- Email: [email protected]
The role of altered [Ca2+]i regulation in apoptosis, oncosis, and necrosis
BB.
Biochi~ic~a et BiophysicaA~ta
ELSEVIER
Biochimica et Biophysica Acta 1313 (1996) 173-178
The role of altered
[Ca2+]i regulation in apoptosis,
oncosis, and necrosis
Benjamin F. Trump *, Irene K. Berezesky Department of Pathology, University of Maryland School of Medicine, 10 South Pine Street, Baltimore, MD 21201, USA Received 17 June 1996; accepted 19 June 1996
Abstract Understanding the processes and events that occur when a cell undergoes a prelethal injury or that lead the cell to death following a lethal injury has been the aim of our research for a number of years. Throughout this period much has been learned, recently at rapid rates, not only by us but by many other investigators as well. Based on the data gathered, we proposed a working hypothesis over a decade ago and have since continually updated it as new experimentation is performed. Our laboratory has focused particularly on the role of cytoplasmic ionized calcium ([Ca2+ ]i) and the effects of its deregulation on prelethal events, including oncosis and apoptosis, and lethal events (necrosis) following cell death. [Ca2+]i appears to be a major link and signalling event. Understanding the mechanisms involved by using a variety of in vivo and in vitro models, coupled with state-of-the-art methodologies, should now allow us to prevent cell death by killing cells when necessary through gene therapy and cancer chemotherapy. Keywords: Cytosolic calcium; Cell injury; Cell death; Cell signalling; Gene regulation
1. Introduction The purpose of this paper is to consider the role of altered ionized intracellular calcium ([Ca 2+ ]i) regulation in the pathogenesis and altered physiology of cell injury and cell death, including apoptosis, oncosis, and necrosis. We will consider the responses of mammalian cells to lethal injury and the role of calcium regulation in relation to other intracellular signalling in the events that follow a lethal injury.
2. Definitions of cell injury and cell death Following a lethal injury, three phases of cellular reaction occur: the prelethal phase; the phase of cell death; and the phase of necrosis (for reviews, see refs. [1,2]). The prelethal phases include changes that occur prior to the 'point of cell death' and are potentially reversible. Two principal patterns of change are currently identified, namely, oncosis and apoptosis.
Abbreviations: CPP32, apopain; CrmA, cowpox virus serpin; ICE, interleukin-1/3 converting enzyme; JNK, Jun N-terminal kinase; MAP, mitogen-activated protein; ARP, polyadenylate ribose polymerase. * Corresponding author. Fax: + 1 (410) 7063743.
In oncosis [3-7], the cells undergo rapid swelling with dilution of the cytosol, formation of cytoplasmic blebs at the plasmalemmal surface, dilatation of the endoplasmic reticulum (ER), mitochondrial condensation followed by inner compartment swelling, dilatation of ER and Golgi components, and nuclear chromatin clumping. In the case of types of injury that can be removed, such as complete anoxia or ischemia, all of these changes rapidly reverse to the normal phenotype. In contrast, the cells in apoptosis are typically shrunken with densification of the cytosol, more extreme clumping of nuclear chromatin, formation of even larger cytoplasmic blebs, and nuclear irregularities; changes in the mitochondria are similar with initial condensation followed by swelling [8]. At the time of cellular death, the changes can be dramatic by time-lapse cinematography (Trump, B.F., unpublished observations) and include interruptions of the continuity of the plasma membrane with escape of cytosolic enzymes to the extracellular space and staining of the nuclei with Trypan blue or the fluorescent dye, propidium iodide. In the phase of necrosis, the changes are similar following either apoptosis or oncosis with the exception that, in the case of apoptosis occurring in vivo, many of the changes occur in fragments of cells or intact cells within the phagolysosome of adjacent parenchymal or mesenchymal cells.
0167-4889//96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. PH S 0 1 6 7 - 4 8 8 9 ( 9 6 ) 0 0 0 8 6 - 9
174
B.F. Trump, L K. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
In necrosis, the changes continue to progress [1,2]. Interruptions in plasmalemmal continuity are seen and often specialized surface structures, such as microvilli, are totally distorted and form complex interlocking whorls resembling myelin figures. The blebs still persist, although many blebs have undergone fragmentation and appear as nearby debris. The nuclear chromatin clumping continues and, in addition, karyolysis begins with increased dissolution of nuclear DNA. The mitochondria remain markedly swollen, and contain one or two types of densities. The first type, so-called flocculent densities, are common to all types of necrosis and consist of aggregated dense intramatrical material composed of denatured matrical proteins; their accumulation is a reliable morphologic indicator of cell death and necrosis. In the second type, precipitates of calcium phosphate occur following types of injury that do not directly interfere with mitochondrial calcium uptake, and, in large amounts, transform to calcium hydroxyapatite crystals.
3. Experimental approach We have followed an experimental approach to elucidate the role of calcium [4,5,7] and cell injury as follows: a series of models, including toxic and others, were chosen to represent the major types of initial injury seen in human disease. These models have included ischemia; anoxia; treatment with carcinogens; and treatment with model toxins with primary characterized interactions such as inhibition of electron transport and oxidative phosphorylation; changes in cell membrane permeability, such as activation of the C5-9 components of complement; use of ionophores including those for calcium; modification of calcium transport in internal stores, such as thapsigargin; oxidant injury using hydrogen peroxide or superoxide generated by the xanthine/xanthine oxidative system; and transfection of genes to modify the reaction to injury or to modify cellular defenses. In many cases, these have been studied both in vivo and in vitro whenever possible. In general, however, to utilize modem technology including fluorescent probes and related technologies, many of the experiments have been carried out on primary cultures or cell lines in vitro. Much of the emphasis of our laboratory has been on studies of the proximal tubule of the kidney; however, we will compare these findings with those experiments performed by other laboratories and those using other cell systems.
4. Relationship of changes to altered [Ca z + ]i 4.1. Source of increased [Ca 2 +]i
Following injury, increases of [Ca2+] i can be rapid, occurring in some cases within seconds. The increased
[Ca2+] i can result from influx of calcium from the extracellular space, from redistribution of intracellular stores, or often from both sources including so-called calcium-induced calcium release [9]. Increased levels of [Ca2+] i are commonly measured by preloading the cells with fluorescent probes such as Fura 2 for digital imaging fluorescence microscopy (DIFM) or Fluo-3 and related compounds for confocal microscopy [10,11]. The [Ca2+]i can be imaged or ratioed and measurements plotted directly using a photon counter. 4.2. Oncosis 4.2.1. Swelling Cytoplasmic swelling in oncosis is marked and begins at an initial rate which is dependent on the type of injury. With direct damage to the cell membrane, e.g., complement or mechanical damage, this occurs very rapidly; with inhibition of energy metabolism limiting the supply of ATP and inhibiting the Na+/K+-ATPase, the change occurs progressively, but more slowly [12]. The swelling is due primarily to increased Na ÷ accompanied by increased C1- and followed by water with accompanying loss of K ÷. As [Na2+] i increases in many cells including the renal epithelium, N a + / C a 2÷ exchange is limited and [Ca2+]i increases as well. Obviously, with inhibition of ATP synthesis in mitochondria and glycolysis, the plasmalemmal Ca2+-ATPase is also inactive and, with direct membrane damage, direct Ca 2÷ influx from the extracellular space Occurs.
