Quantitative Studies of Liver Atrophy After Portacaval Shunt in the Rat

Journal of Surgical Research 131, 225–232 (2006) doi:10.1016/j.jss.2005.11.587

Quantitative Studies of Liver Atrophy After Portacaval Shunt in the Rat Abed Almuti Zaitoun, M.D., Ph.D., F.R.C.Path,*,1 Gustav Apelqvist, M.D., Ph.D.,† Hanan Al-Mardini, BS.c, Ph.D.,‡ Trevor Gray, C.Sci,* Finn Bengtsson, M.D., Ph.D.,§ and Christopher Oswold Record, Ph.D., D.Phil., F.R.C.P.‡ *Department of Histopathology, Queen’s Medical Centre, University Hospital, Nottingham, United Kingdom; †Department of Clinical Pharmacology, Lund University, Lund, Sweden; ‡Department of Medicine, The Royal Victoria Infirmary, Newcastle-upon-Tyne United Kingdom; and §Division of Clinical Pharmacology, Department of Medicine and Care, Faculty of Health Sciences, Linköping University, University Hospital, Linköping, Sweden Submitted for publication June 23, 2005

Background. It is well known that portacaval shunting ultimately leads to a decrease in liver volume and hepatic function, but the mechanism is uncertain. The aim of the present study was to evaluate the effect of portacaval shunting (PCS) upon the morphological changes that occur in the liver in rats after port caval anastomosis. Materials and methods. Sixty-six male rats underwent either PCS (n ⴝ 35) or sham operations (n ⴝ 31). Hormone levels were determined in blood samples taken just before removal and weighing of the livers. Hematoxylin and eosin-stained sections were used for quantitative morphometric analysis. Apoptosis, mitosis, and cellular organelles also were assessed quantitatively. Results. There was a significant reduction in the liver mass together with testosterone levels in PCS rats in comparison with sham rats. The distance between presinusoidal and postsinusoidal vessels was reduced from 500 ␮m in the sham rats to 299 ␮m in the PCS rats (P ⴝ 0.000001). Within the same group, there was a significant reduction in the area of hepatocyte nuclei in zone 3 in comparison with zone 1. Electron microscopy revealed a highly significant (P ⴝ 0.0007) reduction in the membrane-bound cytoplasmic organelles of zone 3 hepatocytes in PCS rats in comparison with the sham rats. Apoptosis was increased in zone 3 in PCS rats (P ⴝ 0.00001), whereas in zone 1 of the same group, there was an associated increased in

1

To whom correspondence and reprint requests should be addressed at Department of Histopathology, University Hospital, Queen’s Medical Centre, Nottingham, NG7 2UH, UK. E-mail: [email protected].

mitosis (P ⴝ 0.000001). Overall, the degree of apoptosis was in excess of mitosis, resulting in a general loss of liver mass. Conclusion. Morphometric analysis at cellular and subcellular levels confirms the morphological findings of liver atrophy in PCS rats. The mechanism of atrophy is a complex one. Portacaval shunting leads to hepatic atrophy that, in turn, results in microcirculatory and hormonal changes that further contribute to liver cell loss in this animal model. © 2006 Elsevier Inc. All rights reserved.

Key Words: apoptosis; liver atrophy; morphometry; portacaval shunt rat; ultra-structure. INTRODUCTION

The creation of a portacaval shunt (PCS) often is accompanied by hepatic encephalopathy. The pathogenesis of this induced condition is still obscure, although it is probably multifactorial in origin [1]. We [2] and others [3] have studied the physiological and histological changes in the small intestine of the port caval shunted rat and found that a decrease in jejunal permeability was probably related to reduction in mucosal area and villi of the small intestine. In a recent study, we [4] found that PCS, in the absence of liver dysfunction, produces testicular atrophy by reductions in mitosis, maturation arrest, and increased apoptosis of the germinal epithelium of the semineferous tubules. Several authors [5–10] have studied the morphological changes of the liver in PCS operated rats. Alteration in both the cellular and nuclear morphology was observed, together with the loss of glucose-6phosphatase, mainly in the centrilobular areas of the liver [10]. In previous animal studies [11–17], several

225

0022-4804/06 $32.00 © 2006 Elsevier Inc. All rights reserved.

