Abstract: Growth of new blood vessels (angiogenesis) is essential for embryo development as well as for wound healing and progression of a number of diseases such as cancer, inflammatory conditions, eye diseases, psoriasis, and rheumatoid arthritis in the adult. Current paradigms explain blood vessel growth entirely by sprouting angiogenesis or by vessel splitting through so called intussusceptive angiogenesis. However, these mechanisms are mainly derived from experiments on the developing embryo while less is known about angiogenesis in the adult during, e.g., wound healing, tumor growth, and inflammation. Recently we showed that blood vessel growth in the adult can be induced and directed by mechanical forces that naturally develop during healing or remodeling of tissues. In contrast to sprouting and intussusception, the new biomechanical hypothesis assumes that functional blood vessels are passively translocated which, if found generic, may drastically change the approach for developing anti- and pro-angiogenic therapies in the treatment of a variety of diseases.

Introduction

Biomechanical regulation of tissue growth and remodeling has received limited attention as compared to models based on chemoattraction where cells migrate toward gradients of growth factors. There is however an increasing awareness that biomechanical forces are of fundamental importance during embryo development as well as in postnatal life. Recently we proposed a new mechanism of blood vessel growth in healing tissues, where biomechanical translocation of the preexisting vasculature is responsible for the initial rapid formation of functional vessels in granulation tissue (Kilarski et al., 2009). Vessels elongated, enlarged, and expanded as vascular loops with functional circulation within the growing granulation tissue. The tensional forces that physically guided and mediated neovascularization were generated by proto- and myofibroblast mediated contraction of the provisional matrix and the surrounding tissue. This mechanism suggests that classical models of vessel expansion, like sprouting and intussusception, are not essential for neovascularization.

Mechanisms of vessel growth in the embryo versus postnatal life

Growth of blood vessels has been mainly studied using embryonic systems, like the Zebrafish, avian chorioallantoic membrane, and the mouse neonatal retina. Easy access to blood vessels and the possibility to visualize even capillaries are some of the advantages of these systems. In addition, the developing vascular network follows a reproducible pattern where vessels are formed de novo together with the surrounding tissue in a manner that is spatially and temporally reproducible. In contrast, neovascularization in postnatal life during, e.g., tissue healing or tumor growth, is stimulated by unpredictable liberation of local factors and occurs in already differentiated tissues. Besides the work of Clarks in the early 20th century (Clark, 1918; Clark and Clark, 1940), the unique and complicated nature of postnatal neovascularization has only rarely been the subject of systematical long-term studies (Zawicki et al., 1981; Dudar and Jain, 1983; Kilarski et al., 2009). However, due to striking differences between developmental angiogenesis as compared to neovascularization of, e.g., healing tissues or tumors in adults, it is likely that these processes are regulated by different mechanisms. To give an analogy, the growth of long bones during embryonic development takes place at the epiphyseal growth plate by a finely balanced cycle of cartilage growth, matrix formation, and calcification. However, healing of a bone fracture in an adult occurs through a completely different mechanism by calcification of granulation tissue (callus) that fills the fracture. Next we will discuss major differences between developmental angiogenesis and adult neovascularization during tissue healing.

The vascular system in the embryo arises through a process called vasculogenesis where blood vessels are formed de novo. Newly formed blood vessels sprout and intussusceptively split to form a vascular bed that matures into a functional cardiovascular system. This process is spatially and temporally controlled by genes that regulate the pattern of the developing vascular system. For example, in Zebrafish we know the exact time post fertilization when endothelial cells start to sprout between somites from dorsal aorta to fuse with dorsal longitudinal anastomotic vessels. In addition, growth of vessels during development is paralleled by maturation of surrounding tissues and organs. The situation is the opposite during wound healing in adults where vessels grow against necrotic or injured tissues within the expanding granulation tissue. Randomly organized, relatively large and densely packed vessels in granulation tissue do not follow any predefined genetically controlled pattern and instead their distribution is the result of stochastic processes. The term “granulation” was introduced in 1865 by the surgeon Theodor Billroth and it refers to the emerging vitreous, translucent tissue that is composed of tiny and light red “grains.” Each of these granules contains a vascular tree with loop-shaped vessels at their surface. It is formed after tissue injury, when the inflammatory reaction declines and the healing space is invaded by hypertrophic contractive fibroblasts, smooth muscle actin-positive myofibroblasts, and a network of enlarged and functional vessels. These vessels are transient structures that undergo substantial remodeling and pruning when tissue barriers are restored and the final scar is formed.