4.2.2. Bleb formation Cytoplasmic bleb formation occurs very early during the development of oncosis. These blebs are typically clear and often have a band of actin at the base. Using time-lapse cinematography, these blebs can be seen to form and reform several times during the immediate injury response. Morphologically, these blebs are associated with substantial modification of the cytoskeleton, including both tubulin and actin. By electron microscopy, the blebs are clear with no evidence of cytoskeletal elements between the remaining cytosol and the plasma membrane. Using immunocytochemistry, it is apparent that the actin appears to detach from the cell membrane forming a band across the base of the bleb. As these blebs are observed to expand and contract and often detach, this is considered in our hypothesis to relate to contraction and relaxation of actin and modification of tubulin prior to release. It is our hypothesis that these actin changes are related to protease activity, including Ca 2+- activated proteases, as pretreatment with protease inhibitors greatly modifies the bleb formation [13]. In addition calcium-binding proteins, such as calmodulin and S-100 proteins, may contribute. A threshold of 300-400 mm [Ca 2+ ]i has been suggested for bleb formation [14].
B.F. Trump, LK. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
4.2.3. Mitochondrial changes The initial condensation of mitochondria has been correlated with inhibition of ATP synthesis. The later inner compartment swelling is the consequence of the mitochondrial permeability transition which is induced by increased [Ca2 + ]i.
175
tion appears to be of different pattems and different cell types and may or may not always be calcium-regulated. Intemucleosomal fragmentation may be a late event. An early formation of large fragments apparently requires protease activity in addition to DNAse.
4.4. Necrosis 4.2.4. Nuclear changes In the early stages of oncosis, there is evident nuclear chromatin clumping along the nuclear envelope and around the nucleolus, resulting in clarification of the remaining nucleoplasm. This aggregation is clearly reversible and restoring or adding the injury can result in several reversals of this phenomenon. Although the mechanism of this phenomenon is presently unknown, it is clear from a number of our experiments that reduction of intracellular pH can result in nuclear chromatin condensation and since reduction of cytosolic pH is a very common early reaction, we hypothesize that pH reduction through protein DNA interactions results in this change. 4.3. Apoptosis 4.3.1. Cellular shrinkage Cellular shrinkage undoubtedly is due to loss of ions, predominantly K + and CI- followed by water ion regulation and has been shown to require phosphorylated ion transport systems; furthermore, the K ÷ channels in many cells are calcium-sensitive. Our hypothesis is that early increases of [Ca 2+ ]i stimulate K + loss and shrinkage prior to cell death. It has been shown by Bortner and Cidlowski [15] that lymphocytes expressing Bcl-2 become resistant to dexamethasone-induced apoptosis, while lymphocytes exposed to hypertonic conditions without additional Bcl-2 initiate DNA fragmentation. Thus, this area of volume regulations may be an important defense and control mechanism involving cell death, but requires additional study. 4.3.2. Cytoplasmic blebs The bleb formation in apoptosis also undoubtedly involves regulation of cytoskeletal membrane attachment, but the blebs are different from those seen in oncosis in that they contain abundant cytoplasm and cytoplasmic organelles rather than being watery. Further detailed investigations are needed on the role of cytoskeletal calcium interactions. 4,3.3. Mitochondrial changes Mitochondrial changes are similar to those seen with oncosis except that the shrinkage of mitochondria in the condensed phase is somewhat more severe. Mitochondria undergo swelling prior to cell death in apoptosis. 4.3.4. Nuclear changes Nuclear changes include extreme chromatin clumping and irregularity of nuclear contour. The DNA fragmenta-
4.4.1. Plasmalemmal interruption and blebbing The interruption of plasmalemmal continuity which results in trypan blue and propidium iodide staining is deemed to be secondary to Ca 2+ activation of phospholipase A phosphorylation through the action of kinases such as MAP kinase. In time-lapse cinematography of apoptosis and oncosis, we have observed that one or more large terminal blebs often form around the time of cell death. These may well have a relationship to the breaks in plasmalemmal continuity seen by electron microscopy. 4.4.2. Mitochondrial changes In necrosis, the mitochondria remain swollen and exhibit aggregation of matrical proteins, so-called 'flocculent densities'. In addition, there is fragmentation of the outer membrane. In certain types of injury, calcium phosphate deposits occur in the matrix in addition to the flocculent densities. This involves active uptake of Ca 2+ by the mitochondria from the cytosol and, therefore, is limited to injuries such as complement activation or other primary modification of the plasmalemma which leave the mitochondrial transport capability intact for a time. 4.4.3. Nuclear digestion As the cell passes the 'point of no return', the clumped nuclear chromatin is progressively digested, resulting in blurred profiles in electrophoresis and loss of staining with cationic probes or dyes. This has been termed 'karyolysis' as ultimately only a nuclear ghost remains.
5. Genetic regulation and signalling 5.1. Pro-death genes 5.1.1. Immediate-early genes The induction of a series of intermediate-early genes including c-fos, c-jun, NFKB, and c-myc characterize many types of acute lethal injury. These genes can be induced within 15 min or less, as in the case of c-fos and c-jun, and are evidently induced even in cells which are destined to die within 1 or 2 h. Following at least some types of injury, e.g., oxidant stress, the induction of these genes is Ca2+-dependent [16]. The function of these early genes, apart from inducing later genes including metalloproteases,
176
B.F. Trump, I.K. Berezesky/ Biochim&aet BiophysicaActa 1313 (1996) 173-178
is not clear in the sense that in some cases these may lead to pathways of cell death while in others they may be preparatory for cell division and regeneration in the event that cell death does not occur. C-jun, in particular, has been identified as a pathway toward apoptosis and under some conditions overexpression of c-myc may be as well. This is further emphasized in the discussion of the kinase pathways below. Work with antisense RNA in some systems has shown that failure to express c-jun can protect against cell death. Somewhat later, there is Ca2+-dependent induction of hsp70 which appears to protect against subsequent lethal injury [17].
