Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Oct 1.
Published in final edited form as: Biomaterials. 2010 Jul 31;31(30):7805–7812. doi: 10.1016/j.biomaterials.2010.07.010

Sequential Delivery of Vascular Endothelial Growth Factor and Sphingosine 1-Phosphate for Angiogenesis

J E Tengood, K M Kovach, P E Vescovi, A Russell, S R Little
PMCID: PMC2924947  NIHMSID: NIHMS222495  PMID: 20674008

1. Introduction

As the primary source of oxygen and nutrients for cells, the vascular system serves a critical role in normal wound healing[1-4]. It could be argued that establishing this supply is an essential first step to regenerative medicine. To this end, progress has been made towards the promotion of angiogenesis in vivo by delivery of various angiogenic growth factors. Yet, delivery of a single factor alone (such as vascular endothelial growth factor, VEGF), is known to be associated with weak and leaky vessels[5]. Consequently, it has been hypothesized that a combination of angiogenic growth factors might be the key to inducing functional, mature angiogenesis that integrates with the existing vasculature[5]. Yet, the process of angiogenesis is an organized series of events, beginning with vessel destabilization, and followed by endothelial cell proliferation and migration, and lastly vessel maturation[5]. During these events, it is thought that different angiogenic factors become important at different points in time[6].

Many factors have shown to be important regulators of angiogenesis, including vascular endothelial growth factor (VEGF), basic fibroblast growth factor, platelet derived growth factor and epidermal growth factor[5]. Certain factors have already been identified as playing a roles in a specific stage of angiogenesis, such as endothelial cell migration and proliferation[5], vascular network maturation[7] and induce a proangiogenic phenotype in endothelial cells[8]. Of these factors, VEGF and sphingosine-1-phosphate (S1P) are two with well documented and distinctive roles. Although VEGF is known to mediate the recruitment of endothelial cells[5], it has been observed that S1P (an angiogenic factor shown to stabilize intracellular junctions and decrease permeability of endothelial cells[9-10]), inhibits the recruitment of these endothelial cells. Furthermore, an examination of S1P and VEGF signaling in endothelial cells suggests that there is a preferred sequence of factor presence and absence during the formation of mature vasculature[11-14]. In light of these data, it is reasonable to speculate that the a logical strategy to stimulate growth of neovasculature would be to first induce recruitment of endothelial cells through VEGF (without inhibition from S1P), followed by the onset of endothelial cell arrangement and mural cell recruitment due to subsequent presence of S1P (without inhibition from VEGF). In other words, exhibiting control over the absence of a given angiogenic factor may be just as important as control over the presence of that factor in a given stage of angiogenesis. VEGF and S1P are an example of factors in which their temporal presence may affect their action on a particular physiological process.

Controlled release is one viable strategy for achieving temporal presentation of small molecules and proteins in a format that can be applied therapeutically. Yet, to date, achieving such a complex release profile has proven elusive[15-17]. For instance, dual delivery of basic fibroblast growth factor (bFGF) and VEGF[18] as well as angiopoietin-1 and VEGF[19] have been explored previously. In these studies, angiogenic growth factors were loaded into the same scaffold so that release of these factors occurs simultaneously (e.g. dual delivery). In addition, several attempts have been made adjust the release of two factors independently (VEGF and PDGF[16, 20]), where each growth factor is loaded into a different scaffold (i.e. each factor is provided its own “resistance” to release over time). Accordingly, VEGF and PDGF were released at different rates, leading to some observable differences in response[16, 20]. Yet, to study systems where the function of a growth factor may inhibit the function of another (e.g. angiogenesis), it would be desirable to develop a model where temporal separation of biomolecule release can be easily tuned.

Here, we describe a sequential delivery model based upon a porous hollow fiber that extends into an acellular site (in vitro or in vivo), permitting external control over presence and absence of angiogenic factors at any time. In this model, a hollow fiber membrane separates the angiogenic factor “reservoir”, which resides in the lumen of the fiber, from a scaffold for cellular infiltration. Due to the ease of accessibility to the hollow fiber lumen, this system is extremely modular, allowing for a quick change in factor delivery at any point in time. The fiber wall microstructure can be controlled through the hollow fiber fabrication process to ensure that large proteins can be effectively released over time to the surrounding matrix[21-24]. We have used this model to study the hypothesis that the sequence and delivery schedule of VEGF and S1P will impact the significance and maturity of angiogenesis, based on evidence that the presence of one factor might inhibit the performance of another factor.

2. Materials and Methods

2.1. Hollow Fiber Fabrication and Characterization

Cellulose acetate hollow fibers were prepared using a double injection nozzle (14G/20G) and two syringe pumps (Braintree Scientific). Twenty percent cellulose acetate (30kD, Aldrich) in a DMSO/acetone/isopropanol/water [49:15:15:1 weight%] was pumped through the outer core of the nozzle at 1.5mL/min and deionized water was pumped through the center core at 10mL/min. The cellulose solution and deionized water were extruded into a deionized water bath where the cellulose solution precipitates in the form of a porous hollow fiber, as previously described[25], creating a flexible hollow fiber membrane capable of implantation into an animal. Hollow fibers were sterilized with UV light and stored in deionized water for future use. Lyophilized hollow fiber cross sections were sputter coated with 3.5nm of gold-palladium and imaged at 5kV using a JEOL 9335 SEM.