226

JOURNAL OF SURGICAL RESEARCH: VOL. 131, NO. 2, APRIL 2006

authors have described the ultrastructural morphometric alterations of the hepatocytes of the liver and the zona granulosa of the adrenal glands. These alterations included the reduction in the number of hepatocytes, atrophy of cytoplasmic organelles, and the decrease in the volume of nuclei with widening of the sinusoidal spaces [11–14]. In this study, we report the changes occurring in the liver of PCS and sham-operated control rats. The biochemical changes in corticosteroid levels were measured in combination with the use of several quantitative techniques at light and electron microscopy to study the morphological changes. In addition, we also quantified apoptosis in the pericentral areas and mitosis in periportal areas of the liver. MATERIALS AND METHODS Male Sprague–Dawley rats initially weighing 200-300 g were used for the studies (Möllegaard Breeding & Research Centre A/S, Skensveel, Denmark). The rats were housed in groups of three in standard animal cages and kept under standard laboratory conditions. They had free access to rat pelleted chow containing 18% crude protein (Altromin No. 1324; Chr. Peterson A/S, Ringstead, Denmark) and tap water. The Animal Ethics Committee at Lund University in Sweden approved this study. For the PCS operation, a permanent end-to-side PCS was created according to Lee and Fisher [18] with minor modifications. In brief, the rats were operated under ether anesthesia. The abdomen was opened by a midline incision, and the portal and the inferior caval vein were exposed. The portal vein was clamped and ligated at the liver hilus and subsequently divided. The portal vein stump was then anastomosed using a continuous polygylcolic acid suture (9-O Vicryl®, Ethicon, Sommerville, NJ) to an elliptical opening in the inferior caval vein just cranial to the renal veins. Peroperatively, the patency of the anastomosis was evident from the rapid return of congested blood from the intestines through the PCS and absence of intestinal cyanosis before closure of the abdomen. The abdominal incision was closed in two layers using absorbable sutures (Vicryl® 3-O, Ethicon). The operating time was approximately 30 min with a clamping time of about 10 min. The sham-operated control rats underwent a similar abdominal surgery procedure, as did the PCS rats, i.e., handling of viscera, dissection, and clamping of the portal and caval veins but without construction of the portacaval anastomosis. Total operating time and clamping of vessels was equal in the two procedures. Individual body weights were measured at the time of surgery, thereafter weekly and finally at the time of sacrifice, 5 to 7 weeks after the operation. Just before the rats were sacrificed, blood samples were taken and immediately centrifuged, frozen, and stored at ⫻20°C. Finally, the rats were sacrificed and the livers removed and weighed before fixation in formalin.

Hormone Determinations Plasma testosterone, estradiol, and luteinizing hormone (LH) hormones were measured in duplicate by radioimmunoassay using commercial kits (Bayer Group Diagnostics, Tarrytown, NY). The detection limit for testosterone was 1.735 nmol/L, for estraolial was 18.35 pmol/L, and for LH was 0.3 IU/L.

Light Microscopy Preparation Livers were fixed in 4% buffered formal saline fixative, sliced at 2-mm thickness, and embedded in paraffin wax after traditional

overnight machine processing for histological examination. Hematoxylin and eosin (H&E)-stained sections at 5-␮m thickness were used for morphological and morphometric analysis.

Ultrastructural Preparation The rat livers were excised and fixed in formal saline as described for light microscopy. Several 1-mm thick slices of liver where taken from each formalin-fixed sample and trimmed to approximately 5 ⫻ 4 ⫻ 1 mm. After washing in cacodylate buffer, the tissue was post fixed in 1% osmium tetroxide for 1 h then washed in distilled water before dehydrating in ascending grades of ethanol, followed by acetone. The tissue was then infiltrated with TAAB epoxy resin (TAAB, Berkshire, England) and polymerized overnight in 8-mm flat embedding capsules (TAAB). After trimming, 0.5-␮m thick sections were cut on a Leica ultramicrotome using glass knives. The sections were mounted onto glass slides and stained with 1% toluidine blue in 1% borax at 90°C. All of the toluidine blue–stained sections were examined under a light microscope and areas containing both a well-defined portal tract (zone 1) and central vein (zone 3) were marked. The relevant blocks were then trimmed to the selected areas and 80-nm thick sections cut on a Leica ultramicrotome and mounted onto 300 mesh copper grids before staining with uranyl acetate and lead citrate [19]. These sections were then examined in a JEOL 1010 electron microscope at 80 Kv.

Morphometric Measurements The line length measurements between portal tracts and central veins were performed using a commercially available interactive overlay measuring system (Prodit 5.2 BMA, Amsterdam The Netherlands). The procedure used a 4X objective on a Nikon Eclipse E600 microscope, and the microscopical image was captured by a video camera and displayed on the computer screen before measurements were made. At least 20 measurements from each liver were obtained. The area measurements of hepatocytes were made using the Prodit 5.2 system in zone 1 (periportal), acinus of Rappaport [20], and zone 3 (pericentral) of the liver. For each hepatocyte, total cellular area, nuclear area, and nucleolar area were measured. The outline of each hepatocyte initially was demarcated followed by the outline of the nucleus using a 40X objective. A separate measurement for nuclei and nucleoli was performed using the same methodology but using a 100X oil objective. At least 50 nuclei from each zone (total ⬎100) were measured from each liver.

