Sjnw1019-04-2301.tex

J Mater Sci: Mater Med (2007) 18:545-550DOI 10.1007/s10856-007-2301-9 Effects of sterilization on an extracellular matrix scaffold:
Part II. Bioactivity and matrix interaction
Jason Hodde . Abram Janis . Michael Hiles
Received: 8 February 2005 / Accepted: 20 March 2006 C Springer Science + Business Media, LLC 2007 Abstract Small intestinal submucosa (SIS) has been suc-
cal applications [1-4] and has also been reported to stimulate cessfully used to treat a variety of damaged or diseased tis- the repair of chronic venous ulcers and other non-healing sues in human patients. As a biologic scaffold, SIS stimulates wounds [5]. In its natural form, SIS consists of several types repair of damaged or diseased tissues and organs with tissue of collagens [6], with smaller amounts of glycosaminogly- that is similar in structure and function to the material it was cans [7], glycoproteins [8], and growth factors [9,10].
meant to replace. To meet clinical safety requirements, bio- Retention of the non-collagenous matrix components in logic materials from animal tissues must undergo processing their natural state is essential to the maintenance of scaffold treatments to minimize host immune response and to elim- bioactivity, but biologic scaffolds used to stimulate wound inate the possibility of disease transmission. The effect of healing undergo a variety of potentially damaging process- peracetic acid disinfection, lyophilization, and ethylene ox- ing treatments designed to minimize host immune response ide sterilization on the in vitro bioactivity of the processed SIS and reduce endotoxin and bioburden levels to insure prod- was therefore examined in murine fibroblasts and pheochro- uct safety. These processes often include processing steps mocytoma (PC12) cells. Specifically, the ability of processed that subject the scaffold to acids, enzymes, or other chemical SIS to support fibroblast attachment, to stimulate PC12 cell treatments that can denature growth factors and inhibit the differentiation, and to upregulate fibroblast VEGF secretion ability of structural proteins to interact with the recipient's was examined. Fibroblasts attach to the sterilized SIS, re- cells. For example, crosslinking agents such as glutaralde- main viable, and more than double their secretion of VEGF hyde or hexamethylene diisocyanate (HMDI) are often used as a result of interacting with the SIS matrix components.
during the processing of these biologic scaffolds to increase Additionally, PC12 cells exhibit increased neurite outgrowth implant strength and reduce their antigenicity, but these com- following stimulation by SIS matrix proteins versus controls.
pounds also reduce the ability of cells to interact with the We conclude that a biologic scaffold can be prepared for hu- treated material [11], to cause cutaneous sensitization [12], man use and still retain significant bioactivity.
and to lead to calcification [13]. Enzymes such as trypsin,amylase, and neuramidase are used to reduce rejection po-tential of the scaffold, but they remove potentially valuable 1 Introduction
globular proteins, such as growth factors, that contribute tothe bioactivity of the matrix. Oxidizing agents such as hydro- Retention of the bioactivity within collagen-based biomate- gen peroxide or peracetic acid (PAA) are often used as disin- rials is essential if the material is to act as more than just fectants during processing, but have the undesirable effects an inert scaffold. Small intestinal submucosa (SIS), a natural of oxidizing the matrix and reducing the structural integrity biologic scaffold material, has been used in numerous surgi- of the collagen fibers. Oxidative destruction of glycosamino-glycans fractures them, impairs their ability to interact withgrowth factors [14], and subsequently increases the suscep- tibility of growth factors to denaturation.
Cook Biotech Incorporated, 1425 Innovation Place, WestLafayette, IN 47906 Even though these processing steps are destructive to scaf- fold bioactivity, studies have suggested that the composition J Mater Sci: Mater Med (2007) 18:545-550 and activity of growth factors bound in the matrix can be mM sodium pyruvate, 100 units/ml penicillin, 100 µg/ml retained if individual protein components are sequestered streptomycin, and 5% fetal bovine serum. Cells from limited naturally [15] or in combination with other carriers [16-19].
passage numbers were used for all experiments.
