ± 0.07	0.14 ± 0.04	0.14 ± 0.05	0.182
Collagen phase slope − Col-E (MPa)	3.19 ± 0.60	3.34 ± 0.50	3.54 ± 0.64	3.88 ± 0.91	0.500

Statistical significant were observed in transition stress (σ_Τrans), ultimate tensile strain (ε_UTS), p < 0.05

Convent. Cryopr; UAs conventionally cryopreserved UAs

Cytotoxiciy assay

Evaluation of the cytotoxic effect that may be possibly caused by the remaining decellularization or cryopreservation solutions was performed by contact and tissue extract cytotoxicity assay. There was no observable evidence of cytotoxicity in native, decellularized, vitrified and conventionally cryopreserved UA segments, compared to positive control, when contact cytotoxicity assay performed (Fig. 4). Further assessment for any possible residual remnants of decellularization or cryopreservation solutions was conducted by extract cytotoxicity assay. No evidence of any toxic effect was observed in all samples, when compared to positive control group (Fig. 4).

Fig. 4.

A–J Cytotoxicity assay was conducted by seeding MSCs P3 on human UA samples. A–H Conatact and tissue extract cytotoxicity assays on non decellularized (A, B), decellularized (C, D), vitrified-stored for 6 months (E, F), conventionally cryopreserved samples (G, H). Non decellularized (native) sample served as negative control and MSCs P3 with 1.5 mM SDS served as positive control. Images were acquired with original magnification 10×, scale bars 100 μm

Implantation of vitrified umbilical artery

The structure–function evaluation of the vitrified UAs was performed by implantation to porcine model. The vitrified UA characterized by sufficient suture retention strength for the current procedure without causing any difficulties to the surgeon (Fig. 5). All animals survived postoperatively (n = 3). However, at the day of the graft harvesting, clot and thrombosis formation were observed in all vessel grafts. This phenomenon was caused due to the lack of an organized endothelium and the exposure of collagenous vessel wall to platelets.

Fig. 5.

Replacement of porcine common carotid artery with vitrified umbilical artery. A, B Vitrified umbilical artery before and after implantation to porcine model

The histological examination of the explanted vessel grafts by H&E, showed the successful cell infiltration and alteration of implanted UA’s ECM similar to porcine common carotid artery (Fig. 6). Orcein stain revealed the development of elastin fibers in the explanted vessel graft (Fig. 6). Specifically, the explanted vascular grafts characterized by internal elastic lamina formation in the tunica intima, curled elastic fibers in tunica media and a high number of tightly compacted elastic fibers in tunica adventitia. In addition, presence of vasa vasorum was observed in UA after 30 days of implantation, similar to native porcine carotid artery (Fig. S5).

Fig. 6.

A–I Histological analysis of the implanted umbilical artery in the porcine model. A, D and G Vitrified UA before implantation stained with H&E and Orcein staining. B, E and H Implanted UA in porcine model stained with H&E and Orcein staining. C, F and I Native porcine common carotid artery stained with H&E and Orcein staining. Images A–F acquired with original magnification 10×, scale bars 100 μm. Images G–I acquired with original magnification 40×, scale bars 50 μm

The developed internal elastic lamina of the implanted UA, can be observed better by 3D reconstruction models, where indirect immunofluorescence again elastin was performed (Fig. 7). The vitrified UAs (before implantation) completely lacked of well-organized elastic fibers, a structure that is characteristic in elastic vessels such as the common carotid artery. Internal elastic lamina was produced and specifically observed to the explanted vessels as indicated by the immunofluorescence results. Porcine common carotid artery is characterized by similar formations, with the only exception, the presence of more straightforward elastic fibers in tunica media, and a greater number of tightly compacted fibers in tunica adventitia when compared with those of explanted vessel grafts.

Fig. 7.

