Fibroblast growth factor 2 is a key determinant of vascular sprouting during bovine luteal angiogenesis
Abstract
Fibroblast growth factor (FGF) 2 and vascular endothelial growth factor (VEGF) A are widely recognized as pivotal regulators of luteal angiogenesis, the process by which new blood vessels form within the corpus luteum. However, despite their acknowledged importance, the precise and coordinated roles these growth factors play in orchestrating this complex biological event have not been fully elucidated. To address this knowledge gap, the current study embarked on a comprehensive investigation of the temporal and spatial progression of endothelial network formation. This was achieved by culturing mixed cell populations, carefully isolated from early bovine corpora lutea, on a fibronectin substrate. These cultures were maintained in the presence of both FGF2 and VEGFA over an extended period, ranging from 6 hours to 9 days, allowing for a detailed observation of the developmental stages of vascularization.
Initial observations, confirmed by von Willebrand factor (VWF) immunohistochemistry, a specific marker for endothelial cells, revealed that endothelial cells initially proliferated and aggregated into distinct cell islands during the first three days of culture (days 0–3). This phase was then succeeded by a transformative period, spanning days 3 to 6, characterized by robust vascular sprouting, where the cell islands began to extend and differentiate into more tubule-like structures. By the conclusion of the 9-day culture period, these structures had developed into extensive and intricately interconnected endothelial networks, closely mimicking the complex vascular architecture found in vivo.
To precisely delineate the roles of FGF2 and VEGFA during these identified developmental windows, mixed populations of luteal cells were subjected to specific pharmacological interventions. Cells were treated with either SU1498, a selective inhibitor of VEGF receptor 2 (VEGFR2), or SU5402, a specific inhibitor of FGF receptor 1 (FGFR1), or left untreated as controls. These treatments were applied during distinct temporal windows: days 0–3 (representing island formation), days 3–6 (representing vascular sprouting and tubule initiation), or days 6–9 (representing network development). Quantitative analysis of the total area occupied by endothelial cells revealed a notable lack of effect with SU1498 treatment across any of the chosen time windows, suggesting that VEGFA signaling through VEGFR2 may not be the primary determinant of overall endothelial cell area in this model.
In stark contrast, treatment with SU5402 elicited a profound and maximal reduction in the total area of endothelial cell networks, particularly when administered during the days 3–6 window. During this critical period of tubule initiation, SU5402 treatment led to an impressive mean reduction of 81% in endothelial cell network area compared to control cultures (P < 0.001). Furthermore, the inhibitory effects of SU5402 during the days 3–6 window extended beyond just the total area. It dramatically reduced the total number of branch points (P < 0.001) and significantly diminished the degree of branching observed per individual endothelial cell island (P < 0.05). Importantly, these reductions in branching and network formation occurred in the absence of any significant changes in the mean area of individual endothelial cell islands, indicating a specific impact on the architectural organization rather than initial cell aggregation. These compelling findings strongly suggest that FGF2 plays an indispensable and critical role as a key determinant of vascular sprouting, thereby highlighting its essential contribution to the physiological development and function of the corpus luteum.
Introduction
The corpus luteum (CL), a transient endocrine gland formed after ovulation, necessitates an exceptionally rich and extensive vascular supply to adequately support its remarkably rapid growth and its crucial steroidogenic function, particularly the production of progesterone. Insufficient progesterone output has been consistently linked to compromised embryo development and an elevated incidence of pregnancy failure in bovine species. Angiogenesis, the intricate biological process involving the formation of new blood vessels from pre-existing vasculature, is a tightly regulated phenomenon. This complex cascade is meticulously controlled by a delicate balance between both pro-angiogenic factors, which promote vessel formation, and anti-angiogenic factors, which inhibit it.
Vascular endothelial growth factor (VEGF) A stands as one of the most potent and extensively studied stimulators of vascular endothelial cell proliferation and migration, fundamental processes in angiogenesis. Its critical role in promoting vascularization has been widely acknowledged across various physiological and pathological contexts. In the bovine corpus luteum, VEGFA is consistently found at high concentrations, and notably, these levels remain relatively stable throughout the various stages of luteal development. This sustained presence underscores its continuous importance. Furthermore, in vitro studies have demonstrated that VEGFA directly stimulates luteal angiogenesis, reinforcing its local pro-angiogenic actions. Compelling evidence from several species, including the cow, indicates that the strategic inhibition of VEGFA activity during the periovulatory period—the critical window around ovulation—leads to a marked suppression of luteal function and a significant reduction in luteal vascularity when studied in vivo. These observations collectively underscore the essential role of VEGFA in the successful establishment and maintenance of luteal vasculature and function.
Fibroblast growth factor (FGF) 2 is another prominent polypeptide growth factor recognized for its diverse range of pro-angiogenic actions. Its influence extends to stimulating the proliferation of endothelial cells, including those specifically derived from bovine corpora lutea. The expression profile of FGF2 in the bovine reproductive system is particularly noteworthy; it is found at exceptionally high levels during the critical follicular-luteal transition, the period when the follicle transforms into the corpus luteum. Moreover, direct experimental manipulation, involving the local neutralization of FGF2 through the injection of FGF2-specific antibodies into the developing bovine CL, has been shown to profoundly alter both luteal growth and its functional capacity. This was evidenced by measurable reductions in both luteal volume and the rate of steroid synthesis, reinforcing the critical involvement of FGF2 in the structural and functional integrity of the corpus luteum.
