Abstract
We investigated the relationship between DNA degradation and lysosome activity (loss of lysosomal integrity) in necrotic cell death induced by carbon tetrachloride (CCl4) and dimethylnitrosamine (DMN): coagulation necrosis and hemorrhagic necrosis, respectively. TdT-mediated dUTP–biotin nick end-labeling (TUNEL) and enzyme histochemistry for acid phosphatase were performed in both models and results were analyzed by light microscopy, electron microscopy, and confocal laser scanning microscopy (CLSM). In the CCl4-injected liver, TUNEL staining was closely associated with release of lysosomal enzymes into the cytoplasm, and intranuclear deposition of lysosomal enzymes took place at an early stage of subcellular damage. In the DMN-injected liver, TUNEL-positive nuclei tended to have well-preserved lysosomes and centrally localized TUNEL signals. It was assumed that acute hepatocellular damage in the CCl4-injected liver would be characterized by necrotic cell death with lysosome activation and that damage in the DMN-injected liver would be necrotic cell death without lysosome activation. In the DMN-injected liver, DNA degradation may be selectively induced in the nuclear center, in which heterochromatin (including inactive chromatin) is believed to be a target. We concluded that necrotic cell death, i.e., DNA degradation, would be at least divided into two types, with/without association with lysosome activation, represented by necrotic cell death in the CCl4-injected liver and that in the DMN-injected liver.
N
Both carbon tetrachloride (CCl4) and dimethylnitrosamine (DMN) produce centrilobular necrosis of fairly rapid onset (Butler 1968; Reynolds and Ree 1971; Pritchard et al. 1987; Pritchard and Butler 1989; Lertprasertsuke et al. 1991; Ray et al. 1992; Fukuda et al. 1993; Hashimoto et al. 1995; Shikata et al. 1996; Oyaizu et al. 1997). These models have been investigated from a variety of viewpoints, such as morphological alterations and hematological events (Butler 1968; Hirata et al. 1989; Pritchard et al. 1987; Pritchard and Butler 1989). Although it has been believed for some time that the rat liver injected with CCl4 is a model of necrotic cell death (Butler 1968; Reynolds and Ree 1971; Pritchard et al. 1987; Lertprasertsuke et al. 1991; Fukuda et al. 1993; Hashimoto et al. 1995), recently Shi et al. (1998) reported that this agent also induces apoptosis of hepatocytes. DMN is known to induce both necrotic and apoptotic cell death (Butler 1968; Pritchard et al. 1987; Pritchard and Butler 1989; Ray et al. 1992; Shikata et al. 1996; Oyaizu et al. 1997). Lipid peroxidation is believed to play a key role in the membrane-mediated chromosome damage caused by CCl4 (Lertprasertsuke et al. 1991). In contrast, DNA damage has been suggested to contribute to DMN-induced necrosis (Kamendulis and Corcoran 1994) and may be an important action of this alkylating agent (Ray et al. 1992).
TdT-mediated dUTP–biotin nick end-labeling (TUNEL) was initially designed to detect apoptosis, a form of programmed cell death (PCD), by Gavrieli et al. (1992), and the method opened the way for various studies (Yasuda et al. 1995; Umemura et al. 1996). In addition to TUNEL, enzyme histochemistry using acid phosphatase (ACPase), one of the representative enzymes included in the lysosomes, is of use in the observation of subcellular damage (Gomori 1952; Barka and Anderson 1962). The lysosome was designated as a “suicide bag” by de Duve (1959), and the loss of lysosomal membrane integrity is usually followed by chromatin destruction and cell lysis.
In this study, the distinctive TUNEL pattern prompted us to explore the possible differences between the microenvironments of DNA degradation in CCl4- and DMN-induced liver injury. To determine the morphological changes, the pattern of DNA degradation detected by TUNEL (Gavrieli et al. 1992) was visualized by confocal laser scanning microscopy (CLSM) (Itoh et al. 1997; Yasuda et al. 1997) and analyzed by electron microscopy. To determine whether lysosomes were activated in cell death, the deposition of ACPase was observed by light microscopy, CLSM, and electron microscopy. Here we discuss the differences between two types of necrotic cell death related to TUNEL staining pattern and lysosome activation.
