Acellular porcine and kangaroo aortic valve scaffolds show more intense immune-mediated calcification than cross-linked Toronto SPV valves i
http://www.100md.com
《血管的通路杂志》
a Department of Cardiac Surgery, 5K12, University Hospital Ghent, De Pintelaan 185, BE-9000 Ghent, Belgium
b Veterinary Medicine, Ghent University, Ghent, Belgium
c Histology and Human Anatomy, Ghent University, Ghent, Belgium
d Department of Cardiac Surgery, The Prince Charles Hospital, University of Queensland, Brisbane, Australia
Presented at the 55th International Congress of the European Society for Cardiovascular Surgery, St Petersburg, Russian Federation, May 11–14, 2006.
Abstract
Aim of the study: A major limitation of currently available bioprosthetic valves is their propensity to calcify. At present, one approach in tissue-engineering, uses decellularized, xenogenic scaffolds that are implanted, with the expectation of complete matrix repopulation in vivo. Whether or not such a decellularized matrix will be sufficiently endowed to prevent calcification is unknown. Materials and methods: This study examines the calcification potential of xenogenic biological scaffolds from two species, namely pigs (n=3) and kangaroos (n=3) in the sheep model and compared them to a commercially available glutaraldehyde treated porcine bioprosthetic valve (Toronto SPV) (n=3). Results: Valves and matrices were explanted after 120 days. Histologically (H&E and Von Kossa stain) more calcium was found in the acellular matrices. The mean calcium content (mg/g-dw) of the Toronto SPV valve leaflets was 2.63 mg/g-dw compared to 43.81 mg/g-dw (P=0.12) in kangaroo and 105.08 mg/g-dw (P=0.004) in porcine matrices. On electron microscopy calcific deposits were located between as well as in close association with the collagen fibers in all tissue. In contrast to the cross-linked gluteraldehyde fixed bioprostheses both matrices showed strong immune IgG reaction. Conclusion: Toronto SPV valves calcified significantly less than the tested biological matrices irrespective of species of origin. Surprisingly, xenogenic decelullarized scaffolds are inherently prone to calcification due to a strong immunogenecity.
Key Words: Xenogenic scaffolds; Cross-links; Calcification; Immune response
1. Introduction
The application of tissue engineering to cardiac valves attempts to produce a living organ structurally and functionally comparable to the native valve. An engineered aortic valve could be capable of growth and, like normal native aortic valves, it would maintain its characteristics by regenerating its extra-cellular matrix. The majority of bioprosthetic valves implanted currently are cross-linked with glutaraldehyde. The Toronto SPV valve represents the latest generation of glutaraldehyde treated bioprosthetic heart valves manufactured from porcine aortic valves. The limited durability of bioprosthetic valves has been attributed to altered mechanical properties, antigenic properties of the cells, glutaraldehyde interactions and the calcification potential of cell membrane [1–5].
Current approaches to tissue valve engineering include both the use of decellularized porcine xenogenic tissue and bio-resorbable synthetic scaffolds which are either seeded with cells in vitro before implantation or implanted un-seeded, with expected in vivo repopulation by host cells [6,7]. Acellular biological matrices are devoid of cells and theoretically do not harbor the potential of calcification due to remaining cellular material as do currently available bioprosthetic valves. Moreover, acellular matrices, if not treated with glutaraldehyde, would avoid glutaraldehyde interactions implicated in calcification. Tissue engineered heart valves constructed from xenogenic tissue have already been implanted in humans. The results of these clinical implants however, have been catastrophic for patients due to valve failure and emphasize the need to fully understand the pattern of failure of these constructs [8,9].
2. Aim of the study
In presently used bioprosthetic valves, calcific deterioration remains the major cause of valve failure [10,11]. While a functional tissue-engineered valve is expected to remain vital and regenerate its matrix and thus not calcify, such regeneration is not instantaneous. Recently, we have reported on the hydrodynamic evaluation of kangaroo acellular matrices that suggest superior hydrodynamic properties compared to porcine matrices [12]. Consequently, kangaroo aortic valves might be considered a potential alternative source for tissue valve engineering. The aim of this study was to compare the inherent calcification potential of xenogenic matrices from two species, namely kangaroo and pig, to that of a routinely used glutaraldehyde treated porcine bioprostheses, the Toronto SPV valve.
3. Materials and methods
3.1. Valve procurement and scaffold preparation
St Jude Medical Corporation (St Jude Medical, St Paul, MN, USA) donated the three Toronto SPV valves of 21 mm internal diameter used in this study. Porcine valves were obtained from the slaughterhouse of the Department of Animal Production, Ghent University. Immediately after slaughter, hearts were retrieved and placed on wet ice for transportation. The aortic valves were dissected at the Laboratory of Experimental Cardiac Surgery, University Hospital Ghent as 4-cm long conduits with a 2-mm rim of myocardium and stored in a cold preserving solution of isotonic saline. The kangaroo valves were obtained from Eastern Gray Kangaroos (Macropus Giganteus) under license in Queensland, Australia. Valves were procured at the Queensland Heart Valve Bank, University of Queensland, under similar conditions. Two additional decellularized valves of each type (kangaroo and porcine) were used to check the effectiveness of the decellularization procedure. Acellular matrices are typically prepared by cell lysis in hyper- and hypotonic solutions, with subsequent enzymatic digestion and detergent extraction. Matrices were prepared using a patented detergent-enzymatic protocol, which has been successfully used and reported in the literature [13,14].
