Role of Hemodynamics in Pulmonary Vascular Remodeling
Implications for Primary Pulmonary Hypertension

MITCHELL D. BOTNEY

Respiratory and Critical Care Division, Washington University Medical Center, St. Louis, Missouri

	INTRODUCTION

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

Primary pulmonary hypertension (PPH) is a disease characterized clinically by the progressive increase in pulmonary artery impedance, which ultimately produces right ventricular failure and death. At the time most patients with PPH present to clinical attention, pathologic lesions include pulmonary artery medial hypertrophy, adventitial thickening, and neointimal lesions. Neointimal lesions are new vascular structures composed of smooth muscle cells (SMC) and extracellular matrix located on the lumenal side of the inner elastic lamina or, if the vessel is too small to have an inner elastic lamina, between the endothelium and the medial layer of the vessel. The terms "neointima" and "neointimal lesions" are used interchangeably in this review. Nearly all active remodeling (as defined by extracellular matrix gene expression) in patients with severe pulmonary hypertension occurs in the neointima (1). This observation underscores the importance of understanding how pulmonary vascular neointimal lesions develop but the pathogenesis and natural history of neointimal lesion formation in PPH are poorly understood. On the basis of animal studies, and by analogy to systemic arterial disease, it seems likely that endothelial injury is required (6). Whether injury occurs secondary to infection (7), ingestion of toxins (8), or activation of autoimmunity (9) remains speculative.

Although vascular injury appears necessary to initiate remodeling, other factors besides injury are likely required before pulmonary vascular neointimal lesions develop since the pulmonary artery histology in traditional animal models, in which remodeling is initiated by monocrotaline or hypoxic injury, does not develop lesions that resemble the neointimal lesions of PPH. Hemodynamic forces and injury-induced biochemical mediators are important components in mechanisms of systemic vascular remodeling, including atherosclerosis, restenosis after angioplasty, and neointimal lesions following vascular bypass surgery. New studies suggest hemodynamic factors may be as important in the pulmonary vasculature as they are in the systemic vasculature.

	WHY IS THE PULMONARY ARTERY NORMALLY SPARED FROM NEOINTIMAL HYPERPLASIA OR ATHEROSCLEROTIC DISEASE?

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

Systemic vessels develop neointimal lesions in response to injury in patients, whereas the pulmonary vasculature is spared from neointimal disease, even in the presence of old age, severe hypercholesterolemia, or diabetes. Medial hypertrophy typically develops in animal models of pulmonary vascular remodeling whereas neointimal hyperplasia forms in the systemic vasculature of animals after injury. Heath, Glagov, and Wissler each independently noted the decreased frequency of neointimal disease in the pulmonary artery and suggested that one explanation for the differences between the pulmonary and systemic artery patterns of remodeling is that the substantially lower blood pressures in normal pulmonary arteries are somehow protective (10). The occurrence of pulmonary artery neointimal lesions and even frank atherosclerosis only in the presence of severe pulmonary hypertension, regardless of etiology, is consistent with that hypothesis but until recently experimental evidence that hemodynamic factors could influence the pattern of pulmonary vascular remodeling was missing.

	MEDIAL HYPERTROPHY CAN BE CONVERTED INTO A NEOINTIMAL PATTERN IN EXPERIMENTAL ANIMALS BY COUPLING VASCULAR INJURY WITH INCREASED BLOOD FLOW

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

Recent experiments with the monocrotaline injury model of pulmonary vascular remodeling began to supply that evidence because a medial hypertrophy pattern of remodeling could be converted into a neointimal pattern by a change in pulmonary vascular hemodynamics (13, 14). Early pathologic changes after monocrotaline injury include pulmonary artery endothelial swelling and blebbing (15), and inflammatory infiltrates (16). Later changes include medial hypertrophy, increased vascular tropoelastin, and type I collagen synthesis and deposition (15, 17). Monocrotaline injury is also associated with changes in several biochemical mediators (18, 19) as well as changes in endothelial barrier function (20). Neointimal lesions normally do not develop.