2.2. In Vitro Release

Wells of a 6-well cell culture plate were filled with 3mL Dulbecco’s phosphate buffered saline, or PBS (Invitrogen) and a cellulose hollow fiber was cut to fit the well and injected with 10μL of rmVEGF (R&D Systems) and Fluorescein (Sigma) using a 28G½ insulin syringe (1/2 cc Lo-Dose U-100 insulin syringe, Becton Dickinson and Co.). Hollow fibers were injected first with VEGF (100μg/mL) and subsequent release into a PBS bath was measured by sampling the supernatant and measuring using a VEGF ELISA kit (R&D Systems). After 24 hours, the fiber was rinsed five times with PBS and lumen contents were replaced with an aqueous solution of fluorescein (1800μM). Again, release was measured by sampling the supernatant and measuring fluorescence emissions every hour on a plate reader (SpectraMaxM5, Molecular Devices).

2.3. Murine Matrigel Plug Assay

Growth factor reduced Matrigel (500μL) was injected into the subcutaneous space on the dorsal side of C57BL/6 mice (8-10 weeks old, Charles River) on both the left and right flank, following anesthesia with 2-3% inhaled isoflurane. After five minutes (to permit gelling), a 12G needle was used to thread cellulose hollow fibers through the skin and Matrigel plugs. Hollow fibers were fixed in place using tissue glue and an Elizabethan collar was used to prevent mice from extracting the hollow fiber. On the day of implantation and every day for the next 6 days, hollow fibers on the left side were injected with sterile saline, as an internal negative control, and hollow fibers on the right side were injected with 10μL of an angiogenesis promoting factor: 100μg/mL VEGF (R&D) and/or 1800μM S1P. For mice in the sequential delivery groups, factor switching occurred on the third day after implantation, following five rinses with saline. Seven days post-implantation, implants were extracted, fixed in 2% paraformaldehyde for 5 hours and 30% sucrose overnight and snap-frozen in liquid nitrogen. Frozen sections (8μm) were stained with Hemotoxylin and Eosin (H&E) and analyzed for endothelial cell migration and vessel formation.

2.4. Immunofluorescence

Frozen Matrigel Plug sections (8μm) were incubated with primary antibodies rabbit anti-CD31 (Abcam) and Cy3-conjugated mouse anti-α-smooth muscle actin (Sigma) and secondary antibody goat anti rabbit Alexa Fluor 488® (Jackson Immuno). Sections were also counterstained with Hoescht (Sigma) to identify all mononuclear cells. Images of CD31 labeled cross-sections were taken at 40x. These images were analyzed using threshold analysis on Metamorph to quantify the percent of each image occupied by CD31 staining. These values were averaged to obtain a representative percent for each cross-section and normalized to the internal positive control in which only saline was delivered.

2.5. Statistical Analysis

ANOVA was performed when assays contained more than one experimental group, as in the tubular formation assay (n=3) and Murine Matrigel plug assay (n=3). Pilot studies and a power analysis were performed to determine N for in vivo experiments. Subsequently, a post hoc multiple comparison test was performed to compare means of different experimental groups (Holm-Bonferroni, α=0.05, k=4).

3. Results

3.1. Hollow Fiber Fabrication

To test our hypothesis, we required a delivery system capable of true, sequential release. A hollow fiber based system (in which both ends extend out from the site of delivery) would effectively accomplish this task as long as the wall porosity was made to be large enough to facilitate protein delivery. Given that commercially available fibers typically have smaller pores that do not permit protein delivery over the required time scales, we chose to fabricate fibers in-house using a double injection extrusion/precipitation method. Cellulose was chosen as a non-biodegradable, but biocompatible material. An SEM image of the hollow fiber wall shows the complicated pore structure consisting of both macropores (>10μm) and micropores (<1μm), where the micropores (being the rate limiting portion of delivery) control the rate of delivery from the lumen of the fiber to the surrounding environment (Figure 1a). A higher magnification SEM image shows the interconnected pore structure (less than 1μm) of the cellulose hollow fibers (Figure 1b). The hollow fiber wall thickness was 114.45±10.7μm and the inner diameter was 863.49±66.89 μm.

Figure 1.

Figure 1

Open in a new tab

Scanning electron images of cellulose hollow fiber fabricated using a double extrusion nozzle (14G/20G) with a 20% cellulose solution, cellulose flowing at 1.5mL/min and water flowing at 10mL/min. (a) Hollow fiber wall depicting porous structure of hollow fiber from lumen (L) outward. The edges of the wall display marcopores (denoted as M) around 10μm in width and 30-50μm in length. (b) The microporous voids (denoted as μ) of the remaining scaffold are less than 1μm.

3.2. Sequential Delivery of Molecules of Relevant Size

A hollow fiber-based release system was chosen to present factors sequentially because of the precision afforded through external regulation of the lumen contents over time. For the purpose of ensuring that fibers are capable of sequential control, we chose to modulate the presence/absence of two factors in the lumen of the fibers over time: 1) vascular endothelial growth factor (VEGF, 45kDa) and 2) fluorescein (FITC, 376Da) as an easily detectible molecule of similar size and hydrophobicity to S1P (379Da). This was chosen instead of fluorescently labeled S1P as labeling would significantly increase the molecular size of the molecule (by approximately 100%). Specifically, porous fibers were loaded with VEGF for an initial period of release, rinsed and then subsequently loaded with fluorescein. Egress of these molecules through the fibers and into a surrounding saline solution is represented in Figure 2. Importantly, when factors are exchanged (corresponding with saline flushing prior to administration of a new factor, depicted by the dotted line), VEGF release decreases and fluorescein is subsequently detectable in the supernatant. These results suggest that our fibers are readily capable of sustained release of a growth factor sized protein over at least 24 hours as well as sequential delivery of two factors, as determined empirically.

Figure 2.

Figure 2

Open in a new tab

Release profile of sequentially delivered factors from a cellulose hollow fiber, where dotted line represents the time at which fiber was rinsed. Following injection of VEGF (100μg/mL), release is sustained for 24 hours before the fiber is rinsed five times with PBS. VEGF release drops after rinsing at 24 hours. Injection of FITC (1800μM) occurs at 24 hours, where release is sustained for 24 hours.

