Introduction

Older dogs are the most rapidly growing segment in many veterinary clinic populations. Better and more sophisticated treatment for primary medical conditions has lead to longer-lived dogs. Many of these older dogs will be affected by non-specific signs associated with cognitive changes. Diagnosis cannot be made on the basis of non-specific signs, alone. For example, cognitive changes associated with age can also be signs of separation anxiety. When signs of separation anxiety appear in older dogs they may be associated with anticipatory anxiety. It is on this observation the development of the concepts associated with "cognitive dysfunction" in older dogs have been predicated. Failure of function or behaviors associated with anxiety are not uncommon in older dogs. In one of the first studies on older dogs 13/26 dogs 10 years or older were diagnosed with separation anxiety (i.e., the behaviors occurred only in clients' absence), and 6 were attributed to breakdown of house training (i.e., "cognitive dysfunction") that did not meet the criteria to make a diagnosis of separation anxiety. Older dogs have changing physical and emotional needs; accommodating these needs and treating the dogs with anti-anxiety medications can help modulate symptoms, although the course of whatever the underlying condition is may be inexorable.

Cognitive dysfunction is best defined by the following conditions: a change in interactive, elimination, or navigational behaviors, attendant with aging, that are explicitly not due to primary failure of any organ system. It should be noted that this is a definition for a phenotypic diagnosis. It is non-specific as far as underlying neurocellular mechanism is concerned. Recent advances have indicated that afflicted dogs may have variable amounts of cellular and neurochemical changes that are, themselves, rooted in different molecular processes. Age-related declines are generally associated with vulnerability of the cholinergic neurons. Such vulnerability could be the level of the cholinergic neuron, itself, the neurotrophic support system, cytoskeleton alteration, target loss, and vascular dysfunction. We know so much about canine behavioral changes that occur with age because dogs are good neuroanatomical and neurobehavioral models for humans. A summary of some of the more recent findings for aged humans and canines may suggest how our future understanding will develop.

Behavioral similarities shared by aged dogs and humans: Non-specific signs associated with age in dogs can include (1) changes in reactivity to routine stimuli; these changes are usually thought to involve a decrease in reactivity, but in some case an increase in reactivity involving startle occurs, (2) decreases in responsiveness to sensory stimuli, (3) alterations in sleep-wake cycles, and routines in diehl-based behaviors, (4) changes in elimination behaviors ranging from sporadic inappropriate elimination to incontinence, (5) changes in the ability to respond to requests or commands, (6) changes in the ability to problem solve (e.g., increased frequency of getting stuck in corners or "lost" in the yard or house, and (7) changes in general affiliative and social behaviors. Similar changes, allowing for a species with spoken language, occur in humans.

Neuropathological changes shared by aged dogs and humans: Changes in brain pathology that aged dogs share with aged humans with a diagnosis of one form of human dementia (Alzheimer's disease) include (1) thickening of the meninges and dilation of the ventricles, (2) age-related gliosis, (3) vascular changed, (4) diffuse plaques, and (5) amyloid deposition. Dogs do not seem to form the true neurofibrillary tangles that humans do.

Neurochemical changes seen in canine and human aging: Most of what we know about neurochemistry in these conditions is the result of studies examining correlation. For example, in most neurodegenerative disorders, activity of the enzyme monoamine oxidase B (MAO - B) is elevated and levels of acetylcholine are either primarily or secondarily lowered. In the case of Alzheimer's disease and canine cognitive dysfunction, the increased activity of MAO-B appears to correlate with depletion of dopamine and loss or decreased activity of dopaminergic neurons.

One mechanistic hypothesis is a neuroregulatory one. Because of the increased activity of MAO-B, dopaminergic neurons are unable to maintain production and so undergo some atresia. Normal dopaminergic function is essential for basal ganglia function. Degeneration of dopaminergic neurons is invariably associated with motor and cognitive defects. In part, this decreased metabolism may be responsible for further cellular deterioration and apoptosis. Bombardment with highly reactive free radicals can both induce and augment this effect. Free radicals increase with cellular degeneration, which may then cause DNA fragmentation, which further worsens the functioning of cells. In fact, enzymes called capsases are increased in and around amyloid containing plaques and tangles, and may be essential in encouraging DNA fragmentation and in promoting degeneration of mitochondrial and endoplasmic reticulum products. All of these processes are capable of leading to cellular apoptosis. Free radical production accentuates this process. Compounds that thwart free radical production or destroy them (e.g., superoxide dismutase) hinder the process of apoptosis or programmed cell death and may augment cognitive function.