Rat intra-acinar pulmonary arteries develop neointima if monocrotaline injury is coupled with increased blood flow after contralateral pneumonectomy whereas only medial hypertrophy develops after identical injury with normal flow (14). Immunohistochemical studies show that these neointimal lesions are composed of SMC because they are immunoreactive to an anti--smooth muscle actin antibody. Severe right ventricular hypertrophy (RVH), consistent with the presence of vascular occlusion in the right lung, develops in animals with a neointimal pattern of remodeling. Right lung histology from animals treated with monocrotaline + pneumonectomy is comparable to animals treated with pneumonectomy + monocrotaline, demonstrating that neointimal injury develops independently of the sequence of injury and increased flow. With time, neointima is also observed in more proximal elastic pulmonary arteries suggesting a distal to proximal migration of the conditions required for pulmonary artery neointimal lesions to develop.

Neointimal lesions also develop in monocrotaline-injured elastic pulmonary arteries after exposure to systemic hemodynamic conditions following a left subclavian-left pulmonary artery anastomosis, but not with either monocrotaline injury or anastomosis alone (13).

	HOW DOES INCREASED BLOOD FLOW ALLOW SMC NEOINTIMAL FORMATION AFTER ENDOTHELIAL INJURY?A HYPOTHESIS

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

We propose a "2-hit" model to explain the pathogenesis of pulmonary artery neointimal formation (Figure 1). According to this model, both endothelial injury and changes in pulmonary artery hemodynamics must occur for neointima to form. If the injury is limited and the endothelium can quickly repair itself, then either no remodeling or minimal remodeling occurs without leading to clinical disease. In patients with PPH, sufficient endothelial injury and/or a persistent response despite subsidence of the initiating injury probably occurs with subsequent development of medial hypertrophy. As medial hypertrophy progresses, the hemodynamic conditions change sufficiently over time so that an appropriate hemodynamic milieu is created which initiates a switch in the pattern of vascular remodeling and the development of neointimal lesions. The utility of the monocrotaline + pneumonectomy animal model is that the hemodynamic conditions can be surgically superimposed with the vascular injury within a time period that allows experimental study.

View larger version (19K): [in this window] [in a new window]   Figure 1.   A general model of primary pulmonary hypertension. The initial response to endothelial injury includes release of inflammatory mediators which leads to vasoconstriction, medial SMC hyperplasia and proliferation, and extracellular matrix (ECM) synthesis. According to this hypothesis, the response of the pulmonary artery to most injuries is confined to the medial layer. This medially limited response is distinct from the response of systemic arteries which respond with neointimal proliferation. However, as medial hypertrophy develops, pulmonary artery resistance increases and compliance decreases. Local blood flow diminishes, but shear stresses increase, at sites of injury and remodeling. With time, the local hemodynamic status is sufficiently altered so that a neointima forms. Thus, mediators of vascular "injury" are necessary and sufficient for the development of medial hypertrophy. These mediators are necessary but not sufficient for neointimal formation. Additional factors, induced by changes in pulmonary hemodynamics, are required for neointimal formation.

	FUTURE DIRECTIONS

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

Injury

Although "injury" is considered a prerequisite for remodeling, the term remains vaguely defined. Significant insights into atherogenesis have arisen from understanding how oxidized lipids chronically injure the systemic vasculature; no doubt, similar important insights would be gained from defining the molecular mechanisms of pulmonary vascular injuries leading to pulmonary vascular remodeling and hypertension. Therefore, future studies should focus on defining pulmonary vascular injury in molecular terms. Is injury by monocrotaline (or other toxins) similar to immune-mediated injury or viral infection? Is there a stereotypical response to injury? What are the similarities and differences downstream from different types of injury? Is defining the initial injury relevant by the time patients present to clinical attention? Does "genetic susceptibility" mean susceptibility to injury, or a particular kind of response to injury?

Similarly, how does the endothelial cell respond to injury in vivo? Responses of cells in vitro differ from responses in vivo (21), emphasizing the need for in vivo rather than in vitro studies. New techniques are being developed to more precisely measure those responses in vivo (22). Is the response "dysfunctional" or an appropriate function in the new milieu in which the cell finds itself, i.e., an appropriate adaptive response? What is the impact of endothelial cell "dysfunction" on the underlying vascular smooth muscle cells? Based on a growing body of evidence, significant changes occur in which endothelial mediators (cytokines, growth factors) are released after injury, but other responses have not been well-studied.