5.1.2. p53 - gadd and waf
Among the functions of wild-type p53 is induction of mitotic arrest and acute cell death [18]. Microinjection of p53 into cultured cells commonly results in apoptosis and cell killing. Mutant p53, as seen in many forms of human neoplasia, does not exhibit this property. The precise mode of action of p53 is similarly not well understood. Although the relationship between p53 status and [Ca2+] i is incompletely understood, Weinberg et al. [19] noted that in combination with v-rasHa, keratinocytes deficient in p53 shared decreased Ca2+-induced terminal differentiation. Among the later genes induced by activated p53 are gadd45 and waf(p21) [20,21]. In some experiments, waf(p21) has been shown to be involved in initiation of cell killing, though its exact relationship to the final or execution phase is not known.
5.1.3. Proteases and phospholipases
Beginning with the identification of the pro-death gene ced -3 in the nematode C. elegans, an entire family of proteases has been identified which appear to be late in the pathway leading to cell death [22-24]. The role of these proteases and cytoskeletal alterations and bleb formation was mentioned above. It has been shown that protease inhibitors can protect against both apoptosis and oncosis. The exact relationships in the final execution phase remain under study.
5.1.4. The Degenerin gene family
Although most of the programmed cell death in C. elegans development involves apoptosis, mutations in two genes, mec-4 and deg-1, result in oncosis followed by cell death (see review by Lints and Driscoll [25]). Lints and Driscoll [25] have suggested that these degenerin proteins might constitute an ion channel which, in the presence of mutations, may remain open. In this way, therefore, it would be comparable to activation of C5-9 or the effects of PCMBS in mammalian cells. Although the characteristics of the channel are presently unknown, the model proposed by Lints and Driscoll would clearly permit the
rapid entry of Ca 2+ from the extracellular space and result in the consequences mentioned above.
5.1.5. Others
There is an expanding list of other genes whose function appears to relate to cell death. One of these, ced-4, described in nematodes has been suggested to be a calcium-binding protein with an EF hand-like domain [26]. Its exact function and relationship to [Ca 2÷ ]i remain to be determined. Other calcium-binding proteins, including the S-100 family, have been observed to initiate apoptosis [27-29].
5.2. Anti-death genes
Experiments in nematodes have also indicated the presence of anti-death genes. These led to the identification of ced-9 and later its mammalian homologue, bcl-2, which now are seen as part of a rapidly growing family of genes determining cell survival or cell death [30]. The gene bcl-2 has received considerable attention regarding its mode of action. The dimensions of its protective action remain to be completely defined but we suggest that it protects against both apoptosis and oncosis. Mechanisms that have been suggested for its protective action include anti-oxidant properties, interference with signalling pathways, and modification of intracellular calcium regulation [31,32]. We have studied NRK cells (normal rat kidney proximal tubular epithelium) transfected with bcl-2 in order to examine the reaction to injury. Clearly it is protective against oxidant injury and clearly the response of [Ca2+] i is greatly modified [33]. We suggest that bcl-2 modifies the regulation of [Ca2+]i possibly involving Ca transport systems in mitochondria, plasma membrane, or ER.
5.3. Cell death signalling pathways
Clearly, the signalling pathways that lead through oncosis or apoptosis to cell death need to be determined. Progress is so rapid in this field that it is difficult to make any predictions. However, our current knowledge of immediate-early genes and protein kinase pathways provide important clues as does the determination of the morphology seen in apoptosis and oncosis. Experience with timelapse cinematography of oncosis and apoptosis clearly show marked morphologic changes during the earliest phases. Linkage between plasmalemmal and cytoskeletal change may well provide a direct linkage between morphology and signalling. Faux and Scott [34] recently pointed out that 'approx. 300 of the anticipated 2000 protein kinase and 1000 phosphatase genes have been identified' and proposed a morphologic/functional model of signalling that depends on kinase anchoring and scaf-
B.F. Trump, L K. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
~ Plasma Membrane Permeability
I
/
Mitochondrla ~
Swelling ..9_.._ (Oncosls) ~ [Na+]I
;,
Endonuc eases t DNA strand breaks ~
i '
A 53
I
~ [ ATP ] i
PARP '
t GrowthFactor Endoplasmlc Retlculum
.~_[~ [Ca2+] i i
;
Proteases [
177
~
I
K+ "1 .. Shrinkage Cl'-]erlrlux'-~'(Ap°pt°sls) I CytTkeleton Blebs
1'
"÷
Kinases-JNK,
MAP, ealmodulln [
I I
Transcription Phosphollpase A2 Factors [ DNA Repa ir fos,jun,myc = Poly ADP rlbosylatlon ~ CrmA = war 1(p21) ? .L m Proteases = gadd45 "-] T ced-3 bcl 2 ICE/CPP32 I Cell Death I (ced 9)
Fig. 1. Diagram illustrating our current working hypothesis of the major events leading from cell injury to cell death, focusing on the role of lea 2+ ]i- Some of the major categories of injury are shown at the top of the figure. Decreased and absent [ATP]i commonly results from complete ischemia, anoxia, or treatment with inhibitors of ATP synthesis including potassium cyanide, carbon monoxide, fluoride, and iodoacetic acid. Increased plasma membrane permeability commonly results from the immune response, e.g., activation of complement or T-cell release of perforin, or from treatment with a variety of toxins including mercuric salts, ionophores, or mechanical damage. Growth factor deprivation is often studied in vitro though in vivo it does occur with neuronal damage. In oncosis, the early events involve cell swelling because of increased [Na+]i accompanied by C1- and water. This can result either from increased plasmalemmal permeability or decreased [ATP]i resulting in inactivation of Na +, K+-ATPase. Increased [Na+]i can also contribute to increased [Ca2+ ]i through decreased Na+/Ca 2+ exchange. In apoptosis, the initial events include cellular shrinkage, implying loss of ions in particular K + and CI-. The mechanism of the shrinkage is not presently known although increased [Ca2+ ]i might contribute through activation of K + channels. The initial increase of [Ca2+ ]i can result from influx from the extracellular space or from redistribution from intracellular stores. Following the increase, several prinicipal pathways seem to le~d to the reversible and irreversible changes as shown. Also illustrated in the diagram are the events that follow Ca2+-mediated activation of protein kinases, endonucleases, proteases, and phospholipases as discussed in the text. Our current hypothesis of altered gene expression related to cell death is shown near the bottom of the diagram.
fold proteins. Of the anticipated 1000 phosphatase genes, we have observed that okadaic acid provides an excellent model of apoptosis and that modification of phosphatases 1 and 2A result in dramatic morphologic and biochemical changes [35,36].
6. Hypothesis and conclusions Our current model of lethal cell injury and cell death is illustrated in Fig. 1. Deregulation of [Ca2+]i is shown as central to the signalling involved following several classes of lethal injury: inhibition of ATP synthesis as in the case of anoxia, ischemia, or mltochondrial inhibitors; increased plasmalemmal permeability as in the case of complement activation, cell-mediated killing, or many chemical toxins; or withdrawal of growth or trophic factors. In oncosis, the early cellular changes include increased [Na÷]i with concomitant increase o f water content and cellular swelling. This results from loss of A T P to power the N a ÷ / K + - p u m p s a n d / o r increased N a ÷ influx through the damaged plasmalemma. In apoptosis, the cells become shrunken and dense due to water loss, presumably the secondary result of efflux of intracellular ions such as K ÷ and C1-. This probably implies that A T P concentrations and synthesis remain active longer in this type of prelethal injury.