In recent advancements, our research team successfully developed a novel primary cell culture system specifically derived from early corpora lutea. This innovative system accurately recapitulates the process of luteal angiogenesis in an in vitro setting. A key advantage of this model is that it allows for the co-culture of various cell types naturally present in the corpus luteum, thereby mimicking their complex interactions. This co-culture environment not only facilitates the production of progesterone in response to luteinizing hormone (LH), a physiological characteristic of the CL, but also robustly supports the formation of endothelial cell networks. Crucially, this system demonstrates a high degree of responsiveness to known angiogenic stimuli, making it an ideal platform for studying the intricate mechanisms underlying luteal vascularization. Previous work utilizing this system has already demonstrated that continuous inhibition of either FGF2 or VEGFA signaling significantly impedes the formation of luteal endothelial cell networks. However, despite these foundational insights, a fundamental question has remained unanswered: Are there specific, critical periods and distinct processes during the course of luteal angiogenesis that are acutely dependent on the continuous stimulation by these crucial growth factors? Addressing this specificity is vital for a comprehensive understanding of luteal vascular development.
The overarching aim of the present study was twofold. First, it sought to precisely determine the spatial and temporal patterns governing endothelial cell network growth using this refined in vitro model of bovine luteal angiogenesis. This involved meticulous observation of the developmental stages of vascularization over time. Second, the study aimed to systematically investigate which of these identified stages of growth were most susceptible or sensitive to targeted inhibition of VEGFA and FGF2 signaling. Our working hypothesis posited that the inhibition of FGF2 activity would prove most critical during the very early periods of luteal angiogenesis, particularly during initial cellular organization and sprouting. Conversely, we hypothesized that VEGFA activity would be a continuous requirement, essential throughout all stages of the complex angiogenic process, from initial cell aggregation to the formation of mature, extensive networks.
Results
Development of Endothelial Cell Networks Over Time
The meticulous observation of endothelial cell growth throughout the culture period, from the earliest time point of 6 hours up to 9 days, consistently revealed the presence of endothelial cells, as verified by immunostaining for von Willebrand factor (VWF). This specific marker allowed for precise identification and tracking of the endothelial cell population. Initially, at the nascent stages of culture, endothelial cells were primarily identified as discrete islands, each comprising several cells. Multiple such islands were discernible across each field of view, typically maintaining separation from one another and from other non-endothelial luteal cell types. Over the first day in culture, these endothelial islands demonstrated a progressive increase in size, often expanding to encompass 20 or more endothelial cells by day 2. Within these islands, the endothelial cells exhibited a closely apposed arrangement and presented a characteristic polygonal morphology, indicating a compact cellular organization.
As cell proliferation continued in subsequent days, the endothelial cell islands became increasingly densely packed, and notably, they began to be enveloped by other surrounding luteal cell types, suggesting interactions within the mixed cellular environment. A significant developmental transition was observed with the emergence of tubule-like extensions, which began to sprout outwards from the established endothelial islands. These nascent structures frequently displayed simple branch points, signifying the initial stages of network formation. In the later stages of culture, a notable reduction in the number of isolated endothelial islands was observed. Instead, the majority of endothelial cells had reorganized into extensive and highly complex networks. These intricate structures were characterized by numerous branch points and extensive interconnections between adjacent networks, bearing a striking resemblance to a primitive capillary plexus, a hallmark of early vascular development.
Based on these detailed temporal observations of endothelial cell growth from day 0 to day 9, specific time windows were strategically chosen for the application of growth factor inhibitors. The first defined window, spanning days 0 to 3, was predominantly associated with the initial period of endothelial island formation and proliferation. The subsequent window, encompassing days 3 to 6 of culture, marked a period of significant cellular reorganization and phenotypic change. During this time, the distinct endothelial cell islands began to undergo a transformation, gradually being replaced by the emerging tubule-like structures. This window was specifically characterized by the dynamic processes of vascular sprouting and the critical initiation of endothelial tubules. The final window, from days 6 to 9, represented primarily a period of extensive network development, wherein the nascent endothelial cell networks matured, expanding in size and increasing in complexity to form the intricate vascular architectures observed.
Sensitive Periods During Endothelial Network Formation; Response to Timed Inhibitor Treatments
In the control wells, which received no inhibitory treatment, extensive and well-formed endothelial cell networks were consistently observed after 9 days in culture, serving as a baseline for comparison. Quantitative analysis of the von Willebrand factor (VWF) staining, a marker for endothelial cells, revealed a highly significant effect of treatment with 1 mM SU5402, a specific inhibitor of Fibroblast Growth Factor Receptor 1 (FGFR1), on the total area occupied by endothelial cell networks (P < 0.001).
Specifically, treatment with SU5402 during either the days 0–3 or the days 3–6 windows resulted in a substantial reduction in the total area of endothelial cell networks when compared to control levels (P < 0.05). In contrast, treatment with SU5402 during the later days 6–9 window did not induce a statistically significant alteration in the endothelial cell network area (P > 0.05). The most pronounced inhibitory effect was observed when SU5402 was administered during the days 3–6 window, resulting in a maximal reduction of 81% in the total area of endothelial cell networks. Treatment during the days 0–3 window also caused a notable reduction, approximately 64%, compared to controls. This highlights the critical importance of FGF2 signaling during these earlier stages of network formation.
