[Frontiers in Bioscience 3, d25-43, January 1, 1998]|
HOW DO INTRACELLULAR PROTEOLYTIC SYSTEMS CHANGE WITH AGE?
Ana Maria Cuervo and J. Fred Dice
Department of Physiology Tufts University School of Medicine, Boston, USA
Received 12/5/97 Accepted 12/9/97
4. INTRACELLULAR PROTEOLYTIC SYSTEMS AND AGING
Senescent human fibroblasts in culture have been very helpful in identifying the defective proteolytic systems with age. General techniques for the study of intracellular protein degradation in cultured cells, such as metabolic radiolabeling of intracellular proteins or microinjection of radiolabeled proteins, and further analysis of their intracellular breakdown have demonstrated that total protein degradation was considerably slowed in old fibroblasts when compared with young cells (19, 50-53). The addition of specific protease inhibitors to the culture medium and/or the analysis of protein breakdown under specific cellular conditions has been used to implicate the protease or proteases responsible for the decrease in protein degradation rates with age.
In recent years, several cell-free models have been developed to analyze independently the function of some of the major proteolytic systems of cells. Thus, ubiquitination of proteins can be accomplished in vitro with rabbit reticulocyte or Xenopus egg extracts (54, 55), the 20S proteasome has been isolated from several different organisms and its proteolytic activities carefully characterized (56, 57), endosomal/ lysosomal fusion can be reproduced with isolated vesicles in vitro (58), and direct transport of proteins through the lysosomal membrane can be followed in isolated lysosomes (59-61).
4.1. Cytosolic proteolytic systems: Proteasomes and calpains
4.1.1. Ubiquitin-proteasome system
Several proteases have been described in the cellular cytosol. The 20S proteasome, a multicatalytic protease complex, is without doubt one of the most studied and best characterized cytosolic proteases. The 20S proteasome is normally found associated with several heterogeneous proteins that form one or two 19S caps at the ends of the 20S structure, that regulate its proteolytic activity (62). This complex is called the 26S proteasome (reviewed in ref. 63). The nature of many of these regulatory subunits remains unclear, but ATPase, ubiquitin-binding, and ubiquitin isopeptidase activities have been detected (64).
The 26S proteasome has been considered mainly responsible for the degradation of short-lived and abnormal proteins in the cytosol and nucleus (reviewed in ref. 65). However, recent evidence for the participation of this proteolytic system in basal protein turnover of cytosolic proteins has also been presented (66). In addition, not only cytosolic proteins, but also previously compartmentalized proteins, can be redirected to the cytosol where they are degraded by the 26S proteasome (67). Degradation of proteins by the proteasome usually requires a previous covalent modification of the substrate protein in order to be recognized by the proteolytic complex. This modification, known as ubiquitination, consists of the covalent attachment of a 8.6 kDa protein (ubiquitin) to the epsilon-amino group of a lysine residue of the substrate protein (reviewed in ref. 68). Covalent attachment of ubiquitin to the protein substrate is mediated by a complex group of enzymes (activating enzyme (E1), carrier/conjugating enzymes (E2) and conjugating enzymes (E3)) (see figure 1) (69). Ubiquitin can also be covalently attached to itself to form a ubiquitin chain on a substrate protein that is then targeted for degradation by the 26S proteasome (70). Degradation of polyubiquitinated proteins seems to be facilitated by the direct interaction of some subunits of the proteasome with the multiubiquitin chain (71, 72). Other enzymes are able to remove ubiquitin from proteins (deubiquitination), allowing it to recycle (73, 74). Ubiquitination is required for degradation of many important regulatory proteins such as cyclins, c-fos, and p53 (68).
Figure 1. The ubiquitin-proteasome system. Most short-lived and abnormal proteins in the cell cytosol are degraded by the 26S proteasome, directly or after a covalent modification known as ubiquitination. The multiple enzymatic steps and several enzymes required for the ubiquitination of proteins are summarized in this figure. See the text for details.
Although ubiquitination is the most generalized target signal for protein degradation by the 26S proteasome, ubiquitin-independent degradation of some proteins by the proteasome has also been reported (i.e. ornithine decarboxylase) (75). On the other hand, monoubiquitination of some membrane proteins in their cytosol-exposed region targets them for degradation by the vacuolar (lysosomal) system in yeast (76, 77).
