|Home | About | Journals | Submit | Contact Us | Français|
Neovascularization is essential for the survival and successful integration of most engineering tissues after implantation in vivo. The objective of this study was to elucidate possible mechanisms of phthalimide neovascular factor 1 (PNF1), a new synthetic small molecule proposed for therapeutic induction of angiogenesis. Complementary deoxyribonucleic acid microarray analysis was used to identify 568 transcripts in human microvascular endothelial cells (HMVECs) that were significantly regulated after 24-h stimulation with 30 μM of PNF1, previously known as SC-3–149. Network analysis tools were used to identify genetic networks of the global biological processes involved in PNF1 stimulation and to describe known molecular and cellular functions that the drug regulated most highly. Examination of the most significantly perturbed networks identified gene products associated with transforming growth factor-beta (TGF-β), which has many known effects on angiogenesis, and related signal transduction pathways. These include molecules integral to the thrombospondin, plasminogen, fibroblast growth factor, epidermal growth factor, ephrin, Rho, and Ras signaling pathways that are essential to endothelial function. Moreover, real-time reverse-transcriptase polymerase chain reaction (RT-PCR) of select genes showed significant increases in TGF-β-associated receptors endoglin and beta glycan. These experiments provide important insight into the pro-angiogenic mechanism of PNF1, namely, TGF-β-associated signaling pathways, and may ultimately offer new molecular targets for directed drug discovery.
The Clinical Goals of Tissue Engineering are to repair, replace, or regenerate damaged or lost tissues in the body. In recent years, exciting new advances in stem cell technologies, scaffold biomaterials, and growth factor therapies have foreshadowed a new era in regenerative medicine as tissue engineering draws nearer to widespread clinical use. However, significant challenges must be addressed before the promise of tissue engineering is fully realized.
One particularly important challenge for the advance of tissue-engineering applications is the development of effective therapies to promote neovascularization of engineered tissues.1–5 The formation of new blood vessels via angiogenesis is critically important for many clinical applications,6–11 particularly tissue engineering applications in which significant nutritive and metabolic demands are created. Currently, research strategies for therapeutic neovascularization have emphasized focal delivery of growth factors9–13 and the investigation of their role in the regulation of microvascular pattern formation. However, although peptide-based growth factors have enormous promise for developing effective strategies in tissue engineering, several potential drawbacks have been reported, such as high costs associated with recovery and purification of recombinant proteins and susceptibility to aggregation and degradation.12,14 Processes such as sterilization, solubilization, and storage that are required to deliver the proteins efficiently in the clinical setting may also exacerbate the latter effects.15 Moreover, size, hydrophilicity, and physical and chemical lability can adversely affect their pharmacokinetic and pharmacodynamic properties.16 Indeed, clinical trials have highlighted potential alterations in efficacy due to mode of growth factor delivery.17
More recently, the importance of naturally occurring small-molecule regulators of angiogenesis has been increasingly recognized. For example, insights obtained from lysophospholipid growth factor receptor knockout animals showing severe abnormalities in vascular development have fueled significant interest in lysophospholipid growth factors such as sphingosine-1-phosphate for use as an angiogenic agent to treat ischemic diseases.18–20 In other reports, dimethyl-oxalylglycine21,22 and other naturally derived alkaloids, terpines, and phenolic compounds have been discovered that inhibit tissue oxygen homeostatic agent hypoxia inducible factor-1 inactivation, resulting in a pleiotropic angiogenic response. These and other studies signal the advent of new small-molecule strategies for induction and therapeutic manipulation of angiogenesis.
Our group recently reported the development of the first synthetic small-molecule inducer of angiogenesis, phthalimide neovascular factor (PNF1, formerly known as SC-3–149).23 In vitro evaluation of the compound demonstrated that it is effective in enhancing proliferation, survivability, and in vitro capillary network formation by endothelial cells.23 Preliminary in vivo studies also showed early-stage angiogenic response induced by the drug in rat mesenteric windows. These studies suggested that PNF1 and its analogues might serve as promising new angiogenic agents for targeted drug delivery and therapeutic angiogenesis in tissue engineering.
