Pathological changes consistent with apoptosis have been described in the PNS in models of diabetic neuropathy. Several pathological changes are observed in the diabetic DRG and dorsal roots (Fig. 2). Condensation of chromatin, shrinkage of the nucleus and cell cytoplasm, with preservation of the cytoplasmic membrane are seen in both DRG neurons as well as in satellite cells or SC from diabetic but not from control animals (27). These apoptotic changes occur side-by-side with healthy cells, consistent with single cell deletion. The changes are evenly distributed throughout the DRG neurons and in addition are observed in dorsal root SC. Frequent large vacuoles are observed, evenly distributed throughout the cytoplasm and many correspond to enlarged ballooned Mt with disruption of the inner cristae structure. The abnormal vacuoles are most prominent in cells showing early apoptotic changes such as mild chromatin aggregation with neuronal and cytoplasmic shrinkage. Although occasional vacuolation is observed in DRG cells from control animals, the Mt structure appears normal. After acute induction of hyperglycemia in animals, there is evidence of chromatin compaction, shrinkage of the nucleus and cytoplasm with preservation of the cell membrane in the presence of intact lysosomes. These changes are more severe than in the chronically hyperglycemic rodent (27). Similar to the changes observed in vivo, high glucose induces apoptosis in vitro: the degree of cell injury is dependent on the concentration of glucose present and occurs rapidly. In vitro, DRG cultured in extra glucose show compaction of chromatin into uniformly dense masses, deletion of single cells, cell shrinkage into apoptotic bodies, retention of membrane integrity with some membrane blebbing, and presence of intact lysosomes. This effect cannot be explained merely by glucose-induced hyperosmolarity, as similar changes are not observed when equiosmo-lar concentrations of mannitol or sodium chloride are used (27).
One of the areas of controversy has been whether there is classical apoptosis in the PNS and if loss of DRG neurons by apoptosis is responsible for the observed neuropathic deficits observed in both animals and humans. Most studies to date indicate there is activation of caspases in DRG neurons both in vitro and in vivo (16,20,27-31). There is also evidence of neuronal nuclear DNA fragmentation using in vivo studies with rigorously applied controls showing convincing positive TdT-mediated dUTP-biotin nick end labeling (TUNEL) staining with DNAse (27-29). Most studies indicate some loss of DRG neurons (27,31,77), and in particular there is a statistically significant loss of large DRG neurons (31). In one study, using rigorous counting techniques of DRG nuclei in 6-12 pairs of sections from the whole DRG, it was concluded that there was no loss of neurons in the DRG from diabetic animals (77). In fact, this study showed a 14.5% decrease in the mean number of neurons (determined by nuclei) per ganglia in 12-month diabetic compared with control animals. However, there was a large variance between animal groups and this likely resulted in inadequate power to detect a statistical difference. In fact, there is clear evidence that not all neurons are affected equally and that there is variability in the degree of neuronal loss in experimental animals similar to the observed variability in the severity of neuropathy in humans. Thus, the concept that a similar diabetic insult will always result in neuronal injury in all DRG neurons equally, and will produce the same degree of loss of sensory neurons is simplistic. Eventhough, there is evidence of DRG neuronal loss, the number of DRG neurons showing evidence of caspase-3 cleavage or TUNEL staining appears to be higher than the measured loss of neurons. This might occur because activation of caspases does not invariably result in neuronal death, or that there is an intrinsic capacity for repair within the neuron resulting either from DNA repair or by activation of neurotrophic protective signaling pathways (78).
As indicated in the initial description of mechanisms of apoptosis, PCD is a balance between caspase activation and blocking by inhibitors of apoptosis. One possible repair
Fig. 2. Electron micrographs showing representative early apoptotic changes and vacuolation in DRG neurons, and SC in: (I) diabetic (1 month) and control animals. (A) Control DRG neurons and axons, showing normal diffuse chromatin staining in the nucleus (N), Schwann cell (Sh), and normal axons (A) with intact myelin lamellae showing little or no myelin splitting. (B) DRG neuron from a diabetic animal showing coarse chromatin staining, with early aggregation
Fig. 2. Electron micrographs showing representative early apoptotic changes and vacuolation in DRG neurons, and SC in: (I) diabetic (1 month) and control animals. (A) Control DRG neurons and axons, showing normal diffuse chromatin staining in the nucleus (N), Schwann cell (Sh), and normal axons (A) with intact myelin lamellae showing little or no myelin splitting. (B) DRG neuron from a diabetic animal showing coarse chromatin staining, with early aggregation mechanism would be by elevated expression of the DNA repair enzyme PARP (30). DNA repair by PARP-1 is itself a double-edged regulator of cellular survival. When the DNA damage is moderate, PARP-1 participates in the DNA repair process. Conversely, in the case of massive DNA injury, elevated PARP-1 activation leads to rapid and massive NAD(+)/ATP consumption and cell death by necrosis (10).
