Background
Project Description
Achieved Results, Ongoing Research And Future Plans:
/ Sensory neurons
/ Motoneurons
/ Breakdown of synaptic connectivity, gliosis and oxidative stress during aging

Background

Project Description

 The main aim of my research is to investigate aging-related neuromuscular and sensory dysfunctions, in order to reveal the mechanism(s) causing motor incapacitation and sensory deficits, and to identify possible strategies to impede their occurrence during aging. This project which is supported by the Swedish Medical Research council was initiated 1991. We use rodents with impairment of cognition, sensory and motor functions as revealed by standardized behavioral tests. So far, we have mainly utilized rats (see Appendix), but mice are becoming increasingly important to our work.
 The project is focused around some of the key issues in this field of research, such as: (a) is impairment of sensorimotor functions due to aging-related loss of neurons? (b) is plasticity and maintanence of neuronal connectivity altered in senescence? (c) what kind of changes occur in the gene-expression profile of aging neurons? (d) are changes in neurotrophic signaling and/or oxidative stress of significance in the aging process of neurons?
 Our main approach is to use animal models with well-defined functional deficits in senescence. From these animals and controls, tissue is analysed using standard methods (Northern and Western blots; RT-PCR, HPLC/RIA). Cellular analysis is done using in situ hybridisation (mRNA) and immunocytochemistry (protein detection) at the light and electron microscopic levels. Various environmental paradigms are being used (caloric restriction, enriched environment, dietary supplementation as well as hormone/drug/substance treatments) to decode genetic and epigenetic factors, and their interplay, in the aging-process.
 Gene-arrays are now commercially available and we are currently exploiting this tool for gene-expression profiling aging neurons, target tissues and supportive cells (glia cells). Furthermore, we are currently negotiating collaboration with Neuromics Incorporated (Minneapolis MN USA) to take advantage of the fast advancement in protein analysis (proteomics). Undoubtedly, the two latter techniques will be important for our future research.
 We are currently establishing breeding of several transgenic mice, such as the p75 and trk receptor mutations, neurotrophin 3 and BDNF mutations as well as mice over-expressing neurotrophic ligands(obtained from collaborators or Jacksson Lab., USA). The problem associated with moving from the rat to the mouse relates to the fact that the latter species appears less suitable for behavioural analysis and, furthermore, well-established tests were all developed using the rat. However, the gain in getting access to transgenic technology is considerable.

Achieved Results, Ongoing Research And Future Plans

Sensory neurons

 Given their inability to replace themselves, postmitotic cells of the nervous system may be especially susceptible to the cumulative damaging effects of aging. Loss of neurons has generally been considered the main reason for functional incapacitation in senescence. However, sensory neurons show a complex pattern of changes at the cellular level during aging. There is a small unselective loss of neurons, a selective cell body atrophy, abundant axon aberrations, and phenotypic alterations in the expression pattern of neuropeptides and neurotrophic receptors [69, 71, 74, 76, 78, 81, 82, 83]. Thus, cell loss itself may not explain the sensory deficits seen in elderly. Large sensory neurons, however, show overt signs of axon and cell body atrophy paralleled by a decreased expression of neurofilaments, indicating that they are more susceptible to aging than small sensory neurons. It is well established that aging is associated with axon aberrations, such as axon atrophy and dystrophy, de-/dysmyelination, and signs of degeneration in peripheral nerves and spinal roots [Fujisawa, 1988; Knox, 1989,]. Although axon aberrations are evident in unmyelinated as well as myelinated axons, the latter group is much more affected. The loss of axons is, however, small and compares well with the figures available for loss of sensory cell bodies in the dorsal root ganglia (DRGs) [84]. Consistent with this pattern of changes, aging individuals are more affected by disturbed proprioception than by changes in nociception. From a morphological viewpoint, the data also imply that aging is a distal-to-proximal process, being most marked and, perhaps initiated, in the axon domains of the peripheral and central target regions.