Materials and Methods
Animals and Tissue Sampling
Male Wistar rats aged 10 weeks and weighing approximately 300 g, purchased from Imamichi Institute for Animal Reproduction (Tokyo, Japan), were used. A total of 30 rats were injected IP with either 50% CCl4 in corn oil (0.3 ml/100 g/bw) or 1% DMN in 0.9% saline (0.3 ml/100 g/bw). At intervals of 10, 15, 20, 25, or 30 hr after injection, three rats from each group were anesthetized with diethyl ether and then perfused via a canula inserted into the aorta with 0.03 M PIPES buffer, followed immediately by fixation with 4% paraformaldehyde (PFA) in PIPES for 5 min, after which the livers were excised. Two control rats were injected IP with PIPES alone. Slices of the livers were immediately microwaved at 35C for 45 sec (microwave processor H2500; Energy Beam Sciences, Agawam, MA).
Tissue blocks fixed with 4% PFA for 12 hr were embedded in paraffin for H&E staining and Methods I and VI (see below) and processed through a graded series of sucrose dissolved in 0.01 M PBS for 12 hr, followed by rapid freezing in OCT compound (Sakura Fine Technical; Tokyo, Japan) and storage at −80C for Methods II, III, IV, and VII (see below).
Method I. TUNEL
Paraffin blocks were cut into 4-μm-thick sections and were adhered to 3-aminopropyltrimethoxysilane coated-glass slides (Superfrost S8443; Matsunami Glass, Osaka, Japan). After deparaffinization, sections were treated with proteinase K (Sigma Chemical; St Louis, MO) in PBS (20 μg/ml) for 15 min at room temperature (RT). After washing sections with PBS, endogenous peroxidase activity was blocked using 0.3% H2O2 in methanol for 30 min at RT. TdT and biotin-16-dUTP were purchased from Boehringer Mannheim Biochemicals (Mannheim, Germany). The following TUNEL procedure was described previously in detail (Gavrieli et al. 1992). Sections were stained with 0.025% 3–3′-diaminobenzidine tetrahydrochloride (DAB) in Tris-HCl buffer and counterstained with 5% methyl green, which was processed with purification by chloroform. As a negative control, sections were incubated in the absence of TdT.
Method II. Acid Phosphatase Staining
Frozen blocks were cut into 10-μm-thick sections and stained for ACPase activity. The substrates used were 1.0 mM naphthol AS-BI phosphate sodium salt according to the method of Barka and Anderson (1962) and 4.1 mM citidine-5′-monophosphate according to that of Gomori (1952).
Method III. Double Staining with TUNEL and ACPase
To demonstrate ACPase activity in the hepatocytes with TUNEL-positive nuclei, 10-μm-thick frozen sections were stained for ACPase by the method of Barka and Anderson (1962) immediately after TUNEL staining. Sections were counterstained with 5% methyl green.
Method IV. CLSM for Detection of ACPase Staining
Frozen blocks were cut into 20-μm-thick sections and stained for ACPase activity by the method of Gomori (1952), followed by counterstaining with 5% methyl green. For subcellular analysis, the LSM-410 (objective lens Plan-apochromat, ×64 oil, NA 1.4; Carl Zeiss, Jena, Germany) was used in the reflection mode (confocality 460 nm) or the fluorescence mode (confocality 755 nm) with LSM software version 3.92. An argon ion laser was used at 488 nm for reflection mode and He–Ne laser at 543 nm for fluorescence mode. The LSM-410 was combined with image analysis (Interactive Build Analysis System; Carl Zeiss). Reflection mode was used to detect signals of lead for ACPase staining, and fluorescence mode was used to detect those of methyl green for TUNEL.
Method V. Electron Microscopy for Detection of ACPase Staining
After the process with the microwave described above, small pieces of the liver were fixed with 4% PFA in PIPES at 4C for 4 hr, followed by fixation with 1% glutaraldehyde in PIPES. After washing with 10% sucrose in Tris-malelate buffer (TMB), these liver tissues were embedded in 4% agarose in TMB. Tissue blocks were cut into 20-μm-thick sections using a Tissue Sectioner (Sorvall; Newton Connecticut) and stained for ACPase activity according to the method of Gomori (1952). Sections were treated with 1% OsO4 in cacodylate buffer for 1 hr, routinely dehydrated through a series of graded ethanol and acetone, and embedded in Epon 812. Semithin sections of the liver cut into 1-μm-thick pieces were stained with toluidine blue for light microscopy and trimmed for electron microscopy. Ultrathin sections were stained with lead nitrate and examined under a transmission electron microscope (JEM-1200EX; Jeol, Tokyo, Japan).