Some growth factors, such as FGF-2, are inherently more sta-ble than others (such as VEGF) and are able to retain some 2.3 Procurement and Processing of Small Intestine of their activity under acidic and oxidizing conditions even in the absence of stabilizing agents [20, 21].
The purposes of this study were to determine if processed Sections of porcine jejunum were subjected to treatment SIS was able: (1) to support fibroblast attachment and pro- with a dilute concentration of PAA in water for two hours liferation; (2) to stimulate the differentiation of rat PC12 at room temperature, as described elsewhere [22]. Follow- pheochromocytoma cells; and (3) to stimulate fibroblasts to ing exposure to the disinfectant, the SIS was prepared as secrete VEGF, a potential mechanism for promoting angio- previously described by mechanical delamination [23]. The genesis and wound healing in vivo.
prepared SIS (SISPAA) was either stored at 4◦C in sterilecontainers prior to further evaluation or frozen at −80◦Cand lyophilized overnight (SISLYO) to produce a dry sheet.
2 Materials and methods
Following lyophilization, the SISLYO was packaged into gaspermeable pouches and sterilized with ethylene oxide (EO) gas. EO-sterilized SIS (SISEO) was stored sterile at roomtemperature prior to evaluation.
PAA was obtained from FMC (Chicago, IL). Human recom-binant basic fibroblast growth factor (FGF-2) was purchased from Boehringer Mannheim (Indianapolis, IN) and used at10 ng/ml. Mouse nerve growth factor (NGF) was from Roche SISPAA, SISLYO, or SISEO were placed into polypropylene Diagnostics (Indianapolis, IN) and used at 50 ng/ml. Neu- cell culture inserts and evaluated in triplicate for their abil- tralizing anti-human FGF-2 and neutralizing anti-rat NGF ity to support the attachment and viability of Swiss 3T3 fi- antibodies were purchased from R&D Systems. The anti- broblasts. Three individual assays were performed. The SIS FGF-2 was used at 40 µg/ml and the anti-NGF was used at samples were equilibrated in PBS, pH 7.4 for 30 minutes 150 ng/ml. Human FGF-2 and mouse VEGF multiplex mi- at 37◦C and 5% CO2. The PBS was aspirated and complete crosphere analyte sets (Flourokine MAP system) were pur- DMEM containing 0.5% bovine serum albumin (attachment chased from R&D Systems and used with a mouse base kit medium) was added to the wells. Fibroblasts were harvested also from R&D Systems. AlamarBlue was from Biosource from their tissue culture flasks, counted, and suspended in International (Camarillo, CA). Unless otherwise noted, all attachment medium at a concentration of 125,000 cells/ml.
other chemicals and cell culture reagents were from Sigma A total of 50,000 cells were added to each SIS substrate.
Samples were allowed to incubate at 37◦C and 5% CO2 for60 minutes for attachment to occur. After 60 minutes, the SIS substrates were inverted and centrifuged at 250 x g for 5 min-utes. The attachment medium was carefully aspirated so the Swiss 3T3 fibroblasts were purchased from ATCC and cul- cells that had attached to the SIS would remain undisturbed.
tured as directed in Dulbecco's modified Eagle's medium The SIS substrates were inverted to their original orienta- (DMEM) supplemented with 4 mM Lglutamine and adjusted tion and treated with 10% alamarBlue solution at 37◦C and to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 1.0 5% CO2 for 18 h. Following incubation, 100 µl samples of mM sodium pyruvate, 100 units/ml penicillin, 100 µg/ml alamarBlue solution from each of the substrate wells were streptomycin, and 10% bovine calf serum.
