Extended Release of rhPDGF-ββ and rhBMP-2 From the Heparin Binding Site of Fibrin: An In Vitro Pilot Study

Initiation of the wound healing cascade starts with the formation of fibrin from fibrinogen.¹ Fibrin along with platelets enables hemostasis to occur at the wound or surgical site.1 This provisional matrix allows cells from the immune system as well as mesenchymal stem cells to migrate through and complete the healing response.

Tissue engineering is the creation of new tissue for the therapeutic reconstruction of the human body by the deliberate and controlled stimulation of selected target cells through a systematic combination of molecular and mechanical signals.² Engineering techniques can be used that mimic the critical aspects of the natural healing processes—the “wound healing cascade”—by providing suitable biochemical and physico-chemical factors that can enhance tissue regeneration.³

The earliest sequence in the wound healing cascade starts with the formation of fibrin along with platelet activation. Blood technologies such as solid platelet-rich fibrin (PRF) or leukocyte PRF (700 g-force for 8 minutes),⁴ liquid PRF (500 g-force for 5 minutes),⁵ and concentrated PRF (2,000 g-force for 8 minutes)6 have entered the dental market exploiting the endogenous release of platelet growth factors. Unfortunately, none of these technologies can induce bone formation. In 1996, an article by Odrljin et al reported that fibrin possesses a promiscuous binding site in its E domain.⁷ Martino and colleagues later demonstrated in an in vitro study that this binding site has the capability to bind to many growth factors, including platelet-derived growth factor-ßß and bone morphogenetic protein-2.⁸ Together, both of these growth factors are chemotactic, mitogenic, angiogenic, and morphogenic, and both are crucial for the development of new bone formation. Miron et al showed fibrin to be an effective carrier to release these growth factors.^9,10

Growth factors act in the molecular signaling of tissue engineering by releasing chemical signals to the nascent wound region, while fibrin provides the physical matrix allowing cells to migrate through and begin regenerating the damaged or surgical area. Fibrin along with platelets has no osseoconductive properties.¹ By exploiting the heparin binding in fibrin to retain and release growth factors over an extended timeframe,⁸ it is possible to transform this scaffold into a powerful osseoinductive matrix. The strong binding affinity of these two growth factors to fibrin allows the growth factors to be used at much lower concentrations making them more biologically and cost effective for everyday use.

The purpose of the present study was to determine if the central region of fibrin can bind and release recombinant human bone morphogenetic protein-2 (rhBMP-2) and recombinant human platelet-derived growth factor-ßß (rhPDGF-ßß) over an extended timeframe and create a biologic membrane large enough to be used for guided bone regeneration.

Materials and Methods

Ten 9-ml test tubes with Z serum clot activator were used to obtain blood samples from the antecubital vein of one of the authors. Two test tubes with 9 ml of blood were doped with 0.5 ml of rhPDGF-ßß (T1 and T2), two test tubes with 9 ml of blood were doped with 0.25 ml of rhBMP-2 (T3 and T4), and two test tubes with 9 ml of blood had alpha-2 antiplasmin added to the 0.25 ml of rhBMP-2 (T5 and T6). Alpha-2 antiplasmin prevents the degradation of fibrin by inhibiting the action of plasmin, which breaks down the fibrin polymer. Four test tubes with 9 ml of blood and no growth factors added served as controls: C1, C2, C3, and C4 (Figure 1).

The blood samples were centrifuged at 2,700 RPMs for 12 minutes using a PCO2 machine (Intra-Lock, intra-lock.com). Each test tube had its own leukocyte PRF (L-PRF) clot removed by separating it from the red blood cell (RBC) layer using a 21c blade (Figure 2). Each L-PRF clot from the controls and test clots doped with growth factor were placed into separate wells containing 1 ml of Dulbecco’s Modified Eagle’s Medium (DMEM) ATCC 30-2002 with 10% fetal bovine serum and 1% penicillin-streptomycin (Figure 3 and Figure 4). The first sample taken was of the platelet poor plasma after removal of the L-PRF clot. The second sample was from the forcible extraction of the L-PRF clot in a standard clot box (Intra-Lock). All 10 L-PRF clots were placed in wells with 1 ml of growth medium, and the growth medium was removed after 20 minutes, 4 hours, 20 hours, 3 days, and 7 days. At each time interval all the liquid was removed and 1 ml of fresh growth medium was added. In addition, the rhBMP-2 growth medium was also collected at 13 days and 14 days. The L-PRF clots were stored at 37°C at 5% carbon dioxide (CO2) at Enzo Biochemical in Farmingdale, New York.

