Cell-free vascular grafts that grow with the host

Abstract

Cell-free small diameter vascular grafts, based on small intestinal submucosa (SIS) functionalized with heparin and vascular endothelial growth factor (VEGF) manufactured and implanted successfully into the arterial system of neonatal lambs, where they remained patent and grew in size with the host to a similar extent and with similar rate as native arteries. Acellular tissue engineered vessels (A-TEV) integrated seamlessly into the native vasculature and developed confluent, functional endothelium that afforded patency. The medial layer was infiltrated by smooth muscle cells, showed no signs of calcification and developed contractile function. The vascular wall underwent remarkable extracellular matrix remodeling exhibiting elastin fibers and even inner elastic lamina within six months. Taken together, our results suggest that VEGF-based A-TEVs may be suitable for treatment of congenital heart disorders to alleviate the need for repeated surgeries, which are currently standard practice.

Graphical Abstract

ToC Schematic Text

Cell-free vascular grafts containing immobilized VEGF were implanted into the arterial system of neonatal lambs, where they remained patent, underwent significant remodeling and grew with the animal host in length and diameter. Successful size expansion, integration into the native vasculature and development of vascular function may provide off-the-shelf bioengineered arteries for the treatment of congenital heart defects.

INTRODUCTION

Cardiovascular disease is the leading cause of death worldwide, claiming over 17.7 million lives in 2017. Congenital heart disease (CHD) is the leading cause of birth-defect associated infant death. CHDs affect nearly 1% (~40,000 infants) each year in the United States, with the reported prevalence increasing, mostly due to improvement in in-utero diagnosis. There are also about 1 million adults in the United States living with CHD. While some CHD patients can be treated with catheter procedures, severe CHD requires surgical intervention, during which inadequate or misdirected blood flow in and around the heart and lungs must be corrected using synthetic shunts or artificial blood vessels. The majority of infants born with CHD will need surgery immediately or in the first year of life. Unfortunately, additional surgical interventions are required to replace the vascular grafts with larger ones that can accommodate the increased blood flow as CHD-affected infants grow into adults. Therefore, engineering vascular grafts that can integrate into the native vasculature and grow with the patient may provide an alternative treatment to CHD, as it may alleviate the need for repeated surgical procedures ^[1].

A small number of studies demonstrated implantation of large-diameter tissue engineered vessels to neonatal animals and human patients. Polyglycolic acid (PGA) coated with poly-4-hydroxybutyrate based vascular grafts that contained endothelial cells and smooth muscle cells from the animal’s vasculature were implanted in sheep and demonstrated good performance in vivo up to 100 weeks^[2]. In a pioneering study, Shinoka and colleagues implanted poly (l-lactic-co-ε-caprolactone) copolymer based TEV seeded with autologous bone marrow cells into the pulmonary circulation of 25 young patients that were followed up to 11 years post-operatively ^[3] The grafts were successful as evidenced by lack of aneurysm, rupture, infection, or calcification. The main complication was stenosis that occurred in seven patients and was treated with balloon angioplasty^[3c]. The same biomaterial poly (l-lactic-co-ε-caprolactone) was also used to implant vascular grafts in sheep or mice to address growth potential as well as the role of the inflammatory response in regeneration ^[4]. A more recent study employed fibrin and ovine myofibroblasts that were cultured for at least seven weeks in a dynamic environment under flow and pressure, before being decellularized and implanted into neonatal sheep ^[5]. The grafts matured over a period of over 40 weeks and had comparable structure and mechanical properties as native arteries. Although these studies were met with various degrees of success, they either contained autologous cells or required long culture times for manufacturing. Most important, all of them used large diameter grafts (ranging from 16 to 24mm in diameter) and were implanted into the high flow environment of the pulmonary circulation, which may have reduced the likelihood of clotting.

Recently, we reported on the implantability of acellular (A)-TEVs based on SIS functionalized with heparin and VEGF. First, we demonstrated the ability of surface immobilized VEGF to capture EC cells with high specificity and under flow in vitro ^[6]. These results led the way to a subsequent study that utilized the natural biomaterial SIS with immobilized heparin and VEGF in the lumen to engineer a cell-free vascular graft ^[7]. A-TEVs (4.5cm long, 4.5mm diameter) were implanted interpositionally into the carotid arteries of sheep, where they remained patent throughout the 3-month study (n=25 implants, 92% patency). Notably, after only one month in vivo, the grafts developed a confluent endothelium from host cells, even in the middle of the graft, as evidenced by immunostaining for EC markers CD144 (VE-cadherin) and eNOS. In addition, the vascular wall was infiltrated by host SMC exhibiting uniform cell density along the length of the graft (from distal to proximal anastomotic sites) and developed contractile function as evidenced by vasoreactivity measurements. However, the growth potential of these completely biological and cell-free grafts was not evaluated.

