Understanding angiogenesis during aging: opportunities for discoveries and new models

Abstract

Microvascular network growth and remodeling are common denominators for most age-related pathologies. For multiple pathologies (myocardial infarction, stroke, hypertension), promoting microvascular growth, termed angiogenesis, would be beneficial. For others (cancer, retinopathies, rheumatoid arthritis), blocking angiogenesis would be desirable. Most therapeutic strategies, however, are motivated based on studies using adult animal models. This approach is problematic and does not account for the impaired angiogenesis or the inherent network structure changes that might result from age. Considering the common conception that angiogenesis is impaired with age, a need exists to identify the causes and mechanisms of angiogenesis in aged scenarios and for new tools to enable comparison of aged versus adult responses to therapy. The objective of this article is to introduce opportunities for advancing our understanding of angiogenesis in aging through the discovery of novel cell changes along aged microvascular networks and the development of novel ex vivo models.

INTRODUCTION

A common characteristic of aging is impaired angiogenesis, most often defined as the sprouting of new capillaries from existing vessels. Intriguingly, our understanding of how and why angiogenesis is impaired remains underinvestigated. Angiogenesis is a multicellular process that involves the temporal and spatial coordination between endothelial cells, pericytes, growth factors, extracellular matrix, and macrophages (6, 10, 17, 27). Further complicating our understanding are the cellular dynamics that vary at different locations within a microvascular network and vary across different types of tissues.

The complexity of angiogenesis motivates the need to identify the altered cellular and molecular mechanisms associated with aging. The objective of this review will be to emphasize the opportunities for advancing our understanding through the discovery of novel cell changes along aged microvascular networks and the development of novel ex vivo models (Fig. 1). Undoubtedly, therapies aimed at manipulating the microcirculation require the ability to control angiogenesis. However, most therapeutic targets are developed using adult animal models, which does not account for changes resulting from age, for which these pathologies are more frequent. Thus, the development of effective therapeutic strategies, whether blocking angiogenesis for the treatment of cancer or diabetic retinopathy or promoting angiogenesis for myocardial infarction, stroke, or hypertension, requires investigation of responses in aged tissues. As examples of impacting basic science discovery and the development of age-related biomimetic models, we will focus on two findings from our laboratory that support emerging themes from the literature: 1) vascular pericyte coverage is increased along capillaries in aged networks and 2) angiogenesis in cultured aged rat mesenteric tissues can be rescued by media supplementation.

Fig. 1.

Opportunities for understanding angiogenesis during aging. Angiogenesis, characterized by capillary sprouting off existing vessels, involves multiple cell-cell and cell-protein interactions. Main players are endothelial cells, vascular pericytes, growth factors (GF), and the extracellular matrix (ECM). Evidence for impaired angiogenesis during aging motivates the need to identify how these interactions are altered in aged tissues and during age-related pathologies and highlight opportunities for basic science discovery and the development of new biomimetic models for aging research. GF, growth factor.

EVIDENCE FOR ALTERATIONS DURING AGING

Capillary sprouting during angiogenesis is characterized by endothelial cell proliferation, migration, and recruitment. Each of these dynamics has been reported to be impaired with age (19, 21). Impaired angiogenesis associated with aging has been linked to endothelial senescence, and evidence for decreased endothelial cell dynamics in aged scenarios is further supported by alterations in growth factor signaling and the extracellular matrix (15, 34, 35). Regarding growth factors, vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF) expression are decreased, and responses to basic fibroblast growth factor (bFGF) are diminished due to decreased receptor phosphorylation (35). As for the extracellular matrix, the matrix metalloproteinases (MMPs) responsible for breaking down the extracellular matrix have been shown to be increased, whereas MMP inhibitors have been reported to be decreased (14, 35). In aged skin, extracellular matrix, particularly collagen, has been described as more disorganized (14, 60).