3.3. Endothelial Cell Recruitment and Vessel Formation

A modified murine Matrigel plug assay was utilized to measure angiogenesis in response to various delivery regimens in vivo. Specifically, a subcutaneous Matrigel plug serves as a cell-free matrix that is amenable to cellular invasion. A fiber is threaded through this plug to create a source for factor release to surrounding cells. The ends of the hollow fiber remain exposed, giving access to the contents of the lumen of the fiber (and consequently what is released into the cell-free matrix) over the course of experimentation. We explored delivery of: 1) VEGF alone (Figure 3b), 2) S1P alone (Figure 3d), 3) VEGF followed by S1P (Figure 3c), 4) S1P followed by VEGF (Figure 3e), and 5) dual delivery of VEGF and S1P (Figure 3f). Each experimental group contained an internal negative control where saline alone was administered through an implanted fiber (Figure 3a) over the course of experimentation (7 days). In the sequential delivery groups, factor exchange (when relevant) occurred at 3 days post-implantation (as endothelial cell recruitment and vessel formation has previously been observed as early as 2 days in murine Matrigel plugs[26]). Hemotoxylin and Eosin stained sections (Figures 3a-3f) reveal detectible cellular infiltration in all groups (purple nuclear stain). However, cellular infiltration into the Matrigel is more prevalent in the plugs in which an angiogenic factor has been delivered (Figures 3b-3f). Importantly, in the plugs where VEGF delivery was followed by S1P delivery, H&E staining not only reveals denser cells, but the presence of red blood cells are indicative of functional angiogenesis within the Matrigel plug (Figs. 3c and 3g). This same result (the presence of red blood cells surrounded by mononuclear cells in a tubular formation) was sometimes seen in plugs in which VEGF or S1P were delivered alone or together, but with much less frequency than in the group where VEGF delivery was followed by S1P delivery, as depicted in Figure 3.

Figure 3. Sequential delivery of VEGF and S1P results in cellular recruitment and functional angiogenesis in vivo.

Figure 3

Open in a new tab

(a-f) H&E images of murine matrigel plugs (scale bar=500μm). (a) Saline. (b) VEGF (100μg/mL). (c) VEGF (100μg/mL), followed by S1P (1800μM). (d) S1P (1800μM). (e) S1P (1800μM), followed by VEGF (100μg/mL). (f) VEGF (100μg/mL) and S1P together (1800μM). (d) Magnification of blood vessels observed when delivery of VEGF (100μg/mL) was followed by delivery S1P (1800μM), dotted line in (c) (scale bar=50μm).

Similar results were observed in the CD31 stained Matrigel plug sections (Figures 4a–4f). Generally, CD31+ staining was more prevalent in groups where angiogenesis promoting factors were delivered as compared to internal negative controls. However, greater amounts of CD31+ staining were observed in plugs where VEGF-then-S1P or VEGF alone was delivered as compared to all other groups (Figures 4b and 4c). Additionally, we observed that in groups where VEGF delivery was followed by S1P delivery, endothelial cells had arranged into tubular structures that appear larger than that of a capillary, indicating that this delivery schedule is not only capable of promoting angiogenesis in the acellular matrix on the capillary level, but also a larger, more developed vascular network (Figure 4c)[27].

Figure 4. Delivery of VEGF followed by S1P results in a greater recruitment of CD31+ cells in vivo than other delivery schedules.

Figure 4

Open in a new tab

(a-f). Immunoflourscent staining of CD31 (green) and nuclei (blue) in Matrigel plug crossections, scale bar=100μm. (a) Saline. (b) VEGF (100μg/mL). (c) VEGF (100μg/mL), followed by S1P (1800μM). (d) S1P (1800μM). (e) S1P (1800μM), followed by VEGF (100μg/mL). (f) VEGF (100μg/mL) and S1P (1800μM). (g) CD31 quantification based on Metamorph threshold imaging and normalization to a saline injected plug. Percent areas of images covered by CD31 staining are averaged across all plugs. Negative control plug percent areas (saline injection, left flank) for each mouse was subtracted from the Experimental Group percent areas (right flank) for a normalized percent area for each mouse. *significantly different when compared to all other groups (ANOVA, followed by Holm-Bonferroni correction for t-test of multiple comparisons, k=4, α=0.05)

A semi-quantitative method for endothelial cell migration was also performed using CD31 staining of Matrigel plug sections. The percent area of images that were labeled with Alexa Fluor 488 (secondary antibody) was used to quantify CD31 expression in each sample. Images representing the entire periphery of the plug were recorded, and an average percent area was determined (Figure 4g). It is evident that statistically more CD31+ cells are observed in sections of the Matrigel plug treated with the VEGF-then-S1P regimen than in any other experimental group.

3.4. Vasculature Maturation Index

A quantitative method was used for determining the maturation level of a vessel using CD31 and αSMA staining of Matrigel plug explants (CD31 is present on endothelial cells and αSMA is present on mural cells). The colocalization of these two cell types is indicative of mature vessels[28]. Five, 60x areas in which CD31+ cells have arranged in a capillary-like structure were examined, and the percent of αSMA+ colocalization was recorded as the maturation index[28]. In general, fluorescent images illustrate that αSMA colocalization with CD31 can be seen in Matrigel plugs in the following groups: VEGF-then-S1P (Figure 5b), S1P (Figure 5d) and S1P-then-VEGF (Figure 5e). A magnified image of αSMA+ vessels from the VEGF-then-S1P group (Figure 5f) shows αSMA staining surrounding the CD31+ vessels. In the plugs where only VEGF was delivered, we see only CD31 positive cells and no αSMA positive cells (Fig 5a). When VEGF and S1P are delivered together (dual delivery), both CD31 and αSMA positive cells have migrated into the Matrigel plug, but we did not observe substantial co-localization of these cells (representative image shown in Figure 5c). The maturation index (percent of vessels co-localized with αSMA+ cells) is highest when sequential delivery is utilized, specifically when VEGF delivery is followed by S1P delivery (Figure 5g).

