Abstract
Tissue repair is a dynamic compensatory cell proliferation and tissue regeneration response stimulated in order to overcome acute toxicity and recover organ/tissue structure and function. Extensive evidence in rodent models using structurally and mechanistically diverse hepatotoxicants such as acetaminophen (APAP), carbon tetrachloride (CCl4), chloroform (CHCl3), thioacetamide (TA), trichloroethylene (TCE), and allyl alcohol (AA) have demonstrated that tissue repair plays a critical role in determining the final outcome of toxicity, i.e., recovery from injury and survival or progression of injury leading to liver failure and death. Tissue repair is a complex process governed by intricate cellular signaling involving a number of chemokines, cytokines, growth factors, and nuclear receptors leading to promitogenic gene expression and cell division. Tissue repair also encompasses regeneration of hepatic extracellular matrix and angiogenesis, the processes necessary to completely restore the structure and function of the liver tissue lost to toxicant-induced initiation followed by progression of injury. New insights have emerged over the last quarter century indicating that tissue repair follows a dose response. Tissue repair increases with dose until a threshold dose, beyond which it is delayed and impaired due to inhibition of cellular signaling resulting in runaway secondary events causing tissue destruction, organ failure, and death. Prompt and adequately stimulated tissue repair response to toxic injury is critical for recovery from toxic injury. Tissue repair is modulated by a variety of factors including species, strain, age, nutrition, and disease condition causing marked changes in susceptibility and toxic outcome. This review focuses on the properties of tissue repair, different factors affecting tissue repair, and the mechanisms that govern tissue repair and progression of injury. It also highlights the significance of tissue repair as a target for drug development strategies and an important consideration in the assessment of risk from exposure to toxicants.
Keywords
Introduction
For a number of years, the fate of a chemical and its beneficial or destructive effects in the body of a living organism were estimated based solely on the established rules of toxicokinetics and toxicodynamics. It was observed that toxic chemicals, similar to the pharmacological agents, follow the rules of absorption, distribution, metabolism, and excretion (Rozeman and Klaassen, 2001). Central to the toxic actions of any chemical was the metabolism of the compound by the drug metabolizing enzymes (DMEs), such as cytochrome P450. After the toxicant is absorbed and distributed in the body, it undergoes metabolism to generate water-soluble metabolites, which would be easily excreted from the body (deBethizy and Hayes, 2001). It was observed that metabolism of toxic chemicals also resulted in generation of highly reactive metabolites and free radicals that attack the cellular macromolecules and inflict tissue injury (deBethizy and Hayes, 2001). While this generalized mechanism may initiate injury it was understood that continuation or progression of injury occurs through other mechanisms (Mehendale, 1991, 1994; Soni and Mehendale, 1998). Liver is the main site of drug and toxicant metabolism since the hepatocytes are a reservoir of microsomal and cytosolic, phase I and phase II drug-metabolizing enzymes (deBethizy and Hayes, 2001). This has made liver a prime target for chemical-induced injury. The degree of liver injury was thought to be proportional to the generation of reactive metabolites of the chemical via DME-mediated metabolism. It is now known that such oversimplified concepts overlook the determining effects of biological responses to toxic injury that control the final toxic outcomes. Very little was known about the opposing toxicodynamic response of tissue repair following chemical-induced liver injury (Mehendale et al., 1994; Plaa, 2000; Plaa and Charbonneau, 2001).
