|
|
||||||||
Tissue Engineering Center, University Department of Surgery, Royal Free and University College Medical School, University College London and The Royal Free Hospital, London, UK
2Correspondence: University Department of Surgery, Royal Free and University College Medical School, University College London, The Royal Free Hospital, Pond Street, London NW3 2QG, UK. E-mail: A.Seifalian{at}RFC.UCL.AC.UK
| ABSTRACT |
|---|
|
|
|---|
Key Words: hybrid graft anticoagulant peptide endothelial cells spacer arm covalent bonding x-ray photoelectron spectroscopy
| INTRODUCTION |
|---|
|
|
|---|
We have developed a compliant graft (MyoLinkTM) based on poly(carbonate-urea)urethane chemistry that has compliance similar to the human artery (6)
. The MyoLink graft has a honeycomb structure that allows it to maintain compliance and pulsatile flow in vivo through a mechanism of wall compression that accommodates increases in volume without the need for external dilation. This is unlike previous polyurethane based grafts that exhibited a compliance mismatch caused by perivascular ingrowth (7)
. The graft has undergone in vitro degradation tests and has been implanted in a dog model for 36 months, demonstrating very high biostability (8
, 9)
. This graft is already in use as an A-V fistula for hemodialysis access and is undergoing a phase I clinical trial as a peripheral vascular bypass graft.
To further improve the patency of prosthetic grafts by reducing the thrombogenic effects, lining the lumen of the graft with autologous endothelial cells (EC), a process known as seeding, has been advocated (10)
. This process has been applied clinically using expanded polytetrafluoroethylene (ePTFE) graft (11)
. However, this required weeks of culture before use because of the large number of cells needed for seeding since a portion of the cells was washed off once exposed to physiological blood flow.
The design of the MyoLink graft with the honeycomb structure makes it particularly suitable for seeding and therefore tissue engineering (12)
. We have already established that the MyoLink graft has better cell retention than ePTFE and have identified optimal seeding density and incubation times for this conduit (12
, 13)
. However, there is still a need for a substance to facilitate the attachment of EC to the graft that is inherently highly nonthrombogenic.
Fibronectin is a glycoprotein (
250,000 Da) with a modular structure. It contains domains for binding fibrin heparin, gelatin, collagen, EC, and has cysteine and arginine-glycine-aspartate (RGD) residues. We have already shown it significantly improves EC retention (12)
, but fibronectin is susceptible to hydrolysis (14)
.
To reduce such hydrolysis, we used only the RGD peptide sequence and heparin, for which two donors are present in the native parent fibronectin (15
, 16)
. Heparin was used to provide an additional antithrombogenic effect (17
, 18)
. We have also been looking at covalent grafting of RGD and heparin to prevent desorption of the moieties over time from the surface and thus provide a highly nonthrombogenic surface with cell attachment properties potentially better than the parent molecule fibronectin. The commonly used reaction schemes to covalently attach the latter have involved coupling of the moiety sequence to a polymeric surface that contains, for example, hydroxyl, thiol, carboxylic, or amine groups. However, this imposes steric constraints that affect both the affinity and specificity of the ligand, especially when more than one moiety is used (19
, 20)
. The harsh chemicals used in these reactions affect the activity of the moiety sequence and create competing side reactions that affect the overall grafting yield (21)
. To overcome the above problems, we used a new grafting technique using spacer arms that reduces steric hindrance caused by the proximity of the ligand to the rigid surface of the polymer backbone (22)
.
In this study, we describe the use of this new bonding technique to immobilize covalently RGD, heparin (Hep), and RGD/Hep onto the surface of MyoLink. We provide experimental data to demonstrate the effect of these biological moieties on EC attachment and retention and have assessed cell viability under arterial shear stress.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Covalent bonding of peptides onto the graft using a low sterically hindered spacer arm technique
The moieties were covalently attached to the surface of MyoLink to enhance the anticoagulant properties and cell attachment. All three steps are water based to prevent hydrolysis, oxidation, or any cross-reactions occurring to either the moiety or polymer, including surface functionalization, grafting, and bonding. The functionalization step uses an isovaleric-based solution to generate free radicals in order to abstract hydrogen from the surface of the polymer via a complex sequence of hydroperoxide-based reactions. The grafting step uses the hydroxide groups generated on the surface to allow attachment of the spacer arms. Grafts are heated to
70°C with stirring, then a ceric ion solution is added to induce radical polymerization and incorporation of the spacer arm. The final stage involves the attachment of the moiety, achieved by adding an acidified carbodiimide solution with the moieties of interest under nitrogen with stirring at pH 4.5 for 4 h.
