Blood flow forces liver growth

Increases in biomechanical forces in the liver’s blood vessels have now been shown to activate two mechanosensitive proteins. The proteins trigger blood-vessel cells to deploy regenerative factors that drive liver growth.
Sina Y. Rabbany is in the DeMatteis School of Engineering and Applied Science, Hofstra University, New York, New York 11548, USA, and in the Division of Regenerative Medicine, Ansary Stem Cell Institute, Weill Cornell Medicine, New York.

Search for this author in:

Shahin Rafii is in the DeMatteis School of Engineering and Applied Science, Hofstra University, New York, New York 11548, USA, and in the Division of Regenerative Medicine, Ansary Stem Cell Institute, Weill Cornell Medicine, New York.

Search for this author in:

The molecular pathways that initiate and sustain liver growth during development and after injury are orchestrated in part by a balanced supply of stimulatory and inhibitory factors secreted from specialized liver sinusoidal endothelial cells (LSECs), which line the organ’s blood vessels14. But it is unclear how the liver vasculature senses the need to produce these endothelial-cell-derived (angiocrine) growth factors, such as hepatocyte growth factor (HGF) and Wnt proteins, to guide proper organ growth4. In a paper in Nature, Lorenz et al.5 show how mechanical forces created by the passage of blood through the liver activate signalling pathways that promote the production of angiocrine factors and the proliferation of the organ’s main cell type, hepatocytes, in mice.

Mechanosensing in the liver depends on the amount of blood delivered by the portal vein and the hepatic artery, and on the tensile strength of blood-vessel walls, which is imparted by collagen fibres. The net blood flow (perfusion) subjects LSECs to two major forces6. First, mechanical distortion and tension of the vessel wall owing to blood pressure results in cyclic stretch in the cells. Second, friction arising from viscous blood flow over the vessel wall causes fluid shear stress. These synergistic biomechanical forces lead to upregulation of various mechanosensing proteins, inducing LSECs to produce angiocrine factors such as nitric oxide and reactive oxygen species that act to modulate the vasculature, together with ‘stronger’ angiocrine factors that stimulate hepatic regeneration. However, the mechanism(s) by which biomechanical forces activate the strong angiocrine function of LSECs to choreograph hepatic proliferation have not been elucidated7.

Lorenz et al. set out to investigate these mechanisms using mouse embryos removed from mothers and cultured ex vivo. They first observed that an increase in the liver’s growth rate over different developmental stages correlated with enhanced blood perfusion through the organ. Most proliferating hepatic cells in the developing liver were localized to regions that had been perfused, and the researchers found that the level of vascular perfusion correlated with the level of activation of two receptor proteins on LSECs that sense and respond to force — integrin β1 and vascular endothelial growth factor receptor 3 (VEGFR3). In turn, these proteins promoted secretion of the key angiocrine factor HGF (Fig. 1).

The authors then modified perfusion rates in cultured embryos by using drugs to halt or increase the fetal heartbeat. Blocking liver perfusion reduced HGF secretion and resulted in diminished hepatic growth — as did deleting the genes that encode integrin β1 or VEGFR3 in LSECs in embryos in vivo. By contrast, enhancing the rate of blood perfusion induced the secretion of HGF, and this was again mediated by integrin β1 and VEGFR3.

Lorenz et al. then turned to livers removed from adult mice and cultured ex vivo. They increased perfusion in the livers by injecting a buffer solution or by removing 70% of the liver, which redirects a large volume of fluid to the organ’s remaining lobes at high pressure. They measured the perfusion rate using an imaging technique called contrast-enhanced ultrasonography. Enhanced perfusion led to increased LSEC diameter, increased blood volume and flow and so increased shear stress, leading to higher activation of integrin β1 and VEGFR3.

Together, these experiments provide evidence that, in mice, activation of mechanosensors in blood vessels in both the fetal and adult liver triggers angiocrine signalling to promote hepatocyte proliferation — presumably, to enable liver growth during embryonic development and maintenance and regeneration in adults. Next, Lorenz et al. turned to human cells cultured in vitro. Here, too, mechanical stretching of LSEC-like cells or antibody-dependent activation of integrin β1 led to a robust increase in secretion not only of HGF, but also of other angiocrine factors. These secreted factors promoted the proliferation and survival of human hepatocytes grown in vitro. Finally, the authors showed that, in metabolically healthy people, increases in systemic blood pressure correlated with significantly larger livers.