The extraordinary ability of liver to regenerate upon surgical resection or tissue injury has been known since prehistoric times (Michalopoulos and DeFrances, 1997). Liver regeneration has been studied in detail in a variety of models with two-thirds partial hepatectomy in rodents serving as the principal model system (Fausto et al., 1995; Taub, 1996). Studies have revealed the details of the intricate signal transduction network consisting of chemokines, cytokines, growth factors, and hormones that governs liver regeneration following surgical removal of liver in PH (Fausto et al., 1995). Investigations by us and others during the last quarter century have revealed that a similar dynamic regeneration response or tissue repair occurs following cell death and tissue injury after exposure to toxic chemicals (Dalhoff et al., 2001; Lockard et al., 1983a, 1983b; Mangipudy et al., 1995b; Mehendale, 1991; Shayiq et al., 1999; Soni et al., 1999). Upon infliction of toxic injury, a cascade of distress signals is triggered (Figure 1), which stimulates surrounding healthy cells to divide in order to replace the dead cells (Apte et al., 2002, 2003; Dalhoff et al., 2001; Gardner et al., 2003; Shankar et al., 2003b; Tomiya et al., 1998). However, such promitogenic signaling is inhibited in case of high-dose exposures resulting in inhibition of tissue repair (Apte et al., 2002, 2003; Mangipudy et al., 1995b; Rao et al., 1997). These findings have been critical in understanding, for the first time, the underlying mechanism of the dose-response phenomenon of compensatory tissue repair. Further investigations have revealed that tissue repair is affected by a variety of factors including species (Cai and Mehendale, 1990, 1991a) and strain (Kulkarni et al., 1996), age (Cai and Mehendale, 1993; Dalu and Mehendale, 1996; Murali et al., 2004; Sanz et al., 1998), nutrition (Chanda and Mehendale, 1994, 1995), caloric restriction (Ramaiah et al., 1998b), and disease conditions (Sawant et al., 2004; Wang et al., 2000a; Shankar et al., 2003c). Although tissue repair has been studied in other tissues such as blood (Sawant et al., 1999; Sivarao and Mehendale, 1995), lung (Barton et al., 2000), and kidney (Vaidya et al., 2003), this review focuses primarily on liver. These studies indicate that ability to mount an effective tissue repair following toxicant exposure can impact the final outcome viz. survival or death following toxic exposures. The detailed study of tissue repair following toxicant-induced injury led us to propose a 2-stage model of toxicity.
Two-Stage Model of Toxicity
Numerous studies have established the determining effect of compensatory tissue repair in the final outcome of toxicity i.e., progression or regression of injury (Anand et al., 2003a, 2003b; Cai and Mehendale, 1991b, 1993; Calabrese and Mehendale, 1996; Chanda and Mehendale, 1994, 1995, 1996; Dalhoff et al., 2001; Ramaiah et al., 1998a, 1998b; Shankar et al., 2003a, 2003b, 2003c; Soni et al., 1999). These studies emphasize the existence of 2 distinct stages of toxicity (Figure 1). Stage I is the inflictive stage in which toxic chemicals initiate injury through well-established mechanisms (Figure 1) tempered via the net effect of bioactivation and detoxification processes, while stage II is the progression/regression phase of injury corresponding with the absence/presence of compensatory tissue repair, respectively. Cell replacement and tissue repair stimulated after the low-to-moderate doses of the toxicants restrain injury resulting in recovery (Mangipudy et al, 1995b; Rao et al., 1997); while high doses of toxicants inhibit compensatory tissue repair leading to unrestrained progression of liver injury and animal death (Calabrese and Mehendale, 1996; Magnipudy et al., 1995b; Mehendale, 1991; Rao et al., 1997). The 2-stage model of toxicity emphasizes the critical role of opposing interplay of progression and regression of acute toxic injury in determining the final outcome (Mehandale, 1995a).
Tissue Repair Follows a Dose Response
Intuitively, one would expect that compensatory tissue repair response obeys the cardinal rule of dose response, just as the toxic action of chemicals does. Indeed, time course studies employing increasing doses of hepatotoxicants have revealed that tissue repair follows the golden rule of toxicology, dose response (Anand et al., 2003a, 2003b; Calabrese and Mehendale, 1996; Mangipudy et al., 1995b; Rao et al., 1997). It has been observed that tissue repair increases in a dose-dependent fashion until a threshold dose is reached. Low-to-moderate doses stimulate tissue repair, which is possible only until a threshold dose with one proviso. With each increment in the dose of a toxicant, there is a corresponding delay in the onset of tissue repair (Mangipudy et al., 1995b). Timely onset of tissue repair is important because during the time lost before the onset of tissue repair, tissue injury progresses (Chanda and Mehendale, 1995; Mangipudy et al., 1995b; Mangipudy and Mehendale, 1998; Sawant et al., 2004; Wang et al., 2000a). To an extent up to the threshold dose, higher incremental response in tissue repair more than compensates the delayed onset due to incremental increase in the toxicant dose as illustrated by thioacetamide-induced hepatic injury-tissue repair model (Mangipudy et al., 1995b). At high doses beyond the threshold, tissue repair is inhibited and what little compensatory tissue repair does occur, is much delayed and too little to arrest the accelerated progression of injury leading to organ failure and death (Mangipudy et al., 1995b). This concept works for all hepatotoxicants tested thus far.