Assessment of bonding using X-ray photoelectron spectroscopy (XPS)
After bonding and exposure to shear force, vascular grafts were assessed by XPS measurements made with an SSI M-probe spectrometer operating at a base pressure of 3 x 10-9 torr. The samples were irradiated with Al K
X-rays (1486.6 eV) using a spot size of 1000 x 400 µm and 180W power. Survey spectra were recorded with pass energy of 150 eV from which surface chemical composition was determined. High-resolution spectra 25 or 50 eV were run and their chemical states were determined.
Cell culture and radiolabeling
Human umbilical vein endothelial cells (HUVEC) were cultured as in a previous protocol (12)
. The cells were used when confluent on third passage. After trypsinization and resuspension in complete tissue culture medium, a cell count was obtained and cells were radiolabeled with 1.8 MBq 111In-oxine (Amersham International, Amersham, Bucks UK) per 106 cells (incubated at 37°C, 15 min). The cell suspension was washed with complete medium and centrifuged at 300 g for 7 min. The washing procedure was repeated twice. The resultant cell pellet was resuspended in complete medium and diluted to the concentration necessary to achieve a seeding density of 2 x 105 cells/cm2. A diagrammatic representation of the graft is shown in Fig. 1
.
|
Effect of shear force on cell retention by pulsatile arterial flow
We have developed and validated a flow system able to simulate in vitro the pulsatility and flow waveform including reverse flow, pressures, and degree of oxygenation and pH of physiological femoral artery circulation in vivo (23)
. We used this model to accurately determine EC adhesiveness and retention on seeded vascular grafts exposed to physiological shear stress. The system has been described (12
, 23)
. The model consisted of a variable-speed electromagnetic centrifugal pump, flexible plastic tubing and reservoir, flow waveform conditioner, Maxima hollow fiber oxygenator supplied with 95% air/5% CO2, an outflow resistance, a 6 mm caliber tubular flow probe and Transonic Medical Flowmeter system, and a Millar Mikro-tip catheter transducer. The circuit was circulated with perfusion solution (12)
. The hemodynamic parameters calculated for this flow circuit are given in Table 1
. Shear stress at the wall of the graft and its compliance was monitored with a Doppler ultrasound.
|
Radioactivity from the 111In-oxine-labeled HUVEC cells was measured using a gamma camera scanner linked to a image processing system. The section of the circuit containing the vascular grafts was positioned over the gamma camera and imaged throughout the perfusion period. An on-line workstation recorded all images within 64 x 64 matrices. The initial 6 images were each acquired over 5 min, with 22 subsequent images acquired over a 15 min span and used to generate time-activity curves corrected for background, spontaneous indium leakage, and isotope decay (half-life of 111In=68 h). Cell attachment (CA) with respect to time was calculated from the equation below, where (t)n is time in minutes, (t)0 is immediately before initiation of flow, and CPMG and CPMB are the counts per minute over the graft and background computed from analysis of dynamic scintigraphy images. Grafts (n=6) were perfused for 6 h.
![]() |
Cell viability postexposure to physiological shear force of flow
Viability of seeded cells postexposure to flow shear stress were assessed using Alamar blueTM assay, as other commonly used viability assays such as MTT interact with the poly(carbonate-urea)urethane polymer, generating formazin crystals and killing off any attached cells. The technique has been described in detail elsewhere (24)
. After exposure to shear stress, grafts were placed in a container and the Alamar blueTM assay (Serotec Ltd., Kidlington, Oxford, UK) was added; absorbance was monitored at 4, 8, 16, and 24 h by removing a 50.0 µl aliquot into a 96-well plate and absorbance read spectroscopically at wavelengths of 570 and 630 nm (Labsystems Multiscan MS UV visible spectrophotometer). Segments were subjected to scanning electron microscopy (SEM) for the presence and morphological appearance of seeded cells.
Data analysis and statistical methods
The experiments were repeated six times. Data are presented in mean ± SD. The time-activity retention data were fitted with one-phase exponential decay. Comparisons between groups were made by 1-way ANOVA (Kruskal-Wallis) test with using Dunns comparison test.
| RESULTS |
|---|
|
|
|---|
|
Since the moieties Hep and RGD both contain sulfur, their presence on the surface of MyoLink is quantified in Table 2
, with none being present in the native polymer itself. Distribution of the moieties was uniform between the random points analyzed. No degradation of the MyoLink polymer was detected postbonding and the bioactivity of each peptide covalently bonded was very high since sulfonamide nitrogen (NSO3) spectra typically showed the moiety in question was attached at an energy of 400 eV (representative of covalent attachment) vs. a value of 399.4 eV typical for an ionic attachment. Sulfur is present in two states: NSO3 and OSO3; the NSO3 signal originates solely from the heparin bonded to the surface. The presence of RGD was indicated by the new state of nitrogen detected in the RGD and RGD/Hep samples, respectively, and not in the heparin grafts at the binding energy of 399 eV, and by the increase in surface composition of nitrogen, namely, 2.02 and 2.06% vs. unbonded graft and heparin, respectively.