Regarding the proteolytic core of the proteasome, at least three different neutral peptidase activities have been described (78). Recently, specific inhibitors for this multicatalytic complex have been developed (79-81). The physiological activators described for the 20S proteasome are located at the 19S ATPase complex. In addition, in the presence of detergent, fatty acids, or polylysine, the activity of the proteasome significantly increases (82), but the physiological significance of such regulation is not known. The proteasome normally exists in a latent state in the cell (65) which may be important in preventing uncontrolled proteolysis (78). Lately, several authors have demonstrated that the subunit composition of the proteasome changes depending on specific stimuli, and that these altered subunits result in changes in the various proteolytic activities (83).
A decrease in a non-acidic protease activity with age has been reported in several types of cells (84, 85). The 26S proteasome is one possible candidate for that proteolytic activity. However, no clear evidence for such age-related changes in proteasome activity has been found. A detailed analysis of each of the peptidase and proteolytic functions of the 20S proteasome along with its intracellular levels in senescent cells has been performed by Shibatani and Ward (86). These authors, working with rat cytosolic extracts, found no differences with age in the intracellular levels or the total proteolytic activity of the 20S proteasome (tested with casein, a well established substrate). Only after SDS stimulation was an increase in the chymotrypsin and trypsin-like activities found in the aged hepatocytes. Under those conditions the peptidylglutamyl peptide-hydrolyzing activity of the 20S proteasome was clearly decreased, but that has been shown to be of minor importance compared with the other proteasome activities, at least in yeast (87, 88). Therefore, Shibatini and Ward (86) concluded that the reported changes in 20S proteasome activity with age are not sufficient to account for the decreased protein degradation in old cells.
Similar results were obtained by Sahakian et al. (89) working with purified rat liver 20S proteasome instead of cellular extracts. In addition, the decrease in non-acidic protease activity with age, first described in rat heart and liver, seems to be restricted to some organisms and tissues since recent studies have shown that this activity is not modified in houseflies or in rat brain with age (90).
In the absence of changes in the proteasome activity, a reduction in protein ubiquitination could also result in reduced intracellular protein breakdown. Several changes in intracellular protein ubiquitination with age have been reported. Studies in senescent human fibroblasts revealed that aged cells have less free ubiquitin and more ubiquitin-protein conjugates than young cells (91). Similar changes have also been observed in lens (92) and in brain (93). However, levels of ubiquitin mRNA do not change with age, and there is no difference in the ability of senescent fibroblasts to degrade ubiquitinated proteins (91). The fact that changes in protein ubiquitination do not always indicate changes in protein degradation rates may reflect the participation of ubiquitin in intracellular processes other than protein degradation (94-96). Although these results indicate that a reduction in the ubiquitin-proteasome pathway is not a common finding in aging, this proteolytic pathway may decline in certain tissues under particular conditions. For example, as described in detail below, under oxidizing conditions, a failure in the ubiquitination process in senescent eye lenses seems to contribute to the accumulation of abnormal proteins (97).
Calpains, or calcium-dependent proteases, constitute the other major cytosolic proteolytic system (reviewed in ref. 98). The activity of these neutral, thiol proteases is tightly regulated by intracellular calcium levels. Micromolar calcium concentrations activate micro-calpain, but millimolar calcium concentrations are required for milli-calpain activation (99). Where in the cell millimolar calcium concentrations may be achieved remains an unsolved problem. Translocation of calpains to the cell membrane and their limited autolysis once there result in calpain activation (100, 101) (figure 2). Calpains partially degrade membrane and cytoskeletal proteins and several membrane-associated enzymes (102-104). The limited proteolysis of certain membrane proteins by calpains has been shown to be necessary for the process of membrane fusion (105). Two of the best characterized substrates for the calpain system are the band 3 protein (an anion exchanger) in the erythrocyte membrane (99) and protein kinase C (106).
Figure 2. Calcium-dependent proteases. Changes in intracellular calcium levels modulate the activity of two cytosolic thiol proteases (calpains). Those proteases are usually present in an inactive form associated with calpastatin, an intracellular inhibitor. After translocation to the cell membrane and limited autolysis, calpains are activated. Calpains are responsible for the degradation of certain membrane and cytoskeletal proteins. See the text for details.
