Crosslinking heparin to collagen scaffolds for the delivery of human platelet-derived growth factor

Crosslinking Heparin to Collagen Scaffolds for the Delivery ofHuman Platelet-Derived Growth Factor

Bo Sun,1* Bing Chen,2* Yannan Zhao,2 Wenjie Sun,2 Kaoshan Chen,1 Jing Zhang,2 Zhanliang Wei,2

Zhifeng Xiao,2 Jianwu Dai2

1 College of Life Sciences, Shandong University, Jinan 250100, China

2 Key Laboratory of Molecular Developmental Biology, Institute of Genetics and Developmental Biology,Chinese Academy of Sciences, Zhongguancun, Beijing 100090, People’s Republic of China

Received 15 October 2008; revised 8 February 2009; accepted 2 March 2009Published online 29 May 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.b.31411

Abstract: Platelet-derived growth factor (PDGF) plays an important role in tissue

regeneration and wound repair. However, the lack of effective delivery and the efficient

targeting specificity limits its clinical applications. Here, heparin possessing PDGF binding

domain was crosslinked to the collagen-based demineralized bone matrix (DBM) for the

delivery of human PDGF(HC-PDGF). In in vitro experiments, heparin improves the binding of

PDGF to collagen. In vitro activity assay indicates that collagen-heparin-PDGF (CH-PDGF)

promotes human fibroblasts to proliferate on collagen gel. In addition, HC-PDGF stimulates

cells to migrate into DBM scaffolds after implantation. The histological analysis shows that

HC-PDGF promotes vascularization of the implants. In summary, heparin-DBM/PDGF could

prevent the diffusion of PDGF, prolong its activity, and promote the cellularization and

vascularization of the scaffold. ' 2009 Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater

91B: 366–372, 2009

Keywords: platelet-derived growth factor (PDGF); demineralized bone matrix (DBM);

heparin; cellularization; vascularization

INTRODUCTION

During tissue regeneration and wound repair, the vasculari-

zation and cellularization of the implants are two important

aspects. Vascularization is important in several biological

processes and pathological conditions including tissue

regeneration, wound healing, and ischaemic heart dis-

eases.1,2 To promote the vascularization and cellularization

of the scaffold, many approaches have been explored. The

most common approach is to add growth factors to bioma-

terials.3–6 Platelet-derived growth factor (PDGF) is consid-

ered critical since it can promote the biomaterial

vascularization and cellularization and wound healing.7,8

However, growth factors usually lose their target specificity

and activity due to their rapid diffusion.9 Therefore,

approaches must be taken to retain the concentration and

activity of growth factors at the target site.

Heparin, a sulfated polysaccharide, has binding ability

for several growth factors including PDGF.10 It has been

added to biomaterials to prevent the diffusion of growth

factors and keep their activity to promote the wound heal-

ing.11 Demineralized bone matrix (DBM), a collagen-based

scaffold, has the excellent clinical potential for wound

repair.12–14 Recently, DBM was incorporated with heparin

to increase its mechanical property as well as specific

BMP2 binding ability to accelerate bone formation.15

Here, heparin was crosslinked to DBM for the delivery

of PDGF to promote DBM cellularization and vasculariza-

tion. It was speculated that this system would be a target

repair system for wound repair and tissue regeneration. The

specific binding of PDGF to the heparin crosslinked colla-

gen scaffolds was tested. Human fibroblast proliferation

in vitro as well as the vascularization and cellularization of

the scaffold in vivo were also examined.

MATERIALS AND METHODS

Preparation of Heparin-Crosslinked DBM (HC-DBM)

DBM was obtained from Zhenghai Biotechnology (Shan-

dong, China), which was prepared from bovine sponge bones.

Bo Sun and Bing Chen contributed equally to this work.Correspondence to: J. Dai (e-mail: [email protected])Contract grant sponsor: NSFC; Contract grant numbers: 30688002, 30600304Contract grant sponsor: Ministry of Science and Technology of China; Contract

grant number: 2006CB943601Contract grant sponsor: Chinese Academy of Sciences; Contract grant number:

KSCX2-YW-R-133Contract grant sponsor: K. C. Wong Education Foundation

' 2009 Wiley Periodicals, Inc.