Vessel branching (sprouting) and vessel splitting

The term “sprouting” is frequently used in the literature as a synonym of angiogenesis or vessel expansion despite that a scientific definition of “sprouting” has not been clearly formulated. The first description of sprouting dates back to the Greek physician Galen (circa 130-200 A.D.) (Patan, 2000) who compared the growth of the developing embryo “that grows along with the umbilical veins” to a plant. The first observations of sprouting on the chick chorioallantoic membrane were done by Fuelleborn 17 centuries later in 1895. It is important to note that even though the process of sprouting is easy to imagine as vessels can “divide, branch and proliferate into sets of progressively smaller divisions between which the flesh of the liver is deposited” (Galen, De foetuum formatione) sprouting may not necessarily be the simplest mechanism to explain tissue remodeling in postnatal life.

Two different models of vascular branching have been described. First, endothelial sprouting which is a mechanism characteristic for the formation of the primary vascular plexus during embryo development. Endothelial cells sprout from the primitive vascular bed by guided migration of specialized tip cells and proliferation of cells in the sprout stalk (Gerhardt et al., 2003). Despite being considered the dominant mechanism for angiogenesis, it is important to emphasize that capillary sprouting has never been continuously followed in adults from activation of endothelial cells to the formation of specialized vessels. The second model is described as vessel budding, which has been observed during wound healing and tumor growth, where blunt-ended vessels grow from pre-existing vessels towards the avascular zone (Ausprunk and Folkman, 1977). In contrast to endothelial sprouting, buds are functional vessels of different sizes that are constantly perfused with blood and their ultra structural architecture is preserved during growth. How these heterogenous multicellular structures would be translocated into the avascular zone, connect, and later differentiate into functional circulation has not been described.

Longitudinal splitting of vessels has been named intussusceptive angiogenesis (Burri and Djonov, 2002) and was shown to occur during wound healing, tumor vascularization, and growth of the endometrium during the female menstruation cycle. Two distinct mechanisms of vessel splitting have been proposed. Endothelium-dependent intussusception is characterized by intraluminal invaginations caused by endothelial cells that are followed by lumen division and subsequent recruitment of perivascular cells. The same net result is achieved during intussusceptive growth initiated by connective tissue cells that wrap around and contract vessels which leads to narrowing and eventually splitting of vessels (endothelium independent intussusception). This mechanism has been described in tissues over-expressing vascular endothelial growth factor (VEGF) and in vascular tumors, like glioblastoma (Pettersson et al., 2000). Intussusceptive vessel growth is achieved at lower energy costs as compared to vessel branching by sprouting or budding (Burri and Djonov, 2002), as constantly functional vessels split to form the highly vascular granulation tissue. However, it is not known how intussusceptively split vessels would be translocated from the wound periphery to the avascular zone.

Properties of granulation tissue that are not explained by sprouting or intussusception

Although being a broadly accepted mechanism of vessel growth there are few basic studies describing post-developmental sprouting in detail. For example it is not known how sprouts connect to each other, how many sprouts succeed in finding their counterparts (sprouting efficacy), and how and when newly formed capillaries and its originating mother vessels differentiate into functional capillaries, veins, and arteries.

The sprouting model of vessel growth does not explain the existence of a clear interface that is found in wounds between the fibrin clot or necrotic tissue and the invading granulation tissue. If sprouting angiogenesis was the exclusive mechanism then a density gradient of vessels would be expected with narrow sprouts at the leading edge and progressively larger and more frequent vessels towards the mother vessel. In contrast, intussusceptive angiogenesis with formation of constantly functional vessels is compatible with the existence of a sharp demarcation between granulation tissue and surrounding tissue but a mechanism for fast translocation of split vessels has not been described.

Re-wounding (or disruption of healing wounds) is a procedure that has been used for reconstruction of tissue defects when transferring autologous tissue since the days of early plastic surgery in India and Egypt (Cherry and Hughes, 2003). In a study where the rate of granulation tissue formation was quantified it took 5 days to fill an experimental skin wound to 50% with granulation tissue (McClain et al., 1996). The granulation tissue was then removed at day 7 by curettage. Surprisingly the wound was filled to 50% with new granulation tissue already within 24 hours and to 100% within 2 days. This phenomenon can not easily be explained by sprouting as one would expect re-vascularization to take the same time. However, the mechanism that we propose where activated fibroblasts and myofibroblasts contract the wound and physically pull vessels into the wound provides such an explanation. Since the wound and surrounding tissue is already populated with activated fibroblasts and myofibroblasts, at the time of re-wounding, neovascularization gets “jump started” and is achieved during a shorter period of time.