3D reconstruction models of elastin immunofluorescence in vitrified, implanted UA and porcine common carotid artery. A Indirect immunofluorescence against elastin on vitrified umbilical artery B, implanted umbilical artery and C porcine common carotid artery. Implanted UA and porcine common carotid artery characterized by internal elastic lamina as indicated by the arrows, while the vitrified UA totally lacked this formation. Original magnification 5×, scale bars 500 μm

Furthermore, the number of elastic fibers of tunica media and adventitia were able to be measured microscopically in each sample. The number of elastic fibers in tunica media and adventitia in vitrified UA (prior to implantation) was 9 ± 2 and 2 ± 1 respectively (Fig. 8). Implanted UAs were consisted of 19 ± 3 in tunica media and 14 ± 3 in tunica adventitia elastic fibers (Fig. 8). In porcine common carotid artery, the number of elastic fibers was 15 ± 2 for tunica media and 19 ± 2 for tunica adventitia (Fig. 8). Statistically significant differences were observed in the numbers of elastic fibers both in tunica media and adventitia in all samples (p < 0.05) Total number of elastic fibers from both tunica media and adventitia layers for vitrified UA (prior to implantation), implanted UA and porcine common carotid artery were 11 ± 2, 33 ± 4 and 34 ± 3 respectively (Fig. 8). Statistical significant difference in the total number of elastic fibers was observed between vitrified UA, implanted (p < 0.001) and vitrified UA, porcine carotid artery (p < 0.001), while no statistical difference was occurred between implanted and carotid artery. Morphometric analysis was also included measurement of inner (lumen) diameter and thickness in proximal, middle and distal points of all samples. The inner diameter and thickness of the vitrified UA under zero pressure conditions were 1.0 ± 0.1 mm and 0.5 ± 0.1 mm, respectively (Table S3, S4). After the implantation, the UA was distended until reaching 1.5 ± 0.7 mm for inner diameter, while the thickness didn’t alter significantly (0.5 ± 0.1 mm). Porcine common carotid artery characterized by an inner diameter of 2.1 ± 0.1 mm and thickness of 0.5 ± 0.1 mm, indicating no statistical difference when compared to vitrified UA before and after implantation (Fig. 8 and Table S3, S4).

Fig. 8.

Morphometric analysis of vessels. A Representable histological images of vitrified UA (before implantation, B implanted UA and C porcine common carotid artery stained with H&E that were used for inner diameter and thickness measurements. Original magnification 2.5×, scale bars 500 μm. D Inner (Lumen) Diameter of the vitrified UA (before implantation), implanted UA and porcine common carotid artery. Statistical significant difference in inner diameter between all samples, p < 0.05. E Thickness of the vitrified UA (before implantation), implanted UA and porcine common carotid artery. F Number of elastic fibers (tunica media) in vitrified UA (before implantation), implanted UA and porcine common carotid artery. Statistical significant difference was observed in the number of elastic fibers measured in tunica media between all samples, p < 0.001. G Number of elastic fibers (tunica adventitia) of vitrified UA before implantation, implanted UA and porcine common carotid artery. Statistical significant difference in the number of elastic fibers measured in tunica adventitia between all samples, p < 0.05. H Total number of fibers of vitrified UA before implantation, implanted UA and porcine common carotid artery. Statistical significant difference was observed in the number of elastic fibers between all samples, p < 0.001