An increase in proteolysis of band 3 protein by calpains in erythrocytes of old individuals has been described (107). This increased degradation of band 3 protein correlates with a higher translocation of calpains from the cytosol to the cell membrane. A diminution of thiols in senescent cells, as a consequence of free radical reactions and oxidative damage (108), may cause enhanced calpain translocation to the cell membrane (100). The increase in degradation of band 3, and probably other membrane proteins, modifies the stability of the erythrocyte membrane and reduces erythrocyte life span in aged animals and humans (109). It is likely that this increase in degradation of membrane proteins described in erythrocytes is also common for most cell types. For example, as described below, the abnormal proteolysis and neuronal degeneration in Alzheimer's disease may be, in part, a consequence of calpain activation (110).
The age-related changes in the calpain system seem not to be related with changes in the proteolytic activity of calpains themselves, because when calpains from young individuals were added to membranes from old erythrocytes an increase in membrane protein degradation was also observed (109). Those authors suggest the existence of specific changes in cellular membrane proteins with age which result in an increased affinity of those membranes for calpains thereby bringing the protease close to its potential substrate proteins. Age-related changes in the substrate susceptibility to proteases should also be considered. Several protein modifications make proteins more susceptible to proteolytic attack. Those susceptibility changes have been extensively studied in the brain, and we will review them in more detail in the section of protein degradation in the central nervous system. Finally, possible modifications in the activity of calpastatin, the physiological inhibitor of calpains, with age need also to be considered (101).
4.2. Lysosomal function in senescent cells
Lysosomes are organelles that contain a powerful mixture of proteases, peptidases, and other hydrolases capable of degrading most intracellular and extracellular macromolecules (reviewed in ref. 111). The proteases responsible for the degradation of proteins inside lysosomes are called cathepsins (reviewed in ref. 112). In general, lysosomes have been considered to be responsible for the continuous basal turnover of most intracellular proteins (mainly long-lived proteins) in liver, kidney and certain other tissues.
Several types of evidence have identified lysosomes as one of the major proteolytic systems affected by age: (i) predominantly, degradation of long-lived proteins is retarded in senescent fibroblasts (13, 53) and in livers of aged rats (113); (ii) in most cells there is an age-related increase in number and size of lysosomes (114); (iii) storage bodies similar to those described in senescent cells become evident in young cells after inhibiting lysosomal proteolysis with cathepsin inhibitors (115).
When considering the age-related reduction in lysosomal function, it is necessary to separately analyze the different mechanisms by which proteins can be transported into lysosomes. The best characterized of these mechanisms are: vesicular fusion (endocytosis and crinophagy), macroautophagy, microautophagy and the direct protein transport through the lysosomal membrane (reviewed in ref. 23) (figure 3).
Figure 3. Lysosomal pathways of protein degradation. Intra- and extracellular proteins can reach the lysosomal matrix by different mechanisms: vesicular fusion (endocytosis (1) and crinophagy (2)), macroautophagy (3), microautophagy (4), and direct transport through the lysosomal membrane (5). See the text for details.
4.2.1. Fusion of lysosomes with vesicular compartments
Primary lysosomes are able to completely degrade the protein content of several kinds of intracellular vesicles after membrane fusion. Vesicles containing secretory proteins originate at the Golgi apparatus, and they normally release their protein content to the extracellular medium after fusion of the vesicles with the plasma membrane. Under specific conditions, when a decrease in protein secretion is required, 20-70% of those vesicles directly fuse with lysosomes in a process called crinophagy, after which secretory proteins are completely degraded (116, 117). In spite of the several methods developed to analyze this vesicular traffic (reviewed in ref. 118), the intrinsic mechanisms regulating this fusion process are still unclear.