Lorenz and colleagues have used sophisticated approaches to link mechanical forces to the induction of angiocrine-mediated liver development and growth. However, several issues remain unresolved. For instance, the cyclic stretch that LSECs undergo in vivo when the vessel widens after exposure to accelerated perfusion is biaxial — that is, the cell is stretched both along the direction of the vessel and sideways. By contrast, Lorenz and colleagues’ ex vivo and in vitro experiments imparted only uniaxial cyclic stretch8. This difference might bias the signalling and angiocrine outputs the group observed. Whether other vascular mechanosensor receptors have a role in the induction of angiocrine factors also needs to be elucidated9.

In addition, the role of this biomechanically responsive pathway during injury remains to be dissected. Excessive increases in shear stress (for example, as a result of acute loss of liver mass) could be detrimental, leading to suboptimal liver regeneration. Lorenz et al. also did not directly assess whether lack of biomechanical activation of integrin β1 and VEGFR3, as might occur in diseases such as diabetes, would lead to decreases in the liver’s regenerative potential1,2.

In future, the ideal magnitude of cyclic stretch or shear stress required to initiate the physiological induction of angiocrine factors should be studied. The recruitment of circulating endothelial progenitor cells (EPCs), which are thought to supply the liver with HGF, could also be affected by shear-dependent activation of LSECs, further altering the liver’s supply of angiocrine factors10. Indeed, how increased biomechanical forces alter the delivery of regenerative modulators to the liver, including circulating EPCs, inflammatory cells and platelets, to drive liver growth without encouraging scarring, needs further investigation.

Exactly how do integrin β1 and VEGFR3 upregulate angiocrine factors? It is plausible that fluid shear stress induces integrin-mediated nuclear localization of specific transcription factors and so promotes the expression of angiocrine-factor genes24. Furthermore, integrin-mediated modulation of the elasticity of the extracellular matrix around hepatocytes in response to shear stress could also modulate hepatocyte proliferation. But what about VEGFR3? Proteins of the VEGFR family are activated by phosphorylation. Biomechanically independent phosphorylation of VEGFR2 on LSECs activates the protein AKT, which recruits the transcription factor Id1 to DNA, inducing the expression of Wnt2 and HGF genes2. But the mechanism by which phosphorylation of VEGFR3 turns on angiocrine factors is unknown.

These questions notwithstanding, Lorenz and colleagues’ work takes into consideration the complexity of the biophysical environment to which LSECs are exposed in vivo, and so solves a mystery that has puzzled liver biologists for decades. The development of strategies that precisely regulate the magnitude of shear stress and cyclic stretch in the liver vasculature might restore angiocrine-dependent regenerative functions of the liver in pathological conditions, such as in cirrhosis, hepatitis and vascular abnormalities. This could in turn open the door to more-effective therapeutic liver regeneration.

Nature 562, 42-43 (2018)

doi: 10.1038/d41586-018-06741-2
Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up


  1. 1.

    Rafii, S., Butler, J. M. & Ding, B.-S. Nature 529, 316–325 (2016).

  2. 2.

    Ding, B.-S. et al. Nature 468, 310–315 (2010).

  3. 3.

    Hu, J. et al. Science 343, 416–419 (2014).

  4. 4.

    Rocha, A. S. et al. Cell Rep. 13, 1757–1764 (2015).

  5. 5.

    Lorenz, L. et al. Nature 562, 128–132 (2018).

  6. 6.

    Rabbany, S. Y., Ding, B.-S., Larroche, C. & Rafii, S. in Mechanical and Chemical Signaling in Angiogenesis (ed. Reinhart-King, C. A.) 19–45 (Springer, 2012).

  7. 7.

    Song, Z. et al. Semin. Cell Dev. Biol. 71, 153–167 (2017).

  8. 8.

    Wang, J. H.-C., Goldschmidt-Clermont, P., Wille, J. & Yin, F. C.-P. J. Biomech. 34, 1563–1572 (2001).

  9. 9.

    Baeyens, N., Bandyopadhyay, C., Coon, B. G., Yun, S. & Schwartz, M. A. J. Clin. Invest. 126, 821–828 (2016).

  10. 10.

    DeLeve, L. D. J. Clin. Invest. 123, 1861–1866 (2013).

Download references