A classic example of the dose dependency of tissue repair is thioacetamide-induced liver injury and tissue repair (Mangipudy et al., 1995b). Unlike many other hepatotoxicants, thioacetamide offers the advantage of large window of time (3.5 to 7 days) before liver failure and death of animals. This is a distinct advantage over the other classic hepatotoxicants such as CCl4, acetaminophen, CHCl3, etc. where animals die from lethal doses within 12 to 24 hour (Anand et al., 2003a; Mehendale and Klingensmith, 1988; Rao et al., 1997; Shankar et al., 2003a). With thioacetamide, both the incline and decline slopes of injury can be examined. Thioacetamide is eliminated with a t1 / 2 of 2.5 hour (Chilakapati et al., 2002; Porter et al., 1979). Male SD rats were exposed to 4 increasing doses of thioacetamide, 50, 150, 300, and 600 mg/kg. Changes in liver injury and tissue repair were measured following exposure to thioacetamide over a time course of 0 to 96 hour. Surprisingly, liver injury induced by the first 3 doses of thioacetamide did not yield a dose response over a 6-fold range. No deaths occurred with these 3 doses. After administration of high dose (600 mg/kg), initiation of injury was significantly lower during the early time points. However, injury aggressively progressed only beyond 48 to 60 hour after the administration of thioacetamide, well after complete elimination of this toxicant (Figure 2). With this high dose of thioacetamide, 90% mortality was observed. Tissue repair response (3H-thymidine incorporation and PCNA analysis) indicated that it was inhibited and much delayed after this high dose (Figure 2). A negligible increase in tissue repair was observed as late as 72 hour following thioacetamide administration, which was too late and too little to rescue the rats from aggressive expansion of injury, liver failure, and death. These data demonstrate that tissue repair functions in a dose-dependent fashion until a threshold dose (somewhere between 300 and 600 mg TA/kg in this case) and is inhibited beyond the threshold dose (Mangipudy et al., 1995b).
The dose-dependent increase in tissue repair has been established with a number of toxicants (Table 1) such as CCl4 (Rao et al., 1997), chloroform (CHCl3) (Anand et al., 2003a), 1, 2, dichlorobenzene (Kulkarni et al., 1996, 1997) trichloroethylene (TCE), and allyl alcohol (AA) (Soni et al., 1998; Soni and Mehandale, 1998). Furthermore, studies have established that mixtures of toxicants also stimulate a dose-dependent tissue repair (Anand et al., 2003b; Soni and Mehandale, 1998). Studies with binary, ternary, and quaternary mixtures such as TCE + CHCl3, TCE + CHCl3 + AA, and TCE + CHCl3 + AA + TA suggest that tissue repair is stimulated with the low doses of mixtures and is inhibited at higher doses (Soni and Mehandale, 1998), suggesting that dose-response relationships for compensatory tissue repair are preserved for mixtures of toxicants just as for the individual compounds (Soni and Mehandale, 1998).
Although a variety of toxicants individually and in mixtures (Table 1) stimulate tissue repair in a dose-dependent fashion, the exact mechanisms are still under investigation. Detailed analysis of cell division cycle following toxicant exposure over a time course indicates that high doses inhibit cell cycle progression, especially between the G1 and S phase of cell cycle (Chanda et al., 1995; Chanda and Mehendale, 1994, 1995; Mangipudy et al., 1995a, 1995b, 1998; Rao et al., 1997; Sawant et al., 2004; Soni et al., 1999; Thakore and Mehendale, 1991, 1994; Wang et al., 2000a, 2001). The cytokine/growth factor mediated signaling and expression of other genes such as cyclin D1 involved in the cell cycle and the effect of high-dose treatment on these factors are of continuing interest (Shankar et al., 2003b).
Tissue Repair as a Determinant of Final Outcome of Toxicity
Although time-course studies on tissue repair following various doses of toxicants indicated that tissue repair plays an important role in the final outcome, i.e. survival vs. death in many studies (Calabrese and Mehendale, 1996; Mehendale, 1991; Soni et al., 1999), conclusive evidence comes from interventional studies employing 2 opposing strategies: (1) antimitosis studies where tissue repair was deliberately inhibited, and (2) preplacement of tissue repair in auto- and heteroprotection studies.