pipetted into a 96-well plate (Corning Costar, Cambridge, Rat Pheochromocytoma (PC12) cells were purchased from ATCC (Manassas, VA) and maintained as directed in Because alamarBlue is a metabolic dye that is reduced lin- RPMI 1640 cell culture medium (Invitrogen Corporation, early with cell metabolic activity, a ratio of the reduced form Carlsbad, CA) supplemented with 5% fetal bovine serum to the oxidized form can be used to measure cell viability. Ab- (FBS), 10% horse serum (HS), 100 units/ml penicillin, and sorbance was measured and 570 nm (reduced form) and 600 nm (oxidized form). The percent of alamarBlue reduction in NIH 3T3 fibroblasts were purchased from ATCC and cul- each SIS well was measured and compared against control tured as directed in Dulbecco's modified Eagle's medium wells in order to correlate to cell number. The results pre- (DMEM) supplemented with 4 mM Lglutamine and adjusted sented represent the percentage of seeded cells that attached to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 1.0 to the substrate for 60 minutes and remained metabolically J Mater Sci: Mater Med (2007) 18:545-550 active during the 18-hour incubation period. Pairwise com- At the end of 24 h, the level of VEGF in the collected cell parisons were made between groups using a 2-tailed t-test.
VEGF and FGF-2 in the cell culture supernatants were mea- Conditioned cell culture medium was prepared from SISEO.
sured using multiplex technology. Briefly, VEGF standards Briefly, serum-free RPMI 1640 cell culture medium was in- diluted in assay buffer and collected samples (200 µl) were cubated with the processed SIS at a concentration of 1 g (wet added to wells of a pre-wetted 96-well filter plate. The sam- weight) / 7.5 ml of medium for 48 h at 37◦C. The SISEO was ples were incubated with 50 µl of the antibody-coupled mi- removed and the medium was filtered through a 20 µm filter crospheres (anti-VEGF and anti-FGF-2 coupled beads) at to remove any particulate matter. The SISEO conditioned cell 25◦C for 3 hr on a plate shaker set to 500 rpm. Wells were culture medium was supplemented with 5% FBS, 10% HS, washed 3 times with 100 µl/wash of buffer, and 50 µl of 100 units/ml penicillin, and 100 µg/ml streptomycin.
freshly diluted secondary/detection antibody was added. The Twelve-well culture plates were coated overnight at 37◦C assay plate was incubated at 25◦C with constant shaking for with 1 ml of 0.2 mg/ml type I rat tail collagen (BD Bio- an additional 60 min. The wash step was repeated, 50 µl sciences, Bedford, MA). The collagen solution was aspirated of streptavidin-PE was added to the wells, and the incuba- and the wells were washed once with phosphate buffered tion was continued under constant shaking for an additional saline (PBS). PC12 cells were harvested, counted using a 30 minutes. The wash was repeated a final time, and the hemacytometer, and seeded in 1 ml of growth medium at microspheres were resuspended in 100 µl assay buffer for 2 a density of 20,000 cells per well. Growth factors and/or minutes with constant shaking. Fifty microliter samples were antibodies were then added at the aforementioned concen- then analyzed for median relative fluorescence (RFU) on the trations. All conditions were evaluated in triplicate.
Luminex 100 (Bio-Plex Suspension Array System, Bio-Rad The cells were cultured for 48 h at 37◦C before being vi- Laboratories) according to the manufacturer's instructions, sually examined for differentiation. Cells were considered with settings set to detect 50 events/bead and doublet dis- differentiated if they exhibited at least one neurite-like ex- criminator gates set at 4335 and 7990.
tension at least twice the diameter of the cell body. Three VEGF in the samples was quantitated using a cubic spline separate, 20X fields were examined per well; the percentage curve fit. Pairwise comparisons were made between groups of cells that met the criteria for differentiation was recorded.
using a 2-tailed t-test. Because no standard curve was created Statistical significance was set at p < 0.05 and evaluated for FGF-2, FGF-2 levels in the supernatants could not be calculated, but the median RFU was evaluated for changesin levels over time. Moreover, the absence of fluorescence on the FGF-2 beads in the wells treated with anti-FGF-2 wasverified to validate the neutralizing activity of the antigen- Conditioned cell culture medium was prepared from SISEO.