Further analysis indicated that the observed reduction in the total area of endothelial cell networks following SU5402 treatment was primarily attributable to a significant decrease in the mean number of individual networks formed (P < 0.001). Importantly, this reduction was not a consequence of a decrease in the mean area of each individual network, which remained unaffected by the treatment at any time point (P > 0.05). Complementary measurements also demonstrated that the total length of endothelial cell networks was similarly reduced by SU5402 treatment (P < 0.001), mirroring the pattern of reductions observed in the total area of VWF staining.
In stark contrast to the effects of SU5402, treatment with 2 mM SU1498, a specific inhibitor of VEGF receptor 2 (VEGFR2), aimed at inhibiting VEGFA signaling, did not result in any significant reduction in the total area of VWF staining across any of the time windows examined (P ≥ 0.86). This indicates that, under the conditions of this experiment, VEGFA signaling through VEGFR2 did not play a critical role in determining the overall endothelial cell area or the extent of network formation.
Inhibition of FGF2 Signaling Reduces Endothelial Sprouting
To gain a more precise understanding of how FGF2 signaling influences the dynamic process of vascular sprouting, a specific set of experiments was conducted where cells were treated with SU5402 during the critical days 3–6 window, and then fixed and analyzed on day 6, rather than the extended day 9 culture period. This immediate assessment allowed for a direct evaluation of the acute impact of FGF2 inhibition during the sprouting phase. The results demonstrated that SU5402 treatment during this window led to a substantial reduction of approximately 79% in the total area of endothelial cell staining compared to day 6 controls (P < 0.001). This significant decrease was largely attributed to a 68% reduction in the mean number of endothelial islands observed (P < 0.001). Notably, despite this reduction in the number of islands, the mean area per individual endothelial island remained unaffected by the FGFR1 inhibitor treatment (P ≥ 0.05), suggesting a specific effect on proliferation or aggregation rather than the initial size of the existing islands.
To quantitatively assess the degree of vascular sprouting, a comprehensive branch point analysis was performed. This analysis strikingly revealed a 90% reduction in the total number of endothelial branch points following SU5402 treatment compared to controls (P < 0.001). This profound reduction underscores the critical role of FGF2 in initiating and facilitating the branching process characteristic of vascular sprouting. Furthermore, granular analysis demonstrated that SU5402 treatment also significantly reduced the mean number of branch points measured per individual endothelial cell island, decreasing it from approximately 3.4 to 1.2 (P < 0.05). These combined findings provide compelling evidence that FGF2 signaling is not only essential for the overall formation of endothelial networks but is also a direct and crucial determinant of the intricate process of vascular sprouting and branching during luteal angiogenesis.
Discussion
Our previous research had already established that a broad or “blanket” inhibition of either Fibroblast Growth Factor (FGF) 2 or Vascular Endothelial Growth Factor (VEGF) A signaling resulted in profound and dramatic reductions in the formation of bovine luteal endothelial cell networks. Building upon these foundational findings, the current study aimed to delve deeper into the intricate regulatory mechanisms of luteal angiogenesis. We strategically employed our established in vitro model of angiogenesis, applying specific inhibitors of FGF2 and VEGFA activity at precisely defined time points throughout the cell culture period. This targeted approach allowed us to identify the specific developmental periods during which luteal angiogenesis exhibits heightened sensitivity or dependence on these crucial growth factors.
Our findings revealed that treatment with SU5402, a highly selective inhibitor of FGF Receptor 1 (FGFR1), caused a maximal reduction in the total area of von Willebrand factor (VWF) immunostaining (an 81% reduction) when applied during the days 3–6 window of culture. A substantial, though lesser, reduction (64%) was also observed when SU5402 was administered during the days 0–3 period. Importantly, in contrast, SU5402 had no discernible effect on VWF staining when applied during the later days 6–9 of culture, relative to control groups. These results strongly indicate that the days 3–6 window represents a particularly critical period of sensitivity to FGF2 regulation within this in vitro culture system. This observation aligns well with our previous in vivo findings, which demonstrated that FGF2 exhibits dynamic expression patterns around the follicular–luteal transition, with protein concentrations peaking in the collapsed follicle immediately following ovulation. Therefore, the present study provides compelling additional evidence supporting an indispensable role for FGF2 in the very early and critical stages of bovine luteal angiogenesis.
SU5402 is recognized as a narrow-range tyrosine kinase inhibitor that has been extensively utilized to specifically impede FGF2 signaling. The FGFR gene family comprises four distinct members (FGFR1–FGFR4), and complex alternative splicing events lead to the generation of numerous receptor isoforms, each potentially exhibiting unique binding characteristics. While the detailed interaction of SU5402 with FGFR1 has been thoroughly investigated and characterized, the inherent similarities in the amino acid sequence and structural motifs of the catalytic domains across various FGFRs suggest that SU5402 may also exert inhibitory effects on the phosphorylation of additional FGFRs. Indeed, previous studies have provided evidence that SU5402 can inhibit the phosphorylation of FGFR2 (specifically variant B) in bladder carcinoma cells and can impede the growth of multiple myeloma cells that harbor activating mutations in FGFR3.