Discussion

The successful development of functional small diameter vascular grafts is one of the milestones of the tissue engineering and regenerative medicine. Unlikely, to large diameter vascular grafts (d > 6 mm) that can efficiently be produced from polymer and synthetic materials (Dacron, ePTFe), development of fully functional small diameter vascular grafts is still a demanding process [7]. These grafts used mainly to coronary artery bypass surgeries as a treatment to CVD, one of the leading causes of death, worldwide, but until now are accompanied by a number of complications [1]. Nowadays, the gold standard strategy that is been followed by surgeons is the use of saphenous vein or synthetic vascular grafts. The most severe complications arising from the use of these grafts are thrombus formation and intima hyperplasia that can significantly low the patency rate within the 1st year of implantation [7]. Most research approaches are including the development of an endothelium layer to the graft prior to implantation, increasing its production period even more [7]. In addition, UVs has been used in the past as possible vascular conduits [11]. However, its low applicability during surgery, led the surgeons to abandon its use. On the other hand, UAs represent a better vessel conduit compared to UVs. UAs with lumen diameter 1–3 mm, can be isolated without invasive procedures, characterized by defined tunica intima, media and adventitia layers, similar compliance with coronary arteries and are easier to be handled by surgeons during the anastomosis performance. Furthermore UAs can be decellularized effectively in order to produce biologically based vascular grafts. Taking into consideration the above data, UAs represent good candidates for small diameter vascular graft engineering. To reduce the time needed for vessel implantation, the use of cryopreservation methods has been proposed. However, proper cryopreservation of vascular grafts has not been achieved at the desired level. The aim of this study was to evaluate initially the impact of vitrification approach to decellularized human umbilical arteries and secondly the functional assessment of the vitrified arteries in order to serve as vascular grafts.

Our results showed, that treatment of UAs with a 3 stage decellularization protocol using CHAPS and SDS detergents following by FBS incubation, produced a vascular graft free of any cellular populations while the initial ultrastructure of UA was retained. Furthermore, these produced vascular grafts were capable to properly be stored in liquid nitrogen, using the vitrification method, for a time period of 6 months. More specifically, histological analysis (H&E, Sirius Red, Orcein Stain and SEM) revealed the successful preservation of vessel’s morphological features, including tunica intima, media and adventitia and ECM proteins such as collagen and elastin both in decellularized and vitrified UAs. On the other hand, conventionally cryopreserved vascular grafts failed to retain their initial morphology and characterized by an extensive damage of the ECM. This altered vessel morphology in conventionally cryopreserved UAs was induced due to ice crystals developed during the cryopreservation process [14]. Indeed, the vitrification method offers sufficient protection to graft’s ECM at low temperatures rather than the traditional preservation methods [14]. Traditional preservation methods rely on the use of DMSO for cryoprotection, whilst the vitrification approach uses a combination of cryoprotective agents with high and low molecular weight, offering better internal and external protection of the graft’s ECM [14]. The above results seem to be similar with other studies that have been previously performed in different tissue engineered vascular grafts [16, 17].

Next step of this study was the biochemical analysis of the produced vascular grafts. For this purpose, total hydroxyproline, sGAG and DNA contents were quantified. Total hydroxyproline content was significantly higher in decellularized, vitrified and conventionally cryopreserved UAs when compared to non decellurized UAs. The fact that no statistical differences were observed between decellularized and vitrified UAs, while significant differences observed in decellularized, vitrified and conventionally cryopreserved UAs, further confirming the successful use of vitrification method. In addition, the elevated hydroxyproline levels observed in decellularized and vitrified vascular grafts might be explained due to the loss of water content and cell mass after decellularization process. Vitrification also has a sort of dehydration effect in tissue, thus lowering its water content even more. Similar results have been described by Luo et al. [18], where decellularization of heart valves was performed. Based on the knowledge that sGAGs are important tissue components and their removal may cause structural alterations, sGAG levels were determined. The sGAG levels of all vascular grafts were significantly lower in comparison to non decellularized samples. It is known that SDS, an anionic detergent, could remove sGAGs from tissues, through binding to their negatively charged sites [18]. Sulphated glycosaminoglycans are forming large macromolecules, the proteoglycans, which are responsible for tissue’s collagen orientation, In this case, removal of sGAGs might result to altered collagen orientation, thus causing further damage to tissue ECM. However, in the present study, regardless this significant decrease of sGAG content, no structural alteration in vascular grafts was observed as shown by the histological analysis (H&E, Sirius Red, Orcein stain and SEM).