Briefly, serum-free DMEM cell culture medium was incu-bated with each ECM at a concentration of 1 g (wet weight)/ 7.5 ml of medium for 48 h at 37◦C. The ECM was removed 3 Results
and the medium was filtered through a 0.2-µm filter. Theconditioned cell culture medium was supplemented with 100 3.1 Cell attachment is retained following sterilization units/ml penicillin, and 100 µg/ml streptomycin prior to use.
NIH 3T3 fibroblasts at 75% confluency were harvested Because rapid interaction between the cells and the ECM is from their tissue culture flasks, counted, and resuspended in a desirable characteristic of tissue engineered matrices, the complete DMEM at a concentration of 50,000 cells / well effects of lyophilization and sterilization on host cell attach- in a 24-well plate and allowed to recover overnight. The ment and short term viability were evaluated. SISPAA, SISLYO, media was aspirated and replaced with serum-free DMEM or SISEO were placed into polypropylene cell culture inserts for 24 h prior to the assay. Cells were cultured under one and evaluated for their ability to support the attachment of of three conditions: 1) serum-free DMEM; 2) conditioned, serum-free, cell culture medium; or 3) conditioned, serum- The alamarBlue assay for attachment and viability of fi- free, cell culture medium containing anti-FGF-2 at 40 µg/ml broblasts indicated that 66% of the seeded cells attached to to neutralize FGF-2 activity in the wells. Cells were incubated the SISPAA during the 1-hour incubation period and remained at 37◦C and 5% CO2 for 24 h. During the growth period, 200 viable for an additional 18 h. Additionally, SISLYO supported µl aliquots of media were sampled at 0, 1, 4, 8, 16, and 24 h.
the viability and attachment of 57% of the seeded cells. SISEO J Mater Sci: Mater Med (2007) 18:545-550 % of Cells with Neurites
% of control
PAA Treated SIS
Lyophilized SIS
EO Sterilized SIS
Fig. 1 Cell attachment and viability on SIS at different levels of pro-
cessing indicates that 3T3 fibroblasts are able to attach and remainviable on lyophilized and EO sterilized SIS. Groups are not statistically Fig. 2 Addition of 10 ng/ml of FGF-2 or 50 ng/ml NGF into the cell
different using paired Student's t-tests culture medium of PC12 cells results in neurite formation after 48 h.
Blocking antibodies abolish this effect. Treatment of PC12 with SISconditioned cell culture medium similarly stimulates neurite formation,partially through an FGF-2 dependent mechanism. Lines between barsrepresent significant differences supported the attachment and viability of this cell line equallyas well as the other forms of SIS, with 51% of the seededcells attaching to the substrate in the first hour after seeding on the matrix (Fig. 1). There was no statistical difference in the percentage of viable cells attached to the SIS matricesafter 60 minutes (SISPAA vs. SISLYO, p = 0.170; SISLYO vs.
FGF-2 is known to stimulate fibroblast proliferation and se- SISEO, p = 0.274; SISPAA vs. SISEO, p = 0.090). Because cretion of VEGF [25]. To further test whether FGF-2 ac- no statistical difference between groups was indicated, all re- maining tests were only performed on terminally-sterilized, measured and compared to wells in which FGF-2 activitywas neutralized. Cell culture supernatants from SISEO con-ditioned cell culture medium contained significantly greater 3.2 FGF-2 from SISEO stimulates PC12 differentiation levels of VEGF at 16 and 24 h than either supernatants ob-tained from cells cultured in DMEM alone or from cells PC12 cells form neurite-like extensions in the presence of cultured with SISEOconditioned medium and anti-FGF-2 laminin, NGF and/or FGF-2, but fail to differentiate in the (Fig. 3a). Levels of FGF-2 in the supernatants did not change absence of these added growth factors [24]. In control wells over time (Fig. 3b), showing that the increase in VEGF de- where cells were seeded in complete RPMI 1640 cell culture tected was not secondary to increased endogenous secretion medium, no cells formed neurite-like extensions after 48 h.