All samples were thawed to room temperature before testing. Indirect enzyme-linked immunosorbent assays (ELISA) (RayBiotech, raybiotech.com) for rhPDGF-ßß and rhBMP-2 were used to quantify the concentration of growth factor in picograms.11 All samples were diluted 25-fold for the first assay (10 µl into 240 µl Buffer D from the ELISA kit [RayBiotech]) for rhPDGF-ßß. All other samples were diluted 500-fold for the second assay, because 25-fold dilution samples were outside the range of standards due to the high levels of rhPDGF-ßß.

The same protocol was followed for rhBMP-2 except the dilution was 30,000-fold for the platelet poor plasma and forcible extraction samples. The remaining test samples for rhBMP-2 had a 1,000-fold dilution to ensure the samples did not fall outside the range of standards.

The second part of the research focused on the creation of a biologic membrane that could retain and release growth factors. Six 9-ml plastic test tubes were used to collect blood. The

plastic-lined tubes inhibit coagulation and allow for the separation of the plasma from the RBC layer. The plasma was placed into a surgical bowl and 500 µl of calcium chloride was added to initiate coagulation. The plasma was incubated at 37°C for 30 minutes. The coagulated plasma clot was squeezed using a perforated metal box with a metal lid covering the clot under pressure for 1 minute to produce a biologic membrane approximately 40 mm x 45 mm (Figure 5).

The raw data figures were obtained from the specific retrieval timeframes. The raw data is expressed in pg/ml, and it ignores the length of time the protein was being released since this is not clinically relevant and may be misleading (Table 1 and Table 2).

The pg/hour is the raw data figure divided by the time the clot was in the buffer solution between the two timeframes. It is extremely relevant because it provides a clinical picture of how much growth factor is being released from the fibrin clot per hour. The pg/hour time represents the mean average between two data timeframes.

Problems encountered working with these proteins had to do with their preparations from the manufacturers. All were highly acidic (rhBMP-2, pH 4.5; rhPDGF-ßß, pH 5.5–6.5; alpha-2 antiplasmin, pH 6.5; blood, pH 7.4) and had to be buffered with phosphate-buffered saline to initiate coagulation, otherwise the blood would not have coagulated after centrifugation.

Results

Table 3 expresses the concentration of PDGF-ßß and rhPDGF-ßß released in pg/hour. The control tubes showed endogenous release of PDGF-ßß of 409 pg at 0.167 hours, 44 pg at 2.167 hours, 25 pg at 12 hours, 10 pg at 44 hours, and 8 pg at 116 hours. The test tubes showed exogenous release of rhPDGF-ßß of 680,388 pg at 0.167 hours, 59,246 pg at 2.167 hours, 16,900 pg at 12 hours, 5,137 pg at 44 hours, and 1,583 pg at (116 hours).

Table 4 expresses the concentration of rhBMP-2 released in pg/hour. After wounding has occurred, the activation of platelets causes the release of many growth factors. One growth factor not present or released is BMP-2. In all control samples the small numbers expressed are artifacts and do not represent any rhBMP-2 present. Therefore, the only numbers of clinical significance are the different concentrations between rhBMP-2 plus rhBMP-2 + alpha-2 (a2) antiplasmin. Also notable was the ability of fibrin to retain and demonstrate slow release of growth factor after 13 days.

The tubes with rhBMP-2 alone showed growth factor release of 966,126 pg at 0.165 hours, 101,061 pg at 2.165 hours, 20,882 pg at 12 hours, 6,015 pg at 46 hours, 3,398 pg at 120 hours, 1,934 pg at 240 hours, and 9,606 pg at 324 hours.