In this study, biomechanical analysis of all vascular grafts was performed. Biomechanical analysis showed that UAs becoming stiffer after decellularization and vitrification compared to the native control. At higher strains, the decellularized, vitrified and conventionally cryopreserved UAs demonstrated increased extensibility compared to the native control. Statistically significant differences in the transition stress and ultimate tensile strain were observed among the different UA groups The above results seemed to be in accordance with the study of Tuan-Mu et al. [19]. On the other hand, Gui et al. [12] mentioned that, the stiffer behavior of UAs after decellularization may be occurred due to elastin removal. However, based in the histological analysis with Orcein stain, we didn’t observed any elastin depletion in UAs after decellularization and vitrification approaches, but only in sGAG content. Possibly, the removal of smooth muscle cells and sGAGs due to detergents used in decellularization, could be a reason for stiffer UAs [18]. After decellularization and vitrification, the UAs were characterized by the same amount of collagen, but with less amount of sGAGs, indicated alteration in the tissue architecture. Moreover, increase in hydroxyproline content, altered collagen orientation due to sGAGs removal and possibly crosslinking between collagen fibers, occurred after cryopreservation, could result to stiffer grafts [17]. Specifically, proteoglycans have been reported to contribute to the cross-linking of the collagen fibers, as well as their crimp. Therefore, removal of the proteoglycans could cause a reduction in the crimp of the collagen fibers, making them straighter and forcing them to bear more load at lower strains, thus, making the scaffolds demonstrate a stiffer behavior (characteristic of collagen). At higher strains, however, and after the collagen fibers have fully uncrimped, the removal of the cross-linking proteoglycans would allow the collagen fibers to move more freely, sliding past each other, further contributing to the tissue extensibility [19]. In spite of these differences, the gross biomechanical behavior of the scaffolds was not substantially affected by decellularization and vitrification, as indicated by the lack of any significant differences in the elastic and collagen phase slopes and ultimate tensile strength. The preservation of these parameters indicated that the decellularised and vitrified UA is a potential candidate for vascular reconstructions [12, 20]. Determination of the DNA content in acellular scaffolds is considered to be an important factor for their successful decellularization. The DNA content was reduced over 95% in decellularized, vitrified and conventionally cryopreserved vascular grafts. In addition, this small amount of DNA that was present in the produced vascular grafts were further analyzed by electrophoresis in 1% w/v agarose gel. No distinct bands were able to be observed in all samples except the non decellularized UAs that were characterized by a dense band. There are a number of studies where Picogreen assay has been performed for DNA quantification in decellularized vascular grafts with contradictory results. Picogreen assay could detect only double stranded DNA, while in decellularized scaffolds single and double stranded DNA could be existed due to detergent’s effect on cell removal [18]. In our study, the DNA was measured photometrically, and additional the presence of this low amount of DNA in vascular grafts (< 5%), was further analyzed by DNA electrophoresis, leading to more accurate results.

In addition, the cytotoxicity of the produced vascular grafts needed to be evaluated. The detergents used for decellularization and the cryoprotective agents applied to the cryopreservation approaches may have a cytotoxic effect to the vascular grafts. Increased cytotoxicity levels prevent cell adhesion, migration, differentiation and proliferation in the scaffold, thus becoming it unable to be used for tissue engineering approaches. When a cytotoxic scaffold is implanted to a living host, could cause immune rejection or might be life threaten for the patients. Among the reagents that were used for decellularization and vitrification, SDS and DMSO are exhibiting cytotoxicity. In our study, the vascular grafts were free of any cytotoxic effects as indicated by the migration activity of Wharton’s Jelly MSCs. Two experimental approaches used for the evaluation of cytotoxic effect. Contact cytotoxicity assay revealed the successful migration of Wharton’s Jelly MSCs towards to the scaffolds. In order to determine if SDS or DMSO remained in the scaffold, tissue extract assay was performed. These results didn’t show any alteration on migration activity of MSCs, suggesting the production of non-cytotoxic scaffolds.