Alterations of angiogenic players support the effect of aging on impaired vessel formation and are additionally supported by examples of decreased microvascular vessel density in unstimulated aged tissues (12). Interestingly, however, this causal effect on vessel numbers in tissues is not consistent, as densities have also been reported to be higher and equal to unstimulated aged tissues (11, 24). Although many of these differences can probably be attributed to different tissue types or different animal ages, they do motivate further consideration of whether angiogenesis in aged environments is always impaired. Similar discrepancies have been reported for angiogenic stimulation responses. For example, Rivard et. al. (45) reported a decrease in capillaries in aged mice 28 days post-hindlimb ischemia. In contrast, Westvik et al. (63) reported an increase in the number of capillaries in mice after 7 days. Such apparent differences in results might be explained by the different experimental end points, as Sweat et al. (57) recently demonstrated that conclusions drawn from capillary sprouting response comparisons in the rat mesentery between aged and adult groups were different, depending on whether comparisons were made after 3 or 10 days post-inflammatory stimulation with a mast cell degranulation. Maybe more interesting is that the stimulated aged rat mesenteric microvascular networks displayed “hot spots” or local regions of increased angiogenesis, an observation consistent with increased high-density capillary regions in renal cell carcinoma tumors from older patients compared with adult patients (32).

Consideration of the microvascular density data examples should cause some hesitation in the assumption that angiogenesis is always impaired and emphasize the need for further description of how cellular and molecular alterations are integrated across networks from various tissues. Furthermore, the contrasting interpretations based on vessel count metrics indicate that aged microvascular networks can be stimulated to undergo “normal” levels of angiogenesis. Arthur et al. (1) demonstrated that media supplementation with growth factors can reverse impaired capillary sprouting of aged microvessels. Likewise, angiogenesis is reestablished in vivo by exogenous application and transcriptional initiation of VEGF-A in aged mice (64) and by exercise in models of hindlimb ischemia (9, 29). Such results motivate the determination of what stimulation treatments are sufficient to counterbalance the changes associated with aged scenarios and whether treatment effects are tissue or disease specific.

Questions regarding when angiogenesis is impaired during aging and how it can be stimulated highlight the scientific knowledge gaps and the opportunities for aging research. These opportunities are further appreciated, as we recognize that angiogenesis is just one mode of vessel growth. Aging influences adult vasculogenesis, defined as the de novo formation of vessels as exemplified by the evidence of decreased endothelial progenitor cell (EPC) survival, proliferation, and migration (19, 62). In addition, the number of circulating EPCs is decreased in age-related pathologies (18, 31, 47). Other modes of vessel growth include intussusception, characterized by vessel splitting, and vascular island incorporation, described by the reconnection of disconnected endothelial cell segments to nearby networks (10, 26, 28, 53). To our knowledge, almost nothing is known regarding how these processes are influenced during aging. So, although angiogenesis is commonly considered to be impaired with age, knowledge gaps permeate our lack of understanding regarding how vessels grow or do not grow in aged tissues.

DISCOVERY OF PERICYTE ALTERATIONS IN AGED MICROVASCULAR NETWORKS

The knowledge gaps related to understanding angiogenesis and aging create opportunities for systematically characterizing cellular and molecular players. As an example for impacting basic science discovery, let’s consider vascular pericytes. Pericytes play two roles during angiogenesis: 1) stabilization and 2) promotion of angiogenesis. Vascular pericytes elongate along capillaries and regulate endothelial cell proliferation and migration in adult tissues (16, 27, 39, 52, 55). During capillary sprouting, pericytes are recruited by PDGF/PDGF receptor-β signaling and can physically interact with endothelial cells via integrin binding while playing a role in growth factor presentation (20, 55). Pericytes are also functionally associated with regulating blood vessel permeability and vessel diameter as well as modulation of surrounding extracellular matrix protein deposition (16, 22). Given their importance in regulating endothelial dynamics and capillary function, pericytes have emerged as a therapeutic target. For example, increasing pericyte coverage via angiopoietin-2 (Ang2) inhibition represents an anti-angiogenic tumor therapy, and increasing their coverage in diabetic retinopathy leads to more stable vessels (27, 33, 42).