The focus of this study was to elucidate candidate mechanisms of the drug using transcriptional regulatory network analysis. Transcriptional profiling is a valuable tool for elucidation of mechanisms underlying many biological pathways.24 Genomic signatures that arise from applying complementary deoxyribonucleic acid (cDNA) microarray technologies can be used to discover unknown mechanisms of action of small molecules and drugs, to predict drug efficacy, and to identify potential new therapeutics.24 We performed a comparative cDNA microarray analysis of gene products isolated from cultures of human microvascular endothelial cells stimulated with PNF1 and those stimulated with a vehicle control.
We used network tools to characterize the transcriptional and molecular level effects of PNF1 stimulation of microvascular endothelial cells in an effort to identify potential mechanistic targets of the drug. Examination of data using network analysis tools instead of evaluation of single-gene perturbation revealed key regulatory events in signaling that may not have led to significant changes in messenger ribonucleic acid (mRNA) expression of any particular gene product. We demonstrate that 24-h stimulation of microvascular endothelial cells with PNF1 results in concerted regulation of many gene products that have known regulatory effects on vascular remodeling. The primary networks of gene products revealed in our analysis are centered upon the transforming growth factor-beta (TGF-β) signaling pathway that has known effects on angiogenesis. Identifying a mechanism for the angiogenic effects of PNF1 is essential for appropriate utilization of these PNF1 and related compounds as vascular therapies, and will further understanding of how small molecule inducers of angiogenesis might impact the fields of tissue engineering and regenerative medicine.
Human microvascular endothelial cells (HMVECs) (Cambrex, Walkersville, MD) were cultured in endothelial growth medium 2-microvascular (bulletkit, BioWhittaker, Walkersville, MD) supplemented as directed with 5% fetal bovine serum. Cells were cultured on tissue culture plates (Nalge Nunc, Rochester, NY) at 37°C in a humidified chamber with 5% carbon dioxide.
HMVECs (passage 9) were plated at 2.5×104 cells/cm2 on tissue culture plastic (Nalge Nunc, Rochester, NY) and grown to confluence. After confluence, medium was refreshed, and 30 μM PNF1 or 0.6% dimethyl sulfoxide vehicle control was added to the sample. After 24 h of supplementation, total RNA from the cultures was isolated using an RNeasy kit (Qiagen, Inc., Valencia, CA) according to the manufacturer’s protocol.
RNA samples were prepared for hybridization using the GeneChip One-Cycle Target Labeling and Control Reagents kit (Affymetrix, Santa Clara, CA). Expression profiles were created using Human Genome U133 Plus 2.0 arrays (Affymetrix) with approximately 47,000 transcripts and variants, including 38,500 well-characterized human gene transcripts. Expression data were analyzed using GeneChip Operating Software (Affymetrix), which calculates signal intensities and detection calls along with their related p-values for each probe set representing a gene transcript. The software automatically assigned the change between the experimental and baseline conditions to a standard category (increase, marginal increase, no change, marginal decrease, or decrease), depending on the p-value (calculated using the Wilcoxon signed rank test); the ranges for each call were 0.0000 to 0.0025, 0.0025 to 0.0030, 0.0030 to 0.9970, 0.9970 to 0.9975, and 0.9975 to 1.0000, respectively. Data will be deposited into the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/).
The gene products identified as significantly differentially regulated (up- or downregulated) according to microarray analysis were used for further network and gene ontology analyses. For these analyses, Ingenuity Pathway Analysis (IPA) software (Ingenuity Systems, www.ingenuity.com), coupled with the Ingenuity Pathways Knowledge Base (IPKB), was employed. IPA has been used recently to study genetic pathways and mechanisms of cell and tissue cultures, those stimulated with small molecules,25–28 toxic elements,29 and radiation therapies30 or those associated with pathologies.31 The IPKB offers genome-wide coverage of more than 23,900 mammalian genes; millions of pathway interactions manually curated from the literature; and representation of biological specificity, including species, location, mutations, and experimental conditions.