(
Method VI. CLSM for Detection of TUNEL Staining
TUNEL-stained paraffin sections were used for subcellular observation by CLSM. To enhance the TUNEL signals visualized by DAB, which were identified in reflection mode, sections were reacted with 0.2% OsO4 in cacodylate buffer for 1 hr.
Method VII. Electron Microscopy for Detection of TUNEL Staining
Ten-μm-thick frozen sections were processed by TUNEL staining, trimmed under light microscopic observation, and embedded in Epon 812. Ultrathin sections were stained with lead nitrate and uranyl acetate and then examined under a transmission electron microscope, JEM-1200EX.
Results
Histology
The livers of rats injected with CCl4 or DMN and those of control rats were histologically examined using routinely stained hematoxylin/eosin sections. After CCl4 injection, by 15 hr fatty metamorphosis of hepatocytes occurred in the centrilobular zone and changes in coagulation necrosis became more intense from 20 to 30 hr. Damaged hepatocytes developed ballooning or eosinophilic changes of the cytoplasm and exhibited coarse chromatin granules or lysed nuclei (Figure 1a). Most of the damaged hepatocytes were swollen. No typical apoptotic bodies were identified in these damaged hepatocytes at any time after injection. In DMN-injected rats, by 10 to 15 hr little or no significant histological alterations of hepatocytes were observed. At 20 hr, spotty foci showing mild hemorrhage with sinusoidal dilatation were observed, although these were considerably restricted. By 25 to 30 hr, damaged foci of hepatocytes were expanded, with distinct hemorrhage. These foci involved not only the centrilobular zone but also occasionally the intermediate zone, accompanied by mild to moderate inflammatory cell infiltration (Figure 1c). The nuclei were shrunken and condensed compared with those seen in the CCl4-injected liver. However, no apparent apop- totic features were identified.
TUNEL
Hepatocytes of control rats showed no positive TUNEL staining. In the CCl4-injected liver, at 10 hr no hepatocytes were stained with TUNEL, and at 15 hr TUNEL-positive hepatocytes appearing in the centrilobular zone were fairly sparse. Many TUNEL-positive hepatocytes were scattered in multiple foci by 20 hr (Figure 1b), in which hepatocytes with no positive TUNEL staining were intermingled with TUNEL-positive hepatocytes and revealed ballooning or lysed nuclei. At 25–30 hr, TUNEL-positive nuclei were markedly reduced in number. No TUNEL staining was noted in hepatocytes in the DMN-injected liver by 10 hr, and a few hepatocytes showed positive TUNEL staining by 15–20 hr. At 25–30 hr, there were multiple foci containing many TUNEL-positive nuclei (Figure 1d). No typical apoptotic bodies were identified. The major difference in the TUNEL staining patterns between both models was that TUNEL-positive nuclei were dispersed in the former but concentrated in the latter. Therefore, the number of TUNEL-positive hepatocytes was larger in the CCl4- than in the DMN-injected liver. These nuclei were enlarged in the former and shrunken in the latter.
ACPase Staining
Control livers showed a prominent granular pattern of ACPase distribution in the cytoplasm of hepatocytes along the capillary bile ducts (Figure 1e). In the CCl4-injected liver, by 20 hr damaged hepatocytes in the centrilobular zone showed diffuse ACPase staining in the cytoplasm (Figure 1f). From 25 to 30 hr the ACPase staining became more diffuse, to the extent that the entire cytoplasm was stained diffusely, although faint granularity was noted in some instances. In the DMN-injected liver, at each time point after injection the granularity of the ACPase staining still remained in the cytoplasm of damaged hepatocytes (Figure 1h).
(
Double Staining with TUNEL and ACPase
Many hepatocytes with TUNEL-positive nuclei showed a diffuse ACPase release pattern in the CCl4-injected liver (Figure 1g). In the DMN-injected liver, the granularity of ACPase staining was preserved in TUNEL-positive nuclei (Figure 1i).
