Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Geriatric depression presents an unique set of physiological and biochemical problems that have an adverse impact on therapeutics and outcome. In particular, this patient population displays a poor response to serotonin-specific reuptake inhibitors, the mainstay of depression therapy (Carroll et al., 1981; Aghajanian et al., 1993). Although many reports claim that serotonin-specific antidepressants are equivalent to the tricyclic antidepressants, only phenelzine, nortriptyline, and imipramine have a proven efficacy in placebo-controlled trials in elderly depressed patients (Schneider, 1993; Volz and Moller, 1994), and in fact, the serotonin-specific reuptake inhibitors have not been evaluated adequately in the elderly population, where a relatively high rate of response to placebo is known to occur (Wilcox et al., 1992). These agents, as well as desipramine, have failed to show satisfactory efficacy in geriatric depression (Danish University Antidepressant Group, 1986, 1990; Roose et al., 1994; Nelson et al., 1995).

In part, the differences between responses in elderly and young depressed patients may reflect regulatory differences in the hypothalamus-pituitary-adrenal (HPA) axis. The elderly depressed population shows twice the rate of dexamethasone nonsuppression in association with a greater general severity of the disease (Ritchie et al., 1990). There is close, mutual regulation of serotonergic systems and glucocorticoids (Arora and Meltzer, 1986; Joëls and De Kloet, 1991; Pepin et al., 1992), a relationship pursued in a number of investigations of geriatric depression. Using inhibition of platelet serotonin uptake as an index of tricyclic antidepressant actions, we recently found that elderly nonsuppressors maintained their imipramine effect, whereas suppressors did not (Slotkin et al., 1997a); simultaneously, clinical evaluations demonstrated that nonsuppression could be used as a predictor of antidepressant therapeutic outcome in geriatric depression (Kin et al., 1997). Indeed, in controlled clinical trials with the dexamethasone suppression test, suppressors have a very low rate of specific response to tricyclic antidepressants (approximately 10% above placebo response rates), whereas nonsuppressors have a specific response rate of approximately 60% because of their low rate of placebo response (Carroll, 1989). These results indicate that elderly depression may be a biologically distinct disorder with different underlying neurochemical alterations and, hence, a different therapeutic prospectus. Comparing young and aged rats, we have demonstrated that elevated glucocorticoid levels induce different sets of responses for serotonin transporter expression and function (Fumagalli et al., 1996; Slotkin et al., 1997b), and at the same time, produce changes in postsynaptic cell-signaling cascades that influence not only serotonergic, but also catecholaminergic neurotransmission (Slotkin et al., 1996). These results suggest that HPA axis abnormalities and their impact on serotonergic system are just one component of the age-related effects contributing to biological differences between young and elderly depressives.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Effects of OBX on body weights of young and aged rats. Data represent means and S.E. obtained from 16 to 27 animals in each group. ANOVA for each age group is shown within the panels; asterisks denote individual time points at which the OBX group differs from the corresponding control. Across both ages, OBX had a significant main effect (p < .02) and a significant interaction with days postsurgery (OBX × postsurgery time, p < .0001). The lesioning effect on body weight was significantly greater in the aged cohort (11% average deficit) than in the younger animals (average 4% deficit), as evidenced by a significant interaction of OBX × age × postsurgery time (p < .08).

