Impact of halogenated compounds on calcium homeostasis in hepatocytes.

Halocarbons (CCl4, 1,1-dichlorethylene) cause a wide spectrum of effects and injury in hepatocytes. One early effect of these compounds is the inhibition and destruction of the endoplasmic reticulum (ER) calcium pump. Subsequent to inhibition of this pump, the ER calcium pool is depleted and cytosolic levels of calcium are increased for a prolonged period of time. This effect of halocarbons has been characterized and is similar in vivo and in vitro. The importance of this redistribution of cell calcium in expression of halocarbon injury of hepatocytes has not been fully resolved. Several degradative enzymes (phospholipases, proteases) have been implicated as calcium-dependent mediators in toxicity. Our preliminary studies of the effect of calcium redistribution suggest that activation of a calcium-sensitive endonuclease in liver does not play a central role in initiating the lethal effect of halocarbons on hepatocytes.


Introduction
Cytosolic calcium is an important regulator of the activity of many metabolic and structural proteins. Cells normally maintain cytoplasmic calcium at very low levels. Calcium concentrations briefly rise severalfold in response to physiological stimuli (1). The potential role of disrupted calcium flux in chemically induced liver injury have been examined (2)(3)(4)(5)(6).
Thetlow resting concentration of ionized calcium in cytosol is maintained by active compartmentation processes. The level of ionized calcium in cytosol is determined by the cooperation of at least three compartmentation processes-the plasma membrane Ca2'-stimulated, Mg2"-dependent ATPase, the endoplasmic reticulum (ER) Ca2'-stimulated, Mg2+-dependent ATPase, and the mitochondrial Ca2+/H' antiporter. Microsomal fractions isolated from a number of nonmuscle cells and tissues have an energy-dependent calcium uptake system (10)(11)(12). The calcium pump in this fraction is thought to principally represent activity of the endoplasmic reticulum (ER). Several investigators have noted similarities between these calcium pumps and the skeletal muscle sarcoplasmic reticulum calcium pump. These similarities suggest that the ER (microsomal) calcium pump functions to sequester cytoplasmic calcium and, thus, participates in regulating ionized calcium in the cytoplasm and maintaining a releasable, intracellular pool of calcium.
Reynolds and co-workers (13) first observed a selective loss of calcium from the endoplasmic reticulum of rat liver after administrations of CC14. Subsequently, it was demonstrated that CC14 rapidly causes the inhibition of the ER calcium pump (14). This pump is thought to play an essential role in regulating cytoplasmic calcium and to provide an intracellular pool of calcium in the hepatocyte that can be released in response to exogenous stimuli. When this calcium pump is destroyed, a pool of calcium appears to be released from the ER (15). More recently, we have demonstrated that cytoplasmic concentrations of ionized calcium increase rapidly and for prolonged periods both in vivo (15) and in vitro (16) after an animal or hepatocytes are exposed to CC14 or 1,1dichloroethylene.

The Calcium Hypothesis
The calcium hypothesis of acute cell injury suggests that something happens during toxic injury to overload the normal homeostatic mechanisms maintaining ionized cytosolic calcium in the nanomolar range. The overload could occur by the toxicant damaging one or more of the energy-dependent calcium sequestration mechanisms, or by the toxicant allowing a massive influx of extracellular calcium. Halocarbons appear to inhibit or destroy either the ER or plasma membrane calcium pumps and in some cases to selectively increase membrane permeability to calcium.
A number of toxic events, including exposure to halocarbons, have been shown to elevate ionized calcium in cytoplasm to supraphysiological levels or for supraphysiological times. Currently, the element missing is demonstration of an event common to all toxicant models and resulting from calcium levels outside the normal range.