3.2. Implantation and explantation of valve and matrices
The Ethical Commission for animal experiments of the University of Ghent approved the study (ECP 04/38). Nine Suffolk sheep (40–45 kg) used in this study were obtained from a licensed supplier in Belgium. The anesthetic and operative procedure employed using right-sided heart bypass and pulmonary valve implantation in sheep has been previously reported in detail. In short, the sheep were pre-medicated intravenously with 0.1 mg/kg midazolam (Dormicum, Roche, Brussels, Belgium) and 0.1 mg/kg methadone (Mephenon, N.V. Demolin S.A., Brussels, Belgium). Anesthesia was induced with 2–4 mg/kg propofol (Diprivan, Astra Zeneca, Destelbergen, Belgium) and maintained with isoflurane (Isoflo, Abbott Laboratories Ltd., Queensborough, Kent, UK) in oxygen, combined with infusions of propofol and fentanyl (Fentanyl-Janssen, Janssen-Cilag, Berchem, Belgium). The heart was exposed by a left antero-lateral thoracotomy via the third inter-costal space. Systemic anticoagulation was induced with 3 mg/kg heparin (Heparine Leo, Leo Pharma, Zaventem, Belgium). Right heart bypass was established by pulmonary and right atrial cannulation. The heart was kept normothermic and beating throughout the whole procedure. The pulmonary artery was clamped, transsected and the test valves and matrices were interposed in the pulmonary trunk using running Prolene 5.0 sutures (Eticon, Merelbeke, Belgium) distally and proximally. Through a separate, lower incision in the pulmonary artery, the native pulmonary valve was rendered incompetent by destruction of its leaflets. Animals were weaned from bypass and heparin was neutralized with 3 mg/kg protamine. An inter-costal block using 0.5% bupivacaine + epinephrine (Marcaine, Astra Pharmaceuticals, Brussels, Belgium) was installed before closing the chest with a temporary thorax drainage system in place. Animals were euthanized after 120 days with an intravenous bolus of 50 mg/kg pentobarbital (Natriumpentobarbital, Kela, Hoogstraten, Belgium). All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health NIH (publication No.85-23, revised 1985).
3.3. Explant analysis
3.3.1. Light microscopy
Samples for histology were fixed in 4% phosphate buffered formaldehyde (Merck, Darmstadt, Germany) and embedded in paraffin. Sections of 5 micron thick were cut and stained with Hematoxylin/Eosine (H/E) and with Von Kossa for calcium (Ca).
3.3.2. Electron microscopy
Samples were fixed in 4% phosphate buffered formaldehyde supplemented with Alcian blue to precipitate and preserve proteoglycans during the dehydration. Specimens were subsequently post fixed with 1% osmium tetroxide (OsO4) in phosphate buffer and embedded in epoxy resin. Ultra-thin 60 nm sections were cut and examined with a Jeol 1200 EX-II transmission electron microscope at 80 keV.
3.3.3. Calcium content
Samples were lyophylized and subsequently mineralized by ashing during 3 h at 450 °C. Calcium content is expressed as milligram per gram of tissue dry weight (mg/g-dw).
3.3.4. Antibody titer determination
Frozen samples of leaflets were weighed and mechanically homogenized in 2 ml PBS buffer at 700 rpm for 10 min on ice. The homogenates were centrifuged at 10,000 g at 18 °C for 30 min. One hundred μl of each supernatant was used for protein determination using the BCA reaction (biuret copper reduction – bicinchonnic acid reaction, Sigma). The remaining supernatant was used for antibody IgG titer determination by an ELISA assay. The ELISA described by de Rooster et al. was slightly modified in coating concentrations [15]. Bound antibodies were detected by adding 100 μl of an appropriate dilution of rabbit anti-sheep immunoglobulin antibodies conjugated to horseradish peroxidase in dilution buffer with 1% pig serum (DakoCytomation, Glostrup, Denmark). Absorbance at 405 nm (OD405) was measured spectrophotometrically after 1 h at 37 °C. The cut-off values for the pig and kangaroo heart valve antigens were calculated as the mean OD405-value of the four control sera (dilution 1/10) increased with 2 times the standard deviation. They were 0.660 and 0.750, respectively. The anti-heart valve antibody titer of a sample is the inverse of the highest dilution that still had an OD405 just above the cut-off value.
3.4. Statistical analysis
One-way analysis of variance on ranks was done by Kruskal-Wallis test with post hoc pairwise multiple comparison procedure by Tukey. The level of significance was set at P<0.05 (SPSS 12.0, SPSS Inc., Chicago, IL, USA).
4. Results
4.1. Light microscopy
4.1.1. Hematoxylin-Eosin
H&E stained histological preparations confirmed the decellularization of the kangaroo and porcine matrices as illustrated in Fig. 1a and b. All explants showed fibrous overgrowth extending from the right ventricle towards the leaflets.