Increased [Ca 2" ]i can result from influx from the extracellular space or from redistribution from other intracellular compartments such as the ER. A variety of effects follow such deregulation. These include activation of proteases, endonucleases, and phospholipases, all of which contribute to the cellular changes that occur during the prelethal phase. Some of the major activities are iUustrated. Of special note are the changes that lead to activation of genes a n d / o r gene products that trigger additional events, such as the ced 3 / I C E protease family members, which appear to be close to the final phase of cell death. Also of note is that the lethal Ca2+-induced vents are antagonized by reducing pH [37]. In conclusion, much is being learned about the processes and events that lead to cellular death. Deregulation of [Ca 2+ ]i is an important link and signalling event in cells following a variety of injuries. Further understanding of these links should lead to improvements in prevention of cell death as well as to improved methods for killing cells as in cancer chemotherapy.
Acknowledgements Supported by NIH Grant DK15440 and Navy N0001491-J-1863. This is contribution No. 3817 from the Cellular Pathobiology Laboratory.
178
B.F. Trump, I.K. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
References [1] Trump, B.F., McDowell, E.M. and Arstila, A.U. (1980) in: Principles of Pathobiology, Third Edn. (Hill, R.B., Jr. and LaVia, M.F., eds.), pp. 20-111, Oxford University Press, New York. [2] Trump, B.F. and Berezesky, I.K. (1992) in: Cardiovascular Toxicology, Second Edn., Ch. 4 (Costa, D., Jr., ed.), pp. 75-113, Raven Press, New York. [3] Von Recklinghausen, F. (1910) Untersuchungen uber Rachitis und Osteomalacie. Verlag Gustav Fischer, Jena. [4] Trump, B.F. and Berezesky, I.K. (1992) Curr. Opin. Cell Biol. 4, 227-232. [5] Trump, B.F. and Berezesky, I.K. (1995) FASEB J. 9, 219-228. [6] Majno, G. and Joris, I. (1995) Am. J. Pathol. 146, 3-15. [7] Trump, B.F. and Berezesky, I.K. (1996) New Horiz. The Science and Practice of Acute Medicine 4, 139-150. [8] Wyllie, A.H., Kerr, J.F.R. and Currie, A.R. (1980) Int. Rev. Cytol. 68, 251-306 [9] Thastrup, O., Cullen, P.J., Drobak, B.K., Hanley, M.R. and Dawson, A.P. (1990) Proc. Natl. Acad. Sci. USA 87, 2466-2470. [10] Smith, M.W., Phelps, P.C. and Trump, B.F. (1991) Proc. Natl. Acad. Sci. USA 88, 4926-4930. [11] Smith, M.W., Phelps, P.C. and Trump, B.F. (1992) Am. J. Physiol. 262 (Renal Fluid Electrolyte 31), F647-F655. [12] Papadimitriou, J.C., Drachenberg, C.B., Shin, M.L., Smith, M.W. and Trump, B.F. (1994) Virchows Arch. 424, 677-685. [13] Elliget, K.A., Phelps, P.C. and Trump, B.F. (1991) Cell Biol. Toxicol. 7, 263-280. [14] Phelps, P.C., Smith, M.W. and Trump, B.F. (1989) Lab. Invest. 60, 630-642. [15] Bortner, C.D. and Cidlowski, J.A. (1996) Am. J. Pathol, in press. [16] Maki, A., Berezesky, I.K., Fargnoli, J., Holbrook, N.J. and Trump, B.F. (1992) FASEB J. 6, 919-924. [17] Yamamoto, N., Smith, M.W., Maki, A., Berezesky, I.K. and Trump, B.F. (1994) Kidney Int. 45, 1093-1104. [18] Harris, C.C. (1993) Science 262, 1980-1981.
[19] Weinberg, W.C., Assoli, C.G., Kadiwar and Yuspa, S.H. (1994) Cancer Res. 54, 5584-5592. [20] Coursen, J.D., Wang, X.W., Foruace, A.J., Jr. and Harris, C.C. (1996) Proc. Am. Assoc. Can. Res. 37, 2. [21] E1-Deiry, W.S., Harper, J.W., O'Conner, P.M., Velculescu, V.E., Canman, C.E., Jackman, J., Pietenpol, J., Burrell, M., Hill, D.E., Wang, V., Wiman, K.G., Mercer, W.E., Kastan, M.B., Kohn, K.W., Elledge, J.E., Kinzler, K.W. and Vogelstein, B. (1994) Cancer Res. 54, 1-6. [22] Yuan, J., Shaham, S., Ledoux, S., Ellis, H.M. and Horvitz, H.R. (1993) Cell 75, 641-652. [23] Enari, M., Talanian, R.V., Wong, W.W. and Nagata, S. (1996) Nature Biotechnol. 380, 723-726. [24] Nicholson, D.W. (1996) Nature Biotechnol. 14, 297-301. [25] Lints, R. and Driscoll, M. (1996) in: Cellular Aging and Cell Death, pp. 235-253, Wiley-Liss, New York. [26] Yuan, J. and Horvitz, H. (1992) Development 116, 309-320. [27] Mariggio, M.A., Fulle, S., Calissano, P., Nicoletti, I. and Fano, G. (1994) Neuroscience 60, 29-35. [28] Van Eldik, L.J. and Hu, J. (1996) Biochim. Biophys. Acta 1313, 239-245. [29] Donato, R. (1996) Biochim. Biophys. Acta 1313, 258-267. [30] Hengartner, M.O., Ellis, R.E. and Horvitz, H.R. (1992) Nature (London) 356, 494-499. [31] Hockenbery, D.M., Oltvai, Z.N., Yin, X.M., Milliman, C.L. and Korsmeyer, S.J. (1993) Cell 75, 241-251. [32] Distelhorst, C. and McCormick, T. (1996) Proc. Am. Assoc. Cancer Res. 37, 23. [33] Ichimiya, M., Liu, H., Amstad, P., Chang, S., Berezesky, I. and Trump, B. (1996) Shock, in press. [34] Faux, M.C. and Scott, J.D. (1996) Cell 85, 9-12. [35] Davis, M.A., Smith, M.W., Chang, S.H. and Trump, B.F. (1995) The Toxicologist 15, 1599. [36] Davis, M.A., Smith, M.W., Chang, S.H. and Trump, B.F. (1994) Toxicol. Pathol. 22, 595-604. [37] Pentilla, A. and Trump, B.F. (1974) Science 185, 277-278.