Studies in cell culture and diabetic animals indicate that there is dysfunction of the normal electron transfer chain function and that this is associated with induction of oxidative stress (20). In a DRG cell culture model, the first 2 hour of hyperglycemia are sufficient to induce oxidative stress and PCD (79). On exposure to elevated glucose concentrations, superoxide formation, inhibition of aconitase, and lipid peroxidation occurs within 1 hour of the hyperglycemic stress, and is followed rapidly by caspase-3 activation and DNA fragmentation (20,79). Injury to the neurons can be prevented by the antioxidant a-lipoic acid (79,80), consistent with a model in DRG neurons where oxidative stress induced apoptosis can be ameliorated or prevented by antioxidants. This finding is supported by in vivo data that antioxidants can prevent neuronal and axonal injury in the PNS (11,23,57,64,81,82). In 1-12-month diabetic rats, 8-Hydroxy-2'-deoxyguanosine labeling is significantly increased at all timepoints in DRG neurons consistent with oxidative DNA injury (29). The changes in oxidative injury are coupled with an increase in caspase-3 labeling in acutely and chronically diabetic animals. These apoptotic changes are coupled with loss of DRG neurons that were higher in chronic in comparison with acutely diabetic animals and a progressive sensory neuropathy (29).
Corresponding to this evidence of apoptosis, are changes in the intrinsic pathway of PCD. In the diabetic state, there is evidence of MMD (20,28,83,84). In affected DRG neurons, there is a decrease in BCL-2 levels and translocation of cytochrome-c
Fig. 2. (Continued) in the nucleus (N) and vacuolation (V) in the cytoplasm. Remains of Mt cristae can be seen in many of the vacuoles, a change that is even more apparent at higher magnifications. (C) More severe chromatin aggregation (Ch), vacuolation (V) throughout the cytoplasm, and condensed rough endoplasmic reticulum associated with apparent loss of perikaryeal volume. (D) Further DRG neuronal chromatin clumping (Ch) and early dissolution of the nucleus. There is further ribosomal aggregation (R). Although blebbing of the cytoplasmic membrane is observed, the membrane remains intact with no evidence of inflammation or phagocytosis. (E) End stage changes in the DRG neuron. The nucleus is severely fragmented and the nuclear outline is no longer apparent. There is marked condensation of rough endoplasmic retic-ulum and ribosomes (R). Despite severe blebbing, the neuronal cytoplasmic membrane is still present. An adjacent satellite cell (S) has maintained its integrity consistent with single cell deletion of the DRG neuron during the process of apoptosis. (F) Enlarged Mt from DRG neuron in B showing disruption of cristae. (II) DRG cell culture. Whole DRG were cultured for 48 hours in the conditions described below: (A) Control DRG. Normal nucleus (N) and nucleolus, with evenly distributed chromatin. The satellite cell (S) shows darker staining, but non-aggregated chromatin, and a "cap shaped" nucleus, which is normal for this cell. (B) Twenty millimolar added glucose. There is well-developed chromatin compaction (Ch) in one neuron, and early compaction in another neuron. Both neurons show shrinkage of the perikaryon, but the outer cell membrane is intact. (C) High glucose (150 mM). There is early chromatin compaction (Ch) and shrinking of the cell nucleus and perikaryon similar to the changes seen in B. (Reprinted from ref. 27, with permission from Elsevier.)
from the Mt to the cytoplasm (85). One of the key events preceding apoptosis is a change in the permeability transit pore (PTP) associated with opening of the ANT/VDAC channel spanning the inner and outer Mt membranes (19,20). In hyperglycemic conditions there is early serine phosporylation of BCL-2 followed by a reduction in BCL-2 expression and loss of A¥M (19).
UCPs also have sequence homology to Mt transporters including the BCL proteins suggesting that they might be Mt carriers (70,86,87). When UCP levels are reduced, the ATM is abnormally high, and increases backpressure on the inner Mt membrane proton pumps. These events might further promote induction of PCD. UCP3 expression is rapidly down-regulated by hyperglycemia in diabetic rats and by high-glucose in cultured neurons (16), and maintained by the presence of bongkrekic acid consistent with regulation of UCPs by the ATm (16,19,22). Overexpression of UCP1, -2, or -3 prevents glucose-induced transient Mt membrane hyperpolarization, ROS formation, and induction of PCD (16,22).