Fusion of lysosomes with endosomes, vesicles containing several extracellular and plasma membrane proteins, is fairly well-characterized (for review see ref. 119). Internalization of several membrane receptors after ligand binding takes place by receptor-mediated endocytosis into clathrin-coated vesicles and subsequent traffic of those vesicles through the endosomal/lysosomal pathway. In that process membrane receptors may be recycled to the cell surface while the ligand is completely degraded after fusion of late endosomes with lysosomes. In other cases the receptor is not recycled and is delivered to lysosomes for degradation along with the ligand. Endocytosis is also required for uptake of extracellular nutrients, maintenance of cell polarity, and presentation of antigenic peptides. Extensive reviews of the movement of membrane receptors through different intracellular compartments and the signals that trigger receptor internalization and receptor recycling have recently been published (119, 120).
A decrease in levels of some serum proteins (121) and plasma membrane proteins (122) with age has been described. Those preliminary reports suggested some alterations in the endocytic system responsible for the internalization and degradation of those proteins. However, more direct studies of the endocytic system in senescent cells suggest little change in this process with age. An increase in the internalization of latex beads in senescent fibroblasts has been found (123), and a decrease in the endocytic activity of aged fibroblasts and Kupffer cells has also been reported (124, 125). In contrast, internalization and degradation of low density lipoprotein and epidermal growth factor are unmodified in senescent fibroblasts (126, 127). These discrepant results may be related to the different analyses of the experimental data performed or in cell type analyzed in each study. Thus, Gurley and Dice (128) showed that endocytosis values should be expressed per microgram of cell protein instead of normalized per number of cultured cells, since old cells are much larger than young cells. Under those circumstances there was no difference in endocytosis rates for young and old fibroblasts. In addition, rates of fluid-phase and absorptive endocytosis were not modified with age. Thus, there is not a general failure in lysosomal function during senescence.
4.2.2. Macroautophagy and microautophagy
A large number of intracellular proteins can be degraded in lysosomes by a process called macroautophagy (reviewed in ref. 129). In this process, complete regions of the cytoplasm, including cytosolic proteins as well as entire organelles, are surrounded by a membrane to become double-membraned structures known as autophagosomes or autophagic vacuoles (figure 3). The proteins inside autophagosomes are then completely degraded after fusion with lysosomes. In organs such as the liver, macroautophagy accounts for most intracellular protein degradation, especially during early starvation (130). Factors inhibiting this process include amino acids, insulin, cell swelling, and disruption of the cytoskeleton (131, 132). Among the mechanisms controlling macroautophagy, protein phosphorylation plays an important role (133). Although proteins in the sequestered cytoplasm appear to be taken up nonselectively, under certain conditions selectivity in this degradative process has been demonstrated. For example, peroxisomal autophagy can be preferentially activated under conditions where peroxisomes are no longer needed (134, 135). In addition, selective degradation of specific cytosolic proteins in the yeast vacuole following a novel vesicular pathway has been recently described (136). The fusion of lysosomes with the autophagosome follows similar mechanisms as those previously described for the endocytic and secretory pathways, and evidence of interaction between both processes has been presented (137).
During senescence, severe changes in the autophagosome/lysosomal system have been described. Degradation of several proteins microinjected into the cytoplasm of senescent human fibroblasts is significantly lower than in young fibroblasts (20). For some of those proteins (RNase S protein, lysozyme), their degradation is believed to follow the autophagosome/ lysosomal pathway. The half-lives of some of those microinjected proteins are summarized in table 2. A decrease in autophagosome formation as well as a decrease in the degradative activity of lysosomes would result in the slower degradation of substrate proteins. Since autophagic vacuole formation rates decrease in senescent cells (138), and rates of elimination of autophagic vacuoles also decrease with age (139), reductions in autophagosome formation and fusion with lysosomes both contribute to the age-related decline in macroautophagy.
Table 2. Age-related changes in the half-lives of several cytosolic proteins
A second process of intracellular protein sequestration by lysosomes has been called microautophagy. During microautophagy only small portions of cytoplasm are engulfed by small invaginations in the surface of lysosomes. After internalization of those vesicles and breakage of their surrounding membrane, the cytoplasmic proteins are degraded inside lysosomes (140). In contrast with macroautophagy that is mainly activated under conditions of nutrient deprivation, microautophagy is responsible for the continuous basal degradation of long-lived proteins (141). No information is available concerning rates of microautophagy in aging.