Figure 5. Delivery of VEGF followed by S1P results in greater colocalization of CD31 and αSMA in vivo than other delivery schedules.

Figure 5

Open in a new tab

(a-e) Immunoflourscent staining of CD31 (green), αSMA (red) and nuclei (blue) in Matrigel plug crossections (scale bar=100μm). (a) VEGF (100μg/mL). (b) VEGF (100μg/mL), followed by S1P (1800μM). (c) VEGF (100μg/mL) and S1P (1800μM). (d) S1P (1800μM). (e) S1P (1800μM), followed by VEGF (100μg/mL). (f) Co-localization of CD31 and αSMA when delivery of VEGF (100μg/mL) was followed by delivery S1P (1800μM), dotted line in (b), scale bar=50μm. (g) Maturation index calculated by the percent of CD31+ blood vessel that are co-localized with αSMA staining in areas where CD31+ blood vessels were observed. *significantly different when compared to all other groups (ANOVA, followed by Holm-Bonferroni correction for t-test of multiple comparisons, k=4, α=0.05)

4. Discussion

Controlled release systems capable of delivering single biological factors are common in medical therapies today[29], while controlled release systems capable of delivering multiple factors either simultaneously or sequentially are under development as an active area of current research. Fine control over sequential delivery would yield a number of therapeutic advantages including the added efficiency resulting from more accurately mimicking natural schedules of angiogenic factor presentation in situ. To this end, studies have demonstrated dual protein release through fully implantable hollow fibers and/or scaffolds where the rate of release is controlled by the respective degradation rate of either the hollow fiber or the scaffold (or both) [30-32]. While these systems can effectively deliver a single factor or a combination of factors simultaneously at different rates, these systems are not capable of sequential delivery where the onset of delivery for one factor is accompanied by the simultaneous abrogation of release for the other factor. The goal of our study was to create and utilize a system that is capable of exploring sequential delivery of multiple angiogenic factors to an acellular site that is conducive to endothelial cell invasion.

Porous hollow fibers allow for sequential delivery of multiple factors to the surrounding environment as determined exclusively by the contents of the lumen at any time (as externally controlled by the user). Figure 2 demonstrates the capability of these hollow fibers to sequentially deliver molecules of relevant sizes and hydrophobicity. Further, the hollow fibers fabricated in this study have shown to be effective at delivering angiogenic factors over at least 1.25mm (radius of the Matrigel plug in vivo) at physiologically relevant concentrations in an externally controlled and sequential manner. In vivo, it is possible that these fibers may experience more advanced membrane fouling than observed in our in vitro studies either due to protein accumulation in the Matrigel plug or cell-mediated barrier formation at the surface of the fiber. The material used for the hollow fiber-based model, cellulose, was chosen to mitigate this risk as a biologically inert material[33]. Indeed, at the experimental endpoint of our studies, no cellular infiltration into the membrane or cellular adhesion onto the membrane surface was observed. Furthermore, on the time-scale of our studies, we did not observe that potential hindrances to diffusion were extensive enough to impair the cellular infiltration and vessel formation induced by both single-factor and (to a greater extent) sequential delivery.

A well-controlled delivery system (such as the one described here) is important to studying the effects of angiogenic factors in vivo given that the alternative (bolus injections of “naked” factors) would result in rapid diffusion and exposure of released agents to enzymes and other proteins that can lead to a dramatic loss of bioactivity (e.g. the half-life of VEGF in serum is 33.7 minutes[34]). Hollow fibers, conversely, would sustain the release of angiogenic factors (originating from the fiber and extending out through extracellular matrix) over an extended period of time. Our hollow fiber system (Figure SI1h) allows for external control over delivery to an internal in vivo location. Following a rinsing step, delivery of one factor can be “turned off”, while delivery of another factor is simultaneously “turned on” (Figure 2). This setup allows us to test the hypothesis (for the first time) that sequential delivery will improve angiogenic response.

Angiogenesis is an ideal regenerative process to explore the advantages of sequential delivery due to its well-studied, stage-wise nature[5, 35]. Early stage angiogenic events include destabilization of existing vessels, as well as proliferation, migration and invasion, of activated endothelial cells[5]. VEGF appears to be involved primarily in the initiation of angiogenesis[16], playing a major role in vascular permeability and endothelial cell recruitment[5]. This is consistent with our data indicating that VEGF efficiently recruits endothelial cells to a subcutaneous Matrigel plug (Figure 4b). However (as discussed in more detail below), the promising early angiogenic events observed when VEGF was exclusively delivered did not progress further as to produce detectable maturation events. Similarly, it has been shown elsewhere in long-term clinical trials that delivery of VEGF alone has led to unstable vessels[36-37]. Remarkably, these results are entirely consistent with studies that suggest that VEGF mediates cellular effects that are conducive to early-stage angiogenic events while being (by definition) inhibitory to later stage angiogenesis events. Specifically, VEGF inhibits pericyte coverage of vascular sprouts by suppressing receptors on vascular smooth muscle cells, leading to existing vessel destabilization[12]. Together, these data suggest that VEGF alone is likely insufficient to complete angiogenesis given its dual role as a promoter of endothelial cell function and a negative regulator of vessel maturation[12-14].