Interestingly, insulin-like growth factor (IGF)-I is neuroprotective by blocking Mt swelling, inner mitochondrial membrane depolarization (MMD), and caspase-3 activation. Two important IGF-I signaling pathways, phosphatidylinositol 3-kinase (PI3K) and mitogen-activated protein (MAP) kinase/MAP-extracellular signal regulated kinase (MEK) are implicated in regulation of apoptosis (88,89). In neurons treated with high glucose, the PI3K pathway is the primary pathway regulating BCL-2 and BCL-xL, MMD, and apoptosis (19,90). Although inhibition of MAPK/MEK signaling independently and partially blocks IGF-I inhibition of MMD, these signaling intermediates do not significantly affect IGF-I inhibition of Mt swelling, indicating that IGF-I regulates different components of Mt function through discrete signaling pathways (19). IGF-I stimulation of the PI3K/Akt pathway phosphorylates three known Akt effectors: the survival transcription factor CREB and the proapoptotic effector proteins glycogen synthase kinase-3P and forkhead. IGF-I regulates DRG survival at the nuclear level by increasing accumulation of phospho-Akt in DRG neuronal nuclei, increased CREB-mediated transcription, and nuclear exclusion of forkhead (90). A further mechanism by which IGF-I can reduce apoptosis in DRG neurons is by activating UCPs and in particular UCP3 through a PI3K regulated signaling pathway (91-93). Similar changes have been observed with insulin treatment. Insulin increases the ATM and prevents MMD by activation of the PI3K signaling pathway and phosphorylation of Akt and cAMP response element binding (CREB) (83,84). This in turn is associated with increased ATP levels.
Other signaling pathways might also be involved with regulation of PCD in DRG neurons: diabetes activates all three groups of MAP kinases in sensory ganglia (94). Inhibition of ERK and p38, stress responsive members of the MAPK family, prevents nerve damage. Antioxidants and aldose reductase inhibitors that improve neuronal function in diabetic rats are also able to inhibit activation of ERK and p38 in DRG and increase activation of JNK (94). In DRG from chronically diabetic rats, the p54/56 isoforms of JNK are activated and furthermore activated when these animals are treated with antioxidants. In contrast, cultured DRG neurons die when treated with JNK inhibitors. It is thus likely that activation of C-Jun N-terminal kinases (JNK) because of a combination of raised glucose and oxidative stress serves to protect DRG neurons from glycemic damage (94). Alternative apoptosis pathways such as p53 activation have been shown to occur in an oxygen glucose deprivation model of ischemia observed in diabetic neuropathy (80,95).
Specifically, there is an increase in phospho-p53 levels under hypoxic-glucose deprivation that is associated with DNA damage and cell-cycle disruption. These changes suggest a possible role for p53-mediated apoptosis in diabetic neuropathy (95). DRG neuronal apop-tosis in this system was prevented by increasing the concentration of NGF in the culture medium consistent with observations in both animal model of diabetes and human neuropathy that nerve growth factor (NGF) might protect against apoptosis (96,97).
Metabotropic glutamate receptors (mGluRs) might also regulate CREB signaling intermediates and prevent neuronal cellular injury (98-101). The mGluRs are a subfamily of glutamate receptors that are G protein-coupled and linked to second messenger systems (101,102). In addition to strong mechanism-driven evidence that glutamate carboxypepti-dase (GCP)II inhibitors and mGluR3 agonists are neuroprotective, there are preclinical data that GCPII inhibitors ameliorate diabetic neuropathy in animal models (103). The GCPII inhibitor 2-(phosphonomethyl)pentanedioic acid (2-PMPA) is protective against glucose-induced PCD and neurite degeneration in DRG neurons in a cell culture model of diabetic neuropathy (98). In this model, inhibition of neuronal PCD is mediated by the Group II mGluR, mGluR3. 2-PMPA neuroprotection is completely reversed by the mGluR3 antagonist, (S)-a-ethylglutamic acid, but not by Group I and III mGluR inhibitors. Other mGluR3 agonists, for example, (2R, 4R)-4-aminopyrrolidine-2,4-dicar-boxylate (APDC) and N-acetyl-aspartyl-glutamate provide protection to neurons exposed to high glucose conditions, consistent with the concept that 2-PMPA neuroprotection is mediated by increased N-acetyl-aspartyl-glutamate activity. Furthermore, the direct mGluR3 agonist, APDC prevents induction of ROS (104). Together these findings are consistent with an emerging concept that mGluRs might protect against cellular injury by regulating oxidative stress in the neuron, and might represent a novel mechanism to prevent ROS induced PCD in diabetic neuropathy.
Was this article helpful?