366

Preparation of HC-PDGF was based on the method by

Lin et al.15 with slight modifications. DBM (4 mm 3 4

mm 3 1 mm) was placed in 96-well plates and loaded

with 50 mM 2-morpholinoethane sulphonic acid (MES)

(pH 5.6) by 200 lL/well overnight. Heparin (H-4784;

Sigma, USA) was dissolved with 4 mg 1-ethyl-3-dimethyl

aminopropyl carbodiimide (EDC)/2.4 mg N-hydroxysucci-

nimide (NHS) (39391 and 14405, Sigma, USA) in 10 mL

of 50 mM MES buffer (pH 5.6) with the concentration

series of 0.125–2 mg mL21. The same amount of heparin

was added to 50 mM MES buffer (pH 5.6) without EDC

and NHS. Heparin mixtures (200 lL/well) were dipped in

MES for overnight, and added to the plates, respectively.

The reaction was proceeded for 4 h at 378C. Consequently,

the DBM was washed with 0.1M Na2HPO4 (2 h), 4M NaCl

(three times in 4 h), and distilled water (three times in

1 h), respectively.

Calculation of the Amount of HeparinCrosslinked to DBM

The amount of heparin crosslinked to the DBM was deter-

mined by the toluidine blue method.16 HC-DBM and DBM

(4 mm 3 4 mm 3 1 mm) were immersed into 100 lL

toluidine blue solution (10 mg L21) dissolved with 0.01Nhydrochloric acid containing 0.2 wt % NaCl. After remov-

ing the scaffolds from solution, 50 lL the unreacted tolui-

dine blue in the water phase was determined by absorption

at 620 nm with a plate reader (TECAN, SUNRISE,

Austria). The amount of heparin crosslinked to HC-DBM

and DBM was calculated based on reference standards.

PDGF Bound to DBM and HC-DBM

PDGF was purified by using its his36 tag.17 HC-DBM was

prepared with 2 mg mL21 of heparin. DBM and HC-DBM

(4 mm 3 4 mm 3 1 mm) were placed into the 96-well

cell culture plate (Costar 3599, Corning, USA), respec-

tively. PDGF (100 lL) was added to each well with the

concentration series of 2–130 lg mL21 and incubated for 1

h at 378C. DBM and HC-DBM were then collected to

perform enzyme linked immunosorbent assay (ELISA) to

measure the amount of PDGF bound to DBM and

HC-DBM.

The primary antibody was anti-poly-histidine antibody

(1:1000, Sigma, USA), and an alkaline phosphatase (ALP)-

conjugated goat anti-mouse IgG (1:10,000 dilution, Sigma,

USA) was utilized as the secondary antibody. The method

was based on the reference described before.17

Heparin was also crosslinked to collagen (HC-COLLA-

GEN) prepared from rat-tail collagen.18 The amount of

PDGF bound was also measured.

PDGF Activity in Heparin Crosslinked Collagen

Acid soluble collagen (250 lL/well) was added to 48-well

plates and crosslinked with and without heparin. After

washing three times with PBS, 100 lL PDGF diluted in

PBS was added at 378C for 1 h with the concentration

series of 0, 55, 110, 220 lg mL21. Then 1000 U mL21

penicillin (Gibco), 1000 lg mL21 streptomycin (Gibco)

was added at 378C for 2 h, followed by washing three

times with PBS. After extensive washing, human fibroblasts

were inoculated into each well (5000 cells in 300 lL of

DMEM-10% fetal bovine serum), and the medium was

changed into DMEM-2% fetal bovine serum. After 4 days

of culturing, the cell numbers were determined by methyl-

thiazoletetrazolium (MTT) assay.

Subcutaneous Implantation

Chinese Ministry of Public Health (CMPH) guidelines for

the care and use of laboratory animals has been followed.

DBM and HC-DBM were sterilized by Co60 irradiation.

Male Wistar rats (180–200 g) were anesthetized by injec-

tion of pentobarbital (40 mg kg21). The middle back area

was shaved and disinfected with 75% alcohol. For implan-

tation, three 1-cm incisions were made on the dorsum of

each rat. Through the incision, one subcutaneous pocket

was made by blunt scissors. Three complexes: DBM load-

ing with PBS (DBM/PBS), DBM loading with 12-lg

PDGF (DBM/PDGF), HC-DBM loading with 12-lg PDGF

(HC-DBM/PDGF), respectively, were randomly embedded

into the three subcutaneous pockets in one rat. After

implantation, the rats were housed in separate cages and

given standard food and water. At Days 7 and 14, the rats

were sacrificed and the matrixes were explanted and fixed

in 4% formaldehyde. Sections of the implants were made

H&E staining and immunohistochemical staining with for

histological analysis of cellularization and vascularization.

a-smooth muscle actin (a-SMA) antibody was used for

immunohistochemical staining. Briefly, the samples depar-

affinized, rehydrated and blocked the intrinsic peroxidase

activity by H2O2 (3% solution in methanol, 15 min).