 Sensory neurons synthesize neuromodulatory peptides such as calcitonin gene-related peptide (CGRP), substance P (SP) and galanin, which sub-serve neuromodulatory functions both at the central and the peripheral terminations [reviewed by Hökfelt, 1991]. Sensory neurons of aged rats disclosed a dramatic decrease in CGRP and SP affecting mainly small (B4 binding) sensory neurons [69, 81]. In large myelinated (RT97 positive) sensory neurons of the same individuals, a novel expression of galanin and a robust upregulation of NPY are evident [69, 81]. The down-regulation of, in particular, CGRP but also SP in small nociceptive sensory neurons may have relevance for the observed increase in nociceptive thresholds in senescence [74], since NGF induced hyperalgesi in young animals has been associated with upregulation of CGRP and SP [Lewin, 1993]. It is well established that stimulation of primary afferents can induce cutaneous vasodilatation and plasma extravasation, mediated by SP and CGRP in the cutaneous axon reflex [Delay-Goyet, 1992]. In this context, our findings of a decrease in CGRP in perivascular axons, as well as a decrease in CGRP and SP, in primary sensory neurons of aged rats, fit well with previous descriptions of a weaken axon reflex in senescence [71; Helme, 1985; Khalil, 1994]. A diminished axon reflex will have a negative impact on inflammatory responses and wound healing, i.e. well established traits of aging.
 Evidence are accumulating indicating disturbed neurtrophin signaling as a mechanism in peripheral neuropathies and aging-related aberrations [Barde, 1999]. In the aged rat, the characteristic size distribution of neurotrophin (trk) receptor expressing sensory neurons is maintained, with a selective expression of trkA and trkC in small and large sensory neurons, respectively, and trkB expressed by neurons of all size categories [69, 76]. However, the level of trk expression in aged sensory neurons is downregulated and, furthermore, a covariation between the distribution of symptoms and trk-downregulation is evident with the sensory neurons innervating the hind limbs being more affected than those of the face [76]. Lesion experiments, and transgenic animals, have shown that target-derived neurotrophins can regulate the cognate trk receptor expression in sensory neurons both during development and in adulthood. Thus, since trk downregulation in aging sensory neurons may reflect a failure of targets to synthesize neurotrophic factors, we examined the expression of neurotrophins in peripheral target tissues of aged rats. In target muscles a downregulation of all neurotrophins was observed [82]. Notably, the decreased expression of neurotrophins covaried with the extent of behavioral sensorimotor disturbances among the individuals. These findings imply an attenuated neurotrophin signaling between target tissues and sensory neurons in senescence. In order to resolve if sensory neurons remain dependent on the target for their expression of trk receptors in senescence, we transected peripheral nerves and examined the expression level of trk receptors [76] and found a further downregulation of trk-receptors. Furthermore, the level of neurotrophin receptor expression was found to be very similar in adult and aged rat sensory neurons following axotomy. One interpretation of the latter observation is that the residual expression of trk’s following target disconnection depends on other signals possibly provided through auto/paracrine neurotrophin loops [Acheson, 1995]. If this is the case, then we may conclude that this non-target tissue dependent expression of trk’s is unchanged in senescence, a notion consistent with the observed lack of changes in neurotrophin expression in both DRGs and the spinal cord in senescence [82].
 A decreased neurotrophin-trk signaling may explain the changes in neuropeptide expression during aging. Decreased neurotrophin (NGF)-trkA signaling seems to lower the expression of CGRP and SP in small nociceptive sensory neurons. NGF signaling appears also to be able to suppress galanin and NPY expression, and NPY expression can, furthermore, be restrained by neurotrophin (NT3). Thus, the regulatory changes in neuropeptide expression pattern during aging are fully consistent with a parallel decrease of NFL-trk signaling. NPY upregulation in large myelinated sensory neurons following axon lesioning in adult, and in nonmanipulated-aged rats, may share a common mechanism in a decreased NT3-trkC signaling. During aging there is a slowing of axon conduction velocity of fast conducting sensory neurons [Sato et al., 1985]. This fits well with the abundant morphological aberrations among large myelinated sensory axons in senescence [84]. In adulthood, NT3 has been shown to play a role in the maintenance of muscle spindle afferents and exogenous NT3 can protect axon conduction of large myelinated axons [Munson, 1997]. Furthermore, exogenous NT3 ameliorates the damaging effects on proprioception of cisplatin and high-dose pyridoxin treatment [Konings, 1994; Gao, 1995; Helgren et al., 1997]. Although the cellular mechanism(s) by which NT3-trkC signaling protects large myelinated sensory neurons remains elusive, a common link may be that neurofilament content in adult sensory neurons can be regulated by a neurotrophins. A reduced NT3-trkC signaling in senescence may therefore be of relevance for the axon/cell body atrophy and decrease in neurofilament expression in large sensory neurons with advancing age [74, 76, 82, 83].