View larger version (43K): [in this window] [in a new window]   Fig. 2.   Effects of OBX on brain region weights of young and aged rats. Data represent means and S.E.s obtained from 16 to 27 animals in each group. ANOVA for each age group is shown within the panels; asterisks denote individual regions for which the OBX group differs from the corresponding control. Across both ages and all regions, olfactory bulbectomy had a significant main effect (p < .0005), which was also regionally-selective (OBX × region, p < .0001). The effect on brain region weight was significantly greater in the aged cohort than in the younger animals (OBX × age × region, p < .03). Abbreviations: f/p, frontal/parietal cortex; t/o, temporal/occipital cortex; hip, hippocampus; str, striatum; mb, midbrain; bs, brainstem; cb, cerebellum.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Effects of OBX on tactile startle reactivity and consumption of chocolate milk in young and aged rats. Data represent values obtained from 15 to 24 animals in each group. Startle reactivity was evaluated as the proportion of animals freezing for > 3 s after receiving an air-puff on the back of the neck; comparisons were made by 2. Chocolate milk consumption was measured over a 2-h period when animals were given access to both the milk and water; data are shown as means and S.E.s and comparisons were conducted by ANOVA. Asterisks denote significant differences between the OBX and corresponding control groups. In the sham groups, aging had no effect on tactile startle reactivity but did increase the amount of chocolate milk consumed (main effect of age).

View larger version (53K): [in this window] [in a new window]   Fig. 4.   Effects of OBX on open field behavior in young and aged rats. Data represent values obtained from 14 to 33 animals in each group, measured in a 5-min testing period. ANOVA for each measure is shown within the panels, with asterisks denoting OBX groups that differ significantly from the corresponding sham animals. ANOVA combined across both horizontal (ambulation) and vertical (rearing) activity indicates a significant main effect of OBX (p < .0001), main effect of age (p < .02), an interaction of OBX × age (p < .05) and an interaction of OBX × type of activity (p < .04). For horizontal activity, the effect of OBX (significant main effect) was greater in aged animals than in the young cohort (interaction of OBX × age); in addition, the aged sham-operated animals had lower scores than young sham-operated animals (p < .02). For vertical activity, there was a significant increase caused by OBX (main effect), but the effect was of the same magnitude at both ages and there were no differences between the sham-operated groups. For grooming, the decrease caused by OBX (main effect) was restricted to the aged animals only (OBX × age interaction). Fecal boli in the 5-min test period were increased in the lesioned groups (main effect of OBX) but there was no distinction in this effect in aged versus young rats (no interaction of OBX × age); the aged animals also had a lower overall number of boli (main effect of age).

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Effects of OBX on passive avoidance in young and aged rats. Data represent values obtained from 13 to 17 animals in each group. The training latency, the amount of time required to cross from the light to the dark chamber, was determined on the first day, and animals were tested for avoidance of the dark chamber on the second day. Latency data are means and S.E.s, with comparisons conducted by ANOVA. Avoidance data represent the proportion of animals entering the dark chamber, for which group comparisons were made by 2 analysis. Asterisks denote significant differences between the OBX and corresponding control groups. In the sham groups, aging increased the latency of training time (main effect of age) but did not alter the ability of the animals to learn the avoidance task.

View larger version (46K): [in this window] [in a new window]   Fig. 6.   Effects of OBX on [3H]paroxetine binding to the serotonin transporter in membranes prepared from frontal/parietal cortex of young and aged rats. In the upper left panel, determinations were made in 10 animals from each group, using a single ligand concentration. Data represent means and S.E.s, with comparisons by ANOVA; asterisks denote significant differences between the OBX animals and the corresponding sham group; in addition, the aged group had significantly higher values than the young animals (main effect of age). The remaining panels display Scatchard plots obtained from two additional membrane preparations in each group, with each point representing means and bivariate S.E.s across the two preparations. ANCOVA comparisons for each pair of Scatchard plots are shown within the panels, indicating significant differences in total binding (main effect) but not in slopes. Across all four groups, ANCOVA indicates a significant main effect of OBX (p < .005), a significant main effect of age (p < .0001) and a significant difference in the effect of OBX in young versus aged animals (interaction of OBX × age, p < .0001), again without differences in slope.

