Hepatic response to right ventricular pressure overload Roben G. Gieling, Jan M. Ruijter, Adri A.W. Maas, Marius A. Van Den Bergh Weerman, Koert P. Dingemans, Fibo J.W. ten Kate, Ronald H. Lekanne dit Deprez, Antoon F.M. Moorman, Wouter H. Lamers  Gastroenterology  Volume 127, Issue 4, Pages 1210-1221 (October 2004) DOI: 10.1053/j.gastro.2004.07.057 Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 1 Relation between heart weight (HW) and body weight (BW) in PTB rats. Rats weighing approximately 150 g had their pulmonary trunk banded to 1.1 or 1.2 mm outer diameter (PTB1.1 or PTB1.2) or were sham operated. Heart and body weight are expressed in grams. Animals with a heart/body weight ratio exceeding 0.0027 were considered to have developed cardiac hypertrophy (inset, responders [R, triangles]), whereas those with a lower ratio were nonresponders (inset, NR [squares]). The regression lines relating heart/body weight ratios in responders (y = 0.306 + 0.00427 ×) and nonresponders (y = 0.381 + 0.002241 ×), including shams, as well as the dotted line corresponding to a heart/body weight ratio of 0.0027 are shown. Responders either did (F) or did not (H) have symptoms of heart failure at the time they were killed. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 2 Liver architecture and expression pattern of GS and OAT in the liver of PTB rats. (A) The sections in the left column were stained with H&A and those in the middle and right columns for the presence of GS and OAT, respectively. The hepatic response to PTB was graded in 3 stages, with PTB stage 0 (PTB-0) representing the control condition. Increasing involvement of the liver was demonstrable as a gradual widening of the sinusoids in the pericentral zone and an increase in the connective tissue of the portal tracts and the walls of the central veins (H&A column) and as a decline of the GS- and OAT-positive hepatocytes to a single row around the central vein (PTB-1), to a single but discontinuous row of pericentral hepatocytes (PTB-2), or to the complete disappearance of these GS- or OAT-positive pericentral hepatocytes (PTB-3). In a subset of PTB stage 3 animals, OAT became expressed in the periportal hepatocytes. Note that the histogram of the OD values was linearly stretched for better visualization of the gene expression patterns. Bars = 500 μm. (B) The upper row shows the relation between the cardiac response (x-axis) and the hepatic response (y-axis) to PTB. The hepatic response was based on the expression pattern of GS (left panel) or OAT (right panel). The lower 2 rows show the relation between liver enzyme content as assayed by Western blot on the y-axis and the hepatic response (middle row) or the cardiac response (lower row) on the x-axis for both GS (left panels) and OAT (right panels). Enzyme content in control animals was set at 100%. Note that the decline in both GS and OAT enzyme content corresponded well with the progression in PTB stage but that no difference was observed between cardiac responders without heart failure and cardiac nonresponders. 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 3 Circulating ammonia levels in PTB rats. Rats were classified according to their cardiac response (left horizontal axis) and their hepatic response (right horizontal axis). Columns represent mean venous ammonia levels (±SEM), with the numbers in the columns representing the number of animals studied. Significantly increased ammonia levels were exclusively associated with cardiac failure, showing that this functional condition rather than the disappearance of GS was to blame for the hyperammonemia. 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 4 Expression pattern of CYPred, CPS, and PEPCK in the liver of PTB rats. (A) The sections in the left column were stained for the presence of CYPred, those in the middle column for the presence of CPS, and those in the right column for the presence of PEPCK. The hepatic response to PTB was graded as shown in Figure 2. CYPred expression began to disappear pericentrally and became expressed periportally in PTB stage 2 animals. CPS expression decreased pericentrally, especially in PTB stage 3 animals. PEPCK expression was unchanged. Bar = 500 μm. (B) The relation between liver enzyme content as assayed by Western blot (y-axis) and the hepatic response (upper row) or cardiac response (lower row) (x-axis) for CYPred, CPS, and PEPCK. Enzyme content in control animals was set at 100%. Note that the overall hepatic content of CYPred and PEPCK was unaffected by PTB, whereas CPS content declined, particularly in PTB stage 3 animals. 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 5 Cell death and proliferation in the liver of PTB rats. (A) Apoptosis in the liver as determined by the TUNEL assay. Only one animal of a stage 3 liver was investigated. (B) Apoptosis as a function of time after PTB. No significant trend (y = 0.0021 × + 0.8472) was present, indicating that apoptosis did not develop upon banding or heart failure. (B) Incorporation of BrdU in liver cells after intraperitoneal injection. Note the strong increase in nuclear labeling of responding animals (0 vs. 1 and 2 and nonresponders vs. responders). 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 6 Expression pattern of fibronectin, collagen, and α-SMA in the liver of PTB rats. (A) The left 2 columns in the upper 3 rows show pericentral areas of control (PTB-0) and PTB stage 2 livers, whereas the right 2 columns show the corresponding periportal areas. Note the exclusive periportal expression of fibronectin in control liver and homogeneous expression in PTB stage 2 liver. Note further that collagen type 1 level increases in the walls of central veins and in the portal tracts of PTB rats but that perisinusoidal collagen content remains very low. Finally, note the homogeneous increase in perisinusoidal α-SMA content in PTB stage 2 liver. (B) Hepatic content of α-SMA, fibronectin, and collagen expressed as area fraction in sections (y-axis) as a function of the hepatic response (upper row) or cardiac response (lower row) to PTB (x-axis). The portal tracts and walls of the central veins were excluded from the measurements. Note the pronounced increase in perisinusoidal α-SMA and, to a lesser extent, fibronectin content but the absence of an increase in perisinusoidal collagen content. 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 7 Ultrastructure of the liver of PTB rats. (A) The pictures represent a control (left) and a PTB stage 2 liver (right) to show the absence of a demonstrable basement membrane and the continued presence of fenestrae in the endothelial cells and microvilli on the hepatocytes in PTB stage 2 liver. However, the space of Disse had collapsed. Bars = 1 μm. (B) The decrease in the size of the space of Disse and the hepatic microvilli in PTB livers (y-axis) as a function of the hepatic response to PTB (x-axis). The assessment was performed as described in Materials and Methods. 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions

Figure 8 Presence of ED1- and ED2-positive cells and bile ducts in the liver of PTB rats. (A) The left 2 columns in the upper rows show pericentral areas of control (PTB-0) and PTB stage 2 livers, whereas the right 2 columns show the corresponding periportal areas. Bile ducts (lower row) were visualized by their expression of CK-19. Bars = 500 μm. (B) Hepatic content of ED1- and ED2-positive cells (upper row) and bile ducts (lower row) expressed as area fraction in sections as a function of the hepatic response (left columns) or the cardiac response (right columns) to PTB. 0, 1, 2, 3, PTB stage 0, 1, 2, 3; Sh, sham-operated controls; NR, nonresponders; H, cardiac responders without heart failure; F, cardiac responders with heart failure. Gastroenterology 2004 127, 1210-1221DOI: (10.1053/j.gastro.2004.07.057) Copyright © 2004 American Gastroenterological Association Terms and Conditions