Effect of Halocarbons on Intracellular Calcium
Effect on the Endoplasmic Reticulum Pump and Cytosolic Calcium Results from in vivo experiments are presented in Figure 1. Following a hepatotoxic, but not lethal, dose of CC14, the ER calcium pump activity declined to 35% of control by 30 min. Pump activity remained depressed throughout the experiment and was 6% of control by 8 hr. In livers from rats that received CCl4, phosphorylase a activity was 133% of control by 30 200 , , , 810c 0 min and 150% of control by 1 hr. At 8 hr hepatic phosphorylase a activity remained at 135% of control. Total phosphorylase (a + b forms) activity did not vary significantly over 8 hr in controls or in CC14treated rats (data not shown). Measurement of liver cAMP concentrations demonstrated no statistically significant differences at any time between the treated and control groups (17). These observations allow us to use phosphorylase a activity as a measure of ionized calcium in rat liver cytoplasm. The activity of another endoplasmic reticulum enzyme, glucose-6phosphatase, activity was not significantly affected until 2 hr (71% of control). By 8 hr this enzyme activity was approximately halved. In contrast, this enzyme is not inhibited after exposure of an animal to toxic doses of 1,1-dichloroethylene, a halocarbon that does not initiate lipid peroxidation (18,19). Activity of the plasma membrane enzyme 5'-nucleotidase was not effected by CCl4 administration. To confirm the effect of CCl4 on phosphorylase a levels, glycogen concentrations in liver were examined. Liver glycogen content was reduced to 68% of control as early as 30 min, and glycogen concentrations declined progressively until 24   In liver tissue sampled from control rats, ER calcium pump activity averaged 190 ± 11 nmole calcium/mg protein/30 min, phosphorylase a activity ranged from 44 ± 9.8 to 82 ± 11 nmole phosphate/mg protein/min at 8 AM and 5 PM, respectively: glucose-6-phosphatase activity averaged 1.59 ± 0.04 imole phosphate/mg protein/20 min; 5'-nucleotidase averaged 0.98 ± 0.03 jmole/mg protein/20 min; glycogen content declined from 39 ± 3.4 mg/g liver to essentially zero by 24 hr because food was withheld from all animals after dosing; total liver calcium averaged 55.9 ± 1.6 ,ug/g liver; serum levels of glutamic-pyruvic transaminase averaged 19.6 ± 1.4 IU/L in control plasma. Differences are significant where indicated (p -0.05) by ANOVA (17).   Figure 2. Cultures of rat hepatocytes (18 hr old) were exposed to a toxic concentration of CC14. ER calcium pump activity was depressed to 41% of control at 2.5 min, 19% of control at 5 min and pump activity remained at 10% of control after 30 min. Phosphorylase a activity in cultured hepatocytes was stimulated to 472% of control at 2.5 min and to 392% of control at 5 min. Activity remained elevated beyond 30 min. Conversion of phosphorylase from the b to the a form was complete, as judged from measurement of total a + b activity (data not shown). Stimulation appeared to be greater than had been observed at early times in vivo; however, consistently low control phosphorylase a activity in vitro exaggerated this effect. Glucose-6-phosphatase activity was not significantly inhibited until 20 min after CCl4 exposure. 5'-Nucleotidase activity was never significantly inhibited after CC14 exposure, as in vivo. Glycogen content of cells were depressed to 77% of control by 10 min. Glycogen continued to be mobilized at the same rate until it was 27% of control at 30 min. Later indices of liver cell injury were also examined in hepatocytes cultures. Glutamic-pyruvic transaminase release was massive by 15 and 30 min after CC14 (10-fold greater than control). Hepatocyte calcium (total of ionized and bound) never increased in the in vitro experiments. Only in this determination did the in vitro model depart from results observed with liver tissue from the intact rat. (See below.)

Effect on the Plasma Membrane Pump
Hepatotoxic doses in vivo of CC14, bromobenzene, or acetaminophen produce inhibition of the liver plasma membrane calcium pump (20). After CC14, permeability of plasma membrane vesicles to calcium is substantially increased (21). In contrast, not all hepatotoxins inhibit the ER calcium pump; bromobenzene and acetaminophen inhibit only the hepatic plasma membrane calcium pump (20).