A data set containing gene transcript identifiers and fold-change information (of PNF1-stimulated samples over that of controls) was uploaded into IPA. The software was used to overlay a subset of these genes eligible for generating networks onto a global molecular network developed from information contained in the IPKB. The subset of genes, called focus genes, consisted of genes with associated molecular interaction information within the IPKB. Networks of these focus genes were then generated based on their connectivity, with each network composed of a maximum of 35 gene products, as currently limited by the IPA software for computational reasons (physical memory). These networks were ranked according to a score representing the probability that each isolated network of genes could be achieved by chance alone. A score greater than 3 was considered significant; with this score, there is a 1 in 1,000 chance that the focus genes in the network were isolated because of random chance (scores of ≥ 3 have a 99.9% confidence level of not being generated by random chance alone).
Additionally, IPA was used to characterize suggested molecular and cellular functions for the data set as a means of determining key processes in the cell samples that were regulated by treatment with the drug. The IPA tool isolated 244 gene products eligible for this type of functional analysis. The significance associated with a function is a measure of how likely it is that genes from the data set under investigation participate in that function. The significance is expressed as a p-value, which is calculated using the right-tailed Fisher exact test. In this method, the p-value is calculated by comparing the number of user-specified genes of interest that participate in a given function or pathway with the total number of occurrences of these genes in all functional or pathway annotations stored in the IPKB.
IPA generated graphical representations of the molecular relationships between genes and gene products to further investigate the significantly regulated networks. Genes or gene products are represented as nodes, and the biological relationship between any two nodes (e.g., phosphorylation, transcription) is represented as an edge (line). At least one reference from information stored in the IPKB supports each edge. The intensity of the node color indicates the degree of up- (red) or downregulation (green). Nodes are displayed using various shapes that represent the functional class of the gene product. Edges may be displayed with various labels that describe the nature of the relationship between the nodes (e.g., P for phosphorylation or T for transcription).
cDNA (from 0.01 ug/uL of RNA solution) was synthesized with a Bio-Rad iScript kit (Bio-Rad Laboratories, Hercules, CA) according to the manufacturer’s instructions. Five μL of the subsequent cDNA template was added to a reaction well containing 10 μL each of 0.5-μM solutions of forward and reverse gene-specific primers (Table 1) and 25 μL of iQ SYBR Green Supermix (Bio-Rad), including SYBR green, deoxyribonucleotide triphosphates, Taq enzyme, and reaction buffer. PCR reactions were performed using an iCycler thermal cycler (Bio-Rad); an initial denaturing step was performed at 9°C, before 40 cycles of 95°C for 10 s, x°C for 30 s, and 72°C for 45 s, where x is the specific annealing temperature of the primer set. Reactions were performed for the following genes: TGF-β1, activin-like kinase 1 (ALK1), ALK5, TGF-ß receptor II (TGF-βRII), beta glycan, CD105, CD36, epidermal growth factor receptor (EGFR), and acidic fibroblast growth factor 1 (FGF1). The 18S gene was used as a housekeeping gene.
All primer sequences and primer-specific annealing temperatures can be found in Table 1. All assays were performed in triplicate. Data are reported using means±standard errors of the mean.
To investigate the transcriptional regulation of HMVECs by synthetic small-molecule inducer of angiogenesis PNF1, comparative microarray analysis was performed on control- and drug-treated cultures after 24 h of stimulation. Differential expression between control- and PNF1-stimulated culture samples was found in 568 gene transcripts on the array. The GeneChip Operating Software characterized these differentially expressed gene transcripts as increased, decreased, or marginally increased or decreased; results can be found in Table 2. Approximately 57% of the significantly regulated genes were downregulated, and almost 18% were marginally regulated, as determined according to the corresponding p-values as described in the Materials and Methods section. Table 2 also reports the regulation of the 264 genes available for network analysis and the 244 functional analysis genes (see “Network and gene ontology analyses” below). The percentages of genes whose expression level increased or decreased across these subsets of genes were consistent.