4.1.2. Von Kossa stain
Fig. 2 shows the von Kossa stain of explanted leaflets. Leaflets from the Toronto SPV valves showed significantly less staining compared to either type of matrix (Fig. 2a). Kangaroo matrices (Fig. 2b) stained less intensely compared to porcine matrices (Fig. 2c). In the Toronto SPV explants, the few calcium deposits appeared to follow the distribution of cells. In matrices, calcific deposits were located throughout the whole leaflet and particularly at the free edges.
4.2. Electron microscopy
Electron microscopic sections of explants are illustrated in Fig. 3. Calcific deposits were observed in all valves and matrices. In the Toronto SPV bioprostheses, calcification appeared to be associated with cells and showed a patchy distribution. In the acellular matrices the calcific deposits were distributed throughout the leaflet in close association with the collagen matrix. The larger deposits of calcium were found located between and in association with the collagen fibers. In one preparation in which elastin was observed in an acellular porcine matrix, calcific deposits were also strongly associated with the elastin fibers (Fig. 3c).
4.3. Calcium content
Table 1 illustrates the mean calcium content in milligram per gram of dry weight (mg/g-dw) for each type of tissue. The mean calcium content in Toronto SPV was lower (2.63 mg/g-dw) compared to either kangaroo (43.8 mg/g-dw, with P=0.12) and statistically lower to the porcine (105.8 mg/g-dw, with P=0.004) acellular matrices. Although the significance of the P-value for comparison of means between the porcine and kangaroo matrices appear borderline (P=0.054), bearing in mind that the limited number of observations due to the small sample do not justify a normal distribution in the groups, it was clearly convincing.
4.4. Antibody titer determination
Table 2 shows an absence of IgG response to kangaroo and porcine matrices in the cross-linked Toronto SPV (cut-off =10). By contrast it demonstrates a highly significant increase in IgG antibody titers of both kangaroo (P=0.001) and porcine (P=<0.0001) matrices.
5. Discussion
To our knowledge this is the first study to compare the calcification potential of a routinely used porcine bioprostheses to that of two xenogenic matrices in the right-sided sheep circulatory model. The light microscopic examination of both types of scaffolds in this study confirmed the effectiveness of the decellularization procedure. Both kangaroo and porcine matrices were rendered completely acellular. The most significant finding of this study is that xenogenic matrices calcify significantly more than the Toronto SPV valve, a routinely implanted glutaraldehyde treated porcine bioprosthesis, regardless of whether such matrices were derived from porcine or kangaroo aortic valves.
Several factors have been implicated in the calcification of glutaraldehyde treated porcine bioprosthetic stentless valves. Both the glutaraldehyde treatment as well as cellular components has been implicated in the pathological calcification process. In bioprosthetic valves glutaraldehyde binds covalently to cellular and extra-cellular matrix proteins. However, those initially thermally and chemically stable cross-links become compromised over time. The breakdown of those cross-links results in a leaching out of glutaraldehyde [16]. Such free aldehydes can easily oxidize to carboxylic acid, a potential site for calcium binding [17]. Furthermore, although stable cross-links are considered to reduce immunogenicity, porcine tissue retains a residual ability to trigger an immune response. The immune response activates macrophages, which can turn into an osteoblast calcium depositing phenotype [18]. Moreover, it is well established that cell and cellular debris enhance calcification in biological heart valves, due to a direct relation between specific antibody response and the calcification process [19]. In addition, glutaraldehyde cross-linked valves are and remain non-viable, without opportunity for either growth or tissue renewal which is explained by glutaraldehyde cytotoxicity and the inability of cells to penetrate the cross-linked matrix [17]. In a previous study we failed to detect a difference in calcification potential between kangaroo and porcine aortic valves treated with either high or low concentrations of gluteraldehyde [20].
Xenogenic matrices such as those tested in our study are devoid of cells and are not treated with glutaraldehyde or other cross-linking agent. As such, these constructs might be expected to calcify less than a glutaraldehyde treated porcine bioprostheses. Surprisingly, our xenogenic matrices demonstrated a greater propensity to calcify. Obviously it suggests that their preparation by decellularization and omission of glutaraldehyde treatment is insufficient to completely mitigate calcification. In the present study, we observed deposits of calcium in close association with invasive connective tissue cells as well as in close association with collagen and elastin fibers. Indeed phospholipids and phosphoserine containing proteins have been found in pathologically calcified tissue clearly demonstrating fibrosis at the early sites of calcification. In a recent study, the group of Reider et al. demonstrated residual immunogenicity in porcine xenogenic scaffolds [21]. Compared to decellularized human aortic valves it resulted in a greater monocyte response of U-937 cells, a human monoblastic cell line. The same group has shown that these failed tissue engineered porcine scaffolds possess gal (1,3) gal isotope, known as the major xenoantigen(s) recognized in pigs by human natural antibodies [8,21,22]. In our study we demonstrated unexpected calcification of the xenogenic scaffolds, regardless of their species of origin. The excessive calcification of the decellularized scaffolds is a clear manifestation of an immune response to matrices in which immunogenic moieties persist.