View larger version (43K): [in this window] [in a new window]   Fig. 7.   Effects of OBX on [3H]paroxetine binding to the serotonin transporter in membranes prepared from hippocampus of young and aged rats. Top left: determinations were made in 10 animals from each group, using a single ligand concentration. Data represent means and S.E.s, with comparisons by ANOVA; the lack of significant differences in the hippocampus was distinguishable from the differences found in the frontal/parietal cortex (OBX × age × region, p < .005). The remaining panels display Scatchard plots obtained from two additional membrane preparations in each group, with each point representing means and bivariate S.E.s across the two preparations. ANCOVA comparisons for each pair of Scatchard plots are shown within the panels, indicating significant differences in total binding (main effect) but not in slopes. Across all four groups, ANCOVA indicates a significant difference in the effect of OBX in young versus aged animals (interaction of OBX × age, p < .0001), again without differences in slope. ANCOVA for comparison of effects in hippocampus versus frontal/parietal cortex confirms the presence of a greater effect in the cortex: OBX × region, p < .03; OBX × age × region, p < .09.

View larger version (36K): [in this window] [in a new window]   Fig. 8.   Effects of OBX on [3H]paroxetine binding to the serotonin transporter in platelet membranes prepared from young and aged rats. Top: determinations were made in 10 animals from each group, using a single ligand concentration. Data represent means and S.E.s, with comparisons by ANOVA; there was no effect of OBX but values in the aged group were higher overall (main effect). Bottom: Scatchard plots obtained from two additional membrane preparations in each group, with each point representing the mean value from the two preparations; S.E.s have been omitted for clarity, but were comparable in magnitude to those seen in the Scatchard plots for brain regions. ANCOVA comparisons across all four groups appear within the panel, again demonstrating only a main effect of age but not lesioning, without changes in slope.

View larger version (33K): [in this window] [in a new window]   Fig. 9.   Effects of OBX on levels of the mRNA encoding the serotonin transporter, measured in midbrain and brainstem of young and aged rats. Values are calculated as the ratio of transporter mRNA to the mRNA encoding -actin. Data represent means and S.E.s obtained from 10 animals in each group, with ANOVA comparisons presented in each panel. ANOVA also indicates no significant effects across both regions taken together.

View larger version (53K): [in this window] [in a new window]   Fig. 10.   Effects of OBX on [125I]iodopindolol binding to -adrenergic receptors in membranes prepared from brain regions of young and aged rats. Left: determinations were made in 10 animals from each group, using a single ligand concentration. Data represent means and S.E.s, with comparisons by ANOVA; there was no effect of OBX but values in the aged group were lower overall (main effect), within each region and across both regions taken together (p < .0001). Right: Scatchard plots obtained from two additional membrane preparations in each group, with each point representing the mean value from the two preparations; S.E.s have been omitted for clarity. ANCOVA comparisons across all four groups appear within the panel, again demonstrating only a main effect of age but not lesioning, without changes in slope. Across the two regions, the effect of aging remained significant (p < .0001) and there was a significant difference in slopes between the two regions (p < .0001).

View larger version (31K): [in this window] [in a new window]   Fig. 11.   Effects of OBX on adenylyl cyclase activity in brain regions of young and aged rats, presented as the percent change from values in the corresponding sham-operated animals. Data represent means and S.E.s obtained from 8 to 10 animals in each group. ANOVA across all measures appears within each panel and asterisks denote individual values for which the effect of OBX in the aged rats differs from that seen for comparably lesioned young rats. Across all three regions, the effect of OBX in aged rats can be distinguished from that in young rats (OBX × age, p < .005). The OBX effect differs according to the in vitro conditions used in the cyclase measurement (OBX × measure, p < .0001; OBX × age × measure, p < .0001) and differs among the regions (OBX × age × measure × region, p < .007). Values for sham-operated animals appear in Table 1.