CLSM to Determine ACPase Staining
On three-dimensional subcellular observations in orthogonal sections, a monotonous distribution of methyl green staining was visualized throughout all nuclei, and ACPase staining with a granular pattern was found mainly in the periphery of the cytoplasm in the control livers (Figure 2a). In the CCl4-injected liver at 20 hr, the nuclei showed focal defects of methyl green staining, and ACPase staining was also noted at the rim of the nuclei, indicating that ACPase staining was overlaid on the nuclei (Figure 2b). At 25–30 hr, intranuclear staining of ACPase became more intense in parallel with expansion of the defective area of methyl green staining (Figure 2c). In the DMN-injected liver throughout the period examined, the granular pattern of ACPase staining was found in the cytoplasm mainly along the cell membrane, with no significant defects in methyl green staining in the nuclei (Figure 2d).
Electron Microscopy to Determine ACPase Staining
In the control liver, ACPase staining showed that a few lysosomal particles developed mainly in the periphery of the cytoplasm near the capillary bile duct (Figure 2e). In the CCl4-injected liver at 20 hr, the release of ACPase took place in the cytoplasm, and faint ACPase staining was also observed in the nuclear periphery (Figure 2f), although no significant damage was noted in the nuclear membrane. In the DMN-injected liver, lysosomal particles remained almost completely intact, without release of ACPase into the cytoplasm or the nuclei (Figure 2g).
CLSM to Determine TUNEL Staining
In single-plane CLSM images, TUNEL signals were diffusely distributed in the nuclei and methyl green staining was somewhat multifocally deposited in the nuclei in the CCl4-injected liver (Figure 3a). The red and green signals often overlapped, resulting in the deposition of yellow signals. In the DMN-injected liver, TUNEL signals were largely confined to the central portion of the nuclei and, in contrast, methyl green staining was observed in the nuclear periphery in a ring-like pattern (Figure 3b).
Electron Microscopy to Determine TUNEL Staining
In the CCl4-injected liver, TUNEL signals were irregularly distributed in the nuclei, and most signals tended to be localized along the cell membrane (Figure 3c). In the DMN-injected liver, TUNEL signals were present predominantly in the nuclear central portion (Figure 3d). These features corresponded with those observed by CLSM.
Discussion
The experimental models of acute liver injury, i.e. coagulation necrosis and hemorrhagic necrosis, caused by the injection of CCl4 and DMN, respectively (Figures 1a and 1c), have been widely used for investigation of the process of cell death (Butler 1968; Reynolds and Ree 1971; Pritchard et al. 1987; Pritchard and Butler 1989; Lertprasertsuke et al. 1991; Ray et al. 1992; Fukuda et al. 1993; Shikata et al. 1996; Oyaizu et al. 1997). In the present study using TUNEL and ACPase staining, the relationship between DNA degradation and lysosome activation (loss of lysosomal integrity) in necrotic cell death was comparatively analyzed in both models.
Although previous investigators have reported that these hepatotoxic agents induce not only necrotic but also apoptotic cell death (Pritchard et al. 1987; Pritchard and Butler 1989; Oyaizu et al. 1997; Shi et al. 1998), in our study no typical morphological changes of apoptotic cell death were observed. This may have been due to the differences in experimental design between the present and these previous studies (Pritchard et al. 1987; Pritchard and Butler 1989; Oyaizu et al. 1997; Shi et al. 1998), i.e., the dose injected per body weight and the time course after injection. Apoptotic cell death could be induced considerably sooner after injection in both models. The varied TUNEL patterns in our study were considered to result from differences in severity of subcellular damage followed by DNA degradation (Figures 1b and 1d). Particularly in the CCl4-injected liver, nuclei undergoing intense DNA degradation failed to show TUNEL staining. It is assumed that the different mechanisms of action of the two agents would contribute to the variations in necrotic cell death. CCl4 causes lipid peroxidation, which is followed by the loss of membrane integrity. DMN causes alkylation, which is followed by DNA damage (Ray et al. 1992; Kamendulis and Corcoran 1994). DMN has been shown to damage endothelial cells concurrently with or before significant morphological changes in adjacent hepatocytes (Pritchard et al. 1987; Pritchard and Butler 1989).