One very successful strategy to demonstrate the critical importance of compensatory tissue repair in the recovery from liver injury is to intervene with cell division (Table 2) and tissue repair that oppose progression of injury. Colchicine (CLC) is an antimitotic agent that inhibits cell division by 2 separate mechanisms (Fitzgerald and Brehaut, 1970). First, DNA synthesis is inhibited so that cells cannot enter the S-phase of cell division cycle (Tsukamoto and Kojo, 1989). Second, it also inhibits microtubular formation so that the cells that are in advanced stages of cell division cycle cannot divide (Fitzgerald and Brehaut, 1970). In a classic CLC antimitosis experiment, CLC (1 mg/kg) treatment given at crucial time points well after toxicant-initiated injury (150 and 300 mg thioacetamide/kg) but before or during tissue repair resulted in complete inhibition of cell proliferation and tissue repair. This resulted in conversion of these normally nonlethal doses of thioacetamide (150 and 300 mg/kg) into 100% lethal (Mangipudy et al., 1996). Analysis of tissue repair indicated that CLC inhibited cell proliferation and tissue repair. Consequently the injury progressed leading to liver failure and animal death. Similar results were obtained in another model of toxicity of a combination of chlordecone (CD) and CCl4 (Dalu and Mehendale). Previous studies had indicated that 45-day-old male Sprague to Dawley rats when exposed to CD (10 ppm for 15 days in the diet) + CCl4 (100 μl/kg) exhibit approximately 25% lethality. Treatment of CD + CCl4-exposed rats with CLC (1 mg/kg) resulted in increase in lethality from 25 to 85%, with a significant decrease in tissue repair in the CLC-treated group. Similar increase in lethality was observed in Fisher 344 (F344) rats treated with o-DCB (Kulkarni et al., 1997). Taken together, these data with antimitotic intervention of tissue repair highlight the importance of tissue repair in the final outcome of toxicity.
Another strategy to study the role of tissue repair in the final outcome of toxicity is by preplacement of tissue repair using auto- and heteroprotection models (Dalhoff et al., 2001; Mangipudy et al., 1995b; Mehendale et al., 1994; Shayiq et al., 1999; Thakore and Mehendale, 1991). By administration of low dose of compound “A,” tissue repair is stimulated that further protects against a subsequently administered lethal dose of the same compound “A” (autoprotection) or entirely different compound “B” (heteroprotection). The first small dose of the toxicant initiates promitogenic cellular signals and essentially preplaces tissue repair, which serves to inhibit progression of injury initiated by the subsequently administered normally lethal dose and protects the animals. Autoprotection has been studied using CCl4 (Thakore and Mehendale, 1991), thioacetamide (Mangipudy et al., 1995a), and acetaminophen (Dalhoff et al., 2001; Shayiq et al., 1999) while heteroprotection has been investigated using thioacetamide and acetaminophen combination (Chanda et al., 1995). Preplacement of tissue repair can also be achieved by surgical two-thirds resection of liver by partial hepatectomy before toxicant treatment (Uryvaeva and Faktor, 1976). Liver regeneration after 70% partial hepatectomy protects the animals from a lethal challenge of CCl4 or chlordecone + CCl4 due to attenuation of progression phase of injury (Bell et al., 1998; Cai and Mehendale, 1993; Kodavanti et al., 1989a, 1989b).
In these models of auto-and heteroprotection it should be noted that liver injury initiated by the high dose of toxicants is not diminished by the prior administration of the priming agents (Chanda et al., 1995; Mangipudy et al., 1995a). Even though the same massive and normally lethal liver injury is reached, and is lethal in unprimed animals, the primed animals overcome this injury as a result of sustainable and early onset of tissue repair stimulated due to priming dose.
In essence, in an acute toxicity paradigm, absence or presence of tissue repair response leads to either progression or regression of injury, respectively. Injury regresses upon the onset of timely and robust tissue repair because the dividing/newly divided cells are resilient to progression of injury (Abdul-Hussain and Mehendale, 1992; Bell et al., 1988; Cai and Mehendale, 1993; Dalu and Mehendale, 1996; Kodavanti et al., 1989a, 1989c; Roberts et al., 1983; Ruch et al., 1985). This paradigm highlights the importance of considering tissue repair in biomedicine for potential therapeutic intervention and as a main toxicodynamic factor in risk assessment process.