 In contrast to trk receptors, the expression of p75NTR is slightly increased in senescence [69, 76]. The functional consequence of a increased /sustained p75NTR expression is far from straight-forward considering to the multitude of possible functions ascribed this receptor. An increased/maintained level of p75NTR could possibly represent a compensatory mechanism for the decreased availability of target-derived neurtrophins in aged rats, by means of a positive effect of p75NTR on formation of high-affinity binding sites, retrograde transport and internalization of neurotrophins, or possibly by increasing the selectivity of trk receptors for their preferred ligand(s) [Bothwell, 1995; Curtis, 1995]. That p75NTR knockout mice have an altered response to NGF, but not to other neurotrophins, suggests an effect of p75NTR primarily on nociceptive trkA expressing sensory neurons [Lee, 1992; Davies, 1993]. In these neurons then, the effects of a decrease in trkA may partly be compensated for by a preserved/increased p75NTR expression, whereas no such effect would be anticipated for trkB or trkC expressing sensory neurons. Thus, if this mechanism is in operation it could help to explain the more severe effect on myelinated as compared to unmyelinated primary afferents during aging.
 Still, the more successful aging of nociceptive than proprioceptive sensory neurons may be due to signaling of other neurotrophic factors in adulthood. It was recently shown by several groups that small sensory neurons appear to acquire a dependence on glial cell-line derived neurotrophic factor (GDNF) signaling (GDNF-GFR/RET signaling) during postnatal development and in adulthood [Molliver, 1997]. In aged rats, we have shown that there is a significant upregulation of GFRa1 and RET, while the characteristic distribution of these receptors in distinct populations of DRG neurons appear preserved [78]. In parallel, GDNF expression is dramatically upregulated in target tissues [80]. An increased GDNF-GFR/RET signaling during aging is of considerable interest, since GDNF can protect the B4 phenotype and the axon conduction velocity of small nociceptive sensory neurons, whereas it has a negligible effect on CGRP expression or the slowing of axon conduction velocity in large myelinated fibers following axon lesioning [Molliver et al., 1987; Bennett, 1998; Munson, 1997]. Taken together, an increased GDNF-GFR/RET signaling and a decreased neurotrophin signaling seem jointly to clarify many of the phenotypic changes characterizing aging sensory neurons, and furthermore shed light on why myelinated are more severely affected than unmyelinated sensory neurons.
 From what has been discussed above it may appear as changes in sensory neurons during aging follow a fairly simple mechanistic scheme. In contrast, studies of follicle-sinus complexes (FSC) in mutated mice lacking either neurotrophins or neurotrophin receptors have revealed that each set of sensory nerve ending terminating in the skin require a spatial and temporal unique combination of NFL/receptor signaling to develop normally (Davis et al., 1997; Fundin et al., 1999; Fundin et al., 1997a; Rice et al., 1998). We therefore turned to the Whisker follicle-sinus complexes (FSCs) model to examine sensory nerve-target interaction during aging in closer detail. Although observations are still limited, changes in sensory innervation during aging seem to be not only highly type but also site specific. Furthermore, these alterations appear closely associated with, and are perhaps instigated by, aberrations in target-tissue neurotrophic signaling [83].
 In ongoing and planed studies, we are pursuing the findings of a changed neurotrophic signaling and, moreover, focusing on the interaction of neuron-target cells. In these studies our possibility to use transgenic animals (see above) will be quite important. Since our results, moreover, indicate the significance of target function during aging, more effort is put into understanding changes during aging that occur in target cell populations of the skin and muscle. In this context the effects of meno-, adeno- and adreno-pause are of obvious relevance and we are currently assaying the effects of altered levels of steroid hormones.

 Aging motoneurons also show a distinct pattern of phenotypic changes, while loss of is small (~15% at 30 month of age) despite frequent (~75%) and overt signs of behavioral motor deficits [60]. Aged motoneurons have a preserved cholinergic phenotype and cell body size [60, 77], but disclose a highly specific pattern of regulation of neurotrophic receptors. This pattern shares some similarity but also distinct differences from that expressed by adult motoneurons disconnected from their target.
 In aged rats, skeletal muscles show a downregulation of all four neurtrophins (82) and the degree of decrease appears to be associated with the extent of behavioral sensorimotor impairment among the individuals. Thus, animals with minor symptoms show a preserved expression of the neurotrophin BDNF and only neurotropghin-4 (NT4) disclose a robust downregulation (>40%) also in behaviorally fairly intact aged animals. The level of muscle NT4 expression increases post natum and several lines of evidence indicate that NT4 is positively regulated by neuromuscular junction (NMJ) impulse activity in adulthood [Funakoshi et al., 1995; Griesbeck et al., 1995]. Some evidence point at the intriguing possibility that muscle-derived NT4 may have a strictly local signal-enhancing effect on the NMJ [Wang and Poo; 1997]. It seems quite plausible that the regulation of muscle NT4 in senescence reflects the decreased motor activity typical of aged rats. A decreased motor activity may not necessarily reflect a decreased innervation of the target muscle. Our finding of a concomitant lowering of NGF, NT3 and BDNF in the target would not be expected if neuromuscular aging was due exclusively to a drop out of motoneurons (see above). Rather it suggests a possible incapacitation of the target to maintain adult expression levels of neurotrophins [82].