Morphometric Variables The following variables were used in this study to measure the light microscopy changes: ● Area: the total area of the whole hepatocyte, nucleus, and nucleolus. If there were several nuclei or nucleoli, then those areas were summed. ● Perimeter: the perimeter of whole hepatocytes. ● Diameter: the diameter of nucleus. ● Short axis: the short axis of the nucleus. ● Long axis: the long axis of the nucleus. ● Contour: the contour profile of the nuclear envelope. A value of 1 indicates a smooth surface. ● Roundness: nuclear roundness is the aspect ratio of the nucleus. It is equal to 1 in when a circle. ● Hepatocyte area ratio: the ratio of mean area of whole hepatocytes to the mean area of their nuclei. ● Nuclear area ratio: the ratio of the mean area of nuclei to the mean area of their nucleoli. ● Area difference: the nuclear area minus nucleolar area.

ZAITOUN ET AL.: MORPHOMETRY OF LIVER ATROPHY

227

contact with a portal tract (Fig. 1A and B) and 10 were in direct contact with an associated central vein (Fig. 2A and B). A 50-point square graticule was superimposed over the digital images, and the number of points touching specific cytoplasmic organelles was determined using standard stereological techniques. The organelles were grouped into well-recognizable structures of endoplasmic reticulum, membrane-bound organelles such as mitochondria and peroxisomes and, finally, the organelle-free cytosol. The total cell cytoplasm was determined by addition of the three previous parameters. To make direct comparisons of intercellular changes between the two zones and the two sample groups, the organelle counts were used for comparison and not the cytoplasmic fraction of the organelles.

Statistical Analysis Results were expressed as means and standard error of the means (SE) or mean and the standard deviation of the mean (SD). An unpaired t test or analysis of variance was used for normally distributed data. When Bartlett’s test showed the variances in the sample to differ, the nonparametric Kruskal–Wallis (equivalent to ␹2) was used to obtain the P values. The level of significance (two-tailed) was P ⬍ 0.05. Measurements at light and electron microscopy were performed by individual operators (A.M.Z. and T.G., respectively). In both cases, the intraobserver error was assessed by performing repeated measurements on selected samples. In both cases, the intraobserver error was found to be less than P ⫽ 0.05.

FIG. 1. Electron microscopy images of hepatocytes from periportal area in PCS rat (A) and sham operated rat (B) showing minimal changes to membrane-bound organelles.

Quantification of Apoptosis All measurements were obtained using a 40X objective (diameter 450 ␮m) with 5-␮m thick H&E-stained sections. Using the Prodit image analysis system, apoptotic nuclei were counted in zone 3 (pericentral area) of each section. The mean number of apoptotic cells per unit area of the liver cell parenchyma was then computed and expressed as cells/mm2 of the liver cell parenchyma.

Quantification of Mitoses The same method of quantifying apoptosis was used for the assessment of mitosis. The mitotic index of the hepatocytes was counted per unit area in zone 1 (periportal) of each section. Only mitotic figures in which the nuclear membrane was unequivocally absent and had associated surrounding cytoplasm that was not highly eosinophilic were counted. The mitotic index was expressed as the mean number of mitotic cells/mm2.

Morphometric Measurements at Ultrastructural Level A total of 160 digital images of individual hepatocytes at a magnification of 8000X were used for ultrastructural analysis. Twenty digital electron micrographs from each sample were taken using a Kodak 1.6i Megaplus camera and SIS image analysis software (Soft Imaging System, Germany) attached to a Jeol 1010 transmission electron microscope (Jeol, Japan). Ten were of hepatocytes in direct

FIG. 2. Electron microscopy images of hepatocytes from pericentral zone in PCS rat (A) and sham operated rat (B) showing reduction in the number of membrane-bound organelles.