Sections were treated with pronase (Sigma; 0.1%, 30 min,

at room temperature) and incubated with a-smooth muscle

actin (a-SMA) antibody (1:100 dilution by PBS for 2 h at

378C). A HistostainTM-Plus Kit (Zymed Laboratories, San

Francisco, CA) was used for staining, according to the

manufacturer’s instructions. The histomorphometrical

evaluation was performed using a Nikon calibrated lens

micrometer (Nihon Kogaku, Tokyo, Japan) at 403 magnifi-

cation. Ten areas per specimen were randomly chosen to

quantify numbers of blood vessels, and sixteen areas

(100 3 100 lm2) randomly of HE images were chosen to

quantify the numbers of cells.

Statistical Analysis

All data were expressed as mean 6 standard deviation

(SD). Statistical analyses were completed using Statistics

Package for Social Science (SPSS) software. The signifi-

cance of difference was determined by p values at *, p \0.05 and **, p\ 0.01.

367CROSSLINKING HEPARIN TO COLLAGEN SCAFFOLDS

Journal of Biomedical Materials Research Part B: Applied Biomaterials

RESULTS

The Amount of Heparin Crosslinked to DBM

As shown in Figure 1, using the standard curve, the

amount of heparin bound to DBM with NHS/EDC (HC-

DBM) was significantly more than that without NHS/EDC

(DBM) at the same concentration of heparin. Meanwhile,

the amount of heparin increased gradually along with the

increased concentration of heparin. At the higher concen-

tration of heparin (1 and 2 mg mL21), the adsorbed hepa-

rin to DBM was almost saturated (0.075 6 0.04 lg, 0.075

6 0.058 lg), but adsorbed heparin added 1 and 2 mg

mL21 to HC-DBM was 0.257 6 0.011 lg, 0.361 6 0.047

lg, respectively.

The Amount of PDGF Bound to HC-DBM was Increased

The scaffolds crosslinked with or without 2 mg mL21 of

heparin were prepared with NHS/EDC. As shown in Figure

2, The amount of PDGF bound to scaffolds increased on a

concentration-dependent fashion. The amount of PDGF

bound to scaffolds crosslinked with heparin, was signifi-

cantly increased. The result indicated that scaffolds cross-

linked with heparin could bind to PDGF more efficiently.

The Cell Proliferation Rate Increased on theHC-COLLAGEN/PDGF

The human fibroblasts were cultured in 48-well plates

(5000 cells/well) coated with COLLAGEN/PDGF and HC-

COLLAGEN/PDGF. The initial concentrations of PDGF

were 55, 110, and 220 lg mL21. As shown in Figure 3,

human fibroblasts proliferated faster on HC-COLLAGEN

gel than on COLLAGEN gel at each concentration, espe-

cially at 110 lg mL21 point. The cell proliferation rate

treated with PDGF and HC-COLLAGEN system signifi-

cantly increased compared to the PDGF and collagen sys-

tem. Thus, PDGF bound to the heparin crosslinked scaffold

promoted cell proliferation.

Figure 1. The amount of heparin binding to DBM and HC-DBM. Initial

concentrations of heparin were 0, 0.125, 0.25, 0.5, 1, 2 mg mL21,respectively. Data are presented as mean 6 SEM. **p\0.01.

Figure 2. Binding curves of PDGF in vitro. (A) PDGF binding to

DBM and HC-DBM. Initial concentrations of PDGF were 0, 2, 4, 8,

16.25, 32.5, 65, 130 lg mL21, respectively. (B) PDGF binding to col-lagen and HC-collagen. Initial concentrations of PDGF were 0, 2,

4.175, 8.75, 17.5, 35, 70, 140 lg mL21, respectively. Data are pre-

sented as mean 6 SEM. *p\ 0.05, **p\ 0.01.