In contrast, late stage angiogenesis events include inhibition of endothelial cell proliferation and migration, basement membrane secretion and pericyte recruitment[5]. These events appear to be mediated (at least in part) through S1P and, as stated above, inhibited by VEGF. It has also been shown that elevated levels of S1P can lead to a reduction in endothelial cell migration via rearrangement of their cytoskeleton[38-39]. These observations are consistent with our data, showing the S1P delivery is less effective at recruiting endothelial cells when compared to VEGF (Figure 4g, p=0.023). Rather, S1P is released from activated platelets following injury and has been shown to promote vessel stabilization in vivo[40-41]. Indeed, the importance of S1P in vessel maturation is evident by the fact that knockout of the S1P receptor on endothelial cells S1P1 is embryonic lethal in mice due to severe hemorrhaging[42]. Upon closer inspection, it was observed that these embryos were deficient in mural cells and vascular pericytes, causing microvessels to dilate and rupture[42]. Furthermore, VEGF has been shown to not only upregulate the S1P receptor (S1P1) on endothelial cells[11] but also to increase sphingosine kinase activity[43], leading to the conversion of sphingosine to S1P. For these reasons, it is logical to believe that late stage angiogenesis is characterized not only by the presence of S1P, but also the absence of VEGF.

In addition, productive angiogenesis requires both recruitment of endothelial cells into an acellular site and assembly of these cells into patent, stable vessels. A hallmark characteristic of stable (or mature) vessels is the presence of vascular pericytes supporting the endothelial cell structure[7, 12]. Although microvascular pericytes are poorly understood[44-45], their importance is demonstrated by the pathological phenotypes of mice with poor pericyte development [46-48]. It is known that pericyte function occurs in relatively late microvascular development events[44-45], corresponding to our data that suggests S1P (a factor known for vessel stabilization via activated endothelial cell recruitment of vascular pericytes cells[49]), is best delivered during late angiogenesis development. When examining endothelial cell/pericyte colocalization, it was observed that the highest amount of colocalization occurred when VEGF delivery was followed by S1P delivery (Figure 5). This delivery schedule also resulted in the most endothelial cell recruitment and tubular formation of these endothelial cells (Figure 4). Pericyte coverage of newly forming vasculature provides support and stability for these recruited endothelial cells. As consistent with the literature cited above describing the cellular effects of VEGF and S1P, our results suggest that delivering S1P with VEGF diminishes the effects of both VEGF alone.

Because of the versatility of our experimental model, dosing of VEGF and S1P can be optimized to result in quicker, more stable vessel formation. Our sequential delivery regimen (Figure 3a) was based on reported evidence that endothelial cells can be recruited to a site and form vasculature is as little as three days[26], as well as evidence for appropriate (physiologically relevant) concentrations of S1P and VEGF[40, 50]. However, the cited literature references do not involve support of a growth factor gradient, which may affect the desired dosing. Simply changing the injection timing and concentration can be used to examine the effects of altering the quantities released and the schedule and timing of that release. Additionally, the hollow fiber porosity can be altered by changing key components in the fiber fabrication process, such as cellulose flow rate and cellulose concentration. Furthermore, changing the porosity of the fiber wall leads to a change the rate at which factors are delivered. For these reasons, our model can be used as a versatile tool to examine various delivery schedules for any given set of growth factors delivered sequentially. Information obtained from these studies could pave the way for programming fully injectable, sequential delivery systems, a feat made feasible through recently published mathematical models that can direct the design and fabrication of biodegradable matrices to produce complex controlled release behavior[51-52].

Furthermore, this system can be used to explore sequential delivery of any number of different growth factors for therapeutic responses as well as for studying the biological events leading to stage-wise regeneration of other tissues. To this end, we are currently exploring the delivery of basic fibroblast growth factor, or bFGF, followed by PDGF. These growth factors are also known to be involved with early and late stage angiogenesis events, respectively[16, 53-55]. It has also be observed that bFGF induced tubular structures will regress over time in the absence of other signals[37]. We believe that delivery of bFGF followed by PDGF will result in more mature, stable vessels than delivery of either factor alone as well as dual delivery of these factors. It is also expected that sequential delivery of growth factors will prove to be relevant in other wound healing mechanisms, such as bone healing, in which delivery of an angiogenesis promoting factor like PDGF (that can inhibit osteoblast differentiation) would be followed by delivery of a bone morphogenic protein[56].

CONCLUSION

We have created a system capable of exploring true sequential delivery of angiogenic factors. When using this system to explore sequential delivery of VEGF and S1P for the purpose of promoting angiogenesis, we demonstrated that delivery of VEGF for 3 days followed by delivery of S1P for 4 days resulted in recruitment of more endothelial cells and a higher maturation index than the reverse sequential delivery schedule, single factor delivery or dual delivery. This system can be used to explore any number of delivery schedules, allowing for a facile way to explore different delivery schedules of growth factors in vivo for therapeutic responses as well as for studying the basic biological signals that accompany stage-wise regeneration of tissues.

Supplementary Material

01
02

5. Acknowledgements

This work was supported by the Commonwealth of Pennsylvania Research Development and by the Department of Defense Telemedicine and Advanced Research Center.