Pericytes are known to play a critical role in angiogenesis and represent targets for cell-based therapies, yet almost nothing is known about their function during aging. A PubMed search for “pericytes, aging, and angiogenesis” results in 28 papers, accentuating that an opportunity exists to investigate how this cell type changes in aged microvascular networks. Observations by Hughes et al. (22) in retinas from 3- to 6-mo-old and 22- to 23-mo-old rats suggest that pericytes display higher α-smooth muscle actin (α-SMA) expression and reduced desmin length that wraps around endothelial cells rather than extending along them, but with no significant difference in number. Kador et al. (23) suggests that the number of pericytes in retinas harvested from aged dogs is increased. This increase in pericyte coverage is supported by qualitative observations made by Peinado et al. (40) within the optic chiasm harvested from rats and agrees with the discovery from our laboratory that the percentage of α-SMA and neural/glial antigen 2 (NG2)-positive pericyte coverage along capillaries in aged versus adult rat mesenteric microvascular networks was increased (Fig. 2). The increase in capillary coverage in the rat mesenteric networks is supported by RNA sequencing analysis, indicating an increased expression level of PDGF-receptor-β in aged tissues (57). The sparse descriptive results from the literature and our observations identify an understudied cellular change associated with aging and provoke a novel hypothesis that pericytes in aged networks have an increased stabilization phenotype and decreased proangiogenic function. Although the systematic evaluation of pericyte phenotypic marker expression and functionality across different aged tissues is necessary, careful comparison of results across studies should consider the specific ages and phenotypic dynamics over the time course of angiogenic remodeling. More broadly, the example of vascular pericytes emphasizes the opportunity for both descriptive and mechanistic studies to elucidate altered cell-cell interactions involved in angiogenesis during aging. The lack of information about vascular pericytes parallels the similar gaps in knowledge regarding endothelial cell-smooth muscle cell and endothelial cell-macrophage dynamics.

Fig. 2.

Aged microvascular networks display increased pericyte coverage along capillaries. A–D: representative images from 9-mo-old and aged 24-mo-old Fisher 344 rats labeled for platelet endothelial cell adhesion molecule (PECAM) and α-smooth muscle actin (α-SMA) (A and B) or PECAM and neural/glial antigen 2 (NG2) (C and D). PECAM identifies endothelial cells. α-SMA and NG2 identify perivascular cells, including pericytes along capillaries. Arrow indicates a capillary segment void of pericyte wrapping. E and F: quantification of capillary coverage by α-SMA-positive (E) or NG2-positive (F) pericyte labeling. The number of PECAM-positive capillaries with observable pericyte labeling was quantified across entire tissues (α-SMA: n = 13 tissues from rats for each group; NG2: n = 15 tissues from 7 rats for each group). *P < 0.05, ***P < 0.001 (Student’s t-test). Values are shown as means ± SE. Scale bars, 50 µm.

NEW MODEL FOR AGING RESEARCH

The knowledge gaps in our understanding of the cellular dynamic changes with aging are further appreciated by considering the multisystem interrelationships involved in microvascular network growth. Angiogenesis is temporally and spatially related to the other system level remodeling processes, such as lymphangiogenesis and neurogenesis (54, 56, 58). The multicellular/system complexity involved in not only angiogenesis but the growth and remodeling of a real tissue has motivated the emerging area of biomimetic model development.

The investigation of angiogenesis benefits from the use of in vivo models, all of which can be applied to aging research. These include the corneal micropocket assay, the Matrigel/collagen plug assay, the retinal assay, the dorsal skin chamber model, the rat mesentery angiogenesis assay, the chick chorioallantoic membrane (CAM), and zebrafish (4, 17, 30, 38, 41, 61). Each model offers advantages and disadvantages in relevance and how vessels and individual cells can be observed. The most common limitation for in vivo assays is the necessity for advanced imaging methods to reconstruct intact three-dimensional networks. Although the dorsal skin chamber model and zebrafish come close to meeting this challenge, they suffer from complexities associated with their in vivo environment. For example, angiogenic responses in the dorsal skin chamber model are difficult to decouple from gross wound healing, and the zebrafish model’s relevance is limited by nonmammalian cells. The challenges associated with in vivo experimentation further motivate the development of biomimetic in vitro or ex vivo models that recapitulate complexities associated with physiologically relevant scenarios and enable a more controlled probing of mechanistic interactions.