6. Conclusion
In conclusion, xenogenic scaffolds calcify more than the glutaraldehyde treated Toronto SPV porcine bioprosthetic valve despite their effective decellularization and the absence of glutaraldehyde treatment. Tissue-engineered scaffolds will need to out perform current bioprostheses in their calcification potential in order to take their place in the treatment of heart valve disease. The remaining immunogenicity of such scaffolds might help to explain their calcification. Current approaches to tissue valve engineering include attempts to overcome these obstacles by seeding xenogenic or allogenic scaffolds with autologous cells before implantation and thus shielding them from the host's immune system. Further research into rendering matrices immunologically inert is warranted.
References
Ferrans VJ, Spray TL, Billingham ME, Roberts WC. Structural changes in glutaraldehyde treated porcine heterografts used as substitute heart valves: transmission and scanning electron microscopic observations in 17 patients. Am J Cardiol 1978; 41:1159–1184.
Grabenwoger M, Sider J, Fitzal F, Zelenka C, Windberger U, Grimm M, Moritz A, Block P, Wolner E. Impact of glutaraldehyde on calcification of pericardial bioprosthetic heart valve material. Ann Thorac Surg 1996; 62:772–777.
Thubrikar MJ, Deck JD, Aouad J, Nolan SP. Role of mechanical stress in calcification of aortic bioprosthetic valves. J Thorac Cardiovasc Surg 1983; 86:115–125.
Kim MK, Herrera GA, Battarbee HD. Role of glutaraldehyde in calcification of porcine aortic valve fibroblasts. Am J Pathol 1999; 154:843–852.
Human P, Zilla P. Characterization of the immune response to valve prostheses and its role in primary tissue failure. Ann Thorac Surg 2001; 71:S385–388.
Vesely I. Heart valve tissue engineering. Circ Res 2005; 97:743–755.
Goldstein S, Clarke DR, Walsh SP, Black KS, O'Brien MF. Transpecies heart valve transplant: advanced studies of a bioengineered xeno-autograph. Ann Thorac Surg 2000; 70:1962–1969.
Simon P, Kasimir MT, Seebacher G, Weigel G, Ullrich R, Salzer-Muhar U, Reider E, Wolner E. Early failure of the tissue engineered porcine heart valve SYNERGRAFTTM in pediatric patients. Eur J Cardiothorac Surg 2002; 21:703–710.
Dohmen PM, Lembcke A, Hotz H, Kivelitz D, Konertz WF. Ross operation with a tissue engineered heart valve. Ann Thorac Surg 2002; 74:1438–1442.
Schoen FJ, Levy REJ. Calcification of bioprosthetic heart valves. In: Bodnar E, Frater R. editors Replacement cardiac valves 1992;New York: McGraw-Hill 125–148. In.
Rahimtoola SH. Choice of prosthetic heart valve for adult patients. J Am Coll Cardiol 2003; 41:893–904.
Narine K, Kramm K, Dumont K, Jalali H, Sparks L, Segers P, Verdonck P, Van Nooten G. Hydrodynamic evaluation of kangaroo aortic valve matrices for tissue valve engineering. Artificial Organs 2006; 30:432–439.
Booth C, Korossis SA, Wilcox HE, Watterson KG, Kearny JN, Fisher J, Ingham E. Tissue engineering of cardiac valve protheses I: development and histological characterization of an acellular porcine scaffold. J Heart Valve Dis 2002; 11:457–462.
Zeltinger J, Landeen LK, Alexander HG, Kidd IG, Sibanda B. Development and characterization of tissue-engineered aortic valves. Tissue Engineering 2001; 7:9–22.
De Rooster H, Cox E, Van Bree H. Prevalence and relevance of antibodies to type-I and -II collagen in synovial fluid of dogs with cranial cruciate ligament damage. Am J Vet Res 2000; 61:1456–1461.
Huang-lee LLH, Cheung D, Nimni ME. Biochemical changes and cytotoxicity associated with the degradation of polymeric glutaraldehyde derived crosslinks. J Biomed Mater Res 1990; 24:1185–1201.
McIntosh DB. Glutaraldehyde crosslinks Lys-492 and Arg-678 at the active site of sarcoplasmic reticulum Ca+-ATPase. J Biol Chem 1992; 267:22328–22335.
Mohler ER III, Adam P, McClelland P, Graham L, Hathaway DR. Detection of osteopontin in calcified human aortic valves. Arterioscl Throm Vasc Biol 1997; 17:547–552.
Ketchedian A, Krueger P, Lukoff H, Robinson E, Linthurst-Jones A, Crouch K, Wilfinbarger L, Hopkins RA. Ovine panel reactive antibody assay of HLA response to allograft bioengineered vascular scaffolds. J Thorac Cardiovasc Surg 2005; 129:159–166.
Narine K, Chery CC, Goetghebeur E, Forsyth R, Claeys E, Cornelissen M, Moens L, Van Nooten G. Kangaroo vs. porcine aortic valves. Calcification potential after gluteraldehyde fixation. Eur Surg Res 2005; 37:137–143.
Reider E, Seebacher G, Kasimir MT, Eichmair E, Winter B, Simon P, Weigel G. Tissue engineering of heart valves. Decellularized porcine and human valve scaffolds differ importantly in residual potential to attract monocytic cells. Circulation 2005; 111:2792–2797.