Here we report successful implantation of very small diameter (2.75 mm), VEGF-functionalized SIS based A-TEVs in the arterial system of neonatal lambs. Following implantation, the grafts remained patent for six months - the duration of the study - and a time period equivalent to about five human years, which is typically used as a measure of long-term success of vascular grafts. Similar to adult animals, A-TEVs were endothelialized successfully and developed a functional and contractile medial layer with deposition of new collagen matrix. In contrast to adult animals, A-TEVs in neonatal animals showed development of elastin fibers, and inner elastic lamina (IEL) underneath the lumen. Most notably, the grafts grew in diameter and length to a similar extent and with a similar rate as native arteries during the same time period. The seamless integration into the native vasculature and the gain of vascular function in a short amount of time, coupled with the ease of fabrication may provide an off-the-shelf technology with the potential for treatment of congenital heart defects.

RESULTS

Assessment of growth and patency of A-TEV

A-TEVs were fabricated with biomaterial SIS that was functionalized with heparin and VEGF and implanted into the right carotid arteries of 8–12 week old male lambs, weighing between 40–50 lbs (18–23 kg), using end-to-end interrupted suture technique as shown in Fig. 1A. A total of n=7 A-TEVs were evaluated at 3 and 6 months post implantation (n=3 at 3 months; n=4 at 6 months). At the time of implantation, the inner diameter of the A-TEV measured 2.75 mm. Following implantation and animal recovery, A-TEV grafts were monitored monthly via two-color Doppler ultrasound (Fig. 1B; also see sample ultrasound images obtained throughout the duration of the study in supplementary figure, Fig. S1). Diameter measurements were obtained via ultrasound as a function of time and showed that the graft diameter increased with a constant rate of 0.089±0.009 mm/week increased to 5.2±0.29 mm at 6 months (Fig. 1C).

Figure 1: Graft implantation, patency and growth.

(A) A-TEV sutured into native carotid vasculature using end-to -end interrupted suture technique. Scale bar: 1 cm (B) Color Doppler ultrasound images of A-TEV post-implantation. (C) A-TEV diameter (mm) as a function of time post-implantation as determined by ultrasound. (D) X-ray angiography at 3 and 6 months post-implantation. Anastomosis sites are shown by red dashed lines.

At study end-points of 3 months or 6 months, we performed X-ray angiography (Fig. 1D; the red dashed lines in the figure mark the suture sites) and measured the diameter and length of each A-TEV. (Fig. S2). The length of the graft increased from 4 to 5.9±0.71 cm demonstrate that the graft was capable of growth, not only in the circumferentially but also longitudinally. The weight of each animal was also recorded and the diameter was correlated to animal growth using the power law ^[8]:

$$\frac{d}{d_0}={\left(\frac{w}{w_0}\right)}^n;or n=\frac{Log\left(d/d_0\right)}{Log\left(w/w_0\right)}$$

where, d0, d: the diameter at the beginning and end of the study; w0, w: the weight of each animal at the beginning and end of the study At 3 months, the graft diameter increased with the square root of the animal weight (n_A-TEV=0.5), slightly lower than the native arteries (n_native=0.63). At 6 months, the power decreased (n_ATEV=0.44), but was still similar to the native carotids (n_native=0.46), indicating that similar growth rate of the grafts and native arteries as the animals reached adult size.

Histological Assessment of A-TEVs

Entire ring sections from the middle, proximal and distal end of each A-TEV and native vessel were stained for Hematoxylin and Eosin Y (H&E) and imaged using the Aperio ImageScope to measure the average perimeter and calculate the inner diameter of the graft lumen (Fig. 2A). The luminal perimeters of implanted grafts were similar to those of native arteries but there was a significant increase from 3 to 6 months indicating graft growth during the implantation period (Fig. 2B). Except for one explant, the inner diameters of the all other grafts were very similar to those of native carotids, demonstrating no dilation or stenosis.

Figure 2: Histological analysis of A-TEV explants.

(A) Representative Aperio images of cross sections from the middle of A-TEV explants and native arteries. Scale bar: 2mm (B) Luminal perimeter of A-TEV explants at 3- and 6-months post-implantation. For each graft, the luminal perimeter was averaged over multiple measurements from the middle, proximal and distal sections. (*) indicate statistical significance (p < 0.05, n=3, 3 MO and n=4, 6 MO). (C) Hematoxylin and Eosin staining of explanted grafts. L: denotes lumen. Scale bar: 200 μm (D) Quantification of the cell number in the medial layer of native arteries and A-TEV grafts. (n=3, 3 MO and n=4, 6 MO)

Histological analysis showed that by 3 months post-implantation a confluent endothelium had been formed and the vascular wall was populated by incoming smooth muscle cells (Fig. 2C). All grafts showed similar cell infiltration and cell density approached to that of native tissue in all A-TEVs by 3 months post implantation (Fig. 2D).

Development of endothelial and medial cell layers

Immunostaining confirmed that the graft lumen contained endothelial cells as evidenced by VE-Cadherin and phosphorylated (active) eNOS expression (Fig. 3A). The graft wall was infiltrated by smooth muscle cells (SMC) as evidenced by expression of αSMA and MYH11 (Fig. 3A). SEM showed a confluent endothelium that was aligned in the direction of blood flow (Fig. 3B).