REFERENCES

  • [1].Wu Y, Chen L, Scott PG, Tredget EE. Mesenchymal stem cells enhance wound healing through differentiation and angiogenesis. Stem Cells. 2007;25:2648–59. doi: 10.1634/stemcells.2007-0226. [DOI] [PubMed] [Google Scholar]
  • [2].Kawanabe T, Kawakami T, Yatomi Y, Shimada S, Soma Y. Sphingosine 1-phosphate accelerates wound healing in diabetic mice. J Dermatol Sci. 2007;48:53–60. doi: 10.1016/j.jdermsci.2007.06.002. [DOI] [PubMed] [Google Scholar]
  • [3].Pandya NM, Dhalla NS, Santani DD. Angiogenesis--a new target for future therapy. Vasc Pharmacol. 2006;44:265–74. doi: 10.1016/j.vph.2006.01.005. [DOI] [PubMed] [Google Scholar]
  • [4].Bauer SM, Bauer RJ, Liu ZJ, Chen H, Goldstein L, Velazquez OC. Vascular endothelial growth factor-C promotes vasculogenesis, angiogenesis, and collagen constriction in three-dimensional collagen gels. J Vasc Surg. 2005;41:699–707. doi: 10.1016/j.jvs.2005.01.015. [DOI] [PubMed] [Google Scholar]
  • [5].Bouïs D, Kusumanto Y, Meijer C, Mulder NH, Hospers GA. A review on pro- and anti-angiogenic factors as targets of clinical intervention. Pharmacol Res. 2006;53:89–103. doi: 10.1016/j.phrs.2005.10.006. [DOI] [PubMed] [Google Scholar]
  • [6].Weiss S, Zimmermann G, Pufe T, Varoga D, Henle P. The systemic angiogenic response during bone healing. Arch Orthop Trauma Surg. 2009;129:989–97. doi: 10.1007/s00402-008-0777-5. [DOI] [PubMed] [Google Scholar]
  • [7].Cao R, Brakenhielm E, Pawliuk R, Wariaro D, Post MJ, Wahlberg E, et al. Angiogenic synergism, vascular stability and improvement of hind-limb ischemia by a combination of PDGF-BB and FGF-2. Nat Med. 2003;9:604–13. doi: 10.1038/nm848. [DOI] [PubMed] [Google Scholar]
  • [8].De Marchis F, Ribatti D, Giampietri C, Lentini A, Faraone D, Scoccianti M, et al. Platelet-derived growth factor inhibits basic fibroblast growth factor angiogenic properties in vitro and in vivo through its alpha receptor. Blood. 2002;99:2045–53. doi: 10.1182/blood.v99.6.2045. [DOI] [PubMed] [Google Scholar]
  • [9].Gavrilovskaya IN, Gorbunova EE, Mackow NA, Mackow ER. Hantaviruses direct endothelial cell permeability by sensitizing cells to the vascular permeability factor VEGF, while angiopoietin 1 and sphingosine 1-phosphate inhibit hantavirus-directed permeability. J Virol. 2008;82:5797–806. doi: 10.1128/JVI.02397-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].McVerry BJ, Garcia JG. In vitro and in vivo modulation of vascular barrier integrity by sphingosine 1-phosphate: mechanistic insights. Cell Signal. 2005;17:131–9. doi: 10.1016/j.cellsig.2004.08.006. [DOI] [PubMed] [Google Scholar]
  • [11].Igarashi J, Erwin PA, Dantas AP, Chen H, Michel T. VEGF induces S1P1 receptors in endothelial cells: Implications for cross-talk between sphingolipid and growth factor receptors. Proc Natl Acad Sci U S A. 2003;100:10664–9. doi: 10.1073/pnas.1934494100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Greenberg JI, Shields DJ, Barillas SG, Acevedo LM, Murphy E, Huang J, et al. A role for VEGF as a negative regulator of pericyte function and vessel maturation. Nature. 2008;456:809–13. doi: 10.1038/nature07424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Hendel RC, Henry TD, Rocha-Singh K, Isner JM, Kereiakes DJ, Giordano FJ, et al. Effect of intracoronary recombinant human vascular endothelial growth factor on myocardial perfusion: evidence for a dose-dependent effect. Circulation. 2000;101:118–21. doi: 10.1161/01.cir.101.2.118. [DOI] [PubMed] [Google Scholar]
  • [14].Yancopoulos GD, Davis S, Gale NW, Rudge JS, Wiegand SJ, Holash J. Vascular-specific growth factors and blood vessel formation. Nature. 2000;407:242–8. doi: 10.1038/35025215. [DOI] [PubMed] [Google Scholar]
  • [15].Peattie RA, Nayate AP, Firpo MA, Shelby J, Fisher RJ, Prestwich GD. Stimulation of in vivo angiogenesis by cytokine-loaded hyaluronic acid hydrogel implants. Biomaterials. 2004;25:2789–98. doi: 10.1016/j.biomaterials.2003.09.054. [DOI] [PubMed] [Google Scholar]
  • [16].Richardson TP, Peters MC, Ennett AB, Mooney DJ. Polymeric system for dual growth factor delivery. Nat Biotechnol. 2001;19:1029–34. doi: 10.1038/nbt1101-1029. [DOI] [PubMed] [Google Scholar]
  • [17].Simmons CA, Alsberg E, Hsiong S, Kim WJ, Mooney DJ. Dual growth factor delivery and controlled scaffold degradation enhance in vivo bone formation by transplanted bone marrow stromal cells. Bone. 2004;35:562–9. doi: 10.1016/j.bone.2004.02.027. [DOI] [PubMed] [Google Scholar]
  • [18].Nillesen ST, Geutjes PJ, Wismans R, Schalkwijk J, Daamen WF, van Kuppevelt TH. Increased angiogenesis and blood vessel maturation in acellular collagen-heparin scaffolds containing both FGF2 and VEGF. Biomaterials. 2007;28:1123–31. doi: 10.1016/j.biomaterials.2006.10.029. [DOI] [PubMed] [Google Scholar]