Models that mimic angiogenesis are extremely valuable for elucidating underlying mechanisms and preclinical development of therapies. Historically, in vitro two-dimensional and three-dimensional culture systems have proven crucial for mechanistic investigation of intracellular signaling and cell-cell interactions (1, 25, 59). These “bottom-up” approaches add individual players (i.e., ECM, endothelial cells, pericytes, stem cells) to incrementally increase complexity. Culture or coculture systems, however, are limited in their complexity, and the physiological relevance can be unclear. Recognition of the need to incorporate the multiscale complexity of a real microvascular scenario (i.e., cells, vessels and network) has motivated the application of microfluidic devices (7, 8, 51) and ex vivo tissue explant models such as aortic ring, retina, choroid, and skin explants (13, 37, 46, 48). Specific to aging research, “top-down” tissue culture explant models offer valuable methods for comparing aged vs. adult tissues harvested from pathological animal strains. “Top-down” refers to the goal of maintaining complexity of intact tissues rather than building complexity up player by player. Specific to aging research, the top-down tissue culture explant models offer valuable methods for comparing aged versus adult tissues harvested from pathological animal strains. As an example, Shimada et al. (49) used an aortic ring model to show that the lengths of radial capillary sprouts were decreased in a mutant mouse model of aging vs. wild type mice. Reed et al. (44) reported that aortic rings from 20- to 22-mo-old aged mice exhibited reduced sprouting. Additionally, Shao et al. (48) found that mouse choroid explants have reduced sprouting with an increase in age (P8, P19, P90, P240) with and without retinal pigment epithelium.

Because angiogenesis involves multiple cell types and is related to the growth of other systems like lymphatic networks (16, 56, 58), a need still exists for a model of angiogenesis from intact microvascular networks that more closely reflects an in vivo scenario. No gold standard exists for angiogenic models in aging research. The closest would be the aortic ring model, but in this model sprouting occurs radially from an aortic slice and not a microvascular network where angiogenesis normally occurs. To this end, an opportunity exists to introduce biomimetic platforms that bridge this gap between in vivo and in vitro models, as no model exists that mimics this complexity and enables real-time investigation of cell migration, fate, and function during microvascular network growth during both angiogenesis and lymphangiogenesis.

Recently, our laboratory has introduced the rat mesentery culture model as an ex vivo model to bridge said gap between in vivo and in vitro. It involves harvesting mesenteric tissues from adult rats and culturing them in minimum essential media alone or media supplemented with serum or specific growth factors to stimulate angiogenesis (Fig. 3) (3, 53). A key advantage of our model is its simplicity; the tissue is easy to obtain, easy to culture, self-contained, and does not need to be embedded. In comparison with other models, the rat mesentery culture model offers time-lapse investigation of mechanistic cell-cell interactions at specific locations across blood and lymphatic microvascular networks (2, 58). We have demonstrated that, after culture, mesenteric tissues remain intact and viable. Endothelial cells, smooth muscle cells and pericytes remain present in their in vivo, precultured location. We have also confirmed the presence of lymphatic vessels and nerves (53). In comparison with other tissue explant alternatives, the proposed model is advantageous because it can be used for 1) real-time imaging of the same tissue, 2) quantification of endothelial cell sprouting at specific locations within a microvascular network during growth factor induced angiogenesis, and 3) investigating functional effects of pericytes on endothelial cell sprouting. We have also shown that smooth muscle cells along arterioles maintain their ability to constrict during culture (36) and that lymphatic vessels can be induced to undergo lymphangiogenesis (56). The rat mesenteric model has previously been utilized as medium for cancer cell printing (5, 43). Applied to aging research, media supplementation of pro- or anti-angiogenic factors enables testing on adult and aged tissues (Fig. 3 (3). For example, serum stimulation of adult and aged tissues from Fisher 344 rats displayed similar microvascular density and capillary sprouting levels (Fig. 3). Comparable results can be obtained with bFGF or VEGF media supplementation (data not shown) and suggest that angiogenic responses can be rescued in aged tissues. The finding that angiogenic responses to treatment can be comparable in aged versus adult tissues is supported by the previously provided examples of exogenous growth factor delivery rescuing angiogenesis in aged tissues (1, 29). This supports the use of the rat mesentery culture model as a new tool for mechanistic aging research and applied preclinical therapy evaluation.

Fig. 3.