This leads to the question of how various factors mediate remodeling. What mediators released by endothelial cells induce vascular SMC growth? How much redundancy is there will blocking one factor suppress remodeling? What is the relationship between mediators of vasoconstriction and pulmonary vascular remodeling? What role does pulmonary artery SMC phenotypic diversity have in response to injury? Can the vasculature "escape" from or develop resistance to therapeutic agents because of phenotypic diversity? We recently demonstrated a significant delay in pulmonary artery neointimal formation by angiotensin-converting enzyme (ACE) inhibition (23) even when started 3 wk after injury was induced. Nevertheless, severe neointimal formation eventually developed despite ACE inhibition. Whether this was the result of insufficient focal ACE inhibition, resistance to ACE inhibition by a subpopulation of neointimal precursor cells, or redundancy of mediators (i.e., participation of ACE-independent pathways) is unknown. Recent studies demonstrating both heparin-sensitive and heparin-insensitive subpopulations of SMC within systemic arteries, and selective growth of heparin-insensitive SMC clones, suggest that understanding pulmonary vascular phenotypic diversity may be essential for developing useful anti-remodeling therapeutic approaches.

Flow

The endothelium is the interface between hemodynamic conditions and the vascular wall and transduces hemodynamic forces into biochemical signals to maintain vascular homeostasis. Systemic arteries tend to normalize to a wall shear stress value of approximately 15 dynes/cm² (24). Thus, neointimal formation is thought to be one mechanism by which the vessel is able to maintain an appropriate mechanical environment. Hemodynamic forces can be resolved into two principal vectors: first, shear stress, a frictional force acting at the interface between flowing blood and the vessel wall; and second, pressure acting normal to the vessel wall, which imposes circumferential stretch to the tissue. It is now well accepted that hemodynamic shear stress acts through the endothelium to regulate both acute vessel tone and chronic restructuring of blood vessels (24). Thus, the endothelium is a complex mechanical signal-transduction interface between flowing blood and the vessel wall.

Three general steps are considered essential for cells to appropriately respond to changes in their mechanical environment. First, cells must be able to sense changes in their mechanical milieu. One current hypothesis is that shear stresses acting at the lumenal surface are mechanically transmitted via the cytoskeleton to focal adhesions tethering the cell to underlying extracellular matrix proteins. Increased shear stresses are perceived as increased tension in the extracellular matrix- integrin-cytoskeleton structure, particularly the focal adhesion complexes, initiating signals that will adapt the cell to its new mechanical environment (24, 25). Second, cells must transduce those altered mechanical stresses into secondary messengers and signals. Defining the intracellular signaling pathways mediating these responses to mechanical stresses is a very active area of research currently. Finally, endothelial cells must respond to those intracellular signals. Many studies have demonstrated a wide variety of cellular responses to mechanical stresses. For convenience, these changes are usually organized by how quickly they occur after shear stress changes (Table 1). A biochemical response, such as changes in signaling cascades, calcium fluxes, or vasodilatation, can be stopped quickly or reversed if the mechanical environment exhibits acute volatility. However, if the new mechanical environment is sustained, more permanent changes occur, such as changes in cell shape and cell proliferation that occur over many hours to days (24).

The role of hemodynamic factors in diseases of the systemic circulation has been long appreciated. Evidence that endothelial cells and hemodynamic factors are important to systemic neointimal lesion formation includes (1) the association between regions of disturbed flow and the distribution of atherosclerotic lesions in large arteries; (2) significant differences in the endothelial actin cytoskeletal structure between areas of laminar and disturbed flow (26); (3) the development of intimal hyperplasia after arterial injury is modified by changes in blood flow in rats (27) and other species; (4) endothelial cell structure, function, and gene expression change as hemodynamic conditions change (24); and (5) activation of endothelial cell intracellular signaling pathways by changes in hemodynamic forces (28).