228

JOURNAL OF SURGICAL RESEARCH: VOL. 131, NO. 2, APRIL 2006

RESULTS

In confirmation of previous work [21], there was a significant reduction of the liver weight in PCS operated rats (5.0 ⫾ 1.1 g SD) in comparison to the shamoperated group (13.7 ⫾ 1.6 g; P ⬍ 0.000001). Plasma Hormones

Plasma hormone levels are shown in Table 1. Testosterone was significantly reduced from 10.1 nmol/L in the sham control group to 2.5 nmol/L in the PCS group (P ⫽ 0.002), whereas significant increases in the plasma estradiol level were measured between the PCS and sham-operated rats, (P ⫽ 0.004). No significant changes were observed between the two groups for plasma LH. Morphological Changes

The area between zone 1 and 3 of the PCS-operated rats were smaller than their respective sham controloperated rats (Fig. 3A and B). The liver cell parenchyma in zone 3 of PCS rats showed a large number of apoptotic cells (Fig. 4A), whereas the liver cell parenchyma in zone 1 of PCS showed some mitotic cells (Fig. 4B) that could not be detected in the sham control rats. Morphometric Findings

Changes in Liver Architecture The mean distance between central vein and portal tract area was reduced from 499.68 ⫾ 58.46 ␮m (SD) to 298.76 ⫾ 42.30 ␮m in PCS rats (P ⫽ 0.000001). Changes in the Total Area of Hepatocytes and Cytoplasm The morphometric findings are shown in Table 2. The total area of hepatocytes, perimeter, cytoplasm, and total cell/nuclear ratio all were significantly reduced in the PCS rats in comparison with the control sham group in both zones 1 and 3. Morphometry also showed highly significant differences between zone 1 and zone 3 in both the PCS and sham groups. TABLE 1 The Mean Values (nmol/L) ⴞ SD of the Means of LH, Testosterone, and Estradiol, in PCS (n ⴝ 10) and ShamOperated Rats (n ⴝ 12) PCS rats LH (IU/L) Testosterone (nmol/L) Estradiol (pmol/L)

0.7 ⫾ 0.2

Sham rats

0.7 ⫾ 0.2 P ⫽ 0.8 (ns) 8.64 ⫾ 12.32 35.12 ⫾ 25.16 P ⫽ 0.002 286.26 ⫾ 66.06 190.84 ⫾ 22.02 P ⫽ 0.0004

FIG. 3. H&R-stained section of liver from PCS rat (A) and sham operated rat (B) showing portal tract (PT) and central vein (CV). H&E ⫻4. (Color version of figure is available online.)

Changes in the Nuclei Table 3 shows the changes in the nuclei in both groups of animals. Nuclear area, diameter, short axis, and long axis were significantly reduced in PCS in comparison with sham operated rats in both zones. The changes in nuclear contour and roundness were not significant. Changes in Nucleolar Areas The results of nucleolar areas are shown in Table 4. Within zone 1, the nucleolar areas did not alter significantly between the two groups, but there was a significant change (P ⫽ 0.043) for zone 3. There was also a significant change in nucleoli area between the two zones in each group of rats. The area difference between nuclei and nucleoli were significantly reduced between the zones and within each zone between the

229

ZAITOUN ET AL.: MORPHOMETRY OF LIVER ATROPHY

Ultrastructural Results

Table 5 shows the mean values of the stereological counts for endoplasmic reticulum, membrane-bound organelles, free cytosol, and total cytoplasm. The statistical comparison showed that there was a very significant decrease in counts of membrane-bound organelles in zone 3 hepatocytes of PCS operated rats when compared with the sham-operated rats (P ⫽ 0.0007) as can be seen in Fig. 2, A and B. While in zone 1, the ultrastructural changes were minimal, as can be seen in Fig. 1A and B. Other measurements indicated similar changes to that seen at light microscopy, although these changes were not significant because of the relatively small samples used in this part of the study. Apoptosis and Mitosis

Table 6 shows the mean values of apoptosis in zone 3 from both groups. Apoptosis was increased approximately 8-fold in the PCS group in comparison to the sham-operated rats (P ⫽ 0.00001), whereas the number of mitoses in zone 1 was increased 15-fold (P ⫽ 0.00001). DISCUSSION

FIG. 4. Pericentral zone (zone 3) in PCS rat (A) showing apoptosis (arrowed). H&E ⫻20 and periportal (B) showing mitosis. H&E ⫻10. (Color version of figure is available online.)

two groups of rats. The nuclear to nucleolar ratio was only significantly reduced in zone 3 between the two groups (P ⫽ 0.041) and not in zone 1.