Dopaminergic axonal terminals release dopamine in the substantia nigra. For this release to occur a dopamine transporter protein must be functioning properly. This means that the release of dendritic dopamine is carrier-mediated, and not merely modulated through exocytosis. Accordingly, medications that inhibit the dopamine transporter should augment cognition or at lease slow the deleterious effects of cognitive disorders. Newer substances affecting the dopamine transported could act as neuroprotective agents in the early stages of conditions affecting cognition.

Other transporter proteins may also be implicated in cognitive disorders. Decreased expression of the glutamate transporter GLT-1 may be responsible for the extracellular increase in the excitatory amino acid, glutamate, that is found in people with amyotrophic lateral sclerosis.

Alzheimer's disease may differ in form depending on the relationship between B-amyloid precursor protein and an enzyme, presenilin, which appears to induce further changes in amyloid structure. These structural changes are related to patterns in plaques and neurofibrillary tangles. In another neurodegenerative condition, Parkinson's disease, a small neuronal phosphoprotein, "-synuclein, appears to change. Normally this phosphoprotein forms small, sticky clumps that are degraded after being labeled with ubiquinin. This housekeeping system fails in Parkinson's disease and may also be implicated in the lack of repair that occurs with amyloid deposition.

Cognitive assessment: One of the hidden benefits of the research into canine models for human age-related cognitive decline (ARCD) has been the acknowledgment that dogs are, in fact, cognitive individuals. Dr. Bill Milgram's lab at the University of Toronto has spent over a decade assessing effects of age, sex, social group, genetic relatedness, et cetera on performance as measured by a complex, computer driven algorithm for choice experiments that use food rewards.

Drugs that either are or could potentially be useful

Selegiline: This is a selective inhibitor of MAO B receptors (an MAO-I). MAO-Is that affect the B receptor enhance the degradation of phenylethylamine, norepinephrine, dopamine, and tyramine but have little effect on serotonin. Selegiline exerts its effect both by blocking this degradation and inhibiting the re-uptake or recycling of neurotransmitters in the synaptic cleft, resulting in an increased amount of neurotransmitter in the cleft and increased saturation of the post-synaptic receptors. Selegiline most specifically exerts these effects on dopamine receptors. These actions are thought to be both neuroprotective and directly stimulatory for neuron function, in part because dopamine is a precursor to norepinephrine.

Nicergoline: This compound is an ergoline derivative that has alpha-1 adrenergic blocking effects. Accordingly, one of its main effects is to augment cranial and brain blood flow and distribution. Because of the feedback effects on neuronal metabolism, nicergoline, like selegiline, may have a neuro-protective effect. Nicergoline's neuroprotective effects may be more direct: it inhibits lipid peroxide formation, inhibits lipid peroxidation, and may act as a scavenger of free radicals. Long-term treatment with nicergoline has been associated with expression of the constitutive isoform of nitric oxide (NO) synthesis enzyme nNOS in the cerebral cortex and the basal ganglia. Activation of nNOS leads to synthesis of NO, which plays a modulatory role in regulating blood-brain perfusion. Such an effect can be neuroprotective or neurotoxic.

Galantamine: Galantamine is a new acetylcholinesterase inhibitor that potentiates pre-synaptic nicotinic cholinergic neurotransmission.

Clomipramine: As a relatively specific tricyclic antidepressant (TCA) that has 2 functioning metabolites, one that acts as another TCA that primarily affects norepinephrine and another that acts as a specific selective serotonin re-uptake inhibitor (SSRI), clomipramine augments the functioning of both serotonin (particularly with respect to the pre- and post-synaptic 5-HT_1A subtype receptors) and norepinephrine. By blocking the re-uptake of these neurotransmitters clomipramine may alter neuronal metabolism, and through these changes exert a cryo-protective effect. TCAs and SSRIs also exert their effect, in part, by causing the post-synaptic neuron to re-program itself via protein phosphorylation and protein transcription to make new and more efficient receptors. This, in turn, is a protectant effect that allows the cell to better function in a changing environment. TCAs and SSRIs may also exert some of their beneficial effect in conditions involving cognitive change by decreasing the anxiety attendant with the development of these conditions. The link between anxiety and cognitive function is complex and only recently explored. Because of the association of old-age onset separation anxiety and the potential advent of cognitive dysfunction, treatment with clomipramine may be a rational first step when the diagnosis is unclear.