Endothelial cells also contribute to maintaining the underlying SMC in a quiescent state. Normal adult pulmonary artery SMC are not growing, migrating, or synthesizing matrix once normal development is completed. Endothelial cells maintain SMC quiescence by inhibiting SMC proliferation (29) and matrix synthesis (30). Endothelium-derived factors mediating SMC quiescence include nitric oxide which inhibits SMC migration (31) and proliferation (32) and heparan sulfates that inhibit smooth muscle cell growth (33). Flow modulates production of nitric oxide, platelet-derived growth factor, transforming growth factor-, and likely other factors that regulate cell growth, migration, and matrix synthesis (24).

In addition to maintaining SMC in a quiescent state the endothelium serves an important barrier function regulating extravasation of soluble plasma factors (34). Shear stress may modulate endothelial barrier function (35). If the barrier function of the endothelium does not perform appropriately, such as after injury or in the presence of high shear stress, the interstitium can be exposed to high concentrations of plasma factors which can then act to induce quiescent SMCs to begin proliferating and synthesizing matrix.

How do pulmonary artery endothelial cells respond to changes in hemodynamics (shear stress, transmural pressure)? What are the endothelial mechanotransducers and second messengers? How is the response of endothelial cells to injury modulated by hemodynamics: are additional mediators released or is expression of mediator receptors induced, do cytoskeletal changes occur with subsequent changes in cell function? One advantage of the pulmonary compared with the systemic vasculature is that a pneumonectomy can be performed in mice (whereas altering flow conditions in the mouse systemic vasculature appears problematic). The effect of changes in pulmonary hemodynamics after pneumonectomy can be studied in gene "knock-out" or "over-expressing" mice so that general principles of how flow affects cell structure and function can be determined. Another approach is to use subtraction library or differential display techniques to compare the response of the pulmonary vasculature to injury and flow alone versus injury plus flow.

	SUMMARY

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

Both human pathologic studies and animal models have a role in PPH research. PPH research is at a level of sophistication similar to research on atherosclerosis, asthma or chronic obstructive pulmonary disease many years ago. Researchers recognized then that to understand a disease we must study the disease: sophisticated descriptive studies of those diseases performed years ago prompted the questions today being addressed with animal models or cell culture studies. Thus, animal models can have an important function in PPH depending on the appropriateness of that model for the question being asked. Our current understanding of the role of hemodynamics in pulmonary vascular remodeling illustrates the utility of that approach.

	Footnotes

Correspondence and requests for reprints should be addressed to Mitchell D. Botney, M.D., Respiratory and Critical Care Division, Jewish Hospital of St. Louis, 216 S. Kingshighway Blvd., St. Louis, MO 63110.

(Received in original form May 22, 1998 and in revised form August 26, 1998).

Acknowledgments: Supported by NIH Grants HL02425 and HL29594, the Alan A. and Edith L. Wolff Charitable Trust, and a Career Investigator Award from the American Lung Association.

	References

TOP INTRODUCTION WHY IS THE PULMONARY... MEDIAL HYPERTROPHY CAN BE... HOW DOES INCREASED BLOOD... FUTURE DIRECTIONS SUMMARY REFERENCES

1. Botney, M. D., L. R. Kaiser, J. D. Cooper, R. P. Mecham, J. Roby, and W. C. Parks. 1992. Extracellular matrix protein gene expression in atherosclerotic pulmonary arteries. Am. J. Pathol. 140: 357-364 [Abstract].

2. Botney, M. D., M. J. Liptay, L. R. Kaiser, J. D. Cooper, W. C. Parks, and R. P. Mecham. 1993. Active collagen synthesis by pulmonary arteries in human primary pulmonary hypertension. Am. J. Pathol. 143: 121-129 [Abstract].

3. Liptay, M. J., W. C. Parks, R. P. Mecham, J. Roby, L. R. Kaiser, J. D. Cooper, and M. D. Botney. 1993. Neointimal macrophages co-localize with extracellular matrix gene expression in human atherosclerotic pulmonary arteries. J. Clin. Invest. 91: 588-594 .

4. Botney, M. D., L. Bahadori, and L. I. Gold. 1994. Vascular remodeling in primary pulmonary hypertension: potential role for transforming growth factor-. Am. J. Pathol. 144: 286-295 [Abstract].