It is also important to consider that different FGF ligands exhibit varying affinities for the diverse FGFR isoforms. For instance, FGF2 preferentially activates FGFR1C, FGFR3C, and FGFR4, and to a lesser extent, FGFR1B and FGFR2C. In the context of the bovine corpus luteum, it has been previously shown that levels of FGFR3C and FGFR4 expression are negligible. Consequently, the most probable luteal targets for FGF action are likely to be variants of FGFR1 and FGFR2. While FGFR2C mRNA expression has not been observed to vary significantly during bovine luteal development in vivo, it remains unclear whether or to what extent FGF2 might be signaling via FGFR2C within our specific in vitro culture system, or what relative contribution such signaling might make to the overall regulation of angiogenesis in this model. These considerations highlight the complexity of precisely dissecting the full spectrum of FGFR inhibition by SU5402 in a complex biological system.
The carefully chosen time windows for inhibitor treatment in this study were designed to precisely correspond with distinct periods of endothelial cell growth and developmental progression in culture. The initial period, spanning days 0–3, is primarily characterized by the proliferation and aggregation of endothelial cells into dense island-like formations. The subsequent window, between days 3 and 6, represents a phase of rapid and significant cellular reorganization. This is the crucial stage marked by robust vascular sprouting and the initiation of tubule-like growth, signifying the transition towards a more complex vascular structure. The final window, from days 6 to 9, appears to be predominantly a period of further endothelial cell network development, where the nascent networks mature, expand, and increase in their structural complexity. The observation that the maximal inhibitory response to FGF2 occurred during the days 3–6 window, precisely correlating with the period when endothelial tubules are first observed, strongly suggests a critical and direct role for FGF2 in the initiation of endothelial cell sprouting.
Further reinforcing the critical role of FGF2 in endothelial cell sprouting, specific changes in the growth pattern were observed on day 6 after treatment with SU5402 during the days 3–6 period. Inhibition of FGFR1 resulted in a dramatic 90% reduction in the total number of endothelial branch points. This effect was a result of both a decrease in the overall number of endothelial islands and, more specifically, a profound reduction in the mean number of vascular sprouts observed per individual island. Importantly, this occurred without any significant reduction in the mean area of each island, indicating that FGF2 specifically influences the branching morphology and number of sprouts rather than the initial aggregation or size of the cellular units.
The intricate process of vascular network growth necessitates that a subset of cells at the leading edge of an emerging vascular sprout acquire a specialized phenotype, distinguishing them as “tip cells” from the remaining “stalk cells.” These endothelial tip cells then undergo directional growth, propelled by dynamic filopodia extensions and precisely guided by various angiogenic stimuli. As motile tip cells advance, they eventually establish connections with other adjacent vessels, leading to fusion and the formation of a continuous lumen. Concurrently, endothelial cell proliferation is understood to primarily occur within the stalk cell population, providing the cellular mass for vessel elongation. While the specific time periods observed in our in vitro culture system may not directly correlate with identical chronological time periods of luteal development in vivo, these fundamental early steps in angiogenesis are demonstrably reproduced in our luteal model, ultimately culminating in the establishment of complex endothelial cell networks that resemble the physiological vasculature.
Despite the widely acknowledged pro-angiogenic properties of FGF2, there is remarkably limited specific information detailing its precise roles in determining the process of sprouting during angiogenesis. Nevertheless, existing research provides some contextual support. For instance, exposure to FGF2 has been shown to increase both the density and branching complexity of the vascular tree in the chorioallantoic membrane model. Furthermore, transgenic mice engineered to express a dominant-negative form of FGFR1 exhibited a significantly diminished retinal vasculature characterized by reduced branching, underscoring the in vivo relevance of FGFR1 signaling in vascular development. Based on the multifaceted properties of FGF2, its role in sprouting likely encompasses the promotion of endothelial cell proliferation, enhancement of cell motility, or the active extension of filopodia. Clearly, these potential mechanisms warrant further comprehensive investigation to fully unravel the specific molecular and cellular events mediated by FGF2 during vascular sprouting.
Vascular sprouting itself is an exceptionally complex and highly coordinated biological process, involving a multitude of molecular signals and dynamic changes in cellular phenotype. Strong evidence exists for the crucial involvement of the Notch signaling pathway in endothelial sprouting, primarily through its decisive role in determining endothelial cell phenotype and, consequently, the number of endothelial tip cells that emerge. The Notch signaling pathway comprises various Notch receptors (Notch 1–4) that interact with specific Notch ligands, including Jagged (JAG) 1 and 2, and Delta-like (DLL) 1, 3, and 4. All these ligands are transmembrane proteins, implying that Notch signaling is fundamentally mediated by direct cell–cell interactions. It is widely recognized that endothelial tip cells prominently express DLL4, which then engages with Notch receptors on adjacent endothelial cells. This interaction is critical, as it converts these neighboring cells into “stalk cells,” suppressing their tip cell characteristics and promoting their proliferative and elongating functions. Conversely, JAG1 is robustly expressed in stalk cells and acts to antagonize DLL4-mediated Notch signaling, thereby intricately controlling the number of tip cells and, consequently, the number of new sprouts. Thus, the dynamic switching between endothelial tip and stalk cell phenotypes is likely transient and reversible, critically depending on the delicate balance between JAG1 and DLL4 expression. While VEGFA has been demonstrated to promote the formation of endothelial tip cells by inducing DLL4 expression, FGF2 is more likely to exert its effects by up-regulating Notch4 expression and subtly modulating the action of Jagged. Indeed, the FGF2-induced endothelial cell sprouting has been shown to be potentiated by treatment with an antisense oligomer targeting Jagged in bovine microvascular endothelial cells, further supporting a complex interplay.