Biochi~ic~a et BiophysicaA~ta
ELSEVIER
Biochimica et Biophysica Acta 1313 (1996) 173-178
The role of altered
[Ca2+]i regulation in apoptosis,
oncosis, and necrosis
Benjamin F. Trump *, Irene K. Berezesky Department of Pathology, University of Maryland School of Medicine, 10 South Pine Street, Baltimore, MD 21201, USA Received 17 June 1996; accepted 19 June 1996
Abstract Understanding the processes and events that occur when a cell undergoes a prelethal injury or that lead the cell to death following a lethal injury has been the aim of our research for a number of years. Throughout this period much has been learned, recently at rapid rates, not only by us but by many other investigators as well. Based on the data gathered, we proposed a working hypothesis over a decade ago and have since continually updated it as new experimentation is performed. Our laboratory has focused particularly on the role of cytoplasmic ionized calcium ([Ca2+ ]i) and the effects of its deregulation on prelethal events, including oncosis and apoptosis, and lethal events (necrosis) following cell death. [Ca2+]i appears to be a major link and signalling event. Understanding the mechanisms involved by using a variety of in vivo and in vitro models, coupled with state-of-the-art methodologies, should now allow us to prevent cell death by killing cells when necessary through gene therapy and cancer chemotherapy. Keywords: Cytosolic calcium; Cell injury; Cell death; Cell signalling; Gene regulation
1. Introduction The purpose of this paper is to consider the role of altered ionized intracellular calcium ([Ca 2+ ]i) regulation in the pathogenesis and altered physiology of cell injury and cell death, including apoptosis, oncosis, and necrosis. We will consider the responses of mammalian cells to lethal injury and the role of calcium regulation in relation to other intracellular signalling in the events that follow a lethal injury.
2. Definitions of cell injury and cell death Following a lethal injury, three phases of cellular reaction occur: the prelethal phase; the phase of cell death; and the phase of necrosis (for reviews, see refs. [1,2]). The prelethal phases include changes that occur prior to the 'point of cell death' and are potentially reversible. Two principal patterns of change are currently identified, namely, oncosis and apoptosis.
Abbreviations: CPP32, apopain; CrmA, cowpox virus serpin; ICE, interleukin-1/3 converting enzyme; JNK, Jun N-terminal kinase; MAP, mitogen-activated protein; ARP, polyadenylate ribose polymerase. * Corresponding author. Fax: + 1 (410) 7063743.
In oncosis [3-7], the cells undergo rapid swelling with dilution of the cytosol, formation of cytoplasmic blebs at the plasmalemmal surface, dilatation of the endoplasmic reticulum (ER), mitochondrial condensation followed by inner compartment swelling, dilatation of ER and Golgi components, and nuclear chromatin clumping. In the case of types of injury that can be removed, such as complete anoxia or ischemia, all of these changes rapidly reverse to the normal phenotype. In contrast, the cells in apoptosis are typically shrunken with densification of the cytosol, more extreme clumping of nuclear chromatin, formation of even larger cytoplasmic blebs, and nuclear irregularities; changes in the mitochondria are similar with initial condensation followed by swelling [8]. At the time of cellular death, the changes can be dramatic by time-lapse cinematography (Trump, B.F., unpublished observations) and include interruptions of the continuity of the plasma membrane with escape of cytosolic enzymes to the extracellular space and staining of the nuclei with Trypan blue or the fluorescent dye, propidium iodide. In the phase of necrosis, the changes are similar following either apoptosis or oncosis with the exception that, in the case of apoptosis occurring in vivo, many of the changes occur in fragments of cells or intact cells within the phagolysosome of adjacent parenchymal or mesenchymal cells.
0167-4889//96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. PH S 0 1 6 7 - 4 8 8 9 ( 9 6 ) 0 0 0 8 6 - 9
174
B.F. Trump, L K. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
In necrosis, the changes continue to progress [1,2]. Interruptions in plasmalemmal continuity are seen and often specialized surface structures, such as microvilli, are totally distorted and form complex interlocking whorls resembling myelin figures. The blebs still persist, although many blebs have undergone fragmentation and appear as nearby debris. The nuclear chromatin clumping continues and, in addition, karyolysis begins with increased dissolution of nuclear DNA. The mitochondria remain markedly swollen, and contain one or two types of densities. The first type, so-called flocculent densities, are common to all types of necrosis and consist of aggregated dense intramatrical material composed of denatured matrical proteins; their accumulation is a reliable morphologic indicator of cell death and necrosis. In the second type, precipitates of calcium phosphate occur following types of injury that do not directly interfere with mitochondrial calcium uptake, and, in large amounts, transform to calcium hydroxyapatite crystals.
3. Experimental approach We have followed an experimental approach to elucidate the role of calcium [4,5,7] and cell injury as follows: a series of models, including toxic and others, were chosen to represent the major types of initial injury seen in human disease. These models have included ischemia; anoxia; treatment with carcinogens; and treatment with model toxins with primary characterized interactions such as inhibition of electron transport and oxidative phosphorylation; changes in cell membrane permeability, such as activation of the C5-9 components of complement; use of ionophores including those for calcium; modification of calcium transport in internal stores, such as thapsigargin; oxidant injury using hydrogen peroxide or superoxide generated by the xanthine/xanthine oxidative system; and transfection of genes to modify the reaction to injury or to modify cellular defenses. In many cases, these have been studied both in vivo and in vitro whenever possible. In general, however, to utilize modem technology including fluorescent probes and related technologies, many of the experiments have been carried out on primary cultures or cell lines in vitro. Much of the emphasis of our laboratory has been on studies of the proximal tubule of the kidney; however, we will compare these findings with those experiments performed by other laboratories and those using other cell systems.
4. Relationship of changes to altered [Ca z + ]i 4.1. Source of increased [Ca 2 +]i
Following injury, increases of [Ca2+] i can be rapid, occurring in some cases within seconds. The increased
[Ca2+] i can result from influx of calcium from the extracellular space, from redistribution of intracellular stores, or often from both sources including so-called calcium-induced calcium release [9]. Increased levels of [Ca2+] i are commonly measured by preloading the cells with fluorescent probes such as Fura 2 for digital imaging fluorescence microscopy (DIFM) or Fluo-3 and related compounds for confocal microscopy [10,11]. The [Ca2+]i can be imaged or ratioed and measurements plotted directly using a photon counter. 4.2. Oncosis 4.2.1. Swelling Cytoplasmic swelling in oncosis is marked and begins at an initial rate which is dependent on the type of injury. With direct damage to the cell membrane, e.g., complement or mechanical damage, this occurs very rapidly; with inhibition of energy metabolism limiting the supply of ATP and inhibiting the Na+/K+-ATPase, the change occurs progressively, but more slowly [12]. The swelling is due primarily to increased Na ÷ accompanied by increased C1- and followed by water with accompanying loss of K ÷. As [Na2+] i increases in many cells including the renal epithelium, N a + / C a 2÷ exchange is limited and [Ca2+]i increases as well. Obviously, with inhibition of ATP synthesis in mitochondria and glycolysis, the plasmalemmal Ca2+-ATPase is also inactive and, with direct membrane damage, direct Ca 2÷ influx from the extracellular space Occurs.