 The motoneurons of aging rats show a robust downregulation of neurotrophin receptor trkC, evident also in individuals with only minor behavioral deficits [68, 77]. The downregulation may reflect the lowered levels of NT3 in the target since other potential sources of the trkC ligand NT3 such as sensory neurons, the spinal cord, or its motor nuclei, express NT3 at unaltered levels in senescence [77, 82]. However, in contrast to adult motoneurons disconnected from their target, there is also a decrease of trkB expression in aged motoneurons. This downregulation is less marked in animals with mild symptoms of impairment, while animals with severe symptoms show a more robust decrease of trkB [77]. The changed pattern of trkB expression in motoneurons closely mirrors the changes seen for neurotrophin BDNF (trkB ligand) in the target muscles (see above). Furthermore, the levels of trkB ligands in sensory neurons and spinal cord appear not to change with advancing age, arguing against that altered access to alternative sources of ligands is mechanistic in the downregulation of trkB [77, 82]. Thus it seems reasonable to infer that trkB is downregulated in aged motoneurons due to a decreased access to target-derived ligand. Taken together, the pattern of regulation of neurotrophins and trk receptor transcripts in senescence suggest that spinal motoneurons compete for a decreasing amount of target-derived trk ligands. The loss of muscle fibers and the atrophy of remaining fibers during aging will combined impose a reduced target size where the remaining muscle fibers express decreased levels of neurotrophins [82].
 The functional implications of a decreased trkB and trkC signaling in aged motoneurons are not resolved. However, as mentioned above BDNF and NT3 can protect motoneurons axon properties [Munson et al., 1997] and, thus, a decreased BDNF/NT3 signaling may contribute to the processes underlying the axon aberrations and slowing of the motor axon conduction velocity evident in senescence. As mentioned previously neurtrophins can regulate neurofilament expression and some evidence point at a trophic function of these ligands on myelinating cells, both mechanisms appear highly relevant for the changes observed in aging motor axons. Furthermore, motoneurons produce and release several factors that regulate muscle acetylcholin receptor expression (AChR). One such factor, neuregulin (NRG, ARIA; Falls, 1993) increase AchR transcription through activation of muscle expressed erbB receptors [Carraway and Cantley, 1994]. NRG is expressed by motoneurons during development and in adulthood, however, results from our group show a downregulated in senescence [Edström et al., unpublished observations]. Loeb and Fishbach (1997) showed that, in particular, BDNF and NT3 can upregulate NRG in motoneurons. Interestingly, the isoform of NRG most likely to bind to the basal lamina of the NMJ was only upregulated by BDNF. Although direct evidence is lacking, the downregulation of NRG in motoneurons may represent yet another consequence of a decreased neurotrophin signaling in senescence.
 In contrast to the trk’s, p75NTR is markedly upregulated in aged motoneurons [77]. An increased expression is evident in animals with minor as well as severe symptoms [77]. Motoneurons express high levels of p75NTR during development (when programmed cell death occur) and in response to target disconnection in adulthood (Ernfors et al., 1989). The upregulation p75NTR in lesioned adult motoneurons is triggered by a yet undiscovered retrogradely transported signal (Greeson et al., 1993) and that p75NTR levels become normalized upon target re-innervation. In aged rats, axonal severance induces a further upregulation of p75NTR, indicating that the retrograde triggering mechanism present in adult animals is maintained in senescence [77]. Although the mechanism(s) regulating p75NTR expression in aged nonmanipulated as well as in axotomized motoneurons remains obscure, increased levels of p75NTR may work as an enhancer of neurtrophin-trk signaling in a situation of diminished access to target-derived neurtrophins.