Factors Affecting Tissue Repair
A number of physiological factors including species, strain, age, nutrition, caloric restriction, and disease affect the tissue repair response (Table 3). In assessing risk of exposure to drugs and toxicants in the context of public health, rather wide ranging diversity of responses among the general public is a serious problem. While many factors are considered as underlying causes of such unpredictable diversity, the role of tissue repair as a substantial factor has not been considered. The preceding factors that influence tissue repair response may help in explaining the wide ranging interindividual differences in responses to drugs and toxicants.
Species, Strain Difference in Tissue Repair
Investigations with Mongolian gerbils and Sprague–Dawley rats suggested that the LD50 of CCl4 was 35-fold lower (0.08 ml/kg in gerbils vs. 2.5 ml/kg in rats) in the gerbils (Cai and Mehendale, 1990). Further investigations revealed that the high CCl4-induced toxicity in gerbils could be explained by the extremely sluggish tissue repair in the gerbils (Cai and Mehendale, 1991a) (Table 3). Gerbils are also remarkably resistant to chlordecone-amplified toxicity of CCl4 (Cai and Mehendale, 1990, 1991a). Inhibition of the negligible compensatory tissue repair by chlordecone + CCl4 in gerbils is inconsequential. This interaction is not lethal in gerbils because chlordecone-amplified CCl4 toxicity is known to be due to inhibited CCl4-induced increase in compensatory tissue repair (Mehendale, 1994), and tissue repair is minimal in gerbils and occurs too late to be useful.
Similar species difference was noticed between rats and mice under disease conditions (Shankar et al., 2003a, 2003c; Wang et al., 2000a). Streptozotocin-induced type 1 diabetic rats were found to be highly sensitive to thioacetamide-induced liver injury where even a normally nonlethal dose of thioacetamide is lethal in diabetic rats because of compromised tissue repair response (Wang et al., 2000a, 2001). However, streptozotocin-induced type 1 diabetic mice were completely refractory to liver injury induced by a lethal dose of thioacetamide due to their ability to mount effective tissue repair response (Shankar et al., 2003c). A classic example of strain difference in tissue repair is observed between F344 rats and Sprague–Dawley rats when exposed to 1,2 dichlorobenzene (o-DCB) (Stine et al., 1991). It was observed that F344 rats experience high liver injury following exposure to 0.2, 0.6, and 1.2 ml/kg of o-DCB as compared to Sprague–Dawley rats treated with the same doses. However, the mortality induced by o-DCB is not higher in F344 rats, since these rats are capable of mounting a much stronger tissue repair compared to Sprague–Dawley rats (Kulkarni et al., 1996). The significantly higher tissue repair in F344 rats enables them to escape o-DCB-induced liver injury even though it is 10-fold higher than the S-D rats (Kulkarni et al., 1996, 1997) (Table 3).
Age as a Determinant of Tissue Repair
Age is an important determinant of the extent of tissue repair following toxicant exposure. In general, newborn animals are capable of mounting faster and efficient tissue repair during early developing age compared to adults. This has been demon-strated with CCl4, and chlordecone + CCl4-amplified liver injury in 20-day-old neonatal and 2-month-old young adult Sprague–Dawley rats (Dalu et al., 1995a, 1995b, 1996). The 20-day old neonates were found to be resistant to the CCl4 and CD + CCl4-induced liver injury as compared to the 2-month-old young adult rats (Table 3). Further investigations revealed that the mechanism underlying such resistance in the neonatal rats was the ongoing liver cell proliferation in these rats with growing livers. At 20 days following birth, the livers of the young rats are still under development and are able to mount an effective tissue repair. This ability is lost in the adults at 2 months of age when most of the hepatocytes are in quiescence and fail to divide and mount an effective tissue repair. Furthermore, investigations revealed that in the neonatal liver the proto-oncogenes such as TGF-α, c-fos, H-ras, and K-ras were expressed at much higher levels and at much earlier time points following toxicant exposure (Dalu et al., 1995a). These data indicate that higher and timely expression of these proto-oncogenes stimulating a timely tissue repair in the neonate liver play a crucial role in the resistance exhibited by the neonates.