4.2.2. Bleb formation Cytoplasmic bleb formation occurs very early during the development of oncosis. These blebs are typically clear and often have a band of actin at the base. Using time-lapse cinematography, these blebs can be seen to form and reform several times during the immediate injury response. Morphologically, these blebs are associated with substantial modification of the cytoskeleton, including both tubulin and actin. By electron microscopy, the blebs are clear with no evidence of cytoskeletal elements between the remaining cytosol and the plasma membrane. Using immunocytochemistry, it is apparent that the actin appears to detach from the cell membrane forming a band across the base of the bleb. As these blebs are observed to expand and contract and often detach, this is considered in our hypothesis to relate to contraction and relaxation of actin and modification of tubulin prior to release. It is our hypothesis that these actin changes are related to protease activity, including Ca 2+- activated proteases, as pretreatment with protease inhibitors greatly modifies the bleb formation [13]. In addition calcium-binding proteins, such as calmodulin and S-100 proteins, may contribute. A threshold of 300-400 mm [Ca 2+ ]i has been suggested for bleb formation [14].
B.F. Trump, LK. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
4.2.3. Mitochondrial changes The initial condensation of mitochondria has been correlated with inhibition of ATP synthesis. The later inner compartment swelling is the consequence of the mitochondrial permeability transition which is induced by increased [Ca2 + ]i.
175
tion appears to be of different pattems and different cell types and may or may not always be calcium-regulated. Intemucleosomal fragmentation may be a late event. An early formation of large fragments apparently requires protease activity in addition to DNAse.
4.4. Necrosis 4.2.4. Nuclear changes In the early stages of oncosis, there is evident nuclear chromatin clumping along the nuclear envelope and around the nucleolus, resulting in clarification of the remaining nucleoplasm. This aggregation is clearly reversible and restoring or adding the injury can result in several reversals of this phenomenon. Although the mechanism of this phenomenon is presently unknown, it is clear from a number of our experiments that reduction of intracellular pH can result in nuclear chromatin condensation and since reduction of cytosolic pH is a very common early reaction, we hypothesize that pH reduction through protein DNA interactions results in this change. 4.3. Apoptosis 4.3.1. Cellular shrinkage Cellular shrinkage undoubtedly is due to loss of ions, predominantly K + and CI- followed by water ion regulation and has been shown to require phosphorylated ion transport systems; furthermore, the K ÷ channels in many cells are calcium-sensitive. Our hypothesis is that early increases of [Ca 2+ ]i stimulate K + loss and shrinkage prior to cell death. It has been shown by Bortner and Cidlowski [15] that lymphocytes expressing Bcl-2 become resistant to dexamethasone-induced apoptosis, while lymphocytes exposed to hypertonic conditions without additional Bcl-2 initiate DNA fragmentation. Thus, this area of volume regulations may be an important defense and control mechanism involving cell death, but requires additional study. 4.3.2. Cytoplasmic blebs The bleb formation in apoptosis also undoubtedly involves regulation of cytoskeletal membrane attachment, but the blebs are different from those seen in oncosis in that they contain abundant cytoplasm and cytoplasmic organelles rather than being watery. Further detailed investigations are needed on the role of cytoskeletal calcium interactions. 4,3.3. Mitochondrial changes Mitochondrial changes are similar to those seen with oncosis except that the shrinkage of mitochondria in the condensed phase is somewhat more severe. Mitochondria undergo swelling prior to cell death in apoptosis. 4.3.4. Nuclear changes Nuclear changes include extreme chromatin clumping and irregularity of nuclear contour. The DNA fragmenta-
4.4.1. Plasmalemmal interruption and blebbing The interruption of plasmalemmal continuity which results in trypan blue and propidium iodide staining is deemed to be secondary to Ca 2+ activation of phospholipase A phosphorylation through the action of kinases such as MAP kinase. In time-lapse cinematography of apoptosis and oncosis, we have observed that one or more large terminal blebs often form around the time of cell death. These may well have a relationship to the breaks in plasmalemmal continuity seen by electron microscopy. 4.4.2. Mitochondrial changes In necrosis, the mitochondria remain swollen and exhibit aggregation of matrical proteins, so-called 'flocculent densities'. In addition, there is fragmentation of the outer membrane. In certain types of injury, calcium phosphate deposits occur in the matrix in addition to the flocculent densities. This involves active uptake of Ca 2+ by the mitochondria from the cytosol and, therefore, is limited to injuries such as complement activation or other primary modification of the plasmalemma which leave the mitochondrial transport capability intact for a time. 4.4.3. Nuclear digestion As the cell passes the 'point of no return', the clumped nuclear chromatin is progressively digested, resulting in blurred profiles in electrophoresis and loss of staining with cationic probes or dyes. This has been termed 'karyolysis' as ultimately only a nuclear ghost remains.
5. Genetic regulation and signalling 5.1. Pro-death genes 5.1.1. Immediate-early genes The induction of a series of intermediate-early genes including c-fos, c-jun, NFKB, and c-myc characterize many types of acute lethal injury. These genes can be induced within 15 min or less, as in the case of c-fos and c-jun, and are evidently induced even in cells which are destined to die within 1 or 2 h. Following at least some types of injury, e.g., oxidant stress, the induction of these genes is Ca2+-dependent [16]. The function of these early genes, apart from inducing later genes including metalloproteases,
176
B.F. Trump, I.K. Berezesky/ Biochim&aet BiophysicaActa 1313 (1996) 173-178
is not clear in the sense that in some cases these may lead to pathways of cell death while in others they may be preparatory for cell division and regeneration in the event that cell death does not occur. C-jun, in particular, has been identified as a pathway toward apoptosis and under some conditions overexpression of c-myc may be as well. This is further emphasized in the discussion of the kinase pathways below. Work with antisense RNA in some systems has shown that failure to express c-jun can protect against cell death. Somewhat later, there is Ca2+-dependent induction of hsp70 which appears to protect against subsequent lethal injury [17].
5.1.2. p53 - gadd and waf
Among the functions of wild-type p53 is induction of mitotic arrest and acute cell death [18]. Microinjection of p53 into cultured cells commonly results in apoptosis and cell killing. Mutant p53, as seen in many forms of human neoplasia, does not exhibit this property. The precise mode of action of p53 is similarly not well understood. Although the relationship between p53 status and [Ca2+] i is incompletely understood, Weinberg et al. [19] noted that in combination with v-rasHa, keratinocytes deficient in p53 shared decreased Ca2+-induced terminal differentiation. Among the later genes induced by activated p53 are gadd45 and waf(p21) [20,21]. In some experiments, waf(p21) has been shown to be involved in initiation of cell killing, though its exact relationship to the final or execution phase is not known.