Sprouting angiogenesis postulates that the growing sprout is composed of a tip cell and new endothelial cells that are derived from one or a few proliferating stalk cells located at the origin of the growing sprout (Gerhardt et al., 2003). Proliferation hence becomes a bottleneck since it is a relatively slow process as the whole cell-cycle necessary for cell division takes at least 17 hours in fast proliferating mammalian cells. Sprouting cells can thus hardly explain the phenomenon of fast neovascularization. Proliferation that is not limited to the sprout is more efficient for fast growing tissues as in theory only one cell division by every cell of the vessel is sufficient to double vessel length. In agreement, we found proliferating endothelial and mural cells distributed in different vessel types from capillaries to veins and arteries. Interestingly, previous studies on the importance of cell proliferation for wound contraction and vascularization using x-ray irradiation showed that the sensitive period for irradiation was during the time of most intensive fibroblasts proliferation, 1-2 days after wounding which resulted in reduction of both wound contraction and ingrowth of blood vessels. Irradiation at a later time when vessels begin to enter the wound void (5-7 days after injury) had less effect on wound neovascularization, suggesting that proliferation of fibroblasts and not endothelial cells is the critical factor for the initial tissue neovascularization (Grillo and Potsaid, 1961).

Biomechanical regulation of blood vessel growth

The concept that mechanical factors regulate vessel growth was initially postulated by Thoma in 1893 (reviewed in Clark, 1918) and expressed as “Histomechaniche Principien.” These histomechanic rules are:

1. Increase or decrease in the size of a vessel is regulated by the rate of the blood flow.

2. Increase or decrease in the length of a vessel is governed by the tension exerted on the vessel wall in a longitudinal direction by tissues and organs outside of the vessel.

3. Increase or decrease in the thickness of the vessel wall is dependent upon the blood pressure.

4. New formation of capillaries depends upon increase of pressure in the capillary area.

The first and the third of the rules are now broadly accepted as hemodynamic stimulus for arteriogenesis (Clark and Clark, 1940; Thurston et al., 2000). In our recent work (Kilarski et al., 2009) we validated Thoma’s second postulate and showed that tensional forces generated by proto- and myofibroblast mediated contraction of the provisional matrix and the surrounding tissue, and physically guided and mediated neovascularization. Vessels thus expand as an integral part of the growing granulation tissue in response to mechanical stress. Our results suggest that translocation of the pre-existing vasculature is responsible for the initial rapid formation of functional vessels in granulation tissue. Except for cartilage, every tissue contains a dense capillary network that may be quickly recruited through capillary enlargement and elongation to form the vasculature of the healing wound through biomechanical forces. The quick mobilization of functional neovasculature to the sites where they are immediately needed in order to deliver oxygen, nutrients, and defense elements like complement, antibodies, and leukocytes is hypothetically the major advantage of the vessel translocation mechanism over sprouting angiogenesis. Vessel translocation could also explain certain properties of granulation tissue, like the existence of an interface between granulation tissue and the fibrin clot or necrotic zone as well as the ability to quickly recruit new vessels after re-wounding.

Clinical implications

We have described a mechanism of biomechanical regulation of vascularization that has been termed “looping angiogenesis” (Benest and Augustin, 2009) using two different in vivo wound healing models. Further experiments are clearly required to determine the role of this mechanism of vessel translocation in other conditions. There are however a number of clinical situations where biomechanically driven neovascularization might be important.

The treatment of non-healing wounds has lately been markedly improved by the use of vacuum-assisted closure dressings (VAC), which is a procedure where application of a negative pressure over wounds accelerates the formation of granulation tissue (Plikaitis and Molnar, 2006). Our model of neovascularization might explain how a tensional force generated by the negative pressure of the device acts synergistically with wound contracting myofibroblasts.

Coverage of larger skin defects can pose a problem in plastic and reconstructive surgery. This obstacle can be solved by first expanding the surrounding tissue using tissue expanders. These are silicon implants that are inserted under the skin near the area to be repaired and then gradually filled saline over time, causing the skin to stretch and grow. As in our model where traction or pulling expands the vascular network, tissue expanders can achieve the same effect by pushing the tissue which will create a force that makes the pre-existing vascular network to expand.