Observation of the most significant changes in single gene expression, as measured according to microarray, after 30-μM PNF1 stimulation of HMVECs revealed that the drug substantially upregulated insulin-like growth factor (IGF) binding protein 6 (IGFBP6) and angiopoietin-like factor 4 (ANGPTL4) after 24 h. The pro-angiogenic42 IGF family is known to promote migration and tube formation of vascular endothelial cells,43 and IGFBP6 is upregulated in response to hypoxia in cultured endothelial cells.44 Similarly, ANGPTL4 is an important modulator of vascular permeability45 that is produced during ischemia to promote angiogenesis.46 These molecules may ultimately prove to be viable candidates for signaling pathways that control the pro-angiogenic effects of PNF1 in HMVECs.
We employed network tools to examine the coordinated perturbation of entire networks of gene products by PNF1. The microarray-identified genes were further analyzed using IPA to investigate the nature of their biological interaction. Of the 568 genes, 264 focus genes were identified as the subset of gene products with associated molecular interaction information in the IPKB. Exploration of the regulation of the remaining gene products may be performed in future experiments. The focus genes were then mapped to genetic networks as generated by the IPA tool, using known signaling pathways and molecular interactions from the literature to describe functional relationships between the gene products. IPA then associated the networks with known biologic pathways; 16 networks of 35 genes were identified as significant, in that more of the identified genes were present than could be expected by chance (score > 3). See Table 3 for a comprehensive listing of the identified networks. All of the significantly regulated networks were interconnected; that is, each of the networks can be linked to all of the other PNF1-regulated networks via gene product interactions in the IPKB. The networks were arranged according to decreasing score and percentage incorporation of focus genes from the microarray data. For example, 100% of the genes tabulated in network 1 are focus genes, and the network was assigned a score of 56, meaning that there is a 1 in 1056 chance that the collection of genes were regulated together by chance.
The Functions feature of the IPA suite was used to identify particular high-level cellular processes associated with microvascular stimulation of PNF1 after 24 h. Table 4 details each cellular process and its relative significance or the probability that the included genes are involved with the particular function. The most significantly activated cellular processes included cell-to-cell signaling and interaction, cellular assembly and organization, cellular movement, cellular growth and proliferation, and cell signaling, all of which are integral to the process of angiogenesis. PNF1 has previously been demonstrated to significantly promote endothelial cell proliferation, migration, and tubule formation in vitro,23 and here we report that the genes regulated by this small molecule as identified according to this analysis support our previous findings.
Networks 1 and 2 were identified as having substantially higher scores associated with their regulation and larger incorporation of focus genes than the other networks; hence, we chose to focus on these networks for further analysis. Clustering diagrams of each network can be found in Figure 1A and B, respectively. As described previously, nodes (i.e., gene products) are displayed as various shapes, and the edges between them (i.e., biological relationships) are shown as arrows connecting them. The IPA tool included 5 of the gene products in network 2 (Fig. 1B) in the network, but they were not perturbed, as reported by the original microarray results. It is possible that a concurrent signaling mechanism counter-regulated these non-perturbed gene products or that their individual gene regulation was subtle but sufficient to effect a change in downstream regulation.