5.1.3. Proteases and phospholipases
Beginning with the identification of the pro-death gene ced -3 in the nematode C. elegans, an entire family of proteases has been identified which appear to be late in the pathway leading to cell death [22-24]. The role of these proteases and cytoskeletal alterations and bleb formation was mentioned above. It has been shown that protease inhibitors can protect against both apoptosis and oncosis. The exact relationships in the final execution phase remain under study.
5.1.4. The Degenerin gene family
Although most of the programmed cell death in C. elegans development involves apoptosis, mutations in two genes, mec-4 and deg-1, result in oncosis followed by cell death (see review by Lints and Driscoll [25]). Lints and Driscoll [25] have suggested that these degenerin proteins might constitute an ion channel which, in the presence of mutations, may remain open. In this way, therefore, it would be comparable to activation of C5-9 or the effects of PCMBS in mammalian cells. Although the characteristics of the channel are presently unknown, the model proposed by Lints and Driscoll would clearly permit the
rapid entry of Ca 2+ from the extracellular space and result in the consequences mentioned above.
5.1.5. Others
There is an expanding list of other genes whose function appears to relate to cell death. One of these, ced-4, described in nematodes has been suggested to be a calcium-binding protein with an EF hand-like domain [26]. Its exact function and relationship to [Ca 2÷ ]i remain to be determined. Other calcium-binding proteins, including the S-100 family, have been observed to initiate apoptosis [27-29].
5.2. Anti-death genes
Experiments in nematodes have also indicated the presence of anti-death genes. These led to the identification of ced-9 and later its mammalian homologue, bcl-2, which now are seen as part of a rapidly growing family of genes determining cell survival or cell death [30]. The gene bcl-2 has received considerable attention regarding its mode of action. The dimensions of its protective action remain to be completely defined but we suggest that it protects against both apoptosis and oncosis. Mechanisms that have been suggested for its protective action include anti-oxidant properties, interference with signalling pathways, and modification of intracellular calcium regulation [31,32]. We have studied NRK cells (normal rat kidney proximal tubular epithelium) transfected with bcl-2 in order to examine the reaction to injury. Clearly it is protective against oxidant injury and clearly the response of [Ca2+] i is greatly modified [33]. We suggest that bcl-2 modifies the regulation of [Ca2+]i possibly involving Ca transport systems in mitochondria, plasma membrane, or ER.
5.3. Cell death signalling pathways
Clearly, the signalling pathways that lead through oncosis or apoptosis to cell death need to be determined. Progress is so rapid in this field that it is difficult to make any predictions. However, our current knowledge of immediate-early genes and protein kinase pathways provide important clues as does the determination of the morphology seen in apoptosis and oncosis. Experience with timelapse cinematography of oncosis and apoptosis clearly show marked morphologic changes during the earliest phases. Linkage between plasmalemmal and cytoskeletal change may well provide a direct linkage between morphology and signalling. Faux and Scott [34] recently pointed out that 'approx. 300 of the anticipated 2000 protein kinase and 1000 phosphatase genes have been identified' and proposed a morphologic/functional model of signalling that depends on kinase anchoring and scaf-
B.F. Trump, L K. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
~ Plasma Membrane Permeability
I
/
Mitochondrla ~
Swelling ..9_.._ (Oncosls) ~ [Na+]I
;,
Endonuc eases t DNA strand breaks ~
i '
A 53
I
~ [ ATP ] i
PARP '
t GrowthFactor Endoplasmlc Retlculum
.~_[~ [Ca2+] i i
;
Proteases [
177
~
I
K+ "1 .. Shrinkage Cl'-]erlrlux'-~'(Ap°pt°sls) I CytTkeleton Blebs
1'
"÷
Kinases-JNK,
MAP, ealmodulln [
I I
Transcription Phosphollpase A2 Factors [ DNA Repa ir fos,jun,myc = Poly ADP rlbosylatlon ~ CrmA = war 1(p21) ? .L m Proteases = gadd45 "-] T ced-3 bcl 2 ICE/CPP32 I Cell Death I (ced 9)
Fig. 1. Diagram illustrating our current working hypothesis of the major events leading from cell injury to cell death, focusing on the role of lea 2+ ]i- Some of the major categories of injury are shown at the top of the figure. Decreased and absent [ATP]i commonly results from complete ischemia, anoxia, or treatment with inhibitors of ATP synthesis including potassium cyanide, carbon monoxide, fluoride, and iodoacetic acid. Increased plasma membrane permeability commonly results from the immune response, e.g., activation of complement or T-cell release of perforin, or from treatment with a variety of toxins including mercuric salts, ionophores, or mechanical damage. Growth factor deprivation is often studied in vitro though in vivo it does occur with neuronal damage. In oncosis, the early events involve cell swelling because of increased [Na+]i accompanied by C1- and water. This can result either from increased plasmalemmal permeability or decreased [ATP]i resulting in inactivation of Na +, K+-ATPase. Increased [Na+]i can also contribute to increased [Ca2+ ]i through decreased Na+/Ca 2+ exchange. In apoptosis, the initial events include cellular shrinkage, implying loss of ions in particular K + and CI-. The mechanism of the shrinkage is not presently known although increased [Ca2+ ]i might contribute through activation of K + channels. The initial increase of [Ca2+ ]i can result from influx from the extracellular space or from redistribution from intracellular stores. Following the increase, several prinicipal pathways seem to le~d to the reversible and irreversible changes as shown. Also illustrated in the diagram are the events that follow Ca2+-mediated activation of protein kinases, endonucleases, proteases, and phospholipases as discussed in the text. Our current hypothesis of altered gene expression related to cell death is shown near the bottom of the diagram.
fold proteins. Of the anticipated 1000 phosphatase genes, we have observed that okadaic acid provides an excellent model of apoptosis and that modification of phosphatases 1 and 2A result in dramatic morphologic and biochemical changes [35,36].
6. Hypothesis and conclusions Our current model of lethal cell injury and cell death is illustrated in Fig. 1. Deregulation of [Ca2+]i is shown as central to the signalling involved following several classes of lethal injury: inhibition of ATP synthesis as in the case of anoxia, ischemia, or mltochondrial inhibitors; increased plasmalemmal permeability as in the case of complement activation, cell-mediated killing, or many chemical toxins; or withdrawal of growth or trophic factors. In oncosis, the early cellular changes include increased [Na÷]i with concomitant increase o f water content and cellular swelling. This results from loss of A T P to power the N a ÷ / K + - p u m p s a n d / o r increased N a ÷ influx through the damaged plasmalemma. In apoptosis, the cells become shrunken and dense due to water loss, presumably the secondary result of efflux of intracellular ions such as K ÷ and C1-. This probably implies that A T P concentrations and synthesis remain active longer in this type of prelethal injury.