In the present study, we also assessed the histological changes at light and electron microscopy level using quantitative techniques, and we have focused our interest on the effect of PCS on zones 1 and 3. The effect of PCS on the wet weight of the liver was striking and confirms the results of previous work [21]. In addition, there was a significant reduction in the distance between presinusoidal (portal tract) and postsinusoidal (central vein) vessels, which could be explained by atrophy of hepatocytes throughout the three zones of the acinus of Rappaport. It seems logical that the re-

TABLE 2 Quantitative Parameters of Hepatocyte Area, Perimeter, Cytoplasmic Area, and Cell/Nuclear Ratio in Hepatocytes of Zone 1 and Zone 3 in PCS (n ⴝ 29) and Sham Rats (n ⴝ 31)

Periportal (zone 1)

Pericentral (zone 3)

P value between zone 1 and zone 3

Hepatocyte area

Perimeter

Cytoplasmic area

Cell/nuclear ratio

152.25 ⫾ 32.20 (202.69 ⫾ 35.99) P ⫽ 0.000007 103.29 ⫾ 9.59 (150.34 ⫾ 32.71) P ⫽ 0.000001 P ⫽ 0.00001 P ⫽ 0.000001

45.22 ⫾ 4.89 (52.41 ⫾ 4.63) P ⫽ 0.00001 37.17 ⫾ 1.66 (44.89 ⫾ 4.85) P ⫽ 0.000001 P ⫽ 0.000001 P ⫽ 0.000001

113.69 ⫾ 25.99 (155.00 ⫾ 29.00) P ⫽ 0.000008 74.30 ⫾ 7.27 (112.89 ⫾ 26.24) P ⫽ 0.000001 P ⫽ 0.000003 P ⫽ 0.000006

4.02 ⫾ 0.34 (4.42 ⫾ 0.43) P ⫽ 0.001 3.65 ⫾ 0.25 (4.10 ⫾ 0.34) P ⫽ 0.00008 P ⫽ 0.0005

Results are expressed as mean ⫾ SE of the mean. Area is measured in ␮m 2 and perimeter in ␮m. Results for the sham-operated rats are shown in parentheses.

230

JOURNAL OF SURGICAL RESEARCH: VOL. 131, NO. 2, APRIL 2006

TABLE 3 Quantitative Parameters of Hepatocyte Nuclei of in Zones 1 and 3 in PCS (n ⴝ 29) and Sham Rats (n ⴝ 31)

Periportal (zone 1)

Pericentral (zone 3)

P values between zones, 1 and 3

Area

Diameter

Short axis

Long axis

Contour

Roundness

55.52 ⫾ 6.06 (59.48 ⫾ 5.98) P ⫽ 0.014 45.31 ⫾ 6.49 (50.22 ⫾ 4.53) P ⫽ 0.0014 P ⫽ 0.000001 (P ⫽ 0.000001)

8.36 ⫾ 0.42 (8.66 ⫾ 0.42) P ⫽ 0.013 7.53 ⫾ 0.55 (7.94 ⫾ 0.36) P ⫽ 0.0013 P ⫽ 0.000001 (P ⫽ 0.000001)

7.20 ⫾ 0.39 (7.47 ⫾ 0.41) P ⫽ 0.011 6.54 ⫾ 0.49 (6.89 ⫾ 0.33) P ⫽ 0.003 P ⫽ 0.012 (P ⫽ 0.000001)

9.99 ⫾ 0.62 (10.32 ⫾ 0.52) P ⫽ 0.033 8.94 ⫾ 0.68 (9.43 ⫾ 0.37) P ⫽ 0.002 P ⫽ 0.000001 (P ⫽ 0.000001)

1.0359 ⫾ 0.009 (1.0343 ⫾ 0.013) P ⫽ 0.571 1.0404 ⫾ 0.015 (1.0348 ⫾ 0.011) P ⫽ 0.113 P ⫽ 0.194 (P ⫽ 0.859)

1.0177 ⫾ 0.005 (1.0168 ⫾ 0.006) P ⫽ 0.574 1.0195 ⫾ 0.006 (1.0171 ⫾ 0.005) P ⫽ 0.121 P ⫽ 0.213 (P ⫽ 0.856)

Results are expressed as mean ⫾ SE of the mean. Area is measured in ␮m 2 and length in ␮m. Results from sham-operated rats are shown in parentheses.

sulting closer proximity of central to portal areas will lead to an alteration in liver function, as suggested by Dubuisson [14]. We have studied the morphometric changes seen in zones 1 and 3 because hepatocytes from both zones are known to have different activities in relation to the orderly diffusion gradient of metabolites throughout the liver acinus [22–24]. Significant differences of hepatocyte morphology shown by this study are consistent with the different metabolic activities of hepatocytes in the two zones. Furthermore, the results from electron microscopy have supported these findings by showing a highly significant reduction in membrane bound organelles in zone 3 when compared with zone 1 hepatocytes both in control and the PCS groups. Because zone 3 contains mitochondria and peroxisomes, it can be assumed that there also is an associated reduction in metabolic activity within these hepatocytes. These findings confirm previous reports using light microscopy [5, 6, 8, 9] and electronic microscopy [11–17] in this animal model. We also found a reduction in the nuclear area of hepatocytes in zone 1 and zone 3 comparable with results reported by other authors [16, 30, 31]. TABLE 4 Quantitative Parameters of Nucleolar Areas, Nuclear Minus Nucleolar Area (N 1 ⴚ N 2), and Nuclear/ Nucleolar Ratio From Hepatocytes in Zone 1 and 3 in PCS (n ⴝ 29) and Sham Rats (n ⴝ 31)