When the RPMI was supplemented with 50 ng/ml NGF or10 ng/ml FGF-2, many of the cells began to differentiate byforming neurites. Addition of antibodies to the cell culturemedium significantly reduced the degree of differentiation 4 Discussion
observed. Results are displayed in Fig. 2.
Media conditioned with SISEO was tested for its ability Previous studies have shown that it is possible to retain much to promote differentiation of PC12 cells. After 48 h in the of the composition and 3-dimensional architecture of the presence of the SISEO conditioned media, the PC12 cells ECM when it is processed using peracetic acid and ethy- were attached to the collagen-coated plate and 21.3 ± 5.4% lene oxide gas [21]. However, retention of identifiable matrix of them had begun to form neurite-like extensions. When a components does not necessarily mean that they are retained neutralizing antibody to FGF-2 was added to the cell cul- in a form that can interact with cells and cause meaning- ture conditions, differentiation was reduced to 16.6 ± 3.0% ful downstream effects, such as cell differentiation or secre- (p<0.05). A neutralizing antibody to NGF failed to reduce tion of growth factors. Retention of the bioactivity inherent the number of cells displaying differentiation after 48 h.
within the processed ECM is essential to its ability to incite J Mater Sci: Mater Med (2007) 18:545-550 each other or with cells [14, 28]. In this study, we have shown that these processes do not alter fibroblast interaction with and viability upon SISEO. Furthermore, these processes do not alter the bioactivity of FGF-2, a pro-angiogenic growth factor important in wound healing. Specifically, the abilityof SISEO to stimulate the differentiation of PC12 cells was not impaired by lyophilization and sterilization. An activity VEGF (pg/ml)
neutralizing antibody specific for FGF-2 caused a significantdecrease in the percentage of differentiated cells in the pres- ence of SISEO, indicating that some of the neurite formation was due to active FGF-2. We have previously reported that the remaining effect is not due to the presence of NGF in Time (hours)
the SIS, but rather is likely due to laminin present within the The PC12 differentiation assay showed that SISEO re- tained the ability to cause a directly observable effect incells, but we were also interested in the ability of the SISEO to stimulate cells to secrete growth factors. Therefore, we inves- tigated the ability of SISEO to stimulate VEGF secretion by fibroblasts, and showed that conditioned cell culture mediummade from processed SIS was able to stimulate VEGF secre- FGF-2 (Median RFU V
tion. We also showed that this response could be significantly Time (hours)
reduced when an activity neutralizing antibody specific for Fig. 3 (A) SISEO stimulates VEGF secretion by mouse fibroblasts. The
FGF-2 was added to the conditioned cell culture medium, ver- addition of a neutralizing antibody significantly reduced VEGF secre- ifying that the upregulation of VEGF by fibroblasts occurs tion, indicating that SISEO stimulation of VEGF is mediated through at least partially through an FGF-2 dependent mechanism.