Figure 3: Expression of endothelial and smooth muscle markers and cell proliferation in ATEV explants.

(A) Endothelial cell (CD144, eNOS) and smooth muscle cell (α-SMA, MYH11) marker immunostaining of explants and native arteries. Representative images from the middle section of the tissues are shown. Scale bar: 100 μm. (B) SEM image of ATEV lumen at 6 months post implantation. Scale bar: 100 μm (top), 50 μm (bottom). (C) Immunostaining for Ki67 and CNN1 for native arteries and A-TEV explants from middle of the graft at 3 and 6 MO post implantation. Second column shows higher magnification images that correspond to the areas delineated by the dotted squares. Scale bar: 100 μm (left), 50 μm (right). L: denotes luminal surface. (D) Quantification for percentage of Ki67 positive cells in A-TEV explants and native arteries that served as controls. For each graft, % Ki67+ cells was averaged over multiple measurements from the middle, proximal and distal sections. Statistical significance (p<0.05) was indicated by ($) between 3MO and 6MO A-TEV; and by (*, #) between A-TEV and native arteries at 3 and 6 MO, respectively (n=3 at 3 MO and n=4 at 6 MO).

Next, we examined whether the growth of the grafts was accompanied by proliferation of the incoming SMC within the grafts. Immunostaining for Ki67 and calponin (CNN1) showed that only a small fraction of CNN1+ cells were Ki67+ (Fig. 3C). Specifically, 2.45±1.16% of cells in the vascular wall stained positive for Ki67 at 3 months and the percentage decreased to 1.32±0.52% (p<0.05, n=12 fields of view containing 400–1,000 cells in total) at 6 months (Fig. 3D), suggesting that the grafts were approaching homeostasis.

A-TEV vascular wall remodeling

Trichrome staining showed deposition of new collagen throughout the vascular wall of A-TEVs post implantation (Fig. 4A). As evidenced by Picrosirius red staining under polarized light, both collagen I (orange-red color) and collagen III (green-yellow color) deposited inside the A-TEV grafts post-implantation. Although, at 3 months collagen fibers were thin and lacked organization, at 6 months, the collagen content increased and fibers appeared to be aligned circumferentially (Fig. 4B). Quantitative measurements of collagen using the hydroxyproline assay showed similar amount of collagen at 3 and 6 months (Fig. 4C), possibly reflecting the combined collagen of the SIS matrix and the newly deposited collagen by the infiltrating cells.

Figure 4: Collagen deposition in the A-TEV vascular wall.

(A) Masson’s trichrome staining for native artery and explanted A-TEV at 3 and 6 MO post implantation. Representative images from the middle section of the tissues are shown. (B) Picrosirius red staining imaged under polarized light for native artery and explanted A-TEV at 3 and 6 MO post implantation. Representative images from the middle section of the tissues are shown. Scale bar: 200 μm. L: denotes luminal surface. (C) Quantitative measurements of collagen using the hydroxyproline assay for each A-TEV and corresponding native artery (n=3, R43, R46 and R52; for 3 MO, n=4, R59, R60, R61 and R62; for 6 MO).

Verhoeff’s elastin staining (Fig. 5A) showed the presence of unorganized elastin in the vascular wall at 3 months. By 6 months, both the level of elastin as well as its fibrillar organization improved dramatically as evidenced by the dark brown cables and formation of IEL (inner elastic lamina), similar to that of native arteries. Similarly, quantitative measurements using ninhydrin assay showed low levels of elastin at 3 months, and significantly enhanced elastin deposition by 6 months post-implantation (~ 7-fold increase, n=7, p<0.0001, Fig. 5B). Furthermore, immunostaining for mature elastin showed the presence of immature elastin fibers at 3 months but formation of elastin cables as well as the presence of IEL (red arrowheads) at 6 months, indicating significant remodeling and maturation of the vascular wall (Fig. 5C). Finally, Von-Kossa and Alizarin red staining showed no signs of calcification, demonstrating successful remodeling of the vascular grafts (Fig. S3).

Figure 5: Elastin deposition and IEL formation in A-TEV vascular wall.

(A) Verhoeff’s elastin staining for native artery and explanted A-TEV at 3 and 6 MO post implantation. Representative images from the middle section of the tissues are shown. (B) Quantitative measurements of elastin using ninhydrin assay for each A-TEV and corresponding native artery; (*, #) indicate statistical significance between 3 MO A-TEV (*) or 6 MO A-TEVs (#) and corresponding native artery (p < 0.05, n=3 for 3 MO; R43, R46 and R52; and n=4 for 6 MO; R59, R60, R61 and R62). (C) Immunostaining for mature elastin in native artery and explanted A-TEV at 3 and 6 MO post implantation. Representative images from the middle section of the tissues are shown. (A, C) Top images (scale bar: 25 μm) show higher magnification images that correspond to the areas delineated by the dotted squares in bottom images (Scale bar: 200 μm); L: denotes luminal surface.