Kasimir MT, Reider E, Seebacher G, Wolner E, Weigel G, Simon P. Presence and elimination of the xenoantigen Gal (1,3) Gal in tissue engineered heart valves. Tissue Engineering 2005; 11:1274–1280.(Guido Van Nooten, Pamela )
b Veterinary Medicine, Ghent University, Ghent, Belgium
c Histology and Human Anatomy, Ghent University, Ghent, Belgium
d Department of Cardiac Surgery, The Prince Charles Hospital, University of Queensland, Brisbane, Australia
Presented at the 55th International Congress of the European Society for Cardiovascular Surgery, St Petersburg, Russian Federation, May 11–14, 2006.
Abstract
Aim of the study: A major limitation of currently available bioprosthetic valves is their propensity to calcify. At present, one approach in tissue-engineering, uses decellularized, xenogenic scaffolds that are implanted, with the expectation of complete matrix repopulation in vivo. Whether or not such a decellularized matrix will be sufficiently endowed to prevent calcification is unknown. Materials and methods: This study examines the calcification potential of xenogenic biological scaffolds from two species, namely pigs (n=3) and kangaroos (n=3) in the sheep model and compared them to a commercially available glutaraldehyde treated porcine bioprosthetic valve (Toronto SPV) (n=3). Results: Valves and matrices were explanted after 120 days. Histologically (H&E and Von Kossa stain) more calcium was found in the acellular matrices. The mean calcium content (mg/g-dw) of the Toronto SPV valve leaflets was 2.63 mg/g-dw compared to 43.81 mg/g-dw (P=0.12) in kangaroo and 105.08 mg/g-dw (P=0.004) in porcine matrices. On electron microscopy calcific deposits were located between as well as in close association with the collagen fibers in all tissue. In contrast to the cross-linked gluteraldehyde fixed bioprostheses both matrices showed strong immune IgG reaction. Conclusion: Toronto SPV valves calcified significantly less than the tested biological matrices irrespective of species of origin. Surprisingly, xenogenic decelullarized scaffolds are inherently prone to calcification due to a strong immunogenecity.
Key Words: Xenogenic scaffolds; Cross-links; Calcification; Immune response
1. Introduction
The application of tissue engineering to cardiac valves attempts to produce a living organ structurally and functionally comparable to the native valve. An engineered aortic valve could be capable of growth and, like normal native aortic valves, it would maintain its characteristics by regenerating its extra-cellular matrix. The majority of bioprosthetic valves implanted currently are cross-linked with glutaraldehyde. The Toronto SPV valve represents the latest generation of glutaraldehyde treated bioprosthetic heart valves manufactured from porcine aortic valves. The limited durability of bioprosthetic valves has been attributed to altered mechanical properties, antigenic properties of the cells, glutaraldehyde interactions and the calcification potential of cell membrane [1–5].
Current approaches to tissue valve engineering include both the use of decellularized porcine xenogenic tissue and bio-resorbable synthetic scaffolds which are either seeded with cells in vitro before implantation or implanted un-seeded, with expected in vivo repopulation by host cells [6,7]. Acellular biological matrices are devoid of cells and theoretically do not harbor the potential of calcification due to remaining cellular material as do currently available bioprosthetic valves. Moreover, acellular matrices, if not treated with glutaraldehyde, would avoid glutaraldehyde interactions implicated in calcification. Tissue engineered heart valves constructed from xenogenic tissue have already been implanted in humans. The results of these clinical implants however, have been catastrophic for patients due to valve failure and emphasize the need to fully understand the pattern of failure of these constructs [8,9].
2. Aim of the study
In presently used bioprosthetic valves, calcific deterioration remains the major cause of valve failure [10,11]. While a functional tissue-engineered valve is expected to remain vital and regenerate its matrix and thus not calcify, such regeneration is not instantaneous. Recently, we have reported on the hydrodynamic evaluation of kangaroo acellular matrices that suggest superior hydrodynamic properties compared to porcine matrices [12]. Consequently, kangaroo aortic valves might be considered a potential alternative source for tissue valve engineering. The aim of this study was to compare the inherent calcification potential of xenogenic matrices from two species, namely kangaroo and pig, to that of a routinely used glutaraldehyde treated porcine bioprostheses, the Toronto SPV valve.
3. Materials and methods
3.1. Valve procurement and scaffold preparation
St Jude Medical Corporation (St Jude Medical, St Paul, MN, USA) donated the three Toronto SPV valves of 21 mm internal diameter used in this study. Porcine valves were obtained from the slaughterhouse of the Department of Animal Production, Ghent University. Immediately after slaughter, hearts were retrieved and placed on wet ice for transportation. The aortic valves were dissected at the Laboratory of Experimental Cardiac Surgery, University Hospital Ghent as 4-cm long conduits with a 2-mm rim of myocardium and stored in a cold preserving solution of isotonic saline. The kangaroo valves were obtained from Eastern Gray Kangaroos (Macropus Giganteus) under license in Queensland, Australia. Valves were procured at the Queensland Heart Valve Bank, University of Queensland, under similar conditions. Two additional decellularized valves of each type (kangaroo and porcine) were used to check the effectiveness of the decellularization procedure. Acellular matrices are typically prepared by cell lysis in hyper- and hypotonic solutions, with subsequent enzymatic digestion and detergent extraction. Matrices were prepared using a patented detergent-enzymatic protocol, which has been successfully used and reported in the literature [13,14].