Even though we did not investigate whether vessel translocation is utilized by solid tumors to recruit new capillaries, there are similarities between wounds and tumors that may suggest such a mechanism. Tumors are known as “wounds that do not heal” (Dvorak, 1986) and like wounds contain a substantial amount of myofibroblasts and inflammatory cells, which imply that these cells might also have a prominent role in the progression, growth, and spread of cancers. Consequently, fibrin deposits around tumor nodules are found early during tumor growth similar to blood clots formed in wounds. These accumulations of fibrin are invaded by inflammatory cells, macrophages, fibroblasts, and new blood vessels. The provisional fibrin matrix is thereby transformed into a vascularized connective tissue which in wounds is referred to as granulation tissue and in tumors as tumor stroma. A number of clinical trials in cancer patients have been performed using presumed anti-angiogenic substances. These trials have largely been disappointing and have shown no or limited effects in a subset of human tumors and only when combined with chemotherapy. The unfulfilled promises and resistance of tumors to anti-angiogenic therapies might be due to that there are other mechanisms of neovascularization than sprouting and intussusception (Marshall, 2002; Bergers and Hanahan, 2008; Mack, 2009).

References

Ausprunk DH, Folkman J. Migration and proliferation of endothelial cells in preformed and newly formed blood vessels during tumor angiogenesis. Microvasc Res 14(1):53-65, 1977.

Benest AV, Augustin HG. Tension in the vasculature. Nat Med 15(6):608-610, 2009.

Bergers G, Hanahan D. Modes of resistance to anti-angiogenic therapy. Nat Rev Cancer 8(8):592-603, 2008.

Burri PH, Djonov V. Intussusceptive angiogenesis — the alternative to capillary sprouting. Mol Aspects Med 23(6S):S1-27, 2002.

Cherry GW, Hughes M. Increased healing of the disrupted wound — a myth or simply enhanced angiogenesis following wounding? Personal discussions with Tom Hunt from the 1960s to the present. Wound Repair Regen 11(6):401-404, 2003.

Clark ER. Studies on the growth of blood-vessels in the tail of the frog larva — by observation and experiment on the living animal. Am J Anat 23(1):37-88, 1918.

Clark ER, Clark EL. Microscopic observations on the extra-endothelial cells of living mammalian blood vessels. Am. J. Anat. 66(1):1-49, 1940.

Dudar TE, Jain RK. Microcirculatory flow changes during tissue growth. Microvasc Res 25(1):1-21, 1983.

Dvorak HF. Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing. N Engl J Med 315(26):1650-1659, 1986.

Gerhardt H, Golding M, Fruttiger M, Ruhrberg C, Lundkvist A, Abramsson A, Jeltsch M, Mitchell C, Alitalo K, Shima D, Betsholtz C. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol 161(6):1163-1177, 2003.

Grillo HC, Potsaid MS. Studies in wound healing. IV. Retardation of contraction by local x-irradiation, and observations relating to the origin of fibroblasts in repair. Ann Surg 154:741-750, 1961.

Kilarski WW, Samolov B, Petersson L, Kvanta A, Gerwins P. Biomechanical regulation of blood vessel growth during tissue vascularization. Nat Med 15(6):657-664, 2009.

Mack GS. Mixed news for Avastin. Nat Biotechnol 27(6):494, 2009.

Marshall E. Cancer therapy. Setbacks for endostatin. Science 295(5563):2198-2199, 2002.

McClain SA, Simon M, Jones E, Nandi A, Gailit JO, Tonnesen MG, Newman D, Clark RA. Mesenchymal cell activation is the rate-limiting step of granulation tissue induction. Am J Pathol 149(4):1257-1270, 1996.

Patan S. Vasculogenesis and angiogenesis as mechanisms of vascular network formation, growth and remodeling. J Neurooncol 50(1-2):1-15, 2000.

Pettersson A, Nagy JA, Brown LF, Sundberg C, Morgan E, Jungles S, Carter R, Krieger JE, Manseau EJ, Harvey VS, et al. Heterogeneity of the angiogenic response induced in different normal adult tissues by vascular permeability factor/vascular endothelial growth factor. Lab Invest 80(1):99-115, 2000.

Plikaitis CM, Molnar JA. Subatmospheric pressure wound therapy and the vacuum-assisted closure device: basic science and current clinical successes. Expert Rev Med Devices 3(2):175-184, 2006.

Thurston G, Baluk P, McDonald DM. Determinants of endothelial cell phenotype in venules. Microcirculation 7(1):67-80, 2000.

Zawicki DF, Jain RK, Schmid-Schoenbein GW, Chien S. Dynamics of neovascularization in normal tissue. Microvasc Res 21(1):27-47, 1981.

[Discovery Medicine, 8(40):23-27, 2009]