Recently, genetic propensities for rapid or slow metabolism of compounds that alter monoamine function have been explored. For example, debrisoquine hydroxylase is the enzyme responsible for metabolizing the TCA nortriptyline, and many other psychotropic medications. This enzyme is encoded for by the CYP2D6 gene. At least 16 CYP2D6 alleles that affect debrisoquine hydroxylase function have been identified. Impairment of this gene's function leases to greater drug concentrations, given lower dosages. Furthermore, the number of alleles that encode decreased metabolism are important as a gene dosage effect.

New ideas

Sensorimotor impairment that involves locomotion and postural control as well as fine-tuned movements is a hallmark of senescence in humans. In animals, as already discussed, we tend to focus less on these types of evaluations, unless they affect other functions, and we routinely attribute sensorimotor impairment to degenerative musculoskeletal changes attendant with aging. This may mean that we are failing to address an important issue. Skeletal muscles of old humans and rodents show fiber loss and fiber atrophy concomitantly with increases in connective tissue and fat. These may not be changes solely affected by use. It has been suggested that senile muscle atrophy is of neurogenic origin. Instead of being characterized by neuron loss in muscles, aging is associated with loss of neuronal connections, axon dystrophy, myelin aberrations, neuron atropy in certain cell populations and phenotypic changes in gene-expression pattern. Primary sensory neurons synthesize neuromodulatory peptides like calcitonin gene-related peptide (CGRP), substance P, and galanin. In aged rats, primary sensory neurons have greatly decreased expression of CGRP and substance P. These decreases may have profound deleterious effects on nociceptive function. In addition, the level of trk (typrosine kinase) expression, which affects both protein synthesis and RNA function required to make and maintain synaptic receptors, is lowered in aged neurons. In humans there is a further correlation between the distribution of the motor symptoms associated with age and trk-downregulation: sensory neurons innervating the hind limbs are more affected than are those innervating the face. No such comparisons have been examined in dogs, but dogs provide a rich comparative base for this study.

As for motor neurons, a similar pattern may also occur. Even in very old individuals with profound behavioral impairment, the vast majority of motor neurons are still present but seem to be no longer connected to an intact set of target muscle cells. Age-related axon lesions are most prevalent in ventral roots and peripheral nerves of the lumbar, rather than the cervical, spine, and the distal part of the nerve appears to be more affected than the proximal part, as was also true for sensory nerves. The motor neurons, however, are not lost: they have a preserved cholinergic phenotype and have no signs of cell-body atrophy. However, both CGRP and growth associated protein 43 are markedly increased-a regulatory pattern consistent with growth and regeneration. In what appears to be a consistent set of findings, aged motor neurons appear to have decreased levels of neurotrophins in senescence, and this decrease may reflect a failure to maintain previous adult levels of these neurotrophins (e.g., neurotrophic growth factor [NGF], brain-derived neurotrophic factor [BDNF], and specific neurotrophins). This effect may be a general one since the cognitive deficits associated with aging also appear to be associated with neurotrophin dysfunction.

Finally, with aging long-term potentiation (LTP; cellular learning) duration appears to be reduced in age-impaired rats. It is not impaired in age-unimpaired rats. This appears to lead to decreased plasticity in LTP and may primarily be the effect of impaired late (> 4 hours; protein-dependent) LTP. It is important to note that the TCAs and the SSRIs use the same neurochemical and molecular pathways as does LTP, so that some of their effects in behaviorally and physically age-impaired individuals may be non-specific or regulatory ones associated with enhancement of LTP. Our goals should focus on better phenotype discrimination, and on further mechanistic understanding within the different phenotypic presentations of canine cognitive dysfunction.

References

References are available on request.