FGF-2. Control cells seeded in unconditioned serum-free media dis-played significantly less VEGF secretion; * All three groups signifi- The ability of SISEO to stimulate cellular production of cantly different ( p < 0.05) from each other. (B) FGF-2 levels in the VEGF is important because VEGF is strongly expressed culture media remain steady over time, indicating that VEGF secretion in the frontline of repopulating epithelial, stromal and en- is not caused by endogenous FGF-2 production dothelial cells following injury, and is critically important inthe proliferation and migration of multiple cell types during meaningful cell and tissue repair and growth when used to wound repair and tissue regeneration [30]. In diabetic ulcers, VEGF improves wound healing by locally upregulating other Previously, Hodde et al. [22] reported that endothelial growth factors important for tissue repair [31]. VEGF also cells retain their ability to interact with SIS that had been mobilizes and recruits bone marrow-derived cells to the lo- oxidized using PAA and terminally sterilized with gamma cal wound environment, where they are able to contribute radiation. The current study supports those findings using a to blood vessel formation and accelerate diabetic ulcer heal- second cell type, but also using a matrix processed through ing [31]. Importantly, because VEGF is only loosely bound freeze-drying (lyophilization) and terminal sterilization with in the ECM and is susceptible to degradation by PAA and ethylene oxide gas. Significantly, fibroblast attachment and EO gas [21], the ability of other, more stable growth factors viability were not altered on SISLYO when compared to SIS to stimulate endogenous secretion of VEGF is an important that had not been lyophilized. Because lyophilization has the characteristic of an implantable scaffold material. The loss effect of collapsing the 3-dimensional structure of the matrix of VEGF during matrix processing is likely not critical to [21], these results suggest that complete retention of the three the effective bioactivity of the matrix during wound healing, dimensional architecture of the native matrix is not required because we have shown that VEGF is actively secreted by for SIS-cell interaction and retention of viability. The de- fibroblasts in response to interactions with other components gree to which the matrix can be collapsed and yet allow cell attachment has not been investigated.
In this study, we have shown that SIS retains the ability Matrix oxidation by PAA and alkylation by exposure to to support the interaction and growth of fibroblasts. We have EO gas are processing steps that ensure the viral safety and also shown that FGF-2 remains in the matrix in a form that is sterility of the processed ECM [26-28]. However, these pro- able to stimulate the differentiation of PC12 cells. We have cesses have the potential to alter the structure of matrix con- further demonstrated that FGF-2 and other factors present in stituents or render them inactive or unable to interact with the SISEO are sufficient to stimulate fibroblasts to secrete their J Mater Sci: Mater Med (2007) 18:545-550 own endogenous VEGF, abrogating the need to retain VEGF 12. D. Z I S S U , S. B I N E T and J. C. L I M A S S E T , Contact Der- in the processed matrix. We conclude that processing of SIS matitis 39 (1998) 248.
using agents that oxidize and alkylate matrix components S H O M U R A , D. C. M A R C H I O N , J. C. P F A U and C. M.
as required for clinical safety results in a biomaterial that D U R A N , J. Heart Valve Dis. 9 (2000) 570.
retains its intrinsic bioactivity. This intrinsic bioactivity may 14. H. U C H I Y A M A , Y. D O B A S H I , K. O H K O U C H I and K.
help explain the unique healing effects observed when this N A G A S A W A , J. Biol. Chem. 265 (1990) 7753.
material is used to treat chronic wounds and repair damaged I. V L O D A V S K Y , H. Q. M I A O , B. M E D A L I O N , P.
D A N A G H E R and D. R O N , Cancer Metastasis Rev. 15 (1996)
16. P. A. P U O L A K K A I N E N , J. E. R A N C H A L I S , D. M.
S T R O N G and D. R. T W A R D Z I K , Transfusion 33 (1993) 679.
Acknowledgements This study was supported by Cook Biotech In-
17. S. B. N I C O L L , S. R A D I N , E. M. S A N T O S , R. S. T U A N corporated, West Lafayette, IN, USA.
and P. D U C H E Y N E , Biomaterials 18 (1997) 853.
18. P. C. B E R S C H T , B. N I E S , A. L I E B E N D O R F E R and J.
K R E U T E R , Biomaterials 15 (1994) 593.
19. M. C. P E T E R S , B. C. I S E N B E R G , J. A. R O W L E Y and References
D. J. M O O N E Y , J. Biomater Sci. Polym. Ed. 9 (1998) 1267.
20. M. S E N O , R. S A S A D A , M. I W A N E , K. S U D O , T.
K U R O K A W A , K. I T O and K. I G A R A S H I , Biochem. Biophys. M I C H A E L S O N , J. L. G L A S S and D. A. C H O C K , Hernia 6
Res. Commun. 151 (1988) 701.