Periportal area (zone 1) Pericentral (zone 3) P value between zone 1 and 3

Area

Area difference (N 1 ⫺ N 2)

Nuclear/ nucleolar ratio

2.93 ⫾ 0.50 (3.13 ⫾ 0.45) P ⫽ 0.110 2.54 ⫾ 0.5 (2.79 ⫾ 0.42) P ⫽ 0.043 P ⫽ 0.004 (P ⫽ 0.0027)

51.35 ⫾ 5.55 (55.2 ⫾ 5.56) P ⫽ 0.009 42.05 ⫾ 6.05 (46.80 ⫾ 4.2) P ⫽ 0.001 P ⫽ 0.000001 (P ⫽ 0.000001)

16.27 ⫾ 2.56 (17.14 ⫾ 1.91) P ⫽ 0.144 16.60 ⫾ 2.84 (18.04 ⫾ 2.45) P ⫽ 0.041 P ⫽ 0.641 (P ⫽ 0.111)

Results are expressed as mean ⫾ SE of the mean. Area is measured in ␮m 2. Results from sham-operated rats are shown in parentheses.

Apoptosis represents a series of morphological changes showed by controlled cell death. There are many methods for assessing apoptosis. The gold standard for assessing apoptosis is still regarded to be electron microscopy, although the easiest and most common method for a histopathologist is to identify apoptosis on morphology grounds using light microscopy. However, the limitation of this method is the omission of the early stage of apoptosis as identified by other techniques, such as terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate-digoxigenin nick end labeling (i.e., TUNEL). It is recognized that these early transformations are reversible and do not represent true cell death. At present, morphological assessment is the only irrefutable mean of identifying apoptosis. For this reason, we have decided to assess apoptosis using morphology in our study. In a previous study [4], we have shown that testicular atrophy in PCS-operated rats resulted in a reduction in cell production, mainly by apoptosis. The liver differs from the testes because hepatocytes are in the Go phase of the cell cycle (resting or post G1 phase), unlike the germinal epithelium of the testis that is constantly within an active cell cycle phase. However, the volume of germinal epithelium usually remains in a steady state, balanced by the formation of spermatozoa, mitosis, and cell loss. The mechanisms of liver atrophy in PCS-operated rats may be more complex. In the present study, the effect of increased apoptosis was offset by increased mitosis and thus cannot explain the severity of the hepatic atrophy observed. Our findings are therefore consistent with hepatic atrophy being a direct effect of a decrease in hepatic blood flow after the formation of the shunt. This effect is supported by the resultant measured hormonal changes and the increased estradiol demonstrated in this study is comparable to similar findings in man [25–28] and the rat [29]. The accumulation of estradiol in liver disease is a consequence of reduced hepatic clearance of androgens and sex hormones [26] as well as an increase in the nonhepatic peripheral aromatisation conversion of weak androgen to estrogens [29]. These hormonal al-

231

ZAITOUN ET AL.: MORPHOMETRY OF LIVER ATROPHY

TABLE 5 Ultrastructural Quantitative Parameters of Cytoplasmic Organelles From Hepatocytes in Zone 1 and 3 From PCS (n ⴝ 4) and Sham-operated Rats (n ⴝ 4) ER

MBO

C

TC

Group

Zone1 (PT)

Zone3 (CV)

Zone1 (PT)

Zone3 (CV)

Zone1 (PT)

Zone3 (CV)

Zone1 (PT)

Zone3 (CV)

PCS (n ⫽ 80) Sham (n ⫽ 80) Comparison of sham vs. PCS, P Value

3.5 ⫾ 0.33 2.5 ⫾ 0.3

2.75 ⫾ 0.24 2.4 ⫾ 0.27

6.73 ⫾ 0.38 7.73 ⫾ 0.48

4.97 ⫾ 0.46 7.55 ⫾ 0.57

13.1 ⫾ 0.65 14.33 ⫾ 0.94

13.1 ⫾ 0.99 14.15 ⫾ 1

22.8 ⫾ 0.91 24.5 ⫾ 1.33

20.8 ⫾ 1.23 24.1 ⫾ 1.57

0.22

0.32

0.1

0.0007

0.28

0.45

0.3

0.1

Results of stereology using a 50-point grid are expressed as mean counts ⫾ SE of the mean. ER, endoplasmic reticulum; MBO, membrane-bound organelles; C, organelle-free cytosol; TC, total cytoplasm.