Surprisingly, animals during advanced age also exhibit a prompt and timely tissue repair upon challenge with the combination of chlordecone + CCl4 (Mehendale et al., 1999; Murali et al., 2004). F344 rats from 3 age groups, 3, 14, and 24 months, were exposed to the chlordecone + CCl4 combination. The 14- and 24-month-old rats exhibited higher survival and liver tissue repair as compared to the young adult (3-month-old) rats. No difference in the bioactivation of CCl4 was observed in the 14- and 24-month-old vs. the 3-month old rats. Mehendale et al. (1999) reported that protection against chlordecone + CCl4-amplified toxicity was also evident in Sprague–Dawley rats, suggesting that this remarkable resiliency due to very high compensatory tissue repair in liver is not strain-dependent. These data suggest that the tissue repair response is not only intact in the old animals but it is surprisingly enhanced (Table 3). The exact mechanisms and cellular signaling behind this enhanced tissue repair in older rats is currently under investigation.
Taken together, these data indicate that ability to mount tissue repair following toxicant exposure varies among different species, strains, and age groups and may have a significant impact on the drug development and risk assessment process.
Effect of Nutrition on Tissue Repair
It is known that various nutritional factors such as carbohydrates, proteins, and lipids affect the toxic responses. Extensive studies have demonstrated that modulation of the nutritional factors can directly affect the final outcome of toxicity by changing the tissue repair response. Initial studies with “glucose loading” models indicated that 15% glucose supplementation in drinking water for 8 days inhibits compensatory tissue repair following exposure to centrilobular hepatotoxicants such as thioacetamide, CHCl3, and CCl4 (Table 3). Glucose loading had no effect on the CYP450-mediated metabolism of thioacetamide but substantially decreased the compensatory tissue repair response (Chanda and Mehendale, 1995). Glucose loading did not affect insulin levels either. Furthermore, supplementation of diet with palmitic acid, a primary or preferred source of energy to the periportal hepatocytes, along with its mitochondrial carrier,
Caloric Restriction
Another model that demonstrates the effect on tissue repair is diet restriction (DR). Diet restriction, known for its ability to slow down aging, decrease cancer incidence, and other age-associated immunological disorders, is also known to protect from toxicity of various chemicals such as isoproterenol that induces cardiotoxicity, ozone-induced lung toxicity, and thioacetamide-induced liver injury (Ramaiah et al., 2000). Male Sprague–Dawley rats subjected to 35% diet restriction for 21 days exhibited higher survival (70% in DR vs. 10% in ad libitum) following a normally lethal dose (600 mg/kg) of thioacetamide (Ramaiah et al., 1998a, 1998b). Higher survival occurs in DR rats in spite of the higher liver injury due to induction in thioacetamide-bioactivating enzyme CYP2E1 (Ramaiah et al., 2001). The mechanism behind the increased survival in DR rats was the earlier onset and robust tissue repair response (Ramaiah et al., 1998a, 1998b). Further investigations indicated that following thioacetamide administration, promitogenic cellular signaling was promptly enhanced in the DR rats (Apte et al., 2002, 2003; Corton et al., 2004). Various signaling pathways including IL-6-mediated JAK-STAT pathway, TGF-α and HGF-mediated MAPK pathway, and PPAR-alpha mediated signaling pathway were upregulated in the DR rats resulting in a timely tissue repair response (Apte et al., 2002, 2003). Proteomic analysis has identified additional signaling targets in DR rats such as GRP78, arginase, and calmodulin (unpublished data). These investigations have added the enhanced tissue repair response following toxicant exposure to the long list of beneficial effects of DR.
Tissue Repair in Disease Condition
The earlier reports on inhibition of tissue repair by glucose loading raised questions concerning whether diabetes sensitizes liver to drug- and toxicant-induced toxicity. According to the American Diabetes Association, more than 18 million Americans are suffering from diabetes, and is a predominant cause of morbidity and mortality. DB is known for inducing secondary complications including renal, cardiovascular problems, and impaired wound healing. Studies with streptozotocin-induced type 1 (insulin-dependent) DB indicated that the DB rats exhibit inhibited tissue repair response following treatment with a nonlethal dose of thioacetamide (300 mg/kg), leading to 100% mortality (Wang et al., 2000a, 2000b, 2000c). The DB rats experience increased liver injury, partly due to induction in thioacetamide-bioactivating enzymes CYP2E1 (Wang et al., 2001). However, inhibition of CYP2E1 in DB rats by a relatively specific inhibitor, diallyl sulfide (DAS), failed to protect from thioacetamide-induced liver failure and mortality. Although DAS treatment decreased initial bioactivation-based injury (stage 1 of the 2-stage model of toxicity) to the same level as seen in the non-DB rats, only DB rats failed to stimulate an effective tissue repair response, resulting in progression of initial injury, massive liver failure, and death (Wang et al., 2001).