The tubes with rhBMP-2 and rhBMP-2 + a2 antiplasmin showed growth factor release of 984,482 pg at 0.165 hours, 93,106 pg at 2.165 hours, 23,846 pg at 12 hours, 6,346 pg at 46 hours, 3,725 pg at 120 hours, 2,078 pg at 240 hours, and 12,788 pg at 324 hours.

After separating the uncoagulated plasma from the six plastic test tubes, a 40 mm x 45 mm biologic membrane was able to be fabricated using 500 µl of calcium chloride as an activator for clot polymerization (Figure 5).

Discussion

The fabrication of PRF involves the use of a tabletop centrifuge machine (eg, Horizontal PRF Centrifuge [Bio-PRF, bio-prf.com]; IntraSpin [Intra-Lock]; DUO Quattro [Process for PRF, a-prf.com], Salvin Variable Speed Centrifuge [Salvin, salvin.com]). The centrifuge machine allows for the spinning of the blood at different time intervals as well as different g-forces to produce the specific PRF product desired.

The most common area for venipuncture to obtain a blood draw is from the antecubital fossa of the forearm.¹² A tourniquet is applied just above the antecubital fossa to elevate the veins for the phlebotomy procedure. The largest vein is the median cubital vein. A 21-gauge butterfly needle with a vacuum tube is used to fill each 10-ml test tube with blood. The minimum number of tubes to fill is two, and the maximum number that can be filled is 10. It takes approximately 15 seconds to fill each tube with blood. The blood draw can be performed either before or after the patient has been anesthetized in the dental operatory. No anticoagulants are used during the production of PRF. Upon completion of the blood draw, centrifugation must be carried out within 60 to 90 seconds,¹³ otherwise the fibrinogen will start to gel and convert into fibrin.

There are two types of test tubes available for PRF: hydrophilic tubes, which are made from glass or are silica coated, and hydrophobic tubes made from plastic. The hydrophilic test tubes must be filled with blood first because it will clot faster than in hydrophobic tubes, which have no initiator.¹⁴ The hydrophobic test tubes are used to draw out the liquid PRF that can be injected onto bone substitutes to create sticky bone¹⁵ or into any other scaffold desired for bone augmentation. The contents of the hydrophobic tubes can be combined to make a large biologic membrane as well. The hydrophilic tubes produce clots at the end of centrifugation, and these clots are separated from the RBC layer. The clots are cut into small segments and incorporated into the bone graft material or placed over a conventional membrane to aid in soft-tissue healing.

The technique to incorporate rhBMP-2 and rhPDGF-ßß with the different modalities of PRF is easy to duplicate. First, test tubes—whether hydrophobic or hydrophilic—contain the liquid growth factor(s), which should be added to the mix just prior to the blood draw. The liquid growth factor(s) must be at a pH of 7.4 or else the blood will not coagulate. Phosphate-buffered saline, which is manufactured by numerous companies, is readily available from any medical supplier. The dilution factor is a 1:1 ratio of buffered saline solution to liquid growth factor solution. This is the only additional step required to incorporate the growth factors into the fibrin meshwork.

The endogenous release of PDGF-ßß in the present study was comparable to previous notable studies.^9,16-19 The addition of the exogenous source of rhPDGF-ßß amplifies the ability of this growth factor to affect early wound healing and osteogenesis to a greater degree than with platelets alone from a simple blood draw. The fibrin matrix allows for this release to be gradual and steady for at least 2 weeks, which is significantly longer than that which is obtained when spinning blood alone. The incorporation of the two growth factors along with PRF creates powerful mitogens and morphogens that aid in the recruitment of pericytes from budding blood vessels and stem cells from endosseous bone marrow.20 This results in accelerated bone formation in sinus or ridge augmentation procedures.

Osseoconductive bone substitutes become osseoinductive with the addition of rhBMP-2.²¹ The addition of rhBMP-2 to the graft mixture reduces the amount of autogenous bone required for large augmentation procedures.²²

Three systematic reviews on sinus grafting with PRF found little or no clinical advantage of adding L-PRF with bone substitutes for sinus augmentation.^23-25 The major clinical advantages of L-PRF is its use for small sinus tears and its ability to hold graft particles together to prevent their extravasation apically and block the osteum. To enhance the sinus floor elevation procedure, adding rhBMP-2 to L-PRF converts the graft into an osseoinductive graft capable of inducing mesenchymal cells into osteoblasts to form bone at an accelerated rate.