(
Light microscopic observation of ACPase staining indicated that lysosomal integrity was maintained in the DMN-injected liver similarly to that in the control liver (Figures 1e and 1h). However, in the CCl4-injected liver the loss of lysosomal integrity was characteristically indicated by the diffuse ACPase staining pattern in the cytoplasm (Figure 1f). Both liver injuries were compared using double staining by TUNEL and ACPase, and significant differences were observed in the staining patterns (Figures 1g and 1i). DNA degradation represented by TUNEL staining may be closely associated with the release of lysosomal enzymes in the CCl4-injected liver but is probably not associated with that in the DMN-injected liver. On subcellular observation of the CCl4-injected liver using CLSM, ACPase deposition was visualized with loss of lysosomal integrity or polarity, and some ACPase deposits were overlaid on the nuclear periphery (Figure 2b). With the progression of damage in the CCl4-injected liver, intranuclear ACPase deposition become more extensive, and a marked decrease in intact DNA was recognized by deterioration of methyl green staining (Figure 2c). Pearse (1985) reported that methyl green has affinity for intact or double-stranded DNA. In the DMN-injected liver, on the other hand, a fine granular pattern of ACPase deposition was prominently preserved (Figure 2d). These features were significantly supported by the following electron microscopic findings. The lysosome membrane was delineated, and fine granular ACPase deposits were dispersed in the intranuclear periphery and in the vicinity of the nuclear membrane in the CCl4-injected liver. Lysosomal integrity was well preserved in the DMN-injected liver (Figures 2f and 2g). We believe that these electron microscopic findings, particularly in the CCl4-injected liver, indicate the early morphological events of coagulation necrosis followed by DNA degradation in which destruction of organelles was still minimal, if present.
On subcellular observation of TUNEL-positive cells by CLSM, TUNEL deposits were diffusely visualized, with a grossly granular pattern in the CCl4-injected liver (Figure 3a). The intact DNA area was reduced, and foci showing an overlay of TUNEL deposits on the intact DNA were also seen. At the electron microscopic level, the TUNEL deposits were observed to be irregularly distributed in the nuclei (Figure 3c). However, the TUNEL and methyl green staining patterns in the DMN-injected liver differed markedly from those in the CCl4-injected liver (Figure 3b). TUNEL staining was found mainly in the center of the nuclei, with a compactly aggregated appearance, and the methyl green staining tended to be confined to the nuclear periphery, showing a ring-like pattern. TUNEL staining was demonstrated by electron microscopy not to be localized in the nuclear periphery (Figure 3d). Therefore, on the basis of these analyses of necrotic cell death, we hypothesized that two different types of DNA degradation could be discriminated by TUNEL staining, with/without association with lysosomal activation. Acute hepatocellular damage in the CCl4-injected liver would be characterized by necrotic cell death with lysosome activation, and that in the DMN-injected liver could be satisfactorily defined as necrotic cell death without lysosome activation. In the DMN-injected liver, the induction of DNA degradation may be largely limited to the nuclear center, where euchromatin (including active chromatin) is more or less considered to be a target. Therefore, DMN may induce DNA degradation selectively or specifically. In contrast, in the CCl4-injected liver DNA degradation takes place randomly, with no apparent predilection for heterochromatin or euchromatin.
The loss of lysosomal integrity is generally interpreted as an indicator of necrosis in a broad sense (de Duve 1959; Fukuda et al. 1993; Hashimoto et al. 1995). We believe, however, that necrotic cell death would not always be followed by distinct lysosome activation when relatively weak subcellular damage is induced. Therefore, close attention should be paid to interpretation of the results of TUNEL staining, not merely in the discrimination between necrotic and apoptotic cell death. We conclude that necrotic cell death, i.e., DNA degradation, could be divided into at least two types, with/without association with lysosomal activation, represented by necrotic cell death in the CCl4-injected liver and that in the DMN-injected liver, respectively. It remains to be elucidated whether intranuclear deposition of lysosomal enzymes detected as an early event in the CCl4-injected liver is a generalized phenomenon occurring in various types of coagulation necrosis. It is of interest to analyze hemorrhagic necrosis with regard to the presence or absence of lysosomal activation and selective DNA degradation, as represented by the DMN-injected liver. Our study will be of significant benefit for future analyses of various types of pathological and non-physiological cell death.