Assessment of Biomechanical Properties

These results prompted us to examine the mechanical properties and function of vascular grafts. Uni-axial tensile stress test was used to obtain the stress-strain relationship and calculate the Young’s Modulus (YM), ultimate tensile strength (UTS) and tensile strain at break for each graft and the corresponding native arteries. The results of YM and UTS were normalized to area of each ring section, and reported as percentage of the YM or UTS of the native artery of the same animal (Table 1).

Table 1:

Young’s Moduli (YM) and Ultimate Tensile Strength (UTS) of each A-TEV and contralateral native carotid artery.

Animal Information		YM(MPa)		UTS(MPa)
Animal #	Duration of Study	ATEV	Native	ATEV	Native
R43	3 MO	5.31	4.75	1.99	2.8
R46	3 MO	5.23	4.61	N/A	N/A
R52	3 MO	2.77	0.9	1.29	0.5
R59	6 MO	2.14	1.75	1.39	0.8
R60	6 MO	5.72	3.45	2.43	1.49
R61	6 MO	1.88	4.76	1.39	1.74
R62	6 MO	8.26	2.5	7.26	1.68

Pre-implanted A-TEV were significantly stronger and stiffer than native arteries with UTS = 4.67 ± 0.98 MPa and YM = 24.14 ± 1.17 MPa ^[7]. At 3 and 6 months post-implantation, YM decreased significantly from that of pre-implanted A-TEV and was similar to that of native carotids (3MO: A-TEV: 4.44±1.18 MPa, Native: 3.42±1.78 MPa, n=3, p=0.45, 6MO: A-TEV: 4.5±2.65 MPa, Native: 3.11±1.12 MPa, n=4, p=0.37). The ultimate tensile strength (UTS) was similar to native carotids at 3 months (A-TEV: 1.64±0.34 MPa, Native: 1.65±1.15MPa, n=3, p=0.99) and increased further at 6 months (A-TEV: 3.12±2.43 MPa, Native: 1.43±0.37 MPa, n=4, p=0.22), possibly reflecting more complete ECM remodeling. At 3 months, the tensile strain at break was lower than that of native arteries but it significantly increased at 6 months indicating improved remodeling during the implantation period (Fig. 6A).

Figure 6: Mechanical properties and contractile function of A-TEV explants.

(A) Tensile Strain at break of A-TEVs at 3 MO and 6 MO reported as percentage of the corresponding native carotid artery strain at break. For each graft, tensile strain at break was averaged over multiple measurements from the middle section of the tissue. *= p < 0.05, n=3 for 3 MO; R43, R46 and R52; and n=4 for 6 MO; R59, R60, R61 and R62. (B-E) Representative signature forces exerted by A-TEVs in response to vaso-constrictors (B) U46619; (C) Endothelin-1; or (D) KCl. (E) relaxation of half-maximum constricted tissues in response to the ROCK inhibitor Y27632.

A-TEVs develop contractile function

Next, we evaluated the contractile function of A-TEV explants by measuring vascular constriction in response to thromboxane mimetic U46619, Endothelin-1 and KCl; as well as vasodilation in response to the ROCK inhibitor, Y27632. To this end, rings of explanted A-TEVs were each connected to a force transducer in a warm oxygenated bath containing Krebs Ringer solution, and the force of constriction was measured upon addition of vaso-active agonists. Representative force signatures are shown in Fig. 6B–E. The force was then normalized to the area of each ring and reported as percentage of the response of the native carotid from the contralateral side of the same animal (Table 2).

Table 2:

Contractile response of each A-TEV and contralateral native artery in response to the indicated vasoagonists. All forces were normalized to the contact area and reported in KPa.

Animal Information		U46619(KPa)		Endothelin-1(KPa)		118 mM KCI(KPa)		Y27632(KPa)
Animal #	Duration of study	ATEV	Native	ATEV	Native	ATEV	Native	ATEV	Native
R43	3 MO	1622	2685.2	701.8	1974	4882.9	10124.4	2859.3	2553.8
R46	3 MO	2732.6	1942.8	2183.1	1958.9	5325.9	15570.4	1992.6	2814.8
R52	3 MO	358.2	245.2	151.6	545.1	160.3	807.5	N/A	N/A
R59	6 MO	775.8	1020.9	466.6	871.4	451.3	2329	N/A	N/A
R60	6 MO	1085.7	1420.2	1385.34	1481.6	1817	5340.5	750.8	1141.6
R61	6 MO	177	653.7	168.2	404.7	263.3	9632.2	272.3	523.8
R62	6 MO	2273.9	384.3	735.6	N/A	1446.3	6362	1016.8	571