Assessment of cell retention
The results of the physiological perfusion are summarized in Fig. 2
and Table 3
. The maximum cell loss occurred in the first 60 min after exposure to flow. The percent cell attachment of native MyoLink graft at 6 h was 60.47 ± 10.1%. Values for RGD-, RGD/Hep-, and heparin-bonded grafts were 71.2 ± 9.7%, 75.7 ± 2.3%, and 68.3 ± 9.3%, respectively. The result was significantly better for RGD/Hep-bonded graft than native MyoLink (P<0.05).
|
|
Cell viability
Figure 3
is a summary of the Alamar blue viability test. There was significant positive regression progress in absorbance of Alamar blue reading from 4 to 24 h in all seeded grafts, indicating that cells were viable. Significantly higher metabolic activity was recorded on RGD/Hep-bonded grafts than with MyoLink-seeded graft (P<0.01), indicating more cells were present on the RGD/Hep graft after exposure to shear stress. Although there were more cells on the Hep- or RGD-bonded grafts than unbonded MyoLink, this was statistically not significant. This confirmed the cell retention results from radiolabeling. SEM analysis of each bonded graft showed viable cells resembling a confluent endothelium (Fig. 4
).
|
|
| DISCUSSION |
|---|
|
|
|---|
We have developed a new bonding chemistry by surface analysis that enables uniform layers of peptide and heparin to be attached to the surface that allows the inherent bioactivity of each moiety to be retained. Such low sterically hindered bonding allows a covalent linkage to be formed that reduces the loss of cells when exposed to pulsatile blood flow. This technique overcomes the problem of the cells being washed off as has been seen with fibronectin (26)
.
Our results support our hypothesis that a combination of RGD and heparin has better cell attachment (75.7%±2.3%) than RGD or heparin alone; better attachment than in our earlier experiment when the graft was coated with fibronectin (71.2%±9.7%); and significantly better than fibronectin-coated ePTFE (42.0%±8.0) (12)
. The results of RGD/Hep-bonded graft are better than the retention rates seen by RGD-coated ePTFE grafts (75.7±2.3% vs. 62.9±7.5%) (27)
. It is known that using RGD and heparin together produces a synergistic effect (28
, 29)
. This would explain the superior results seen with the RGD/Hep-bonded graft vs. RGD and Hep alone.
Although one must bear in mind that the cells used in this study were HUVEC, which may differ in cell retention from those of adult cells such as those obtained by Zilla and co-workers from human saphenous vein (30)
, we have nevertheless shown a significant difference in degree of cell attachment from any previous published data.
In conclusion, we have shown it is possible to engineer a new type of adhesion molecule that has potential use in the seeding of grafts.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Received for publication November 1, 2001.
Revision received February 11, 2002.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Doraiswamy and R. J. Narayan Vascular tissue engineering by computer-aided laser micromachining Phil Trans R Soc A, April 28, 2010; 368(1917): 1891 - 1912. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. T. Rashid, B. Fuller, G. Hamilton, and A. M. Seifalian Tissue engineering of a hybrid bypass graft for coronary and lower limb bypass surgery FASEB J, June 1, 2008; 22(6): 2084 - 2089. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. J. Turner, M. O. Murphy, C. M. Kielty, C. A. Shuttleworth, R. A. Black, M. J. Humphries, M. G. Walker, and A. E. Canfield {alpha}2(VIII) Collagen Substrata Enhance Endothelial Cell Retention Under Acute Shear Stress Flow via an {alpha}2{beta}1 Integrin-Dependent Mechanism: An In Vitro and In Vivo Study Circulation, August 22, 2006; 114(8): 820 - 829. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. R. Thakrar, V. P. Patel, G. Hamilton, B. J. Fuller, and A. M. Seifalian Vitreous cryopreservation maintains the viscoelastic property of human vascular grafts FASEB J, May 1, 2006; 20(7): 874 - 881. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. M. Nugent and E. R. Edelman Tissue Engineering Therapy for Cardiovascular Disease Circ. Res., May 30, 2003; 92(10): 1068 - 1078. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | SEARCH RESULT |