 GDNF is one of the most potent neurotrophic factor found so far for motoneurons. In adulthood it is present in glia cells and in peripheral target tissues, and the preferred receptor compliment is expressed by motoneurons. In aged rats there is a clear upregulation in spinal motoneurons of both the binding protein GFRa-1 and the signal transducing RET components of the GDNF receptor [78]. The target muscles show a marked upregulation of GDNF [80]. An increased GDNF signaling between the target muscles and the motoneurons may serve as a compensatory mechanism in the context of decreased neurotrophin-trk signaling (in aging motoneurons), promoting axon regeneration (Naveilhan et al., 1997; Trupp et al., 1997), sprouting and muscle fiber re-innervation (Nguyen et al., 1998); and may be mechanistic in the protection of cell body size and cholinergic phenotype. In support of the latter, is Sagot and coworkers (1997) report that in pmn/pmn mice, suffering a lethal mutation causing severe muscle dystrophy and progressive caudal-cranial motoneuron degeneration, exogenous GDNF could prevent cell body atrophy and protect the cholinergic phenotype of the motoneuron but not impede the progressive damage of their axons. It has also been shown that GDNF has inhibitory effects on the generation of radical oxygen species (Irie and Hirabayashi, 1999) and, thus, may work in orchestra with p75NTR (see above) to protect aging motoneurons from oxidative stress insult.
 Our current research is focused on the molecular machinery of neuromuscular junction, the target disconnection process, the mechanism(s) underlying muscle fiber loss and alterations in the gene-expression profile of the muscle fibers. As mentioned above hormones are of obvious interest (GH-IGF; androgens and testosterone) and we are currently examining their significance in this process. In parallel with deepening our understanding of the changed trophic signaling, we are also trying to manipulate trophic signaling pathways with the objective to ameliorate aging-related changes in the neuromuscular system.

 The breakdown of neuron-target connections is not exclusive to the peripheral nervous system and changes in neuron-to-neuron connectivity, including axon aberrations, are evident in systems intrinsic to nervous system (CNS). Thus, we have shown that the bulbospinal serotoninergic (5HT) system shows conspicuous signs of axon dystrophy, degeneration and loss of 5HT boutons [51, 52] depriving the target motoneurons of synaptic input. Using electron microscopy we have shown that motoneurons in aged rats have lost up to 50% of their synapses.[75] This deafferentation is not a uniform process across the motoneuron population. Instead, some neurons seem to lose many contacts while nearby located motoneurons possess a synaptic architecture similar to that seen in young adult animals. The extent of deafferentation seems to correlate to the degree of behavioral motor deficits in the animals [52, 75]. Using high-resolution immunocytochemistry, we have obtained evidence suggesting that aberrant axons often are glutamatergic and very few are GABAergic and/or glycinergic [79]. Furthermore, among these three amino acid transmitters, it is in particular the glutamatergic boutons that appears to be lost with advancing age. One mechanism suggested to induce degenerative changes during aging is cellular accumulation of free radicals (“free radical theory of aging”; e.g. review by Harman, 1981). Thus, aging-related degeneration may, at least in part, be explained by a changed cellular redox status with decreased antioxidant capacity and/or an increased oxidative stress. Glutathione (GSH) is an important free radical scavenger present in the nervous system and by using quantitative immunocytochemistry we have provided evidence that aging is associated with an increased level of total GSH in synaptic profiles of the spinal motor nucleus [79]. The increase is most dramatic in profiles that show morphological features indicating degeneration, but is also apparent in profiles with normal ultrastructure, suggesting that the increase in total GSH precedes the degenerative process or occurs in an early phase of this. Among profiles with no overt sign of degeneration, an enhanced total GSH signal often coincides with an enrichment of glutamate immunoreactivity. This combination is also typical of many degenerating fibres, suggesting that glutamate containing fibres are more prone to degeneration than are GABAergic or glycinergic ones. Our data lend support to the idea that aging is associated with an increased oxidative stress and indicate that the degenerative process differentially affects different transmitter systems.
 In parallel to the loss of axons, astrocyte, both in the gray and white matter, hypertrophy and one of their main cytoskeletal proteins, glial fibrillary acidic protein (GFAP), is upregulated [75, 85]. This is also a selective process, being dramatic in certain regions while it is hardly measurable in other regions. Thus, it was suggested by us that GFAP may be used as a marker to select neurons undergoing deafferentation. In certain regions such as the spinal cord white matter, dorsal and ventral roots, and gracil nucleus there is a dramatic increase in number of activated microglial cells, indicating ongoing fairly massive phagocytosis [85]. In parallel, we observed an upregulation in many astrocytes of clusterin (SPG-2/apoJ/; SP-40,40) a soluble inhibitor of the complement cascade generated membrane attach complex. This suggests that inflammatory processes may be of importance in the aging-related lesions.
 We are currently examining potential mechanism(s) operating the breakdown of central neuron-to-neuron connections. Trophic signaling is of prime interest but oxidative stress and cytoskeletal processes are also being studied. Various environmental paradigms as well as substance treatment are explored aiming at providing external stimuli, reduce metabolism and to support oxidative tolerance.