Notably, the A-TEVs had developed contractile function as early as three months post-implantation. At 3 months, the grafts were responsive to all vasoagonists with the highest response to U46619 (A-TEV: 1,570.92±970 KPa, Native: 1,690.8±932 KPa, n=3, p=0.88), followed by Endothelin (A-TEV: 1,012.17±857.9 Pa, Native: 1,492.66±670 KPa, n=3, p=0.48) and KCl (A-TEV: 3,546.4±2337.7 KPa, Native: 8,834.11±6095.6 KPa, n=3, p=0.23). Following vasoconstriction with U46619 to half maximum level, the tissues relaxed upon addition of the ROCK inhibitor Y27632. At 6 months, the contractility of A-TEV explants was maintained at similar levels. Taken together, results show that the acellular grafts cellularized, developed functional endothelial and medial layers and demonstrated significant ECM remodeling with elastin fibers, resembling native carotids in size, structure and function.

DISCUSSION

We demonstrated that A-TEVs functionalized with heparin and VEGF performed favorably in a large animal postnatal growth model, both in terms of development of vascular structure and function as well as somatic growth. At implantation, A-TEVs measured 2.75 mm in inner diameter and 4 cm in length. At 6 months post-implantation into the carotid arteries of growing lambs, the grafts grew to ~5.2 mm in diameter and ~6 cm in length, which matched the growth of the contralateral native carotid arteries. A confluent endothelium was formed, which prevented thrombosis and maintained vascular tone, likely in response to the shear stress experienced by the graft. Additionally, incoming host cells surrounded the SIS-based graft and formed the vascular wall. Over time, these cells differentiated into CNN1 expressing SMCs, which exhibited very low rate of proliferation, similar to native SMCs. The SMC layer was functional, as demonstrated by vascular contractility, remarkable ECM remodeling and even formation of IEL by 6 months post-implantation, indicating maturation of A-TEVs towards native arteries. Taken together, these results demonstrate that VEGF-based A-TEVs can regenerate in situ and grow with the animal, thereby alleviating the need for repeated surgical procedures as required for current treatment of pediatric congenital heart diseases.

Very few studies have explored the behavior of vascular grafts in growing organisms. In a pioneering study, Breuer and colleagues employed stem cells that were incorporated into PLA/PGA and implanted successfully into a human patient who was followed for more than ten years ^[9]. Another study utilized fibrin-based TEVs that were cultured in a bioreactor for several weeks before being decellularized and implanted into the high flow environment of pulmonary artery of a small number of growing lambs ^{[5, 10]}. Despite these early successes, as of yet no studies have evaluated very small diameter (<3 mm) A-TEVs in the arterial system of a large animal model. In addition, our grafts are truly acellular as they have never been cultured with cells but only functionalized with heparin and VEGF prior to implantation, thereby providing truly off-the-shelf implantable constructs.

Notably, both the diameter and length of implanted A-TEVs increased over time after implantation at a similar rate as the native left carotid arteries, as evidenced by X-ray angiography. The diameter of the carotid artery correlated strongly with age in children ^[11], showing a sharp increase in the first 5–6 years of life, possibly due to maturation of the carotid-cerebral circulatory system to provide adequate supply to the developing brain ^[12]. The diameter grew at a more modest rate between 6 and 15 years of age ^[12]. In our study, the A-TEV and native diameter correlated strongly with animal weight, which increased over time as the animals increased in size. Fitting the arterial diameters and weights with a power-law equation yielded very similar exponents (n_A-TEV=0.44, n_native=0.46), suggesting that both the A-TEV and native artery diameters increased at a similar rate in the first 6 months post-implantation, a time period equivalent to about 5 human years.

Interestingly, by 3 months post-implantation the lumen was covered with mature and functional EC as evidenced by expression of CD144 and phosphorylated eNOS that were not restricted to the anastomotic sites but appeared uniformly throughout the lumen. In a recently published study ^[13], our laboratory showed that A-TEV with immobilized VEGF were endothelialized with VEGFR1-expressing monocytes (MC) on the graft lumen, where they differentiated into functional endothelial cells (EC) that maintained patency. Interestingly, cells on the A-TEV lumen co-expressed both EC and MC specific markers, including endothelial nitric oxide synthase (eNOS), the enzyme responsible for production of the vasodilatory agent, nitric oxide. Biochemical factors such as immobilized VEGF and biophysical forces, such as shear stress from flowing blood, upregulated arterial and downregulated venous genes in MC-derived ECs, suggesting that the arterial microenvironment might have played significant role in the maturation and functional regeneration of the MC-derived endothelium ^[13]. We expect that a similar mechanism may be at work in the endothelialization of A-TEV in neonatal animals.