Figure 3. The cell proliferation-promoting activity of COLLAGEN/PDGF and HC-COLLAGEN/PDGF. The initial concentrations of

PDGF were 55, 110, 220 lg mL21. Data are presented as mean 6

SEM. *p\ 0.05, **p\ 0.01.

368 SUN ET AL.


HC-DBM/PDGF Promoted Cellularization andVascularization In Vivo

At day 7 after implantation, the scaffolds were retrieved

and examined the cellularization by H&E staining. As

shown in Figure 4, more cells migrated into HC-DBM/

PDGF scaffolds [Figure 4(C,F)] than that of DBM/PDGF

scaffolds [Figure 4(B,E)], only few cells migrated into

DBM/PBS scaffolds [Figure 4(A,D]. The cells migrating

into scaffolds were counted [Figure 4(G)]. The number of

cells were 20 6 12 in DBM/PBS, 34 6 5 in DBM/PDGF,

58 6 9 in HC-DBM/PDGF, respectively. The data sug-

gested that there was a significant difference among these

three different scaffolds because of different amount of

PDGF bound to the scaffold.

At days 7 and 14 after implantation, images of HC-DBM/

PDGF and DBM/PDGF with immunohistochemical staining

showed apparent neovascularization [Figures 5(A–G) and

6(A–G)]. As shown in Figure 5(H), the number of blood

vessels at day 7 in HC-DBM/PDGF (24 6 2) was about

the same as DBM/PDGF (22 6 5), and there was no signifi-

cant difference between them. However, as shown in

Figure 6(H), the number of the blood vessels at day 14 in

HC-DBM/PDGF (21 6 3) was significantly more than that

in DBM/PDGF (10 6 2), and both of them were more than

that in DBM/PBS control (5 6 1).

DISCUSSION

During wound repair and tissue regeneration, an effective

growth factor delivery method is very important. Heparin-

growth factor delivery system is an effective approach. In

recent studies, heparin was crosslinked to several scaffolds,

such as fibrin matrix.19 In this study, the DBM-based colla-

gen scaffold was chosen. The amount of heparin bound to

Figure 4. Histological examination for cellularization of DBM and HC-DBM scaffolds at 7 days after

implantation by HE staining. Red fibers represented the DBM; black dots represented cell immigrated.(A,D) DBM/PBS; (B,E) DBM/PDGF; (C,F) HC-DBM/PDGF; 10 3 20 in A,B,C; 10 3 40 in (D–F); (H) num-

bers of cells migrated in per 100 3 100 lm2 area of DBM/PDGF and HC-DBM/PDGF. DBM without

growth factors was a negative control. Data are presented as mean 6 SEM. **p \ 0.01. [Color figure

can be viewed in the online issue, which is available at www.interscience.wiley.com.]



HC-DBM was more than that bound to DBM, although

both of them increase on a concentration-dependent fash-

ion. It was essential and effective to crosslink the heparin

to DBM by NHS/EDC. Heparin was crosslinked at a con-

centration of 2 mg mL21 to collagen-based scaffolds to

bind PDGF more effectively.

Recent studies have shown that the simple adsorption of

growth factors to the scaffolds limits the process of wound

healing because the growth factors diffuse away from the

scaffolds and lose their activity rapidly.20 In the in vitrostudy, the adsorption of PDGF to HC-DBM and HC-COL-

LAGEN through its heparin-binding domain was higher

than that to DBM and COLLAGEN, respectively. After 4

days of culturing, at the same concentration of PDGF, the

proliferation of cells on HC-COLLAGEN was significantly

faster than on the COLLAGEN which reflected that PDGF

bound to scaffolds still retained its bioactivity.

Studies have shown that activated scaffolds could

induce cell migration and tissue regeneration.21,22 The cel-

lularization and vascularization of PDGF-loaded HC-DBM

were evaluated after subcutaneous implantation. More

cells grew into HC-DBM/PDG than DBM/PDGF. This

indicated PDGF loaded on DBM could easily diffuse

away from the implant, while HC-DBM could retain

Figure 5. Histological examination of the vascularization of scaffolds at 7 days after implantation by

immunohistochemical staining. (A) The appearance of scaffolds, from left to right: DBM/PBS, DBM/PDGF, HC-DBM/PDGF. (B,E) DBM/PBS; (C,F) DBM/PDGF; (D,G) HC-DBM/PDGF;10 3 20 in B, C, D;

10 3 40 in E, F, G. (H) Numbers of blood vessels at 403 magnification of DBM/ PDGF and HC-

DBM/PDGF. DBM without growth factors was a negative control. Data are presented as mean 6

SEM. *p \ 0.05, **p \ 0.01. [Color figure can be viewed in the online issue, which is available atwww.interscience.wiley.com.]