Interestingly, type 1 DB mice were resilient to thioacetamide and APAP-induced hepatotoxicity, due to increased tissue repair (Shankar et al., 2003a, 2003b, 2003c). The increased tissue repair in DB mice was partly explained by timely signaling via PPAR-α, stimulating cell cycle genes such as cyclin D1. Further investigations with type 1 DB rat model have revealed that the mechanism behind inhibition of tissue repair following thioacetamide challenge is down-regulation of MAPK pathway and NF-κB-mediated signaling.
To study the modulation of tissue repair in type 2 diabetes, which afflicts 90% of all diabetic patients, a high fat diet plus streptozotocin-induced model of type 2 diabetes (noninsulin-dependent diabetes) was developed (Sawant et al., 2004). Studies with these diabetic rats revealed inhibition of tissue repair following CCl4-induced hepatotoxicity (Sawant et al., 2004). Mechanistic studies suggest that the mechanism behind inhibition of tissue repair in diabetes is down-regulation of MAPK and NF-κB signaling pathways. In the type 1 rat model, tissue repair is inhibited in the absence of insulin, and in the type 2 model even in the presence of insulin, tissue repair is inhibited (Sawant et al., 2004; Wang et al., 2000a). Taken together, these data provide substantial evidence that diabetic animals are highly susceptible to toxicant-induced liver injury due to inhibition of compensatory tissue repair response (Table 3).
Progression of Injury
Several studies have shown that acute toxic injury develops into 2 distinct stages: stage 1-Initiation of injury, and stage 2-Progression of injury as illustrated in Figure 1 (Calabrese and Mehendale, 1996; Chanda and Mehendale, 1996; Lawson et al., 2002; Luster et al., 2001; Mangipudy et al., 1995a, 1995b; Mehendale, 1991, 1994, 1995a; Ramaiah et al., 1998b, 2001; Rao et al., 1996; Slater, 1984; Soni et al., 1998). Extensive evidence suggests that tissue repair is a dynamic opposing force that curtails stage 2 or progression of injury from developing into an organ failure (Mehendale, 1995b). While much is known about the endogenous bioactivation mechanisms that lead to infliction of cellular and tissue injury, mechanisms of progression of injury have remained obscure. It is thought that this temporal increase in injury over time is due to slower production of reactive metabolites from the residual parent compound over a time course. However, toxicokinetic studies disprove this notion. The toxicokinetics of the model hepatotoxicants such as thioacetamide, CCl4, and APAP indicate that most of the toxicants are excreted from the body within the first 24 hour by conjugation reaction mediated by phase 2 DMEs and other elimination processes (Mehendale et al., 1999; Porter et al., 1979; Shankar et al., 2003a). However, the time course of injury suggests that the liver injury increases and progresses well beyond the 24 hour (Mangipudy et al., 1995b; Rao et al., 1997; Shankar et al., 2003a). This observation suggests that the progression of injury, initiated by the toxicants, takes place in toxicant-independent fashion. In the literature, 3 mechanisms have been proposed as the possible explanations for progression of injury: (1) Contribution of inflammatory cells (Czaja et al., 1994; Laskin and Pendino, 1995; Piguet et al., 1990); (2) Production of free radicals (Poli, 1993; Slater, 1984); and (3) Leakage of degradative enzymes from the dying and injured cells (Cotran et al., 1999; Poli et al., 1987). Activated resident Kupffer cells and the neutrophils recruited at the site of parenchymal liver injury are considered as the primary culprits in damaging surrounding healthy cells as the result of nonspecific action (Laskin and Pendino, 1995; Luster et al., 2001). However, recent evidence suggests that the contribution of the inflammatory cells does not or is not sufficient to mediate progression of injury (Ju et al., 2002; Kutina and Zubakhin, 2000; Lawson et al., 2002). The second leading theory regarding progression of injury is production of free radicals and oxidative stress, and subsequent lipid peroxidation that propagates injury (Kellogg and Fridovich, 1975; Mylonas and Kouretas, 1999). Though the antioxidants prevent/delay the tissue damage partially (Blazka et al., 1995; Czaja et al., 1994), progression of injury still occurs. A very recent study by Pawa and Ali (2004) indicates that inhibition of lipid peroxidation by antioxidants only decreases the initial injury of various hepatotoxicants such as CCl4, thioacetamide, and CHCl3. Blocking lipid peroxidation fails to prevent progression of injury and subsequent lethality. These findings indicate that progression of injury occurs independent of free radicals.