  • [19].Riley CM, Fuegy PW, Firpo MA, Shu XZ, Prestwich GD, Peattie RA. Stimulation of in vivo angiogenesis using dual growth factor-loaded crosslinked glycosaminoglycan hydrogels. Biomaterials. 2006;27:5935–43. doi: 10.1016/j.biomaterials.2006.08.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Chen RR, Silva EA, Yuen WW, Mooney DJ. Spatio-temporal VEGF and PDGF delivery patterns blood vessel formation and maturation. Pharm Res. 2007;24:258–64. doi: 10.1007/s11095-006-9173-4. [DOI] [PubMed] [Google Scholar]
  • [21].Hasegawa M, Sudo A, Komlev VS, Barinov SM, Uchida A. High release of antibiotic from a novel hydroxyapatite with bimodal pore size distribution. J Biomed Mater Res B Appl Biomater. 2004;70:332–9. doi: 10.1002/jbm.b.30047. [DOI] [PubMed] [Google Scholar]
  • [22].Nunes SP, Inoue T. Evidence for spinodal decomposition and nucleation and growth mechanisms during membrane formation. J Membr Sci. 1996;111:93–103. [Google Scholar]
  • [23].Shih CH, Gryte CC, Cheng LP. Morphology of membranes formed by the isothermal precipitation of polyamide solutions from water/formic acid systems. J Appl Polym Sci. 2005;96:944–60. [Google Scholar]
  • [24].van de Witte P, Dijkstra PJ, van den Berg JWA, Feijen J. Phase separation processes in polymer solutions in relation to membrane formation. J Membr Sci. 1996;117:1–31. [Google Scholar]
  • [25].Ye SH, Watanabe J, Takai M, Iwasaki Y, Ishihara K. Design of functional hollow fiber membranes modified with phospholipid polymers for application in total hemopurification system. Biomaterials. 2005;26:5032–41. doi: 10.1016/j.biomaterials.2005.01.049. [DOI] [PubMed] [Google Scholar]
  • [26].Akhtar N, Dickerson EB, Auerbach R. The sponge/Matrigel angiogenesis assay. Angiogenesis. 2002;5:75–80. doi: 10.1023/a:1021507031486. [DOI] [PubMed] [Google Scholar]
  • [27].Thurston G, Wang Q, Baffert F, Rudge J, Papadopoulos N, Jean-Guillaume D, et al. Angiopoietin 1 causes vessel enlargement, without angiogenic sprouting, during a critical developmental period. Development. 2005;132:3317–26. doi: 10.1242/dev.01888. [DOI] [PubMed] [Google Scholar]
  • [28].Baudelet C, Cron GO, Ansiaux R, Crokart N, DeWever J, Feron O, et al. The role of vessel maturation and vessel functionality in spontaneous fluctuations of T2*-weighted GRE signal within tumors. NMR Biomed. 2006;19:69–76. doi: 10.1002/nbm.1002. [DOI] [PubMed] [Google Scholar]
  • [29].Sinha VR, Trehan A. Biodegradable microspheres for protein delivery. J Control Release. 2003;90:261–80. doi: 10.1016/s0168-3659(03)00194-9. [DOI] [PubMed] [Google Scholar]
  • [30].Tilakaratne HK, Hunter SK, Andracki ME, Benda JA, Rodgers VG. Characterizing short-term release and neovascularization potential of multi-protein growth supplement delivered via alginate hollow fiber devices. Biomaterials. 2007;28:89–98. doi: 10.1016/j.biomaterials.2006.08.011. [DOI] [PubMed] [Google Scholar]
  • [31].Lazerri L, Cascone MG, Quiriconi S, Morabito L, Giusti P. Biodegradable hollow microfibres to produce bioactive scaffolds. Polym Int. 2005;54:101–7. [Google Scholar]
  • [32].Kawakami O, Miyamoto S, Hatano T, Yamada K, Hashimoto N, Tabata Y. Acceleration of aneurysm healing by hollow fiber enabling the controlled release of basic fibroblast growth factor. Neurosurgery. 2006;58:355–64. doi: 10.1227/01.NEU.0000195095.66984.24. [DOI] [PubMed] [Google Scholar]
  • [33].Doheny JG, Jervis EJ, Guarna MM, Humphries RK, Warren RAJ, Kilburn DG. Cellulose as an inert matrix for presenting cytokines to target cells: production and properties of a stem cell factor-cellulose-binding domain fusion protein. Biochem J. 1999;339:429–34. [PMC free article] [PubMed] [Google Scholar]
  • [34].Eppler SM, Combs DL, Henry TD, Lopez JJ, Ellis SG, Yi JH, et al. A target-mediated model to describe the pharmacokinetics and hemodynamic effects of recombinant human vascular endothelial growth factor in humans. Clin Pharmacol Ther. 2002;72:20–32. doi: 10.1067/mcp.2002.126179. [DOI] [PubMed] [Google Scholar]
  • [35].Freeman I, Cohen S. The influence of the sequential delivery of angiogenic factors from affinity-binding alginate scaffolds on vascularization. Biomaterials. 2009;30:2122–31. doi: 10.1016/j.biomaterials.2008.12.057. [DOI] [PubMed] [Google Scholar]
  • [36].Isner JM, Pieczek A, Schainfeld R, Blair R, Haley L, Asahara T, et al. Clinical evidence of angiogenesis after arterial gene transfer of phVEGF165 in patient with ischaemic limb. Lancet. 1996;348:370–4. doi: 10.1016/s0140-6736(96)03361-2. [DOI] [PubMed] [Google Scholar]
  • [37].Simons M, Bonow RO, Chronos NA, Cohen DJ, Giordano FJ, Hammond HK, et al. Clinical trials in coronary angiogenesis: issues, problems, consensus: An expert panel summary. Circulation. 2000;102:E73–86. doi: 10.1161/01.cir.102.11.e73. [DOI] [PubMed] [Google Scholar]
  • [38].Wendler CC, Rivkees SA. Sphingosine-1-phosphate inhibits cell migration and endothelial to mesenchymal cell transformation during cardiac development. Dev Biol. 2006;291:264–77. doi: 10.1016/j.ydbio.2005.12.013. [DOI] [PubMed] [Google Scholar]