Networks 1 and 2 were then joined via known signaling connections between the 2 pathways for identification of candidate signaling pathways for PNF1 (Fig. 2), because there was a large overlap between the gene sets of the 2 networks. IPA identified 26 molecular interactions between the 2 genetic networks, as denoted by blue connectors in Figure 2. The gene product with the most molecular interactions is TGF-β1. TGF-β signaling has been shown to play a significant role in the development and function of the cardiovascular system and vascular network formation.47–51 The TGF-βRs include the serine-threonine kinase type I (including ALK1 and ALK5) and TGF-βRIIs,52,53 and the TGF-βRIIIs extracellular betaglycan54 and endoglin (CD105).55,56 Additionally, there is mounting evidence that the TGF-βs can signal through multiple other pathways.47
Signaling pathways emanating from TGF-β1 in Figure 2 that might represent the molecular mechanism for the in vitro angiogenic effects of PNF1 on HMVECs are presented in Table 5 and illustrated in Figure 3. Each of these signaling pathways are well known to enhance cellular proliferation and viability significantly or give other molecular cues for stimulation of angiogenesis that parallel results reported after stimulating microvascular endothelial cells with PNF1 as an initial determination of drug function.23 These include molecules integral to the thrombospondin, plasminogen, FGF, EGF, ephrin, Rho, and Ras signaling pathways that are important to endothelial function. As Table 5 demonstrates, each of these signaling pathways and associated gene products has been characterized in the literature as having functions that are implicated in angiogenesis or cellular functions related to the angiogenic process, including maintenance of cell viability and adaptation to hypoxia, stimulation of endothelial proliferation and migration, degradation of surrounding extracellular matrix, increased production of other angiogenic factors, and vascular network formation and remodeling. For example, TGF-β1 has been found to decrease expression of CD36,57 a receptor for thrombospondin and mediator of its anti-angiogenic activity;58–61 significant downregulation of CD36 by PNF1 stimulation may inhibit angiostatic signaling by thrombospondin.
Additionally, TGF-β1 is known to decrease expression of cell division cycle 289,90 (cdc2, also known as p34, a ribosome-binding protein), a trigger for the apoptosis signaling cascade,62,63 in various cell types. PNF1 downregulated cdc2 and, hence, may be responsible for observed increases in endothelial cell viability in nutrient-deprived conditions, as has been shown previously.63 Cdc2 has also been found to bind to FGF1,64 a well-known angiogenic growth factor75 with increased expression after PNF1 exposure. TGF-β1 also stimulates expression of RhoB (a Rho family GTPase),66,67 which PNF1 also upregulated. This intracellular signaling molecule, among other functions, controls stage-specific survival of endothelial cells during vascular development via Akt survival signaling pathway;68 therefore, RhoB activation by PNF1 may also contribute to its effects on in vitro cell viability.
Another potential mechanism for the angiogenic effects of PNF1 on microvascular endothelial cells occurs through EGFR. TGF-β1 has been implicated in the activation and phosphorylation of EGFR,69,70 an angiogenic growth factor.71,72 This study demonstrates that upregulation of EGFR after stimulation with PNF1 may induce stimulation of ephrin receptor B2, which is well known to affect proper vascular development and angiogenesis74,75 via Ras activation.73 TGF-β1 protein also has known interactions with hairy/enhancer-of-split related with YRPW motif 1 (HEY1), a binding protein; literature findings suggest that supplementing various types of cells with TGF-β1 increases mRNA expression of HEY1,76 although our data indicate that HEY1 was downregulated after stimulation with PNF1. This may have important implications for the exact mechanism of action, but downregulation of HEY1 may prove to be important to the angiogenic process, because overexpression is known to block normal vascular network formation in mouse embryos.77 High-mobility group A1 (HMGA1), a DNA binding protein that is involved in apoptotic signaling78 and expression of nitric oxide synthase,79 an important signaling molecule in angiogenesis,80 also interacts with TGF-β1.79 Our data show that expression of HMGA1 was upregulated, which is consistent with PNF1 effects on cell death and in vitro tubule formation. Additionally, TGF-β1 upregulates expression of non-muscle myosin polypeptide 9 (MYH9)67 gene product; this protein fragment has been demonstrated to be necessary for angiogenic nucleolin signaling.81
According to our microarray data and literature findings, TGF-β1 upregulates transcription and expression of urokinase plasminogen activator, also known as uPA,82 which further activates tissue-type plasminogen activator, or tPA.83
Both types of plasminogen activators are highly expressed during vascular sprouting and angiogenesis, because they contribute to the degradation of extracellular matrix that is necessary for new vessel growth.84–86 Finally, according to the IPA tool, TGF-β1 indirectly (through mitogen-activated protein kinase (MAPK8) regulates activity of SMT3 suppressor of mif two 3 homolog 1 (yeast) (SUMO1),87 an enzyme that is implicated in apoptosis signaling in the cardiovascular system. Our data indicate that SUMO1 was downregulated, which parallels in vitro cell viability data for PNF1, and that promyelocytic leukemia (PML) protein, a downstream transcription regulator, was highly downregulated. PML is integral to TGF-β1 apoptosis signaling,88 so its regulation may also contribute to decreases in apoptosis seen after PNF1 supplementation in microvascular cell cultures.