We also demonstrated that by 3 months post-implantation the vascular wall was populated by host SMC, expressing MYH11 and generating contractile force in response to vascular agonists. Host cell infiltration was also uniform and not restricted to the anastomotic sites, in agreement with a previous study from our laboratory demonstrating that the presence of donor SMC was not necessary for recruitment of host SMC and regeneration of a contractile vascular wall ^[14]. The presence of immobilized VEGF in the wall might have facilitated mural host cell recruitment indirectly via recruitment of immune cells e.g. monocytes/macrophages, which in turn release growth factors, such as PDGF and TGF-β that promote differentiation of incoming myofibroblasts into mature and contractile SMC ^[15]. The presence of biophysical forces acting on the wall may have contributed to SMC maturation as well, as many groups including our own have established that SMC maturation and ECM synthesis in blood vessels required the presence of growth factors as well as biophysical forces ^{[10, 16]}. Interestingly, a few of the newly incoming host cells in the periphery of the vascular wall were proliferative, while medial cells that were recruited earlier and were located closer to the lumen did not proliferate but aligned circumferentially, indicating SMC maturation in agreement with development of contractile function.

Although at 3 months post implantation SIS was partially degraded, as it was visibly noticeable from histological evaluation, the grafts maintained robust mechanical properties (UTS and YM) that were similar to those of the growing lambs’ native arteries. By 6 months post implantation most of the SIS had degraded and replaced by native ECM, while mechanical properties continued to be similar to those of native arteries. While SIS degraded, the amount of collagen remained high indicating robust new collagen synthesis and ECM remodeling by host SMC. Few collagen fibers could be observed at 3 months post-implantation; by 6 months, fiber organization and alignment improved but not to the extent of native arteries, suggesting that such level of organization may require longer implantation times perhaps up to a year or longer.

Notably, elastin measurements showed significant elastin synthesis, rarely observed in tissue-engineered vascular grafts. While thin, fractured elastin fibers were observed at 3 months post implantation, by 6 months, elastin was organized in thick fibers and inner elastic lamina (IEL, red arrowheads) was evident underneath the lumen, indicating significant ECM remodeling and signs of adaptation to volume/pressure fluctuations in the growing lamb arteries. The collagen/elastin synthesis and subsequent maturation/cross-linking eventually enhanced the elasticity and mechanical properties of the neo-tissue, ultimately reaching similar levels as native arteries by 6 months post-implantation.

In summary, we demonstrated that SIS-based VEGF-functionalized A-TEVs of very small diameter could be implanted into the arterial system of neonatal lambs, remained patent for up to six months and grew with the host to a similar extent and with similar rate as native arteries. In addition, A-TEVs demonstrated remarkable remodeling, seamless and complete integration into the native vasculature and development of function in a short amount of time, suggesting that they might be suitable for treatment of congenital heart disorders to alleviate the need for repeated surgeries, which are currently the golden standard.

Conclusion

We developed cell-free, VEGF-fortified vascular grafts that were implanted successfully into the arterial system of neonatal lambs, where they remained patent and grew with the animals, similar to native arteries. The seamless and complete integration into the native vasculature and the gain of vascular function in a short amount of time, coupled with the ease of fabrication may provide an off-the-shelf technology for the treatment of congenital heart defects.

ONLINE METHODS

Fabrication of A-TEV

A-TEVs (2.75 mm in diameter and 4 cm in length) were manufactured by first immobilizing heparin on SIS (small intestinal sub-mucosa) provided by COOK Biotech Inc (West Lafayette, IN) with EDC-NHS chemistry, followed by the attachment of VEGF through the heparin binding domain as reported previously ^[7]. VEGF was produced as previously described and tested by ELISA, MTT assay and SDS page to confirm the purity, proper refolding and activity ^[6]. Heparin was immobilized on SIS using the following procedure. SIS tubes were immersed in 20 ml of 50 mM 2-(N-morpholino)ethanesulfonic acid buffer (MES) at pH 1.5 containing 20 mM 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), 20 mM N-hydroxysuccinimide (NHS) and 8000 U/ml heparin sulfate and incubated overnight at room temperature on a rocking platform shaker. In order to remove the unbound heparin and obtain neutral pH, SIS tubes were washed gently in PBS. Next, VEGF was immobilized on SIS by immersing the heparin-bound SIS tubes in 5 ml PBS containing 1 mg of VEGF for 4 hr at 37°C. The graft was then washed with PBS and kept at 4°C. Samples were used for implantation within 24 hr.

Implantation of A-TEV in the Ovine Carotid Artery Model

Procedures and protocols in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of Angiograft, LLC. A total of 8 Dorset cross lambs between the ages of 8–12 weeks and weighing 15–20 kg were used. The grafts were implanted interpositionally (end-to-end) in carotid arteries of lambs by interrupted suture technique as previously described ^[14]. Aspirin (975 mg/day) and warfarin sodium (20–30 mg/day, Barr Laboratories, Pomona, NY) were given 3–5 days before the procedure and continued for 1 week following A-TEV implantation. Anesthesia was induced via Telazol (Tiletamine & Zolazepam) 4mg/kg IV, intubated and maintained with 1.5–2% isoflurane through a 8 mm endotracheal tube using a positive pressure ventilator and 100% oxygen. Heparin (100 U/kg) was administered intravenously to the animals 20 min prior to the clamping of the proximal and distal segments of the carotid artery that was being replaced. At 30 min after the initial bolus and throughout the surgery, heparin was infused at a rate of 100 U/kg/hr, while maintaining ACT (Active Clotting Time) of 400 s. Following unclamping of the artery and restoration of normal blood flow through the A-TEV, the flow rate was monitored for a short period of time before wound closure and animal recovery. Post-operatively, the sheep also received flunixin meglumine (1.1 mg/kg) IM once a day for two days and buprenorphine (0.005–0.01mg/kg) IM twice a day for one day, for reducing discomfort and pain at the suture site.