If, as shown, poor therapeutic outcome is a characteristic of depressed geriatric patients who maintain normal HPA axis reactivity (Kin et al., 1997; Slotkin et al., 1997a), then an animal model of elderly depression that maintains HPA axis integrity would provide a means of distinguishing between the underlying biological differences related to aging of the brain as opposed to those that are glucocorticoid-related. The olfactory bulbectomized (OBX) rat exhibits behavioral and biochemical characteristics that, as in human, are reversed after chronic, but not acute, antidepressant therapy (reviews, Leonard and Tuite, 1981; Kelly et al., 1997). Importantly for our studies, these animals maintain dexamethasone suppression while developing abnormalities of serotonergic and catecholaminergic function that, along with the behavioral abnormalities, resolve with drugs affecting either of these transmitter systems (van Riezen and Leonard, 1990).

Although previous studies have compared OBX effects in juvenile and adult rats (Broekkamp et al., 1986), to our knowledge, no one has explored the OBX model in aging. We have undertaken a comprehensive behavioral and biochemical comparison of the effects of OBX in young and aged rats, with neurochemical determinations focusing on factors that we have found previously to respond differently with age (Slotkin et al., 1989, 1996, 1997a,b; Fumagalli et al., 1996): serotonin transporter expression in brain regions and platelets, and adenylyl cyclase signaling mechanisms and their response to catecholamines. Neurodegeneration, synaptic dysmorphology, and neuronal loss are likely to be present when very old rats are used, obscuring or exacerbating any primary effects on behavior or cellular function (Meister et al., 1995). Accordingly, we have concentrated on 20-month-old rats, rather than examining animals at the very extreme of the life span.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animal Treatments and Behavioral Testing. Studies were carried out in accordance with the declaration of Helsinki and with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health. Male Sprague-Dawley rats (Camm Research Institute, Wayne, NJ) were obtained at 9 weeks or 19 months old and were housed individually with free access to food and water and a 12-h light/dark cycle (6:00 AM to 6:00 PM). Animals were handled and weighed daily from the time of arrival until completion of the study. One week after arrival, animals were anesthetized with 6.5 mg/kg of xylazine and 44 mg/kg of ketamine, given i.p. The top of the skull was shaved and swabbed with an antiseptic, after which a midline frontal incision was made in the scalp and the skin was retracted bilaterally. Burr holes (2-3 mm) were drilled into the skull 2 mm lateral to the bregma suture, after which the olfactory bulbs were severed from the frontal cortex and aspirated according to established protocols (Leonard and Tuite, 1981; Kelly et al., 1997). The cavity was packed with surgical foam, the skin was closed with surgical clips and bupivacaine was applied. The animals were given 40,000 IU/kg of procaine penicillin i.m., and were allowed to recover with warming to maintain body temperature. Sham-operated animals underwent the same procedure except for excision and aspiration of the olfactory bulbs. After surgery, animals were handled and weighed daily.

Tissue Preparations. The day after completion of behavioral testing, animals were anesthetized with sodium pentobarbital (50 mg/kg i.p.) and blood was collected by cardiac puncture using syringes containing 3.8% sodium citrate (pH 7.4) as an anticoagulant, with the volume adjusted to achieve a final concentration of 0.38% after blood collection. As described previously (Moret and Briley, 1991; Slotkin et al., 1991) platelet-rich plasma was isolated by serial centrifugation at 100g (twice), 250g, and 600g, after which all the supernatant fractions were pooled, diluted 50% with calcium-free Krebs-Henseleit medium, and sedimented at 39,000g. The presence of viable platelets was verified under a microscope and platelet membranes were prepared (Moret and Briley, 1991; Slotkin et al., 1991). The membrane suspension was divided into several aliquots that were flash-frozen and stored at 45°C until used.