5. Bahadori, L., J. Milder, L. I. Gold, and M. D. Botney. 1995. Active macrophage-associated TGF- colocalizes with type I procollagen gene expression in atherosclerotic human hypertensive pulmonary arteries. Am. J. Pathol. 146: 1140-1149 [Abstract].

6. Stenmark, K. R., and R. P. Mecham. 1997. Cellular and molecular mechanisms of pulmonary vascular remodeling. Ann. Rev. Physiol. 59: 89-144 [Medline].

7. Mette, S. A., H. I. Palevsky, G. G. Pietra, T. M. Williams, E. Bruder, A. J. Prestipino, A. M. Patrick, and J. A. Wirth. 1992. Primary pulmonary hypertension in association with human immunodeficiency virus infection: a possible viral etiology for some forms of hypertensive pulmonary arteriopathy. Am. Rev. Respir. Dis. 145: 1196-1200 [Medline].

8. Widgren, S. 1977. Pulmonary hypertension related to Aminorex intake. In E. Grundmann and W. H. Kirsten, editors. Current Topics in Pathology. Springer, New York. 1-64.

9. Barst, R. J., E. R. Flaster, A. Menon, M. Fotino, and J. H. Morse. 1992. Evidence for the association of unexplained pulmonary hypertension in children with the major histocompatibility complex. Circulation 85: 249-258 [Abstract/Free Full Text].

10. Heath, D., E. H. Wood, J. W. DuShane, and J. E. Edwards. 1960. The relation of age and blood pressure to atheroma in the pulmonary arteries and thoracic aorta in congenital heart disease. Lab. Invest. 9: 259-272 [Medline].

11. Glagov, S., and A. K. Ozoa. 1968. Significance of the relatively low incidence of atherosclerosis in the pulmonary, renal, and mesenteric arteries. Ann. N.Y. Acad. Sci. 149: 940-955 [Medline].

12. Wissler, R. W., and D. Vesselinovitch. 1990. Atherogenesis in the pulmonary artery. In A. P. Fishman, editor. The Pulmonary Circulation: Normal and Abnormal. University Pennsylvania Press, Philadelphia, PA. 245-255.

13. Tanaka, Y., D. P. Schuster, E. C. Davis, G. A. Patterson, and M. D. Botney. 1996. The role of vascular injury and hemodynamics in pulmonary artery remodeling. J. Clin. Invest. 98: 434-442 [Medline].

14. Okada, K., Y. Tanaka, M. Bernstein, W. Zhang, G. A. Patterson, and M. D. Botney. 1997. Pulmonary hemodynamics modify the rat pulmonary artery response to injury: a neointimal model of pulmonary hypertension. Am. J. Pathol. 151: 1019-1025 [Abstract].

15. Todorovich-Hunter, L., D. J. Johnson, P. Ranger, F. W. Keeley, and M. Rabinovitch. 1988. Altered elastin and collagen synthesis associated with progressive pulmonary hypertension induced by monocrotaline: a biochemical and ultrastructural study. Lab. Invest. 58: 184-195 [Medline].

16. Stenmark, K. R., M. L. Morganroth, L. K. Remigio, N. F. Voelkel, R. C. Murphy, P. M. Henson, M. M. Mathias, and J. T. Reeves. 1985. Alveolar inflammation and arachidonate metabolism in monocrotaline- induced pulmonary hypertension. Am. J. Physiol. 248: H859-H866 .

17. Kameji, K., H. Otsuka, and Y. Hayashi. 1980. Increase of collagen synthesis in pulmonary arteries of monocrotaline-treated rats. Experientia 36: 441-442 [Medline].

18. Arcot, S. S., D. W. Lipke, M. N. Gillespie, and J. W. Olson. 1993. Alterations of growth factor transcripts in rat lungs during development of monocrotaline-induced pulmonary hypertension. Biochem. Pharmacol. 46: 1086-1091 [Medline].