It is also highly probable that FGF2 does not act in isolation but engages in intricate interactions with other pivotal regulatory factors in the angiogenic cascade. Members of the transforming growth factor beta (TGFB) superfamily, for instance, are known to exert significant angiogenic effects, including a synergistic enhancement of endothelial branching when combined with FGF stimulation. Furthermore, other signaling pathways that provide crucial guidance cues for vascular development, such as the Angiopoietin/Tie system, the Slit/Roundabout (Robo) system, and the WNT-signaling pathways, are also likely to integrate their signals with those emanating from FGF pathways, creating a highly interconnected regulatory network that orchestrates precise vessel formation.
In the present study, the maximal reduction in endothelial cell networks in response to FGF2 inhibition was most pronounced between days 3 and 6 of culture. This contrasts somewhat with in vivo data, which indicate that luteal FGF2 content is approximately sixfold lower on day 5 compared to day 2 of the estrous cycle. Any observed differences in timing between in vitro and in vivo models are likely attributable to several factors. These include a potential “lag time” as cells adapt and establish themselves in the culture environment, the relatively broad temporal windows chosen for both inhibitor treatment and timed corpus luteum collection, and expected small variations in the starting biological material, which could subtly shift the exact timing of events in culture relative to the defined treatment windows.
It is important to note that no single inhibitor treatment window was capable of completely abolishing endothelial cell growth, with the maximal reduction observed being 81%. Indeed, none of the individual treatment windows achieved the profound 95% reduction in endothelial cell networks previously observed when SU5402 inhibitor treatment was applied continuously throughout the entire culture period at the same concentration. This suggests that the previously observed near-total inhibition was primarily a cumulative effect resulting from continuous inhibition from days 0 to 6, with relatively less influence exerted during the final days of culture when networks are more established.
The significant reductions in the total area of VWF immunostaining observed after SU5402 treatment primarily reflect a decrease in the absolute number of endothelial cell networks present on day 9, rather than a reduction in the average area of each individual network. This outcome is unlikely to be the result of generalized cellular toxicity, as substantial cell density and evident cell proliferation were still apparent by day 9 in the cultures. Furthermore, it is noteworthy that progesterone production, a key physiological function of the corpus luteum, consistently increases during the culture period and was previously unaffected by SU5402 treatment, strongly suggesting the presence of a healthy and functional steroidogenic cell population. Moreover, there does not appear to be endothelial-specific toxicity, given that treatment with SU5402 during days 6–9 resulted in no reduction in the total area of VWF staining compared to controls. Since significant reductions in the total area of endothelial cell networks were observed for both the days 3–6 and days 0–3 treatment windows, it is also improbable that these reductions are simply a consequence of endothelial cells failing to establish and/or recover from the initial tissue disruption and subsequent inability to proliferate. Instead, a reduction in the number of networks could plausibly be due to an increase in physiological cell death or apoptosis, resulting from a period of deprivation of essential growth factor support. FGF2 has been well-documented to act as a crucial survival factor, providing protection to endothelial cells from programmed cell death in vitro. Therefore, the inhibition of this vital survival signal might logically lead to marked endothelial cell loss. Indeed, endothelial cells may exhibit particular sensitivity to the withdrawal of growth factor support during the critical phase of tubule initiation, making them vulnerable to apoptosis. The necessity for positive survival signals in luteal development has been previously proposed, and apoptotic endothelial cells have been histologically localized in primate corpora lutea following the suppression of VEGFA. Furthermore, the apparent FGF2-independence that develops over time in culture, as observed in this study, might reflect an increasing resistance to endothelial cell apoptosis that typically occurs as nascent vessel-like structures mature and recruit crucial perivascular support cells. While apoptosis is an essential component of structural luteolysis, the programmed regression of the corpus luteum, and is initially evident in luteal endothelial cells, it is interesting to consider that programmed cell death may not solely be involved in tissue regression. Instead, it might also contribute to the precise formation of vascular-like endothelial cell networks in vitro, with evidence suggesting that blocking apoptosis can lead to disrupted vascular growth in vivo. Therefore, the critical balance of pro-apoptotic and anti-apoptotic signals that must be precisely maintained for appropriate luteal vessel development warrants extensive further investigation.
An important observation from this study was that the extent of the reduction in the total area of endothelial cell networks after SU5402 treatment during days 3–6 remained remarkably similar, approximately 80%, regardless of whether the cells were collected and analyzed on day 6 or day 9 of culture, when compared to time-matched controls. While the time course experiment suggested that networks typically continue to develop and expand between days 6 and 9, and thus an expected difference in network area between controls at these time points, the consistent effect of SU5402 indicates a sustained inhibitory impact. This suggests that there is neither a period of rapid “catch-up” or recovery for endothelial cells between days 6 and 9 following treatment, nor is there an additional detrimental effect on endothelial cell growth that persists beyond the defined treatment period.
In contrast to the significant effects of FGF2 inhibition, treatment with SU1498, specifically designed to inhibit VEGFR2 activity, did not result in any statistically significant effect on endothelial cell network growth compared to control groups at any of the investigated time points. This finding stands in notable contrast to our previous study, where a continuous or “blanket” treatment with SU1498 at the same concentration throughout the entire 0–9 day culture period resulted in a substantial 60% reduction in the total area of VWF immunostaining. This discrepancy suggests that the continuous presence of VEGFA signaling inhibition is necessary to elicit a significant effect on overall network formation, and that short, timed inhibition windows may not be sufficient to reveal its full impact.