3.2. Implantation and explantation of valve and matrices
The Ethical Commission for animal experiments of the University of Ghent approved the study (ECP 04/38). Nine Suffolk sheep (40–45 kg) used in this study were obtained from a licensed supplier in Belgium. The anesthetic and operative procedure employed using right-sided heart bypass and pulmonary valve implantation in sheep has been previously reported in detail. In short, the sheep were pre-medicated intravenously with 0.1 mg/kg midazolam (Dormicum, Roche, Brussels, Belgium) and 0.1 mg/kg methadone (Mephenon, N.V. Demolin S.A., Brussels, Belgium). Anesthesia was induced with 2–4 mg/kg propofol (Diprivan, Astra Zeneca, Destelbergen, Belgium) and maintained with isoflurane (Isoflo, Abbott Laboratories Ltd., Queensborough, Kent, UK) in oxygen, combined with infusions of propofol and fentanyl (Fentanyl-Janssen, Janssen-Cilag, Berchem, Belgium). The heart was exposed by a left antero-lateral thoracotomy via the third inter-costal space. Systemic anticoagulation was induced with 3 mg/kg heparin (Heparine Leo, Leo Pharma, Zaventem, Belgium). Right heart bypass was established by pulmonary and right atrial cannulation. The heart was kept normothermic and beating throughout the whole procedure. The pulmonary artery was clamped, transsected and the test valves and matrices were interposed in the pulmonary trunk using running Prolene 5.0 sutures (Eticon, Merelbeke, Belgium) distally and proximally. Through a separate, lower incision in the pulmonary artery, the native pulmonary valve was rendered incompetent by destruction of its leaflets. Animals were weaned from bypass and heparin was neutralized with 3 mg/kg protamine. An inter-costal block using 0.5% bupivacaine + epinephrine (Marcaine, Astra Pharmaceuticals, Brussels, Belgium) was installed before closing the chest with a temporary thorax drainage system in place. Animals were euthanized after 120 days with an intravenous bolus of 50 mg/kg pentobarbital (Natriumpentobarbital, Kela, Hoogstraten, Belgium). All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health NIH (publication No.85-23, revised 1985).
3.3. Explant analysis
3.3.1. Light microscopy
Samples for histology were fixed in 4% phosphate buffered formaldehyde (Merck, Darmstadt, Germany) and embedded in paraffin. Sections of 5 micron thick were cut and stained with Hematoxylin/Eosine (H/E) and with Von Kossa for calcium (Ca).
3.3.2. Electron microscopy
Samples were fixed in 4% phosphate buffered formaldehyde supplemented with Alcian blue to precipitate and preserve proteoglycans during the dehydration. Specimens were subsequently post fixed with 1% osmium tetroxide (OsO4) in phosphate buffer and embedded in epoxy resin. Ultra-thin 60 nm sections were cut and examined with a Jeol 1200 EX-II transmission electron microscope at 80 keV.
3.3.3. Calcium content
Samples were lyophylized and subsequently mineralized by ashing during 3 h at 450 °C. Calcium content is expressed as milligram per gram of tissue dry weight (mg/g-dw).
3.3.4. Antibody titer determination
Frozen samples of leaflets were weighed and mechanically homogenized in 2 ml PBS buffer at 700 rpm for 10 min on ice. The homogenates were centrifuged at 10,000 g at 18 °C for 30 min. One hundred μl of each supernatant was used for protein determination using the BCA reaction (biuret copper reduction – bicinchonnic acid reaction, Sigma). The remaining supernatant was used for antibody IgG titer determination by an ELISA assay. The ELISA described by de Rooster et al. was slightly modified in coating concentrations [15]. Bound antibodies were detected by adding 100 μl of an appropriate dilution of rabbit anti-sheep immunoglobulin antibodies conjugated to horseradish peroxidase in dilution buffer with 1% pig serum (DakoCytomation, Glostrup, Denmark). Absorbance at 405 nm (OD405) was measured spectrophotometrically after 1 h at 37 °C. The cut-off values for the pig and kangaroo heart valve antigens were calculated as the mean OD405-value of the four control sera (dilution 1/10) increased with 2 times the standard deviation. They were 0.660 and 0.750, respectively. The anti-heart valve antibody titer of a sample is the inverse of the highest dilution that still had an OD405 just above the cut-off value.
3.4. Statistical analysis
One-way analysis of variance on ranks was done by Kruskal-Wallis test with post hoc pairwise multiple comparison procedure by Tukey. The level of significance was set at P<0.05 (SPSS 12.0, SPSS Inc., Chicago, IL, USA).
4. Results
4.1. Light microscopy
4.1.1. Hematoxylin-Eosin
H&E stained histological preparations confirmed the decellularization of the kangaroo and porcine matrices as illustrated in Fig. 1a and b. All explants showed fibrous overgrowth extending from the right ventricle towards the leaflets.
4.1.2. Von Kossa stain
Fig. 2 shows the von Kossa stain of explanted leaflets. Leaflets from the Toronto SPV valves showed significantly less staining compared to either type of matrix (Fig. 2a). Kangaroo matrices (Fig. 2b) stained less intensely compared to porcine matrices (Fig. 2c). In the Toronto SPV explants, the few calcium deposits appeared to follow the distribution of cells. In matrices, calcific deposits were located throughout the whole leaflet and particularly at the free edges.