Effect on Mitochondrial Calcium Sequestration
In vivo mitochondrial calcium sequestration is not inhibited by CCI4 (14), and mitochondria accumulate great quantities of calcium parallel to tissue death, thus increaseing the total calcium in the liver (13,14). In hepatocyte culture exposed to a range of CC14 concentrations, mitochondrial as well as ER calcium sequestration could be depressed (Fig. 3). The data presented in this figure suggest that, in our culture . ER and mitochondrial calcium sequestration in material isolated from hepatocytes exposed to CCI4. Hepatocyte cultures were exposed to CC14 (0.01 to 0.2 gL/mL or vehicle for 5 min).
system, exposure of hepatocytes to initial CC14 concentrations than 0.1 gL/mL do not completely model CC14 action in vivo. Inhibition of mitochondrial calcium uptake by CC14 in homogenates of cultured hepatocytes accounts for the lack of calcium accumulation by the injured cells.

Effect of Altered Calcium Homeostasis on Cell Function
The relationship between a sustained increase of cytosolic calcium and toxicity by a number of halocarbons has lead several groups to search for a mechanism or mechanisms by which an increase of calcium could initiate an event or events ultimately leading to death of the hepatocyte. It is assumed that a prolonged increase of ionized calcium to supraphysiological levels may cause a profound and prolonged stimulation of a normal (physiological) process that irreversibly alters the plasma membrane and destroys the semipermeable character of the plasma membrane.
Recently, several groups have examined whether or not phospholipase activation may contribute to the expression of toxicity in hepatocytes. Hydrolysis of phospholipids is increased after a toxicant insult (22)(23)(24)(25). It is widely recognized that calcium plays some role in the control of intracellular proteolysis in several tissues. This system is best characterized in skeletal muscle (26). Activation of nonlysosomal proteases in hepatocytes has been demonstrated following exposure to a toxicant (27).

Effect of Increased Cytosolic Calcium on Endonuclease Activity
Endonucleases are among the hepatic enzymes having activity that can be stimulated by Ca2" (28,29).
Activation of endonucleases has been implicated in thymocyte killing produced by glucocorticoids both in vivo (30,31) and in vitro (32). We have examined whether or not liver endonuclease activity is stimulated by this increase in intracellular ionzed calcium and if such activation could play a role in initiating the ensuing hepatotoxic events.
Endonuclease activation was assessed by separating liver nuclear DNA on agarose gels to determine if DNA was fragmented. No generalized DNA fragmentation was observed in vivo, and in hepatocytes, fragmentation occurred only at very late times following the loss of plasma membrane integrity. Endonuclease activity was further examined at a more sensitive level by specifically monitoring serum albumin DNA. This gene has been shown to have characteristic cleavage sites, in the 5' flanking region, that are hypersensitive to endonuclease action. No changes were observed in the fragment sizes of DNA prepared from rat liver or hepatocyte DNA at early times when toxicity was developing. Thus, our evidence suggests no stimulation of endonuclease activity occurs within a time frame consistent with the rise of intracellular ionized calcium. We conclude that nucleases likely do not mediate the hepatotoxicity accompanying increased cytosolic calcium that follows halocarbon exposure.

Summary and Conclusions
A diversity of toxic insults, including thioacetamide, glactosamine, dimethylnitrosamine, heavy metals, and ischemia are known to alter tissue calcium content significantly. These agents may exert their effects initially by disrupting intracellular calcium homeostasis. Elevated cytosolic calcium concentrations that are supraphysiological, either in duration or in magnitude, may excessively activate a variety of calcium-dependent enzymes found at multiple locations within cells. Calcium-responsive degradative enzymes activated in an uncontrolled manner may damage sites distant (e.g., plasma membrane) from the original cellular injury. Potentially, this scheme (early redistribution of intracellular calcium, producing elevated ionized calcium in cytosol, which causes over-stimulation of calcium responsive enzymes) will prove to be a mechanism for initiating the irreversible development of cellular injury common to a wide variety of toxic substances and target tissues.
The authors thank Cynthia L. Wallace and Cynthia Cavert for their excellent technical assistance. This work was supported by grant ES03437 from the National Institute of Environmental Health Sciences, NIH, Public Health Service.
The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting I I1I I I I the views of the Department of Defense or USUHS. These experiments were conducted according to the principles set forth in the Guidefor Care and Use ofLaboratory Animals, Institute of Animal Resources, National Research Council, DHEW Pub. No. (NIH) 78-23.