To validate results obtained from the microarray experiment, as has been reported in the literature,91 TGF-β1, CD36, EFGR, and FGF1 from networks 1 and 2 were selected for quantitative RT-PCR analysis. As demonstrated previously, TGF-β1 was the gene product with the highest number of signaling interactions in networks 1 and 2. CD36, EGFR, and FGF1 were selected because of their known effects on angiogenic signaling pathways and their high degree of regulation by the novel drug. Because TGF-β1 is known to induce its own receptors,92 we assayed the 5 known receptors as well. These 9 genes had varying magnitudes of differential expression (Table 6). In 8 of 9 (88%) of the assays performed, a gene product that was upregulated in the microarray experiment was also upregulated when using the PCR method. Although RT-PCR did not validate ALK5 signaling perturbation as measured using microarray, the gene was not significantly regulated in either case, showing only a 1.15-fold change according to RT-PCR compared with a negative change in expression value according to microarray.
Target genes specific to TGF-β1R signaling have been investigated previously. Lux et al.33 overexpressed ALK1 in HMVECs via viral transfection to decouple the transcriptional profile of TGF-β1-stimulated endothelial cells from that of ALK1-specific signaling. Table 6 presents the expression levels of the TGF-βRs, as stimulated by low (0.5 ng/mL) and high (4 ng/mL) levels of TGF-β1 and ALK1 overexpression,33 and compares them with results for 30-μM PNF1 obtained via cDNA microarray and real-time RT-PCR analysis. The change in expression of TGF-β1Rs TGFβRII, ALK1, and ALK5 was unremarkable after 24 h, although it is possible that changes in the expression of these genes may occur at earlier timeframes after stimulation with the drug. Moreover, treatment with PNF1 significantly upregulated the beta glycan and endoglin accessory TGF-βRs. Endoglin is primarily expressed in angiogenic endothelial cells and is upregulated during hypoxia93,94 and promotes ALK1 signaling transduction in proliferating endothelial cells.95 Conversely, beta glycan has been found to inhibit angiogenesis in human cancer cells.96 The 2 analogues are known to form complexes with one another on human microvascular endothelial cells.97 It is thought that perhaps the interaction between endoglin and beta glycan may be necessary for maintenance of positive and negative aspects of TGF-β signal transduction on endothelial cells.97
The novel small molecule PNF1 has been shown to stimulate angiogenic activity in microvascular endothelial cells in vitro, including enhanced proliferation, viability, tube formation, and migration.23 Here, our data demonstrate that the stimulation of microvascular endothelial cells over 24 h with PNF1 in its most-effective in vitro concentration (30 μM) produces a significant change in the regulation of their genetic networks and related cell-signaling processes. Although single-gene regulation may prove important in understanding the drug’s mechanism of action, a focus on the concerted regulation of larger networks of genes by an angiogenic agent might lend more valuable insights, because the differential expression of many gene products that are integral to the angiogenic process may be subtle. For example, this type of network analysis could elucidate potential side effects of this or any drug before testing in animal or clinical studies. Thus, we chose to employ network tools to examine the influence of PNF1 on transcriptional regulation of microvascular cells on a larger scale than single-gene product analysis allows.