In vivo studies also provide a complex picture regarding VEGFA’s role. For instance, direct delivery of anti-VEGF antibodies into the bovine corpus luteum for 7 days post-ovulation has been reported to reduce luteal volume by approximately 50% compared to controls. Similarly, in the marmoset, a 10-day treatment with anti-VEGF antibodies also led to a reduction in endothelial cell area by about 50%. These in vivo observations suggest a broader, more constitutive role for VEGFA in luteal development. However, the current study, with its timed inhibition approach, suggests that there might not be one single, discrete period of heightened sensitivity to VEGFA inhibition. Instead, it is more plausible that VEGFA plays a more subtle, constitutive role throughout angiogenesis, or that a cumulative, minor response might occur across each window, which individually is not statistically significant. This aligns with our previous observations of limited temporal variations in luteal VEGFA concentrations throughout the bovine estrous cycle. An important role for VEGFA in luteal angiogenesis is undeniably suggested both by its expression within the corpus luteum and by the markedly compromised luteal form and function observed in response to VEGFA suppression in various studies. While VEGF antagonist treatment has been reported to cause near-total suppression of luteal angiogenesis in the rat, it is important to consider that this might also reflect broader effects on both follicular and luteal development, particularly if the antagonist was administered prior to ovarian stimulation. However, our study observed no effect of VEGF inhibition on endothelial cell area in any of the distinct treatment windows. This aligns with observations in the marmoset, where short-term treatment with anti-VEGF antibodies for 3 days also had no effect on endothelial cell area. An alternative explanation for the lack of effect of short-term SU1498 treatment could be that the endothelial cells possess an inherent capacity to subsequently recover from any transient inhibitory effects imposed during the days 0–3 or 3–6 windows. If such recovery occurs, then no discernible difference in VWF staining would be observed by day 9, effectively masking any short-term impact.
It is also important to remember that VEGF ligands interact with three distinct receptors: VEGFR1, VEGFR2, and VEGFR3. While VEGFR1 is crucial for embryonic blood vessel development, its kinase activity is relatively weak when compared to VEGFR2, which is generally considered the primary VEGFR mediating angiogenic signaling pathways. Indeed, in bovine microvascular endothelial cells, VEGFR2, and not VEGFR1, has been identified as the key receptor responsible for endothelial tube formation. Although mRNA encoding VEGFR3 has been detected in the ovary, including the primate corpus luteum where expression levels peak around the midluteal phase, signaling via VEGFR3 is primarily considered important for the lymphatic vasculature and lymphangiogenesis, rather than blood vessel formation.
In conclusion, this comprehensive study has successfully identified specific and critical windows of sensitivity to FGF2 inhibition during luteal angiogenesis. Our findings robustly support the initial hypothesis that the inhibition of FGF2 activity is most crucial during the early phases of luteal angiogenesis, particularly during the pivotal processes of endothelial cell sprouting and the initiation of tubule formation. This positions FGF2 as a dynamic and acutely important modulator of angiogenesis in the bovine corpus luteum. In contrast, while VEGFA undoubtedly plays an essential role in luteal vascularization, our timed inhibition experiments suggest that its contribution is more constitutive and prolonged throughout the luteal lifespan, rather than being confined to specific, highly sensitive temporal windows.
Materials and Methods
Materials
All materials and reagents utilized throughout this study were procured from Fisher Scientific, located in Loughborough, UK, unless specifically stated otherwise, ensuring a standardized source for common laboratory supplies.
Tissue Collection
Bovine ovaries, serving as the biological source material, were meticulously collected from a local slaughterhouse. Upon collection, they were immediately immersed and transported in warmed 1X Phosphate Buffered Saline (PBS) to the laboratory, a crucial step to maintain tissue viability and integrity. Early corpora lutea, specifically those corresponding to days 1–4 of development, were carefully selected based on well-established morphological criteria that have been previously described and validated. Once identified, any surrounding ovarian tissue that was not part of the corpus luteum was carefully trimmed away to isolate the target tissue precisely.