Increased [Ca 2" ]i can result from influx from the extracellular space or from redistribution from other intracellular compartments such as the ER. A variety of effects follow such deregulation. These include activation of proteases, endonucleases, and phospholipases, all of which contribute to the cellular changes that occur during the prelethal phase. Some of the major activities are iUustrated. Of special note are the changes that lead to activation of genes a n d / o r gene products that trigger additional events, such as the ced 3 / I C E protease family members, which appear to be close to the final phase of cell death. Also of note is that the lethal Ca2+-induced vents are antagonized by reducing pH [37]. In conclusion, much is being learned about the processes and events that lead to cellular death. Deregulation of [Ca 2+ ]i is an important link and signalling event in cells following a variety of injuries. Further understanding of these links should lead to improvements in prevention of cell death as well as to improved methods for killing cells as in cancer chemotherapy.
Acknowledgements Supported by NIH Grant DK15440 and Navy N0001491-J-1863. This is contribution No. 3817 from the Cellular Pathobiology Laboratory.
178
B.F. Trump, I.K. Berezesky / Biochimica et Biophysica Acta 1313 (1996) 173-178
References [1] Trump, B.F., McDowell, E.M. and Arstila, A.U. (1980) in: Principles of Pathobiology, Third Edn. (Hill, R.B., Jr. and LaVia, M.F., eds.), pp. 20-111, Oxford University Press, New York. [2] Trump, B.F. and Berezesky, I.K. (1992) in: Cardiovascular Toxicology, Second Edn., Ch. 4 (Costa, D., Jr., ed.), pp. 75-113, Raven Press, New York. [3] Von Recklinghausen, F. (1910) Untersuchungen uber Rachitis und Osteomalacie. Verlag Gustav Fischer, Jena. [4] Trump, B.F. and Berezesky, I.K. (1992) Curr. Opin. Cell Biol. 4, 227-232. [5] Trump, B.F. and Berezesky, I.K. (1995) FASEB J. 9, 219-228. [6] Majno, G. and Joris, I. (1995) Am. J. Pathol. 146, 3-15. [7] Trump, B.F. and Berezesky, I.K. (1996) New Horiz. The Science and Practice of Acute Medicine 4, 139-150. [8] Wyllie, A.H., Kerr, J.F.R. and Currie, A.R. (1980) Int. Rev. Cytol. 68, 251-306 [9] Thastrup, O., Cullen, P.J., Drobak, B.K., Hanley, M.R. and Dawson, A.P. (1990) Proc. Natl. Acad. Sci. USA 87, 2466-2470. [10] Smith, M.W., Phelps, P.C. and Trump, B.F. (1991) Proc. Natl. Acad. Sci. USA 88, 4926-4930. [11] Smith, M.W., Phelps, P.C. and Trump, B.F. (1992) Am. J. Physiol. 262 (Renal Fluid Electrolyte 31), F647-F655. [12] Papadimitriou, J.C., Drachenberg, C.B., Shin, M.L., Smith, M.W. and Trump, B.F. (1994) Virchows Arch. 424, 677-685. [13] Elliget, K.A., Phelps, P.C. and Trump, B.F. (1991) Cell Biol. Toxicol. 7, 263-280. [14] Phelps, P.C., Smith, M.W. and Trump, B.F. (1989) Lab. Invest. 60, 630-642. [15] Bortner, C.D. and Cidlowski, J.A. (1996) Am. J. Pathol, in press. [16] Maki, A., Berezesky, I.K., Fargnoli, J., Holbrook, N.J. and Trump, B.F. (1992) FASEB J. 6, 919-924. [17] Yamamoto, N., Smith, M.W., Maki, A., Berezesky, I.K. and Trump, B.F. (1994) Kidney Int. 45, 1093-1104. [18] Harris, C.C. (1993) Science 262, 1980-1981.
[19] Weinberg, W.C., Assoli, C.G., Kadiwar and Yuspa, S.H. (1994) Cancer Res. 54, 5584-5592. [20] Coursen, J.D., Wang, X.W., Foruace, A.J., Jr. and Harris, C.C. (1996) Proc. Am. Assoc. Can. Res. 37, 2. [21] E1-Deiry, W.S., Harper, J.W., O'Conner, P.M., Velculescu, V.E., Canman, C.E., Jackman, J., Pietenpol, J., Burrell, M., Hill, D.E., Wang, V., Wiman, K.G., Mercer, W.E., Kastan, M.B., Kohn, K.W., Elledge, J.E., Kinzler, K.W. and Vogelstein, B. (1994) Cancer Res. 54, 1-6. [22] Yuan, J., Shaham, S., Ledoux, S., Ellis, H.M. and Horvitz, H.R. (1993) Cell 75, 641-652. [23] Enari, M., Talanian, R.V., Wong, W.W. and Nagata, S. (1996) Nature Biotechnol. 380, 723-726. [24] Nicholson, D.W. (1996) Nature Biotechnol. 14, 297-301. [25] Lints, R. and Driscoll, M. (1996) in: Cellular Aging and Cell Death, pp. 235-253, Wiley-Liss, New York. [26] Yuan, J. and Horvitz, H. (1992) Development 116, 309-320. [27] Mariggio, M.A., Fulle, S., Calissano, P., Nicoletti, I. and Fano, G. (1994) Neuroscience 60, 29-35. [28] Van Eldik, L.J. and Hu, J. (1996) Biochim. Biophys. Acta 1313, 239-245. [29] Donato, R. (1996) Biochim. Biophys. Acta 1313, 258-267. [30] Hengartner, M.O., Ellis, R.E. and Horvitz, H.R. (1992) Nature (London) 356, 494-499. [31] Hockenbery, D.M., Oltvai, Z.N., Yin, X.M., Milliman, C.L. and Korsmeyer, S.J. (1993) Cell 75, 241-251. [32] Distelhorst, C. and McCormick, T. (1996) Proc. Am. Assoc. Cancer Res. 37, 23. [33] Ichimiya, M., Liu, H., Amstad, P., Chang, S., Berezesky, I. and Trump, B. (1996) Shock, in press. [34] Faux, M.C. and Scott, J.D. (1996) Cell 85, 9-12. [35] Davis, M.A., Smith, M.W., Chang, S.H. and Trump, B.F. (1995) The Toxicologist 15, 1599. [36] Davis, M.A., Smith, M.W., Chang, S.H. and Trump, B.F. (1994) Toxicol. Pathol. 22, 595-604. [37] Pentilla, A. and Trump, B.F. (1974) Science 185, 277-278.