370 SUN ET AL.


PDGF. Thus, it promoted cells to grow into the scaffold

more efficiently.

Angiogenesis is an important aspect during wound heal-

ing and tissue regeneration. At day 7 after implantation, the

immunohistochemical staining for blood vessels showed

that there were no significant difference on the numbers of

new blood vessels between HC-DBM/PDGF and DBM/

PDGF. This perhaps was due to the initial burst release of

PDGF freely bound to DBM at earlier phases. However, at

day 14, DBM/PDGF and DBM/PBS scaffolds became

smaller, while the size of HC-DBM/PDGF scaffolds did

not change [Figure 6(A)]. It might relate to the improve-

ment of rigidity of HC-DBM which prevented the scaffold

degradation.15 While PDGF loaded on DBM diffused

away, the PDGF bound to HC-DBM induced more blood

vessel formation [Figure 6(B–H)]. Moreover, the number of

blood vessels in DBM/PBS and DBM/PDGF at day 14 was

lower than that at day 7, but it almost did not change in

HC-DBM/PDGF [Figures 5(H) and 6(H)]. This maybe

because that PDGF stimulated the expression of PDGF-aand PDGF-b receptors.7 PDGF-a receptor was related to

the angiogenic synergism, and PDGF-b receptor was essen-

tial for vascular stability.23,24 PDGF loaded on DBM scaf-

fold diffused away rapidly so that the blood vessels formed

originally vanished, but PDGF loaded on HC-DBM scaf-

fold could keep an effective concentration to stimulate the

expression of PDGF-a and PDGF-b receptors. Conse-

quently, it improved the stability of blood vessels. These

Figure 6. Histological examination of the vascularization of scaffolds at 14 days after implantation byimmunohistochemical staining. (A) The appearance of scaffolds, from left to right: DBM/PBS, DBM/

PDGF, HC-DBM/PDGF. (B,E) DBM/PBS; (C,F) DBM/PDGF; (D,G) HC-DBM/PDGF; 10 3 20 in B,C,D; 10 3

40 in E, F, G. (H) Numbers of blood vessels at 403 magnification of DBM/ PDGF and HC-DBM/PDGF.DBM without growth factors was a negative control. Data are presented as mean 6 SEM. *p\ 0.05, **p

\0.01. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]



data suggested that DBM scaffold crosslinked with heparin

was an ideal PDGF delivery system for tissue regeneration.

CONCLUSIONS

Heparin was crosslinked to DBM by NHS/EDC (HC-DBM)

for delivering PDGF. PDGF could specially bind to

HC-DBM. PDGF loaded on HC-DBM promoted the cellu-

larization and vascularization in vivo. All data showed that

HC-DBM/PDGF scaffolds could be an effective system for

tissue regeneration.

REFERENCES

1. Patel ZS, Mikos AG. Angiogenesis with biomaterial-baseddrug- and cell-delivery systems. J Biomater Sci Polym Ed2004;15:701–726.

2. Fukuda S, Yoshii S, Kaga S, Matsumoto M, Kugiyama K,Maulik N. Angiogenic strategy for human ischemic heartdisease: Brief overview. Mol Cell Biochem 2004;264:143–149.

3. Smith JD, Melhem ME, Magge KT, Waggoner AS, CampbellPG. Improved growth factor directed vascularization intofibrin constructs through inclusion of additional extracellularmolecules. Microvasc Res 2007;73:84–94.

4. Babensee JE, McIntire LV, Mikos AG. Growth factor deliveryfor tissue engineering. Pharm Res 2000;17:497–504.

5. Chen RR, Mooney DJ. Polymeric growth factor delivery strat-egies for tissue engineering. Pharm Res 2003;20:1103–1112.

6. Whitaker MJ, Quirk RA, Howdle SM, Shakesheff KM.Growth factor release from tissue engineering scaffolds.J Pharm Pharmacol 2001;53:1427–1437.

7. Heldin CH, Westermark B. Mechanism of action and in vivorole of platelet-derived growth factor. Physiol Rev 1999;79:1283–1316.