Application of the rat mesentery culture model for aging research. A: in the rat mesentery culture model, mesenteric tissues can be harvested from adult and aged rats and cultured for comparison of microvascular growth dynamics. Advantages of this tissue culture model include time lapse observation of angiogenesis across intact microvascular networks and the ability to probe cell-cell interactions (2, 54). B: representative images of an aged mesenteric network from an aged 24-mo-old Fisher 344 rat immediately after harvesting and after culturing for 3 days in minimum essential media supplemented with 10% fetal bovine serum (FBS). Platelet endothelial cell adhesion molecule (PECAM) labeling identifies the hierarchy of intact networks, including arterioles (A), venules (V), and capillaries. Angiogenesis in the cultured tissues is supported by the observation of regions with high vascular density (*) and capillary sprouting (arrows). C: data demonstrating the feasibility of comparing angiogenic responses in mesenteric tissues from adult (9 mo) and aged (24 mo) Fisher 344 rats in unstimulated (UN) and stimulated (ST) conditions (n = 8 tissues from 5 rats/group). NS, nos. of vascular segments per area and capillary sprouts per vascular area were not significantly different between adult and aged networks (P > 0.05, 2-way ANOVA, followed by a Bonferroni pairwise comparison test). *Significant difference between unstimulated and stimulated conditions (P < 0.05, 2-way ANOVA followed by a Bonferroni pairwise comparison test). These results suggest that angiogenesis in aged scenarios can be rescued by exogenous delivery of growth cues. Values are shown as means ± SE. Scale bars, 100 µm.

A limitation of the model, potentially influencing long-term cell viability, is the lack of fluid shear stress, which has been implicated in the regulation of angiogenesis (25, 50). Another limitation associated with a lack of perfusion is the absence of circulating cell populations. Nonetheless, we have confirmed that nonperfused cultured tissues remain viable out to ≥7 days (53) and have also observed viable cells along angiogenic endothelial cell segments out to 14 days (data not shown).

The relevance of the rat mesentery for aging research is supported by the association of tissues harvested from 24-mo-old Fisher-355 rats with impaired angiogenesis in response to a chronic inflammatory stimulus (57). Interestingly, despite the impaired response, the vascular densities and sprouting in unstimulated adult and aged tissues are not significantly different (Fig. 3 (57). So although angiogenesis has been shown to be impaired, microvascular networks in the rat mesentery can seemingly undergo “normal” maturation with increased age. This ability is consistent with the capability of aged tissues to undergo angiogenesis in stimulated culture conditions. Thus, although angiogenesis has been shown to be impaired in the aged mesentery tissues, it can also be rescued. This discrepancy emphasizes a major point presented earlier in this article. The evidence for impaired angiogenesis associated with aging can be tissue, stimulation, time, and metric specific. Although future studies will need to evaluate whether mechanisms associated with aged mesenteric tissues are like those in other age-matched tissue scenarios, making sense of age-related angiogenic responses remains difficult and represents a current challenge for aging research. This motivates the need for models that can be used for systematically investigating multiple experiment parameters in a controlled environment.

CONCLUSIONS AND FUTURE OPPORTUNITIES

In multiple age-related pathologies, blocking angiogenesis would be beneficial. In others, promoting angiogenesis would be desirable. The therapeutic need to regulate angiogenesis emphasizes the need to understand better angiogenesis during aging and the development of new aging-relevant experimental models. Although angiogenesis is commonly thought to be impaired during aging, conflicting findings from the literature suggest that the story might be more complicated and raise key questions for aging research. What is the correct age for “old” animals? How can we decouple unhealthy versus healthy aging? Do aged versus adult comparisons temporally depend on experimental end points? Does a decrease in growth factor production necessarily mean that angiogenesis is impaired? Do age-related pathologies affect aging studies? Is impaired angiogenesis tissue specific? Can angiogenesis be rescued in aged tissues? Is impaired angiogenesis causally linked to microvascular dysfunction? For the most part, these questions remain unanswered. Large knowledge gaps exist regarding what angiogenic player dynamics and vessel growth processes are altered during aging, and with these gaps come obvious opportunities to impact aging research. In this article, we used vascular pericytes and the rat mesentery culture model as examples to highlight the opportunities for making basic science discoveries and model development. Considering the importance of angiogenesis, elucidating its relationship with aging represents an intriguing challenge.

GRANTS

The preparation of this publication was supported by the National Institute of Aging under Award No. R01-AG-049821.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

N.A.H., A.D.S.-M., and W.L.M. conceived and designed research; N.A.H., A.D.S.-M., and W.L.M. performed experiments; N.A.H., A.D.S.-M., and W.L.M. analyzed data; N.A.H., A.D.S.-M., and W.L.M. interpreted results of experiments; N.A.H., A.D.S.-M., and W.L.M. prepared figures; N.A.H., A.D.S.-M., and W.L.M. drafted manuscript; N.A.H., A.D.S.-M., and W.L.M. edited and revised manuscript; N.A.H., A.D.S.-M., and W.L.M. approved final version of manuscript.