4.2.3. Direct transport of cytosolic proteins through the lysosomal membrane.
A third mechanism of lysosomal degradation of certain cytosolic proteins occurs after direct transport through the lysosomal membrane (reviewed in ref. 23). This protein transport resembles in many aspects the transport of proteins into other cellular compartments (142). At least two molecular chaperones and a receptor protein, elements also described for other transport systems, are also present in this lysosomal pathway. The sequence of events that directs specific cytosolic lysosomes to their lysosomal degradation by this pathway is summarized in figure 4. The substrate proteins contain in their sequence a consensus peptide motif biochemically related to the pentapeptide, KFERQ (143, 144), that is recognized by a cytosolic chaperone of 73 kDa (hsc73) (143, 145). The binding of hsc73 to that region of the substrate protein directs the complex toward the lysosomal membrane (59). A lysosomal membrane glycoprotein of 96 kDa (lgp96 in rats or lamp2 in humans and mice) acts as a receptor for the substrate protein (146). Assisted by a form of the cytosolic hsc73 in the lysosomal matrix (147, 148), the substrate protein reaches the lysosomal matrix where it is completely degraded. Binding to the lysosomal membrane and uptake into the lysosomal matrix are the rate-limiting steps in this process and are directly related with lysosomal levels of lgp96 and hsc73, respectively (146, 148).
Figure 4. The hsc73-mediated pathway of cytosolic protein degradation. The main components of this system of direct transport of proteins through the lysosomal membrane are: a cytosolic chaperone of 73 kDa (hsc73), a receptor protein at the lysosomal membrane (lgp96) and an intralysosomal form of the cytosolic chaperone (lys-hsc73). The substrate protein after interaction with hsc73 (1) is directed to lysosomes where after binding (2) and transport through the lysosomal membrane (3), it is completely degraded in the lysosomal matrix (4).
This hsc73-mediated lysosomal pathway was first described in human fibroblasts in culture (51, 149, 150) and then characterized in rat liver (60, 151). The direct transport of proteins into lysosomes is activated by serum deprivation of confluent cultured cells (152) or prolonged starvation of intact animals (153, 61). Approximately 30% of the cytosolic proteins are degraded by this pathway (153). Several of the substrates already identified for this pathway are: ribonuclease A (150), glyceraldehyde-3-phosphate dehydrogenase (60), aldolase (60), some subunits of the 20S proteasome (154), transcription factors such as c-fos (155) and IkB (Cuervo, A.M., Hing, M., Lim, B. and Dice, J.F., unpublished results), and alpha-2-microglobulin, a secretory protein that is also located in the cytosol of hepatocytes and kidney proximal tubular cells (Cuervo, A.M., Hildebrand, H., Bomhard, H. and Dice, J.F., unpublished results).
Analysis of the degradation of proteins microinjected into senescent fibroblasts revealed a decrease in the degradation of ribonuclease A, one of the described substrates for the direct lysosomal pathway (152) (figure 5A). In addition, differences between young and old cells in total protein degradation rates were especially evident in confluent cells after serum withdrawal, a condition under which the hsc73-mediated transport of proteins into lysosomes is activated.
Figure 5. Evidence for reduction of the hsc73-mediated pathway with age. Decrease in the degradation of microinjected RNase A in old fibroblast when compared with young fibroblasts, especially after serum deprivation (A), and a reduced ability for the direct uptake of this protein by isolated liver lysosomes from old rats (B) strongly support a reduction of the hsc73-mediated pathway of protein degradation with age.
Working with lysosomes isolated from old rats we have also observed a decrease in their ability to directly take up substrate proteins for this pathway (Cuervo and Dice, unpublished results) (figure 5B). All those results suggest that the direct transport of cytosolic proteins through the lysosomal membrane is impaired in old cells, and this reduction is at least partially responsible for the age-related decrease in total protein degradation. The components of the hsc73-mediated pathway that are affected by age are currently being studied in our laboratory.
In conclusion, the analysis of the effects of senescence on the lysosomal function reveals that, rather than a generalized failure in the lysosomal system with age, specific degradative mechanisms are impaired in senescent cells. The important role of lysosomes in bulk protein turnover in liver, kidney, brain and fibroblasts make them mainly responsible for the age-related decrease in protein degradation in those organs.