8. Futamura A, Izumino K, Nakagawa Y, Takata M, Inoue H,Iida H. Effect of the platelet-derived growth factor antagonisttrapidil on mesangial cell proliferation in rats. Nephron1999;81:428–433.

9. Bowen-Pope DF, Malpass TW, Foster DM, Ross R. Platelet-derived growth factor in vivo: Levels, activity, and rate ofclearance. Blood 1984;64:458–469.

10. Mangrulkar RS, Ono M, Ishikawa M, Takashima S,Klagsbrun M, Nowak RA. Isolation and characterization ofheparin-binding growth factors in human leiomyomas andnormal myometrium. Biol Reprod 1995;53:636–646.

11. Nillesen ST, Geutjes PJ, Wismans R, Schalkwijk J, DaamenWF, van Kuppevelt TH. Increased angiogenesis and bloodvessel maturation in acellular collagen-heparin scaffolds

containing both FGF2 and VEGF. Biomaterials 2007;28:1123–1131.

12. Urist MR. Bone: Formation by autoinduction. Science 1965;150:893–899.

13. Piattelli A, Scarano A, Corigliano M, Piattelli M. Comparison ofbone regeneration with the use of mineralized and demineralizedfreeze-dried bone allografts: A histological and histochemicalstudy in man. Biomaterials 1996;17:1127–1131.

14. Trevisiol CH, Turner RT, Pfaff JE, Hunter JC, Menagh PJ,Hardin K, Ho E, Iwaniec UT. Impaired osteoinduction ina rat model for chronic alcohol abuse. Bone 2007;41:175–180.

15. Lin H, Zhao Y, Sun W, Chen B, Zhang J, Zhao W, Xiao Z,Dai J. The effect of crosslinking heparin to demineralizedbone matrix on mechanical strength and specific binding tohuman bone morphogenetic protein-2. Biomaterials 2008;29:1189–1197.

16. Jeon O, Kang SW, Lim HW, Hyung Chung J, Kim BS. Long-term and zero-order release of basic fibroblast growth factorfrom heparin-conjugated poly(L-lactide-co-glycolide) nano-spheres and fibrin gel. Biomaterials 2006;27:1598–1607.

17. Lin H, Chen B, Sun W, Zhao W, Zhao Y, Dai J. The effectof collagen-targeting platelet-derived growth factor on cellula-rization and vascularization of collagen scaffolds. Biomateri-als 2006;27:5708–5714.

18. Gentleman E, Lay AN, Dickerson DA, Nauman EA, LivesayGA, Dee KC. Mechanical characterization of collagen fibersand scaffolds for tissue engineering. Biomaterials 2003;24:3805–3813.

19. Thomopoulos S, Zaegel M, Das R, Harwood FL, Silva MJ,Amiel D, Sakiyama-Elbert S, Gelberman RH. PDGF-BBreleased in tendon repair using a novel delivery system pro-motes cell proliferation and collagen remodeling. J OrthopRes 2007;25:1358–1368.

20. Yao C, Roderfeld M, Rath T, Roeb E, Bernhagen J, SteffensG. The impact of proteinase-induced matrix degradation onthe release of VEGF from heparinized collagen matrices.Biomaterials 2006;27:1608–1616.

21. Tokuda Y, Toda S, Masaki Z, Sugihara H. Proliferation anddifferentiation of rat dorsal prostatic epithelial cells in colla-gen gel matrix culture, focusing upon effects of adipocytes.Int J Urol 1999;6:509–519.

22. Schor SL. Cell proliferation and migration on collagensubstrata in vitro. J Cell Sci 1980;41:159–175.

23. Lu H, Xu X, Zhang M, Cao R, Brakenhielm E, Li C, Lin H,Yao G, Sun H, Qi L, Tang M, Dai H, et al. Combinatorialprotein therapy of angiogenic and arteriogenic factors remark-ably improves collaterogenesis and cardiac function in pigs.Proc Natl Acad Sci USA 2007;104:12140–12145.

24. Zhang J, Cao R, Zhang Y, Jia T, Cao Y, Wahlberg E. Differ-ential roles of PDGFR-a and PDGFR-b in angiogenesis andvessel stability. FASEB J 2009;23:153–163.

372 SUN ET AL.


Documents

Crosslinking heparin to collagen scaffolds for the delivery of human platelet-derived growth factor