4.2. Electron microscopy
Electron microscopic sections of explants are illustrated in Fig. 3. Calcific deposits were observed in all valves and matrices. In the Toronto SPV bioprostheses, calcification appeared to be associated with cells and showed a patchy distribution. In the acellular matrices the calcific deposits were distributed throughout the leaflet in close association with the collagen matrix. The larger deposits of calcium were found located between and in association with the collagen fibers. In one preparation in which elastin was observed in an acellular porcine matrix, calcific deposits were also strongly associated with the elastin fibers (Fig. 3c).
4.3. Calcium content
Table 1 illustrates the mean calcium content in milligram per gram of dry weight (mg/g-dw) for each type of tissue. The mean calcium content in Toronto SPV was lower (2.63 mg/g-dw) compared to either kangaroo (43.8 mg/g-dw, with P=0.12) and statistically lower to the porcine (105.8 mg/g-dw, with P=0.004) acellular matrices. Although the significance of the P-value for comparison of means between the porcine and kangaroo matrices appear borderline (P=0.054), bearing in mind that the limited number of observations due to the small sample do not justify a normal distribution in the groups, it was clearly convincing.
4.4. Antibody titer determination
Table 2 shows an absence of IgG response to kangaroo and porcine matrices in the cross-linked Toronto SPV (cut-off =10). By contrast it demonstrates a highly significant increase in IgG antibody titers of both kangaroo (P=0.001) and porcine (P=<0.0001) matrices.
5. Discussion
To our knowledge this is the first study to compare the calcification potential of a routinely used porcine bioprostheses to that of two xenogenic matrices in the right-sided sheep circulatory model. The light microscopic examination of both types of scaffolds in this study confirmed the effectiveness of the decellularization procedure. Both kangaroo and porcine matrices were rendered completely acellular. The most significant finding of this study is that xenogenic matrices calcify significantly more than the Toronto SPV valve, a routinely implanted glutaraldehyde treated porcine bioprosthesis, regardless of whether such matrices were derived from porcine or kangaroo aortic valves.
Several factors have been implicated in the calcification of glutaraldehyde treated porcine bioprosthetic stentless valves. Both the glutaraldehyde treatment as well as cellular components has been implicated in the pathological calcification process. In bioprosthetic valves glutaraldehyde binds covalently to cellular and extra-cellular matrix proteins. However, those initially thermally and chemically stable cross-links become compromised over time. The breakdown of those cross-links results in a leaching out of glutaraldehyde [16]. Such free aldehydes can easily oxidize to carboxylic acid, a potential site for calcium binding [17]. Furthermore, although stable cross-links are considered to reduce immunogenicity, porcine tissue retains a residual ability to trigger an immune response. The immune response activates macrophages, which can turn into an osteoblast calcium depositing phenotype [18]. Moreover, it is well established that cell and cellular debris enhance calcification in biological heart valves, due to a direct relation between specific antibody response and the calcification process [19]. In addition, glutaraldehyde cross-linked valves are and remain non-viable, without opportunity for either growth or tissue renewal which is explained by glutaraldehyde cytotoxicity and the inability of cells to penetrate the cross-linked matrix [17]. In a previous study we failed to detect a difference in calcification potential between kangaroo and porcine aortic valves treated with either high or low concentrations of gluteraldehyde [20].
Xenogenic matrices such as those tested in our study are devoid of cells and are not treated with glutaraldehyde or other cross-linking agent. As such, these constructs might be expected to calcify less than a glutaraldehyde treated porcine bioprostheses. Surprisingly, our xenogenic matrices demonstrated a greater propensity to calcify. Obviously it suggests that their preparation by decellularization and omission of glutaraldehyde treatment is insufficient to completely mitigate calcification. In the present study, we observed deposits of calcium in close association with invasive connective tissue cells as well as in close association with collagen and elastin fibers. Indeed phospholipids and phosphoserine containing proteins have been found in pathologically calcified tissue clearly demonstrating fibrosis at the early sites of calcification. In a recent study, the group of Reider et al. demonstrated residual immunogenicity in porcine xenogenic scaffolds [21]. Compared to decellularized human aortic valves it resulted in a greater monocyte response of U-937 cells, a human monoblastic cell line. The same group has shown that these failed tissue engineered porcine scaffolds possess gal (1,3) gal isotope, known as the major xenoantigen(s) recognized in pigs by human natural antibodies [8,21,22]. In our study we demonstrated unexpected calcification of the xenogenic scaffolds, regardless of their species of origin. The excessive calcification of the decellularized scaffolds is a clear manifestation of an immune response to matrices in which immunogenic moieties persist.
6. Conclusion
In conclusion, xenogenic scaffolds calcify more than the glutaraldehyde treated Toronto SPV porcine bioprosthetic valve despite their effective decellularization and the absence of glutaraldehyde treatment. Tissue-engineered scaffolds will need to out perform current bioprostheses in their calcification potential in order to take their place in the treatment of heart valve disease. The remaining immunogenicity of such scaffolds might help to explain their calcification. Current approaches to tissue valve engineering include attempts to overcome these obstacles by seeding xenogenic or allogenic scaffolds with autologous cells before implantation and thus shielding them from the host's immune system. Further research into rendering matrices immunologically inert is warranted.