ternations are known to lead to DNA fragmentation and apoptosis [32]. The increase in apoptosis in zone 3 therefore can be explained by the alteration of the hormone levels in PCS rats. The hormonal induced increase in hepatocyte cell loss by apoptosis presumably induced an alteration to the steady state in the liver that in turn led to the induction of hepatocyte mitosis observed in zone 1. The reductions in the cytoplasmic and nuclear areas are probably a consequence of hepatic dysfunction. With the advance of liver atrophy, we have found the distance between presinusoidal and postsinusoidal vessels is decreased. Dubuisson et al. [12] found that the sinusoid was slightly enlarged and that the Kupffer cell volume density increased after postcaval shunting whereas changes in the volume of endothelial cells and perisinusoidal cells [14] resulted in thickening of the sinusoidal barrier. This may lead to an impaired exchange between the sinusoidal lumen and Disse space. Although the morphometry of sinusoidal spaces has not been studied by us, the finding [5] of these authors [12, 14] strongly indicates that functional abnormalities may be a consequence of liver cell atrophy. These findings are highly relevant to path-physiological alterations observed in patients with cirrhosis because they confirm that such changes do not require the presence of direct damage to hepatocytes by alcohol, hepatic viruses or other noxious agents. Our study, by using morphological, biochemical, and

quantitative techniques at architectural, cellular, and subcellular levels, has confirmed previous morphology and electron microscopy work in PCS rats. The findings from this study also are consistent with hepatic atrophy being a direct effect of a decrease in hepatic blood flow after the formation of the shunt in this animal model. REFERENCES 1. 2.

3.

4.

5.

6.

Castaing D, Beaubernard C, Ariogul O, Gigou M, Franco D, Bismuth H. Liver atrophy and encephalopathy after portacaval shunt in the rat. Eur Surg Res 1982;14:192.

7.

Di Cataldo A, Scilletta B, Li Destri G, et al. Morphologic and endocrine changes in the male reproductive system after portacaval anastomosis in rats. Experimental research. Minerva Chir 1993;48:253.

8.

Thompson JS, Porter KA, Hayashida N, et al. Morphologic and biochemical changes in dogs after portacaval shunt plus bile fistula or ileal bypass: failure of bile fistula or ileal bypass to prevent hepatocyte atrophy. Hepatology 1983;3:581.

9.

Dubuisson L, Vonnahme FJ, Stzark F, Biolac-Sage P, Balabaud C. Hyperplastic foci in the atrophic liver of rats after portacaval anastomosis. Liver 1985;5:21.

TABLE 6 Mean Number of Mitosis/mm 2 ⴞ SE or the Mean in Periportal Area and Apoptosis in Pericentral Area in PCS and Sham Rats Group PCS Sham

Record CO. Neurochemistry of hepatic encephalopathy. Gut 1991;32:1361. Pantzar N, Bergqvist PBG, Bugge M, et al. Small intestinal absorption of polyethylene glycol 400 to 1000 in the portacaval shunted rat. Hepatology 1995;21:1167. Hess F, Jerusalem C, Willemen A. Altered portal pressure secondary to portacaval and portasystemic shunts in the rat: the effect on liver function and intestinal integrity. Eur Surg Res 1982;14:286. Zaitoun AM, Apelquist G, Wikell C, Al-Mardini H, Bengtsson F, Record CO. Quantitative studies of testicular atrophy following portacaval shunt in rats. Hepatology 1998;28:1461. Lanier VC Jr, Buchanan RD, Foster JH. Hepatic morphologic changes following end-to-side portacaval shunt in dogs. Am Surg 1968;34:185.

Mitosis (zone 3 periportal), mean

Apoptosis (zone 1 pericentral), mean

10.

1.16 ⫾ 0.65 (0.2–2.6) 0.08 ⫾ 0.17 (0.0–0.7) P ⫽ 0.000001

2.40 ⫾ 1.27 (0.7–6.8) 0.3 ⫾ 0.34 (0.0–1.3) P ⫽ 0.000001

Evarts RP, Raab M, Marsden E, Thorgeirsson SS. Histochemical changes in livers from portacaval-shunted rats. J Natl Cancer Inst 1986;76:731.

11.

