Longitudinal ultrasound imaging and network modeling in rats reveal sex-dependent suppression of liver regeneration after resection in alcoholic liver disease

Liver resection is an important surgical technique in the treatment of cancers and transplantation. We used ultrasound imaging to study the dynamics of liver regeneration following two-thirds partial hepatectomy (PHx) in male and female rats fed via Lieber-deCarli liquid diet protocol of ethanol or isocaloric control or chow for 5–7 weeks. Ethanol-fed male rats did not recover liver volume to the pre-surgery levels over the course of 2 weeks after surgery. By contrast, ethanol-fed female rats as well as controls of both sexes showed normal volume recovery. Contrary to expectations, transient increases in both portal and hepatic artery blood flow rates were seen in most animals, with ethanol-fed males showing higher peak portal flow than any other experimental group. A computational model of liver regeneration was used to evaluate the contribution of physiological stimuli and estimate the animal-specific parameter intervals. The results implicate lower metabolic load, over a wide range of cell death sensitivity, in matching the model simulations to experimental data of ethanol-fed male rats. However, in the ethanol-fed female rats and controls of both sexes, metabolic load was higher and in combination with cell death sensitivity matched the observed volume recovery dynamics. We conclude that adaptation to chronic ethanol intake has a sex-dependent impact on liver volume recovery following liver resection, likely mediated by differences in the physiological stimuli or cell death responses that govern the regeneration process. Immunohistochemical analysis of pre- and post-resection liver tissue validated the results of computational modeling by associating lack of sensitivity to cell death with lower rates of cell death in ethanol-fed male rats. Our results illustrate the potential for non-invasive ultrasound imaging to assess liver volume recovery towards supporting development of clinically relevant computational models of liver regeneration.

The virulence-associated protein HsvA from the fire blight pathogen Erwinia amylovora is a polyamine amidinotransferase

Studies of virulence determinants in the bacterial phytopathogen Erwinia amylovora, the cause of devastating fire blight disease in apple and pear, have shown that HsvA, a putative amidinotransferase enzyme located in the Hrp pathogenicity island, is required for systemic infection in apple. However, the mechanism by which HsvA contributes to virulence is unclear. To investigate the role of HsvA in virulence, we carried out a series of biochemical and structural studies to characterize the amidinotransferase activity of HsvA. We found that HsvA displays a preference for linear aliphatic polyamines as the amidino acceptor substrate, especially for spermidine and putrescine (Km values of 33 μm and 3.9 mm, respectively). The three-dimensional structure, determined at 2.30 Å resolution using X-ray crystallography, revealed that the overall architecture of HsvA is similar to that of the human arginine-glycine amidinotransferase in the creatine biosynthesis pathway. The active site is located in the core of the protein at the base of a long, narrow substrate access channel. Specific amino acids near the entrance of the channel may serve as major determinants of the substrate specificity, including a glutamate residue at the rim of the channel entrance that appears to be positioned to interact with the distal primary amine in the putrescine substrate as well as the internal and distal amines in the spermidine substrate. These results suggest potential in vivo functions for HsvA as a virulence factor in fire blight and may also provide a basis for strategies to control fire blight by inhibiting HsvA activity.

Ray casting and the enclosing-net algorithm for extracting shapes from volume data

One of the most useful applications of virtual reality is to let doctors view the inside of the human body non-invasively and in real time. In this paper, we first survey the area of virtual reality and volume-visualization techniques. We discuss our ray casting implementation for viewing the medical data which is available as a set of slices from NMR and CT scans. Next we present a new method, the enclosing-net algorithm, for extracting iso-surfaces from volume data. A simple implementation of our technique is described. Since the topology of the extracted surface is well defined and non-ambiguous, the enclosing-net algorithm eliminates a major problem of surface extraction techniques. In addition, the number of polygons representing the surface can be controlled to obtain a finer shape. Therefore real-time interaction is feasible on both low-end and high-end graphics machines, making the enclosing-net algorithm suitable for virtual reality experiments.