A third relatively less-studied theory that may explain the progression of injury is the leakage of degradative enzymes or “death proteins” from the dying and injured cells, which may destroy neighboring cells causing progression of injury. Pathology literature (Poli et al., 1987; Cotran et al., 1999) has provided some evidence for such a mechanism. However, this mechanism has not been explored in a systematic manner. We chose to explore whether such cellular degradative enzymes released upon infliction of injury from necrosed hepatocytes mediate progression of injury (Figure 3).
Our recent studies have provided substantial evidence that cysteine protease, calpain, plays a predominant role in progression of injury (Limaye et al., 2003). Calpain is known to degrade several membrane- and cytoskeletal proteins including fodrin/spectrin, talin, filamin, and other macromolecules pivotal for cellular integrity (Croali and DeMartino, 1991; Saido et al., 1994; Miyoshi et al., 1996; Carragher and Frame, 2002), and thereby may cause progression of injury. In our studies, calpain inhibition using a specific calpain inhibitor, CBZ-VAL-PHE-methyl ester (CBZ) administered 1 hour after CCl4 in rats, led to 50% reduction in CCl4-induced mortality. In order to establish whether this protection is due to inhibition of progressive phase of liver injury (stage 2) a nonlethal dose of CCl4 (2 ml/kg, ip) was used. CBZ was administered to 1 group of rats 1 hour after the injection of CCl4. The other group received only the vehicle (DMSO, 0.2 ml/kg, ip) used for CBZ. Time-course measurements of liver injury assessed by plasma ALT elevation indicated that progression of liver injury initiated by CCl4 was substantially decreased by CBZ intervention. Histopathology of liver also confirmed protection against the progression of injury (Figure 4) (Limaye et al., 2003). Calpain inhibition also protected against acetaminophen-induced progression of injury and subsequent mortality in mice (Limaye et al., 2003). These findings indicate that calpain’s role in progression of injury is neither species-specific nor toxicant-specific. In both cases, calpain inhibitor had no effect on the major bioactivating enzyme CYP2E1. In vitro incubation studies with the micro-somes also did not reveal any change in the catalytic activity of CYP2E1 enzyme even with 500-fold concentration range of CBZ. Covalent binding of 14CCl4-derived radiolabel in rats and 14C-acetaminophen-derived radiolabel in mice was unaltered regardless of the administration of CBZ (Limaye et al., 2003). These observations strongly suggest that calpain inhibitor CBZ does alter the bioactivation of these toxicants tested. Observations such as increase in calpain leakage with increase in liver injury, decrease in calpain-mediated degradation of fodrin, a substrate of calpain, in CBZ-treated rats, and ability of calpain to induce cell death in isolated primary hepatocytes in vitro further support involvement of calpain in progression of injury. Calpain inhibition resulted in prevention of progression of injury, paving the way for tissue repair to take over and restore the dead tissue mass. However, how the dividing cells escape calpain-induced cell death remains to be investigated. Recent study revealing overexpression of calpastatin, an endogenous inhibitor of calpain, may explain the mechanism of resiliency of new cells against progression of injury (unpublished data).
Significance of Tissue Repair
Extensive evidence gathered during the last quarter-century supports the role of tissue repair as an important factor affecting the final outcome of toxic injury. Tissue repair is a dose-dependent dynamic process, affected by several factors including species, strain, age, nutrition, caloric restriction, and disease condition. Various interventional strategies, detailed signal transduction studies, and genomic and proteomic studies have revealed that tissue repair plays a decisive role in determining survival or death of an animal exposed to a toxicant. Recent findings suggesting involvement of calpain and other death proteins in progression of injury aid in our understanding of a general paradigm of acute toxicity (Figure 5). These data argue for consideration of tissue repair as a factor in risk assessment and drug development strategies. Consideration of endogenous compensatory response to toxicity induced by a test chemical would be helpful in resolving imprecise risk assessment issues and may offer explanation for interindividual variation in adverse drug/toxicant effects. Similarly, assessment of tissue repair stimulated by a test compound may provide additional mechanistic information extremely valuable for drug development process. Taken together, these data indicate that assessment of tissue repair initiated by toxicants upon exposure can have enormous impact on public health.