  • [39].Okamoto H, Takuwa N, Yokomizo T, Sugimoto N, Sakurada S, Shigematsu H, et al. Inhibitory regulation of Rac activation, membrane ruffling, and cell migration by the G protein-coupled sphingosine-1-phosphate receptor EDG5 but not EDG1 or EDG3. Mol Cell Biol. 2000;20:9247–61. doi: 10.1128/mcb.20.24.9247-9261.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Lee OH, Kim YM, Lee YM, Moon EJ, Lee DJ, Kim JH, et al. Sphingosine 1-phosphate induces angiogenesis: its angiogenic action and signaling mechanism in human umbilical vein endothelial cells. Biochem Biophys Res Commun. 1999;264:743–50. doi: 10.1006/bbrc.1999.1586. [DOI] [PubMed] [Google Scholar]
  • [41].Wacker BK, Scott EA, Kaneda MM, Alford SK, Elbert DL. Delivery of sphingosine 1-phosphate from poly(ethylene glycol) hydrogels. Biomacromolecules. 2006;7:1335–43. doi: 10.1021/bm050948r. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Liu Y, Wada R, Yamashita T, Mi Y, Deng CX, Hobson JP, et al. Edg-1, the G protein-coupled receptor for sphingosine-1-phosphate, is essential for vascular maturation. J Clin Invest. 2000;106:951–61. doi: 10.1172/JCI10905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Cuvillier O, Rosenthal DS, Smulson ME, Spiegel S. Sphingosine 1-phosphate inhibits activation of caspases that cleave poly(ADP-ribose) polymerase and lamins during Fas- and ceramide-mediated apoptosis in Jurkat T lymphocytes. J Biol Chem. 1998;273:2910–6. doi: 10.1074/jbc.273.5.2910. [DOI] [PubMed] [Google Scholar]
  • [44].Bergers G, Song S. The role of pericytes in blood-vessel formation and maintenance. Neuro Oncol. 2005;7:452–64. doi: 10.1215/S1152851705000232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Betsholtz C, Lindblom P, Gerhardt H. Role of pericytes in vascular morphogenesis. Exs. 2005:115–25. doi: 10.1007/3-7643-7311-3_8. [DOI] [PubMed] [Google Scholar]
  • [46].Hellstrom M, Gerhardt H, Kalen M, Li X, Eriksson U, Wolburg H, et al. Lack of pericytes leads to endothelial hyperplasia and abnormal vascular morphogenesis. J Cell Biol. 2001;153:543–53. doi: 10.1083/jcb.153.3.543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [47].Lindahl P, Johansson BR, Leveen P, Betsholtz C. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science. 1997;277:242–5. doi: 10.1126/science.277.5323.242. [DOI] [PubMed] [Google Scholar]
  • [48].Enge M, Bjarnegard M, Gerhardt H, Gustafsson E, Kalen M, Asker N, et al. Endothelium-specific platelet-derived growth factor-B ablation mimics diabetic retinopathy. Embo J. 2002;21:4307–16. doi: 10.1093/emboj/cdf418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Allende ML, Proia RL. Sphingosine-1-phosphate receptors and the development of the vascular system. Biochim Biophys Acta. 2002;1582:222–7. doi: 10.1016/s1388-1981(02)00175-0. [DOI] [PubMed] [Google Scholar]
  • [50].Kano MR, Morishita Y, Iwata C, Iwasaka S, Watabe T, Ouchi Y, et al. VEGF-A and FGF-2 synergistically promote neoangiogenesis through enhancement of endogenous PDGF-B-PDGFRbeta signaling. J Cell Sci. 2005;118:3759–68. doi: 10.1242/jcs.02483. [DOI] [PubMed] [Google Scholar]
  • [51].Rothstein SN, Federspiel WJ, Little SR. A simple model framework for the prediction of controlled release from bulk eroding polymer matrices. J Mater Chem. 2008;18:1873–80. doi: 10.1016/j.biomaterials.2008.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [52].Rothstein SN, Federspiel WJ, Little SR. A unified mathematical model for the prediction of controlled release from surface and bulk eroding polymer matrices. Biomaterials. 2009;30:1657–64. doi: 10.1016/j.biomaterials.2008.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [53].Nakajima H, Sakakibara Y, Tambara K, Iwakura A, Doi K, Marui A, et al. Therapeutic angiogenesis by the controlled release of basic fibroblast growth factor for ischemic limb and heart injury: toward safety and minimal invasiveness. J Artif Organs. 2004;7:58–61. doi: 10.1007/s10047-004-0252-1. [DOI] [PubMed] [Google Scholar]
  • [54].Unger EF, Banai S, Shou M, Lazarous DF, Jaklitsch MT, Scheinowitz M, et al. Basic fibroblast growth factor enhances myocardial collateral flow in a canine model. Am J Physiol. 1994;266:H1588–95. doi: 10.1152/ajpheart.1994.266.4.H1588. [DOI] [PubMed] [Google Scholar]
  • [55].Magnusson PU, Looman C, Ahgren A, Wu Y, Claesson-Welsh L, Heuchel RL. Platelet-Derived Growth Factor Receptor-{beta} Constitutive Activity Promotes Angiogenesis In Vivo and In Vitro. Arterioscler Thromb Vasc Biol. 2007;27:2142–9. doi: 10.1161/01.ATV.0000282198.60701.94. [DOI] [PubMed] [Google Scholar]
  • [56].Huang Z, Nelson ER, Smith RL, Goodman SB. The sequential expression profiles of growth factors from osteoprogenitors [correction of osteroprogenitors] to osteoblasts in vitro. Tissue Eng. 2007;13:2311–20. doi: 10.1089/ten.2006.0423. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS222495-supplement-01.doc (41KB, doc)
02
NIHMS222495-supplement-02.tif (662.2KB, tif)

RESOURCES