After 24 h of stimulation, PNF1 perturbs a number of signaling pathways that are known to regulate the angiogenic process. Figure 2 demonstrates that the most-targeted regulation by the drug (networks 1 and 2) occurred through a TGF-β1-related signaling pathway. It is important to note that TGF-β1 is a well-characterized growth factor, and therefore more potential interactions may be curated in the database for this molecule than other gene products in the networks. However, statistical examination of the networks validates the importance of the coordinated perturbation of these genes, which share known relationships to TGF-β1 signaling. Network analysis methods employed in this study reveal that the statistical probability of the coordinated regulation of gene products listed in networks 1 and 2 unrelated to PNF1 stimulation is low (1 in 1056 and 1 in 1042, respectively), but although drug supplementation clearly initiated the TGF-β1-centered networks of genes products, the differential regulation of TGF-β1 itself after 24 h was unremarkable—only 1.3-fold according to cDNA microarray analysis. Use of network analysis tools instead of examination of single genes allowed for identification of key regulatory events in signaling that may not have led to significant changes in mRNA expression of any particular gene product. The statistical power afforded by this method of analysis permitted us to focus on distinct sets of gene products that drug stimulation most highly regulated. Additionally, association between the pro-angiogenic effects of PNF1 and a particular pathway (TGF-β signal transduction) allows for the investigation of related gene products, such as receptors, that were not classified in regulated networks at a particular time point but are integral to signaling events that define the mechanism of action of PNF1.
Our microarray data, compared with that of Lux et al.,33 gives no definitive indication of which (if any) TGF-β signaling mechanism our drug initiates. Moreover, it is known that the effect of TGF-β signaling of endothelial cells varies during microvascular network remodeling, although application of PNF1 after 24 h upregulated the 2 accessory TGF-βRs beta glycan and endoglin. The interplay between the 2 molecules is integral to the balance of pro- and anti-angiogenic signaling in microvascular cells; therefore, future examination of the genetic regulation of these receptors by the drug, including their differential regulation at earlier time points, may provide important insight into the mechanism of action of PNF1.
PNF1 also significantly affected the regulation of signaling molecules that are not directly involved with known TGF-βR signaling. For example, PNF1 may alter expression of extracellular matrix proteases or promotion of monocyte recruitment and associated secretion of pro-angiogenic cytokines.98 Members of the Ras and MAPK pathways;47 including K-ras and MAPK4 and MAPK8, were also identified in significantly regulated networks 1 and 2 (Fig. 2). In addition, PNF1 decreased the expression of CD36, the receptor for anti-angiogenic thrombospondin.58–60
In conclusion, we used network analysis tools to elucidate what global networks of transcriptional regulation mediated by PNF1 produce effects in microvascular endothelial cells that are consistent with angiogenic stimulation.23 However, angiogenesis is a complex process that requires a number of cell types and a myriad of complicated, somewhat poorly understood signals, and therefore the implications of cell responses assessed using standard in vitro culture methods on in vivo angiogenesis are indirect. Nevertheless, our studies show that network tools are particularly effective in elucidating coordinated perturbation of entire networks of gene products by PNF1, which would not be possible through the examination of single-gene regulation alone. Moreover, we identified a well-known signal transduction pathway (TGF-β) regulated by PNF1 after 24 h and examined several smaller groups of genes that may ultimately reveal the drug’s mechanism of action. We believe that delivery of PNF1 is a potentially effective strategy to improve tissue repair outcomes, where revascularization is known to be a limiting step to successful healing.
Sources for support of these studies include the National Institute of Arthritis and Musculoskeletal Disease Grant K01AR052352–01A1 to Dr. Botchwey. Kristen Wieghaus was also supported by the U.S. Department of Education’s Graduate Assistance in Areas of National Need funding. Erwin Gianchandani was supported by the National Institutes of Health (NIH) (GM08715/NIH Biotechnology Training Grant).