A-TEV Evaluation

Ultrasound Evaluation of Growth

For ensuring graft patency after implantation, directional color doppler ultrasonography (Acuson Cypress, Siemens Medical Solutions, Mountain View, CA, USA) was performed immediately following surgery, and monthly thereafter. The diameter of the graft at different time points was also measured to evaluate the growth of implanted A-TEV.

Angiographic Assessment at Study End-point

In order to evaluate the diameter and length of the graft at indicated time points (3 and 6 months post-implantation), three-dimensional rotational angiography was performed right before euthanasia. Briefly, animals were anesthetized and both carotid arteries were cannulated proximal to the thoracic inlet using 3mm diameter tubing inserted approximately 4cm cranially followed by contrast (Omnipaque® (iohexol) 240mg/ml, GE Healthcare, USA) injection through both arteries simultaneously. X-ray images were acquired using an electronic x-ray imager (Toshiba Rotanode^™, Toshiba Electron Tubes & Devices Co., LTD, Tochigi, Japan).

Evaluation of Vascular Function and Strength

Response of A-TEV Explants to Vasoactive Agonists

In order to evaluate SMC function in A-TEVs, both A-TEV and native ring sections from the middle of explanted tissues were subjected to vasoreactivity tests as we described previously ^{[7, 14, 17]}. After 3 and 6 months, the animals were euthanized and both carotids were explanted and the lumens were washed with PBS. Tissues were trimmed off from connective tissue around the grafts and cut into ring sections measuring approximately 5 mm in length. Images were taken from the sections for measurement of contact area, which was used to normalize the force.

Tissue rings were mounted each in a bath containing Kreb’s-Ringer solution with constant oxygen supply (94% O₂, 6% CO₂) and connected to a force transducer. Tissue rings were exposed to the following vasoactive agonists: 1) KCl (118mM) to activate voltage dependent calcium channels; 2) Endothelin-1(10⁻⁸M, Sigma); and 3) the thromboxane A2 mimetic U46619 (2×10⁻⁷M, Sigma) to activate endothelin and thromboxane A2 (TXA2) receptors, respectively. Finally, the ROCK inhibitor Y27632 (1×10⁻⁶M, Sigma) was added to pre-constricted A-TEV’s to measure the ability of the tissues to relax and the force of relaxation was reported as % of the maximal constriction. Data was acquired using Powerlab and analyzed using Chart 5 software. All forces were divided by the area of contact and reported in KPa. (at least n=3 rings from each graft were tested).

Mechanical Properties

Ring tissue sections of similar length and diameter from the middle of explanted tissues were mounted onto tester grips of uniaxial Instron tensile tester (Model 3343, 50N load cell, Instron Corporation, Norwood, MA) using stainless steel clamps. Samples were stretched uniaxially until failure. The Young’s modulus was determined by calculating the slope of stress-strain curve and UTS was measured as the highest force at breaking point. Both UTS and Young’s modulus were divided by the area of contact and reported in MPa. Tensile strain at break was normalized and expressed as % of the strain of native carotids.

Histological Analysis

After excising, cleaning and trimming off the tissue at the end of study at 3 and 6 months, graft and native vessels were pressure fixed in 10% formalin, dehydrated in series of graded ethanol solutions and xylene, and then embedded in paraffin. Paraffin blocks were cut into 5 μm sections. After de-paraffinization and rehydration of tissue sections in Xylene and series of graded ethanol, they were stained with Harris Hematoxylin and Eosin (H&E) to evaluate tissue morphology. Tissue sections were also stained for Masson’s trichrome, picrosirius red and Verhoeff’s elastin to assess collagen and elastin deposition, as well as Von-Kossa and Alizarin red to examine possible calcification of the grafts.

Quantification of Collagen and Elastin

The hydroxyproline assay was employed to measure the collagen content of the grafts and native tissues as described previously ^[7]. Tissues were lyophilized, and incubated in 6N HCl for 2 hr at 110°C. Tissues were lyophilized again to remove HCl and resuspended in assay buffer (5% (v/v) monohydrate citric acid, 12% (v/v) trihydrate sodium acetate, 3.4% (v/v) sodium hydroxide and 1.2% (v/v) glacial acetic acid, pH 6.5). The suspension was spun down and 100 μL of each supernatant was loaded into wells of a 96-well plate. Subsequently, 50μL of chloramine-T (62 mM, Alfa Aesar, Ward Hill, MA) was added to each well and incubated for 20 min at room temperature followed by addition of 50μL of Ehrlich’s solution (Sigma-Aldrich) for 20 min at 65°C. The L-hydroxypoline concentration was measured by measuring optical density at 550 nm (background signal at 630 nm was subtracted) (Biotek Synergy 4 Spectrophotometer,Winooski, VT) and using a standard curve relating optical density to concentration (0–20μg/mL of L-hydroxyproline, Sigma-Aldrich). L-hydroxylproline concentration was then converted to collagen concentration using the conversion factor, F = 7.46g collagen per g L-hydroxyproline.