In a systematic review of extraction socket healing, Miron et al showed L-PRF had better healing results early on than natural healing, but not as good as a bone graft to prevent collapse of the socket.²⁶ Temmerman et al and Castro et al both showed L-PRF had better wound healing than natural healing but was unable to counteract ridge resorption.^27,28 The incorporation of rhPDGF-ßß with L-PRF with a rapid resorbing bone graft material will prevent the collapse of the alveolar socket and leave no unwanted graft particles during drilling of the osteotomy.

PRF has many clinical applications for regenerative dentistry. Due to its angiogenic properties, it can be used as a barrier in guided bone regenerative procedures to aid in wound closure. Another scaffold application of PRF is to saturate it with demineralized trabecular bone in the form of a sponge (Figure 6), or it can be incorporated into a 3D scaffold made of hydroxyapatite (Figure 7). Both of these applications would become osseoinductive once rhBMP-2 is added to the PRF. The enhanced gels would not require any additional graft material to be used during ridge augmentation procedures.

Conclusion

The present study demonstrated that the central region of fibrin can bind and release rhPDGF-ßß over 116 hours 200 times more than the endogenous release of PDGF-ßß. The results of this study also showed that rhBMP-2 alone and rhBMP-2 + alpha-2 antiplasmin can be released at 13 days with a concentration of 9,606 pg and 12,788 pg, respectively. The research presented here shifts the emphasis away from the platelet and leukocyte content and focuses instead on the heparin binding site in the fibrin architecture. Fibrin alone has no osseoinductive properties and cannot grow bone. The results of this study demonstrated that once fibrin is doped with rhPDGF-ßß and rhBMP-2 it becomes a powerful scaffold that is mitogenic, morphogenic, and angiogenic. This study also showed that a PRF membrane can be fabricated as large as 40 mm x 45 mm and utilized as a biological barrier. This biologic membrane contains BMP-2 and is osseoinductive, while other membranes can only protect the graft from soft-tissue invagination and are incapable of stimulating stem cells to osteoblasts.

ACKNOWLEDGMENT

The authors thank Mr. Barry Weiner and Dr. Elazar Rabbani for allowing the research to be conducted at Enzo Biochemical. Author contributions: Dr. Coleman was instrumental in the technical aspect of interpreting the ELISA assay data. Drs. Froum and Horowitz helped organize and proofread the clinical research paper. Dr. Scolnick was the principal investigator in regard to the research protocol and authored the clinical research paper. Sole funding for the clinical research project was from Dr. Scolnick.

ABOUT THE AUTHORS

Jeffrey A. Scolnick, DDS
Assistant Clinical Professor, Department of Periodontology and Implant Dentistry, New York University College of Dentistry, New York, New York; Private Practice in Prosthodontics, New York, New York; Diplomate, International Congress of Oral Implantologists

Jack Coleman, PhD
Director of Biochemistry, ENZO Life Sciences, Farmingdale, New York

Stuart J. Froum, DDS
Adjunct Clinical Professor and Director of Clinical Research, Department of Periodontology and Implant Dentistry, New York University College of Dentistry, New York, New York, Private Practice, New York, New York

Robert A. Horowitz, DDS
Clinical Assistant Professor, Department of Oral and Maxillofacial Surgery, New York University College of Dentistry, New York, New York

Figures and Images

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Table 1

Table 2

Table 3

Table 4

Figure 1

Fig 1. Blood samples.

Figure 2

Fig 2. Blood clot collection.

Figure 3

Fig 3. Clots doped with rhPDGF-ββ.

Figure 4

Fig 4. Clots doped with rhBMP-2.

Figure 5

Fig 5. Method used for generation of biologic membrane.

Figure 6

Fig 6. Demineralized sponge.

Figure 7

Fig 7. 3D scaffold.

Table 1

Comparison of Raw Data Vs Release of Protein per Hour for BMP-2.

Table 2

Comparison of Raw Data Vs Release of Protein per Hour for PDGF-ββ.

Table 3

Release of PDGF Over Experimental Timeframe.

Table 4

Release of BMP-2 Over Experimental Timeframe.