Dubuisson L, Bioulac P, Saric J, Balabaud C. Hepatocyte ultrastructure in rats with portacaval shunt. A morphometric study of acinar zones. Dig. Dis Sci 1982;27:1003.

12.

Dubuisson L, Bioulac-Sage P, Bedin C, Balabaud C. Sinusoidal

Minimum and maximum values are shown in parentheses.

232

JOURNAL OF SURGICAL RESEARCH: VOL. 131, NO. 2, APRIL 2006 cells in rats with portacaval shunt: a morphometric study. J. Submicrosc Cytol 1982;14:453.

13.

Dubuisson L, Bioulac-Sage P, Bedin C, Balabaud C. Hepatocyte ultrastructure in the rat after long term portacaval anastomosis: a morphometric study. J. Submicrosc Cytol 1984;16:283.

23.

24.

14.

Dubuisson L, Sztark F, Bedin C, Bioulac-Sage P, Balabaud C. The sinusoidal barrier in rats with portacaval anastomosis: a morphometric study. Virchows Arch A Pathol Anat Histopathol 1985;407:347.

15.

Lombardo DL, Cortesini C. The zona glomerulosa of the adrenal gland in rats with portacaval shunt. An ultrastructural and morphometric study. Int J Exp Pathol 1990;71:639.

16.

Ravelingien N, Kiss R, Pector JC. Hepatic morphonuclear alterations following portacaval shunt in the rat. J. Surg Res 1996;61:491.

27.

17.

Terlunen E, Altenahr E, Becker K, Ossenberg FW. Liver atrophy following portacaval shunt in normal rats—a morphometric study. Res Exp Med (Berl) 1977;170:133.

28.

18.

Lee SH, Fisher B. Portacaval shunt in the rat. Surgery 1961; 50:668.

29.

19.

Robinson G, Gray T. Electron microscopy 2: practical procedures. In: JD, Bancroft A, eds. Stevens Theory and practice of histological techniques, 4th ed. Edinburgh: Churchill Livingstone, 1996:585– 626.

20.

Rappaport AM. The structural and functional unit in the human liver (liver acinus) Anat Rec 1958;130:673.

21.

Cooper AJ. Ammonia metabolism in normal and portacavalshunted rats. Adv Exp Med Biol 1990;272:23.

22.

Galvao-Teles A, Anderson DC, Burke CW, Marshall JC. Biologically active androgens and estradiol in men with chronic liver disease. Lancet 1973;1:173.

25.

26.

30.

31.

32.

Kent JR, Scaramuzzi RJ, Lauwers W, et al. Plasma testosterone, estradiol and gonadotrophins in hepatic insufficiency. Gastroenterology 1973;64:111. Van Thiel DH, Gavaler JS, Lester R, Goodman MD. Alcohol-induced testicular atrophy: An experimental model for hypogonadism occurring in chronic alcoholic men. Gastroenterology 1975;69:326. Van Thiel DH, Gavaler JS, Cobb CF, McClain CJ. An evaluation of the respective roles of portasystemic shunting and portal hypertension in rats upon the production of gonadal dysfunction in cirrhosis. Gastroenterology 1983;85:154. Van Thiel DH, Gavaler JS, Slone FL, Cobb CF, Smith WI, Bron KM, Lester R. Is feminzation in alcoholic men due in part to portal hypertension: a rat model. Gastroenterology 1980;78:81. Hawkins PA, DeJoseph MR, Vina JR, Hawkins RA. Comparison of the metabolic disturbances caused by end-to-side and side-to-side portacaval shunts. J. Appl Physiol 1996;80:885. de Boer JE, Oostenbroek RJ, van Dongen JJ, Janssen MA, Soeters PB. Sequential metabolic characteristics following portacaval shunt in rats. Eur Surg Res 1986;18:96. Lautard M, Lesourd M, Trivin F, Testas P. Changes in the levels of hepatocyte mitochondrial cytochromes after portacaval shunt in the rat. Ann Chir 1984;38:315. Doyle D, Ryan CJ, Benjamin IS, Blumgart LH. Changes in the nuclei of astrocytes following portacaval shunting and portacaval transposition in the rat. Br J Exp Pathol 1978;59:461. Garcia-Moreno L, Arias JL, Lorente L, Aller MA, Vallejo-Seco G, Arias J. Behaviour of nucleolar organizer regions in the different hepatic lobes after end-to-side portacaval shunt in the rat. J Med 1994;25:341. Distelhorst CW. Glucocorticoid-induced apoptosis. In: SC, ed. Kaufmann Apoptosis: pharmacological implications and therapeutic opportunities. New York: Academic Press, 1997:247– 270.