The functionality of the vitrified UAs was assessed by implantation to porcine model as carotid artery interposition graft. After 30 days of implantation the vascular grafts were explanted and evaluated histologically. The explanted UA recellularized successfully by animal’s cellular population and characterized by an extended ECM remodeling. Physiologically, the UAs are muscular vessels bearing only a few elastin fibers. On the contrary, the implanted UAs characterized by a wide number of elastin fibers and moreover the development of internal elastic lamina as it was indicated by histological analysis (H&E, Orcein stain and 3D remodeling constructs). This is a quite remarkable feature of the implanted UAs, suggesting the successful remodeling from a muscular vessel towards to more elastic structure. Under this scope, and in order to better understand the underlying mechanism to this, determination of inner diameter, thickness and the number of elastic fibers was performed. The implanted UAs characterized by an extended inner diameter which was similar to the porcine’s common carotid artery. In addition, the thickness of vitrified UAs didn’t reveal any significant difference before and after implantation.

Moreover, the number of elastic fibers of the implanted UAs was higher in tunica media and lower in tunica adventitia, while carotid artery characterized by totally opposite results. However, the total number of elastin fibers between these vessels was quite similar. These can be explained by the fact, that in vivo remodeling process was on progress after 30 days of implantation and demanded more time for complete remodeling. This remodeling process is attributed mainly by two factors, blood flow and vascular smooth muscle cells (VSMCs). In this way, the UAs implanted in a highly blood flow environment, infiltrated by host’s cellular population and resulted to this remarkable altered of structure function properties. The typical blood pressure in common carotid artery is more than 50 mmΗg [21], while UAs characterized by a blood flow of 20–40 mmHg [22–25]. This might be the explanation for the remodeling of implanted UAs. Evidence of vasa vasorum was also observed in the implanted UAs. Vasa vasorum are considered to be small blood vessels that are involving in the nourishment and blood supply of large vessels such as aorta and carotid artery [26]. The presence of these formations in the implanted UAs showing the successful adaption of the graft from the host, thus holding promising results for the functionality and survivability of this graft, over time. Similar results were shown in the study of Gui et al. [12], where decellularized UA was used as abdominal rat aorta interposition graft. In their study, cell infiltration and alteratiom of ECM similar to our results was observed. In addition, Gui et al. [12] showed and extensive distensibility of the UA upon implantation to rats reached 4.5 mm, whereas in our study the highest inner diameter was 1.4 mm. Also, no significant diameter alteration of the implanted UA to porcine model was observed, unlike to Gui et al’s study [12]. These discrepancies in the results might be explained due to the different animal models, and different approaches in morphological features measurements that were applied between these two studies.

In conclusion, human umbilical arteries might be good candidates for the development of small diameter vascular grafts. These vessels can be isolated without any invasive procedures from the discarded umbilical cord after the gestation. Their length can be reached over 30 cm and no branches alongside are existed [10]. In addition, the UAs can be effectively decellularized, vitrified and stored over a long time period at − 196 °C. An ideal vascular graft could be non-cytotoxic, capable for cell infiltration, ECM remodeling, enough strength to withstand to blood flow, properties that UAs meet. Future studies will involve the in vitro recellularization of the produced vascular grafts combined with successful vitrification of these constructs. Interestingly, these vessels conduits can be produced, repopulated with patient’s own cells, properly stored and implanted back to the patient upon demanding without any time loss, bringing the personalized medicine one step closer to its clinical application.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Compliance with ethical standards

Conflicts of interest

The authors declare no conflict of interest for this study.

Ethical statement

The study protocol involved the use of human umbilical cords. All human umbilical cords were accompanied by informed consent signed by the mothers before the delivery. The informed consent was in accordance with the ethical standards of Greek National Ethical Committee and accepted by the ethical board of Biomedical Research Foundation Academy of Athens (Ref 2578/32). The animal studies were in accordance to the Greek National Ethical Committee and approved by our Institution’s ethical board (No 5959). All care and handling of the animals were provided according to the Guide for the Care and Use of Laboratory Animals of BRFAA, conformed to the Directive 2010/63/EU of the European Parliament.