References
Ferrans VJ, Spray TL, Billingham ME, Roberts WC. Structural changes in glutaraldehyde treated porcine heterografts used as substitute heart valves: transmission and scanning electron microscopic observations in 17 patients. Am J Cardiol 1978; 41:1159–1184.
Grabenwoger M, Sider J, Fitzal F, Zelenka C, Windberger U, Grimm M, Moritz A, Block P, Wolner E. Impact of glutaraldehyde on calcification of pericardial bioprosthetic heart valve material. Ann Thorac Surg 1996; 62:772–777.
Thubrikar MJ, Deck JD, Aouad J, Nolan SP. Role of mechanical stress in calcification of aortic bioprosthetic valves. J Thorac Cardiovasc Surg 1983; 86:115–125.
Kim MK, Herrera GA, Battarbee HD. Role of glutaraldehyde in calcification of porcine aortic valve fibroblasts. Am J Pathol 1999; 154:843–852.
Human P, Zilla P. Characterization of the immune response to valve prostheses and its role in primary tissue failure. Ann Thorac Surg 2001; 71:S385–388.
Vesely I. Heart valve tissue engineering. Circ Res 2005; 97:743–755.
Goldstein S, Clarke DR, Walsh SP, Black KS, O'Brien MF. Transpecies heart valve transplant: advanced studies of a bioengineered xeno-autograph. Ann Thorac Surg 2000; 70:1962–1969.
Simon P, Kasimir MT, Seebacher G, Weigel G, Ullrich R, Salzer-Muhar U, Reider E, Wolner E. Early failure of the tissue engineered porcine heart valve SYNERGRAFTTM in pediatric patients. Eur J Cardiothorac Surg 2002; 21:703–710.
Dohmen PM, Lembcke A, Hotz H, Kivelitz D, Konertz WF. Ross operation with a tissue engineered heart valve. Ann Thorac Surg 2002; 74:1438–1442.
Schoen FJ, Levy REJ. Calcification of bioprosthetic heart valves. In: Bodnar E, Frater R. editors Replacement cardiac valves 1992;New York: McGraw-Hill 125–148. In.
Rahimtoola SH. Choice of prosthetic heart valve for adult patients. J Am Coll Cardiol 2003; 41:893–904.
Narine K, Kramm K, Dumont K, Jalali H, Sparks L, Segers P, Verdonck P, Van Nooten G. Hydrodynamic evaluation of kangaroo aortic valve matrices for tissue valve engineering. Artificial Organs 2006; 30:432–439.
Booth C, Korossis SA, Wilcox HE, Watterson KG, Kearny JN, Fisher J, Ingham E. Tissue engineering of cardiac valve protheses I: development and histological characterization of an acellular porcine scaffold. J Heart Valve Dis 2002; 11:457–462.
Zeltinger J, Landeen LK, Alexander HG, Kidd IG, Sibanda B. Development and characterization of tissue-engineered aortic valves. Tissue Engineering 2001; 7:9–22.
De Rooster H, Cox E, Van Bree H. Prevalence and relevance of antibodies to type-I and -II collagen in synovial fluid of dogs with cranial cruciate ligament damage. Am J Vet Res 2000; 61:1456–1461.
Huang-lee LLH, Cheung D, Nimni ME. Biochemical changes and cytotoxicity associated with the degradation of polymeric glutaraldehyde derived crosslinks. J Biomed Mater Res 1990; 24:1185–1201.
McIntosh DB. Glutaraldehyde crosslinks Lys-492 and Arg-678 at the active site of sarcoplasmic reticulum Ca+-ATPase. J Biol Chem 1992; 267:22328–22335.
Mohler ER III, Adam P, McClelland P, Graham L, Hathaway DR. Detection of osteopontin in calcified human aortic valves. Arterioscl Throm Vasc Biol 1997; 17:547–552.
Ketchedian A, Krueger P, Lukoff H, Robinson E, Linthurst-Jones A, Crouch K, Wilfinbarger L, Hopkins RA. Ovine panel reactive antibody assay of HLA response to allograft bioengineered vascular scaffolds. J Thorac Cardiovasc Surg 2005; 129:159–166.
Narine K, Chery CC, Goetghebeur E, Forsyth R, Claeys E, Cornelissen M, Moens L, Van Nooten G. Kangaroo vs. porcine aortic valves. Calcification potential after gluteraldehyde fixation. Eur Surg Res 2005; 37:137–143.
Reider E, Seebacher G, Kasimir MT, Eichmair E, Winter B, Simon P, Weigel G. Tissue engineering of heart valves. Decellularized porcine and human valve scaffolds differ importantly in residual potential to attract monocytic cells. Circulation 2005; 111:2792–2797.
Kasimir MT, Reider E, Seebacher G, Wolner E, Weigel G, Simon P. Presence and elimination of the xenoantigen Gal (1,3) Gal in tissue engineered heart valves. Tissue Engineering 2005; 11:1274–1280.(Guido Van Nooten, Pamela )