2. A. B. R U T N E R , S. R. L E V I N E and J. F. S C H M A E L Z L E , S H E R M A N and C. J O H N S O N , J. Mater. Sci. Mat. Med. In Press.
Urology 62 (2003) 805.
22. J. P. H O D D E , R. D. R E C O R D , R. S. T U L L I U S and S. F.
3. D. P A V C N I K , B. T. U C H I D A , H. A. T I M M E R M A N S , B A D Y L A K , Tissue Eng 8 (2002) 225.
C. L. C O R L E S S , M. O ' H A R A , N. T O Y O T A , G. L.
23. United States Patent No. 6,206,931.
M O N E T A , F. S. K E L L E R and J. R O S C H , J. Vasc. Surg. 35
G U R O F F , J. Neurosci. 5 (1985) 307.
4. V. M U S A H L , S. D. A B R A M O W I T C H , T. W. G I L B E R T , 25. K. P. C L A F F E Y , K. A B R A M S , S. C. S H I H , L. F.
E. T S U D A , J. H. - C. W A N G , S. F. B A D Y L A K and S.
B R O W N , A. M U L L E N and M. K E O U G H , Lab. Invest. 81
L. - Y. W O O , J Orthopaedic Res. 22 (2004) 214.
5. R. H. D E M L I N G , J. A. N I E Z G O D A , G. D. H A R A W A Y 26. J. P. H O D D E and M. C. H I L E S , Biotechnol. Bioeng. 79 (2002)
and E. N. M O S T O W , Wounds 16 (2004) 18.
6. S. L. V O Y T I K - H A R B I N , A. O. B R I G H T M A N , B. Z.
27. National Toxicology Program. Ethylene Oxide. Rep. Carcinog. 10
W A I S N E R , J. P. R O B I N S O N and C. H. L A M A R , Tissue Eng. 4 (1998) 157.
28. V. L. D E L L A R C O , W. M. G E N E R O S O , G. A. S E G A , J.
7. J. P. H O D D E , S. F. B A D Y L A K , A. O. B R I G H T M A N R. F O W L E 3 R D and D. J A C O B S O N - K R A M , Environ Mol and S. L. V O Y T I K - H A R B I N , Tissue Eng. 2 (1996) 209.
Mutagen 16 (1990) 85.
8. T. B. M C P H E R S O N and S. F. B A D Y L A K , Tissue Eng. 4
29. R. R E C O R D , J. H O D D E and S. B A D Y L A K , in Proceedings of the Biennial Meeting of the European Tissue Engineering Soci- 9. S. L. V O Y T I K - H A R B I N , A. O. B R I G H T M A N , M.
ety, Freiburg, Germany, November 7-10, 2001.
K R A I N E , B. W A I S N E R and S. F. B A D Y L A K , J. Cell. 30. L. G A N , P. F A G E R H O L M and J. P A L M B L A D , Acta Oph- Biochem. 67 (1997) 478.
thalmol. Scand. 82 (2004) 557.
10. J. P. H O D D E , S. F. B A D Y L A K and R. D. R E C O R D , En- dothelium 8 (2001) 11.
K. A. B H A T T , M. C A L L A G H A N , N. B A S T I D A S , S.
11. S. C. R O E , B. K. M I L T H O R P E and K. S C H I N D H E L M , B U N T I N G , H. G. S T E I N M E T Z and G. C. G U R T N E R , Am. Artif. Organs 14 (1990) 443.
J. Pathol. 164 (2004) 1935.

Source: http://www.kci-medical.com.br/midia/pdf/23)%20Hodde_Effects%20of%20Sterilization%20on%20an%20ECM%20Scaffold_PartII_Bioactivity%20and%20Matrix%20Interaction_JMaterSciMaterMed_2007.pdf