For elastin quantification, tissues were lyophilized and dissolved in 0.1N NaOH for 1hr at 95°C. The insoluble fraction containing cross-linked elastin was collected after centrifugation. The pellet was washed with DI water twice and then hydrolyzed in 6N HCl at 110°C overnight followed by lyophilization. The residue was reconstituted in 500 μl of DI water and 100 μl was loaded into 96 well plates. Subsequently, 100 μL of ninhydrin reagent (Sigma-Aldrich) was added to each well and incubated at 65°C for 1hr. The OD was measured at 550 nm (background signal at 630 nm was subtracted) using a Biotek Synergy 4 Spectrophotometer and elastin concentration was calculated from standard curve that was obtained using hydrolyzed elastin (3–100 μg/ml, Alpha elastin, EPC).

Immunofluorescence and Quantification

Following de-paraffinization and rehydration with xylene and series of graded ethanol, 5 μm-thick paraffin tissue sections from the proximal, middle and distal portions of the A-TEV grafts or native arteries were placed on glass slides and subjected to pressure-activated high temperature antigen retrieval. After permeabilization with PBS containing 0.1 % Triton-X tissue slides, tissue sections were incubated overnight at 4°C with the following antibodies: MYH11 (Abcam, ab53219, Rabbit polyclonal, 1:100 dilution), CNN1 (Santa Cruz Biotechnology, sc70487, Mouse monoclonal IgG_1, 1:100 dilution), α-SMA (Cell Signaling Technology, 14968S, Rabbit polyclonal, 1:100 dilution), CD144 or VE-Cadherin (Cell Signaling Technology, 2500S, Rabbit monoclonal, Clone: D87F2, 1:50 dilution), eNOS (Cell Signaling Technology, 32027S, Rabbit monoclonal, Clone: D9A5L, 1:100 dilution) and Ki67 (Thermo Fisher Scientific, PA519462, Rabbit polyclonal, 1:300 dilution) in PBS containing 0.01% Triton-X and 5% goat serum (Sigma-Aldrich) treatment. Subsequently, anti- mouse/rabbit Alexa flour secondary antibodies (568/488, 1:200 dilution, Invitrogen) were added for 1 hr at room temperature. Cell nuclei was counterstained with Hoechst 33342 (Invitrogen, H3570) for 10 minutes at room temperature.

Elastin fibers were stained with elastin antibody (Abcam, ab21610, Rabbit polyclonal, 1:50 dilution) in PBS containing 0.05% Tween-20 and 0.1% BSA overnight at 4°C, following glycine treatment (50 mM in PBS, pH=7.8) for 30 min at room temperature. Secondary antibody and Hoechst 33342 were then added as described above.

To quantify cell density, tissue sections were stained with DAPI and the number of cells was determined by counting using ImageJ software (NIH). For each explanted graft (n=3 grafts at 3 months post-implantation; n=4 grafts at 6 months post-implantation), 10 fields of view were quantified, normalized to the area of the tissue in each field of view and reported as cell/mm². For Ki67 quantification, cells that stained positive for Ki67 and CNN1 were counted and reported as percentage of Ki67 positive cells to total number of the cells.

Statistical analysis

In total n=7 animals were implanted; A-TEV grafts were explanted at 3-months (n=3) or 6 months (n=4) post-implantation, respectively. Each measurement (mechanical properties (UTS, YM), vascular reactivity, collagen and elastin content) was repeated at least three times (n=3) with each of the three rings from the middle, proximal and distal end of each graft. Statistical significance for each experiment was determined by performing unpaired t-test or Two-Way ANOVA (Analysis of Variance) analysis followed by Tukey’s multiple comparison test using GraphPad Prism software and statistical significance was defined as p<0.05. All data expressed as means ± standard deviation (STD).

Supplementary Material

1

Figure S1: Doppler imaging. Color Doppler ultrasound images taken at several times (1 day, 1 MO, 3MO, 6MO) post-implantation to monitor blood flow and measure A-TEV diameter.

Figure S2: A-TEV sizing. Diameter (mm), Length (cm) and Volume (cm³) of grafts over time as measured using X-ray imaging. Corresponding native values are shown with red triangles where indicated.

Fig. S3: A-TEVs show no calcification. Von-Kossa and Alizarin red stained images for native artery and A-TEV explants at 3 MO and 6 MO post-implantation show no signs of calcification.