Luteal Angiogenesis Culture System
The preparation of mixed cell populations, encompassing essential luteal cell types such as steroidogenic cells, endothelial cells, pericytes, and fibroblasts, was performed following a previously described and validated protocol. In brief, the freshly collected corpora lutea were subjected to enzymatic dispersion using a combination of collagenase and DNase digestion. This enzymatic treatment effectively dissociates the tissue into a single-cell suspension. Following digestion, the dispersed cells were thoroughly washed to remove residual enzymes and debris. Subsequently, the cells were cultured in a specialized endothelial basal medium (EBM-2), obtained from Lonza (Verviers, Belgium), which is specifically formulated to support endothelial cell growth. This basal medium was further enriched with a comprehensive array of growth factors and supplements, including human epidermal growth factor, LR3 insulin-like growth factor 1, hydrocortisone, ascorbic acid, gentamicin, and heparin, all added strictly according to the manufacturer’s instructions. Additionally, the culture medium was supplemented with 5 ng/ml Luteinizing Hormone (LH), a crucial hormone for luteal function (specifically, AFP11743B, with a biopotency of 1.06x oLH NIDDK-1-2, kindly provided by Dr. A.F. Parlow, NIDDK, Torrance, CA, USA). To ensure a sterile environment and optimal cellular health, the medium also contained 100 units/ml penicillin and 10 mg/ml streptomycin, acting as broad-spectrum antibiotics. Further nutritional support was provided by 10 mg/ml insulin, 5.5 mg/ml transferrin, and 5 ng/ml selenium, all procured from Sigma–Aldrich Co. Ltd. Crucially, the medium was supplemented with 2% (v/v) fetal bovine serum (Lonza) and, to support angiogenesis, 1 ng/ml VEGFA and 1 ng/ml FGF2 (Lonza) were added to all wells. Cell number and viability were meticulously assessed using the trypan blue exclusion method, ensuring that only healthy and viable cells were used for plating. Cells were then plated at a density of 2 x 10^5 viable cells per well into 12-well plates. For optimal cell attachment and morphology, these wells contained fibronectin-coated cover slips, prepared as previously detailed. The cell cultures were maintained in a humidified incubator at a physiological temperature of 39°C, under an atmosphere of 5% CO2 and 95% air, conditions optimized for bovine cell growth. The culture medium was refreshed at regular intervals, specifically after 1, 3, 5, and 7 days of culture, to ensure continuous nutrient supply and waste removal. At the end of the culture period, cells were fixed in a pre-chilled solution of acetone:methanol (1:1) at 4°C for 5 minutes, preparing them for subsequent immunostaining procedures.
Time Course of Endothelial Cell Growth in Culture
To comprehensively investigate the dynamic progression of endothelial cell network formation over time, luteal cells were prepared and plated following the established protocol described above. For this specific experiment, a total of 5 separate cultures were established, with each culture derived from a single corpus luteum, ensuring biological replicates. From each culture, triplicate cover slips were systematically harvested at predetermined time points: after 6, 12, and 18 hours (designated as day 1) in culture, and subsequently on every successive day until day 9. Upon removal from the culture plate, the cells on these cover slips were immediately fixed in the chilled acetone:methanol (1:1) solution at 4°C for 5 minutes. Following fixation, endothelial cells were specifically immunolocalized using von Willebrand factor (VWF) as a marker, as detailed in a subsequent section. The patterns of endothelial cell growth and network development were then meticulously assessed and documented at each time point.
Effect of Inhibitor Treatments on Endothelial Cell Network Formation
To precisely investigate the impact of targeted FGF2 or VEGFA inhibitor treatments during distinct phases of endothelial cell network formation, luteal cells were prepared and plated using the standardized methodology. Similar to the time course experiment, 5 independent cultures were established, with each originating from a unique corpus luteum, providing sufficient biological replication. Cells were maintained in the standard supplemented EBM-2 medium, which included 1 ng/ml VEGFA and 1 ng/ml FGF2, serving as the control condition. Experimental groups additionally received either 1 mM SU5402, a specific FGFR1 inhibitor, or 2 mM SU1498, a specific VEGFR2 inhibitor. The strategic selection of treatment windows was based directly on the insights gained from the preliminary time course experiment, which identified key stages of endothelial cell growth. Cells were treated in triplicate with the respective inhibitors during one of three specific windows: days 0 to 3 (corresponding to the period of initial island formation), days 3–6 (representing the critical phase of vascular sprouting and tubule initiation), or days 6–9 (encompassing the phase of extensive network development). Following their respective treatment periods, all cover slips were fixed on day 9 in acetone:methanol (1:1) at 4°C for 5 minutes, in preparation for subsequent immunostaining.
SU5402 exerts its inhibitory action by directly interacting with the catalytic domain of FGFR1, thereby suppressing its kinase activity and downstream signaling pathways. This compound has been widely employed as a specific FGF2-antagonist and an effective inhibitor of FGF2-mediated angiogenesis in numerous studies. SU1498, on the other hand, specifically targets and inhibits VEGFR2 phosphorylation, leading to a reduction in endothelial cell proliferation in vitro and the suppression of angiogenesis in vivo. Both SU5402 and SU1498 were used at concentrations (1 mM and 2 mM, respectively) that have been previously demonstrated to effectively inhibit angiogenesis within this specific culture system, in experiments performed under identical conditions. The inhibitors, obtained from Calbiochem (Merck Chemicals Ltd.), were initially dissolved in DMSO and then subsequently diluted into the culture media to achieve a final DMSO concentration of 0.1% (v/v). A corresponding 0.1% DMSO was also added to all control wells to account for any potential solvent effects.
To further refine the investigation into the specific effects of SU5402 treatment during the critical days 3–6 window, an additional set of experiments was performed. Luteal cells were prepared, plated, and treated with 1 mM SU5402 as described above (n=5 cultures). However, for this particular assessment, the cells were collected and fixed immediately following the treatment period on day 6, rather than extending the culture to day 9. This allowed for a direct evaluation of the acute impact of FGF2 inhibition during the sprouting phase, before subsequent maturation. Standard VWF immunostaining and detailed image analysis were then performed on these day 6 fixed cultures.