19. Tanaka, Y., M. L. Bernstein, R. P. Mecham, G. A. Patterson, J. D. Cooper, and M. D. Botney. 1996. Site-specific responses to monocrotaline-induced vascular injury: evidence for two distinct mechanisms of remodeling. Am. J. Respir. Cell Mol. Biol. 15: 390-397 [Abstract].

20. Kido, M., T. Hirose, K. Tanaka, T. Kurozumi, and Y. Shoyama. 1981. Increased alveolar-capillary membrane permeability by monocrotaline. Jap. J. Med. 20: 170-179 [Medline].

21. Liaw, L., and S. M. Schwartz. 1993. Comparison of gene expression in bovine aortic endothelium in vivo versus in vitro: differences in growth regulatory molecules. Arterioscler. Thromb. 13: 985-993 [Abstract/Free Full Text].

22. Taguchi, J., C. E. Murry, B. I. Herren, M. Pech, S. M. Schwartz, and V. Lindner. 1998. A quantitative method for determination of endothelial mRNA expression in vivo. Am. J. Pathol. 152: 903-912 [Abstract].

23. Okada, K., M. Bernstein, W. Zhang, D. P. Schuster, and M. D. Botney. 1998. Angiotensin converting enzyme inhibition delays pulmonary vascular neointimal formation. Am. J. Respir. Crit. Care Med. 158: 939-950 [Abstract/Free Full Text].

24. Davies, P. F.. 1995. Flow-mediated endothelial mechanotransduction. Physiol. Rev. 75: 519-560 [Abstract/Free Full Text].

25. Ingber, D.. 1991. Integrins as mechanochemical transducers. Curr. Opin. Cell Biol. 3: 841-848 [Medline].

26. Kim, D. W., A. I. Gotlieb, and B. L. Langille. 1989. In vivo modulation of endothelial F-actin microfilaments by experimental alterations in shear stress. Arteriosclerosis 9: 439-445 [Abstract/Free Full Text].

27. Kohler, T. R., and A. Jawien. 1992. Flow affects development of intimal hyperplasia after arterial injury in rats. Arterioscler. Thromb. 12: 963-971 [Abstract/Free Full Text].

28. Berk, B. C., M. A. Corson, T. E. Peterson, and H. Tseng. 1995. Protein kinases as mediators of fluid shear stress stimulated signal transduction in endothelial cells: a hypothesis for calcium-dependent, and calcium-independent events activated by flow. J. Biomech. 28: 1439-1450 [Medline].

29. Clowes, A. W., M. A. Reidy, and M. M. Clowes. 1983. Kinetics of cellular proliferation after arterial injury: I. Smooth muscle growth in the absence of endothelium. Lab. Invest. 49: 327-336 [Medline].

30. Powell, R. J., J. Hydowski, O. Frank, J. Bhargava, and B. E. Sumpio. 1997. Endothelial cell effect on smooth muscle cell collagen synthesis. J. Surg. Res. 69: 113-118 [Medline].

31. Dubey, R. K., E. K. Jackson, and T. F. Luscher. 1995. Nitric oxide inhibits angiotensin II-induced migration of rat aortic smooth muscle cell: role of cyclic-nucleotides and angiotensin1 receptors. J. Clin. Invest. 96: 141-149 .

32. Garg, U. C., and A. Hassid. 1989. Nitric oxide-generating vasodilators and 8-bromo-cyclic guanosine monophosphate inhibit mitogenesis and proliferation of cultured rat vascular smooth muscle cells. J. Clin. Invest. 83: 1774-1777 .

33. Castellot, J. J., M. L. Addonizio, R. Rosenberg, and M. J. Karnovsky. 1981. Cultured endothelial cells produce a heparinlike inhibitor of smooth muscle cell growth. J. Cell Biol. 90: 372-379 [Abstract/Free Full Text].

34. van Hinsbergh, W. M.. 1997. Endothelial permeability for macromolecules: mechanistic aspects of pathophysiological modulation. Arterioscler. Thromb. Vasc. Biol. 17: 1018-1023 [Free Full Text].

35. Waters, C. M.. 1996. Flow-induced modulation of the permeability of endothelial cells cultured on microcarrier beads. J. Cell. Physiol. 168: 403-411 [Medline].