Constructing computational 3D liver lobule microstructures lays the basis for understanding the spatial aspects of tissue-level events such as transport and delivery of oxygen within the liver. To satisfy the demand of liver cells for oxygen and nutrient, hepatic sinusoids that perfuse hepatocytes must maintain an appropriate network structure. Here we describe a refinement of the liver portion of our multiscale model consisting of a multi-cell 3D liver lobule that includes blood flow. The human liver is constructed of approximately one million lobules "plumbed" with blood vessels in a parallel arrangement. The parallel nature of the lobules within the liver makes it possible to simulate the entire liver by simulating a small number of lobules. We have used Compucell3D to construct representative portions of a 3D lobule consisting of portal arterial plus venous blood sources, a network of sinusoids (liver capillaries) lined with hepatocytes and a central vein drain. The lobule is stochastically constructed but directed by parameters derived from the analysis of mouse livers. The resulting network was then used to simulate flow of oxygenated blood though the lobule using Kirchhoff's law. A critical assumption in the generation of the model is that the diameter of the sinusoids is constant throughout a lobule. The spatial patterning of sinusoidal network was designed to be optimally favorable for oxygen and nutrient delivery. These assumptions, and Kirchhoff's law, result in the requirement that the linear blood velocity increases from the pericentral region to the central vein of a liver lobule. We have also modeled the diffusion of oxygen out of the sinusoidal blood and its consumption by hepatocytes, using a PDE solver. The combination of the stochastic sinusoid structure, and the changes in blood velocity in different regions of a lobule, give rise to localized differences in oxygen availability.

CC3D; Liver; Simulated 3D liver lobule; Oxygen distribution; blood flow; computational