Immunocytochemistry for VWF
Endothelial cells were precisely identified and localized through immunocytochemistry utilizing von Willebrand factor (VWF) as a highly specific and reliable endothelial cell marker. This methodology strictly adhered to previously validated and extensively described protocols, ensuring consistency and reproducibility of results. The procedure commenced with blocking endogenous peroxidase activity, a common source of non-specific staining, by incubating the cover slips with 3% (v/v) hydrogen peroxide in methanol. Following this, non-specific protein binding, which can lead to background noise, was effectively minimized by a blocking step involving 20% (v/v) normal goat serum, procured from Sigma–Aldrich Co. Ltd. Subsequently, the cover slips were incubated overnight in a humidified chamber with the primary antibody, a rabbit anti-human VWF antibody at a concentration of 4 mg/ml, obtained from Dako UK Ltd., Ely, UK. This crucial step allowed the primary antibody to bind specifically to the VWF protein expressed by endothelial cells. Following incubation with the primary antibody, detection of the bound VWF was achieved using a biotinylated goat anti-rabbit secondary antibody, which specifically recognizes the rabbit primary antibody. This was then followed by the Vectastain ABC method, supplied by Vector Laboratories Ltd., Peterborough, UK, a highly sensitive avidin-biotin complex amplification system. The enzymatic reaction was visualized using diaminobenzidine (DAB), also from Vector Laboratories Ltd., as the peroxidase substrate. This reaction produces a visible brown precipitate at the site of VWF localization, allowing for microscopic visualization and subsequent quantification of endothelial cells.
Image Analysis
Positive staining for von Willebrand factor (VWF), indicative of endothelial cells, was quantitatively assessed using methods previously established and described, with minor methodological refinements to enhance precision. Tissue sections were systematically viewed under 5x magnification utilizing a Leica DM 4000B microscope, manufactured by Leica Microsystems Ltd., Milton Keynes, UK. Digital images were captured with high fidelity using a QImaging micropublisher 5.0 RTV color camera, supplied by QImaging (UK) Ltd., St Helens, UK. All captured images were subsequently transferred to and analyzed using Image Pro-Plus 6.0 software, developed by Media Cybernetics, Wokingham, UK, a robust platform for image processing and quantification.
For each treatment condition within each cell culture, two wells were randomly selected for analysis, ensuring unbiased representation. From each selected well, images were acquired from 20 distinct fields of view. These individual fields were then meticulously tiled together to construct a larger, composite image, which was subsequently subjected to comprehensive digital processing. Within these composite images, areas exhibiting positive VWF staining and displaying a distinct tubule-like appearance, specifically those exceeding an area of 150 square micrometers, were highlighted. These selected areas underwent digital smoothing, and any internal holes or discontinuities in the staining were algorithmically removed to ensure accurate representation of the vascular structures. The sophisticated image analysis capabilities of the software then precisely determined several key quantitative parameters: the total cumulative area of VWF staining across the composite image, the discrete number of individual endothelial cell networks formed, and the specific area and total length of each identified individual network.
In addition to these general network parameters, specialized image analysis was performed to meticulously determine potential changes in the pattern of endothelial cell growth observed specifically on day 6 after treatment with SU5402. This was achieved by employing an automated measure of branching, specifically the “Branch/End” feature available within Image Pro-Plus 6, a method that has been previously validated for quantifying vascular branching morphology. This advanced analysis allowed for a more granular understanding of the impact of FGF2 inhibition on the intricate architecture of the developing vascular network.
Statistical Analysis
All collected data underwent rigorous statistical analysis to determine the significance of observed differences between experimental groups. A randomized block one-way ANOVA (Analysis of Variance) was employed, a statistical technique well-suited for experiments where variation within blocks (in this case, individual cell cultures) is expected. Following the ANOVA, Bonferroni’s multiple comparison test was used for post-hoc analysis. This test is a conservative method that adjusts for multiple comparisons, thereby reducing the likelihood of Type I errors (false positives) when comparing several group means.
These statistical methods were applied to compare various quantitative parameters: the total area and length of endothelial cell networks, the mean number of networks, and the mean area per network between different experimental groups on day 9 of culture. Furthermore, for the specific analysis of day 6 data, comparisons were made for the total area of endothelial cell islands, the mean number of islands, the mean area per island, the total number of branch points, and the mean number of branch points per island. In all analyses, the individual cell cultures served as the blocking variable, effectively accounting for inherent biological variability between different primary cell preparations, while the specific treatment applied was considered the primary factor influencing the response. Prior to statistical analysis, all data sets were meticulously examined for adherence to assumptions of normality and homogeneity of variance. Where necessary, variables were subjected to logarithmic transformation to meet these parametric assumptions, ensuring the validity of the statistical tests. This data transformation was performed using Genstat 12 software, developed by VSN International, Hemel Hempstead, UK.
Declaration of Interest
The authors explicitly declare that there are no financial, professional, or personal conflicts of interest that could potentially be perceived as prejudicing or unduly influencing the impartiality of the research reported within this manuscript. All authors have maintained objectivity throughout the study’s design, execution, analysis, and reporting.
Funding
This research endeavor received financial support from a grant awarded by the Biotechnology and Biological Sciences Research Council (BBSRC), specifically grant number BB/F002998. Additional funding and support were generously provided by Pfizer. These funding sources were crucial for the successful execution of the experimental work.
Acknowledgements
The authors wish to express their sincere gratitude and appreciation to all staff members at the University of Nottingham who provided invaluable assistance with both the collection of samples and the subsequent analysis procedures throughout the duration of this study. Their contributions were instrumental in the successful completion of this research.