[³H]Paroxetine Binding. The cell membrane fraction was prepared from brain regions by techniques described previously (Moret and Briley, 1991). The suspension was then used immediately for [³H]paroxetine binding (Moret and Briley, 1991), using approximately 10 µg (platelets) or 100 µg (brain regions) of membrane protein and paroxetine[phenyl-6'-³H] (specific activity 25.4 Ci/mmol, New England Nuclear Corp., Boston, MA) with or without addition of 100 µM serotonin (Sigma Chemical Co., St. Louis, MO) to displace specific binding. Incubations lasted 120 min at 20°C, and were stopped by addition of 5 ml of ice-cold buffer, followed by vacuum filtration and washing onto Whatman GF/C filters presoaked in 0.05% polyethyleneimine (Sigma). Nonspecific binding was approximately 5% of the total.

Serotonin Transporter mRNA. Determinations of the mRNA encoding the serotonin transporter were conducted in the midbrain and brainstem, the regions containing the highest concentration of the cell bodies that supply ascending and descending serotonergic innervation (Meister et al., 1995; Fumagalli et al., 1996). Total RNA was isolated by the CsCl method and transporter mRNA was evaluated with Northern blots (Fumagalli et al., 1996). All lanes were loaded with approximately equivalent amounts (20 µg) of total RNA, as confirmed by absorbance readings at 260 nm and by the intensity of ribosomal RNA bands. Results were taken only from undegraded samples having a ratio of 28S:18S ribosomal RNA ethidium bromide staining of 2 and for which little or no DNA contamination was present (as demonstrated by the absence of residual ethidium bromide fluorescence at the origin well). Probe hybridization was carried out as described previously, using a cRNA probe (Fumagalli et al., 1996). Antisense probe was obtained by in vitro transcription with T3 polymerase using [³²P]CTP as labeled nucleotide. Blots were quantitated by phosphorimaging (Molecular Dynamics, Sunnyvale, CA; ImageQuant 3.3 software) and were standardized by evaluating the -actin mRNA band (Fumagalli et al., 1996).

-Adrenergic Receptors and Adenylyl Cyclase. Cell membrane fractions were isolated, and receptor binding capabilities assessed, by methods described earlier (Slotkin et al., 1996). Again, the overall strategy was to examine binding at a single ligand concentration in preparations from every animal, followed by Scatchard determinations in membranes pooled from several animals to identify whether binding alterations resulted from changes in K_d or B_max. [¹²⁵I]Iodopindolol (specific activity 2200 Ci/mmol, New England Nuclear) was incubated with the tissue membrane preparation (0.2 mg protein) for 20 min at room temperature and the assay was terminated, and labeled membranes were trapped and counted, essentially as described above for [³H]paroxetine binding. Single ligand concentration studies used 67 pM[¹²⁵I]iodopindolol, whereas Scatchard analyses spanned subsaturating to saturating ligand concentrations. Nonspecific binding (displacement by 100 µM isoproterenol; Sigma) ranged from 5 to 20% depending on ligand concentration.

Data Analysis. Parametric data are presented as means and S.E.s, with intergroup comparisons by multivariate ANOVA (data log-transformed whenever variance was heterogeneous). The factors included age (young versus aged rats), treatment (sham-operated versus OBX, or, in some cases, unoperated versus sham-operated), and brain region (a repeated measure, because more than one region was examined from each animal for each variable). In each case, whenever a significant interaction of OBX × age was found, Fisher's Protected Least Significant Difference was used to identify specific intergroup differences; posthoc analysis was not undertaken when only main effects of treatment or age were detected without an interaction. Main effects were considered significant at p < .05 and interaction terms at p < .1.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

In the first 24 h after surgery, young sham-operated rats showed a small decline in body weight before resuming normal growth (Fig. 1). The young OBX group lost a small additional percentage (4%) but nevertheless resumed a nearly-normal growth rate. Other than the initial postoperative drop, the sham-operated group showed no significant weight differences from unoperated controls (n = 7, data not shown). Aged rats subjected to the sham operation showed the same proportional postoperative decline in body weight as had the younger cohort but recovery to baseline levels took longer. The OBX procedure in the aged animals produced a more persistent and pronounced fall in body weight (11%), although by the time of behavioral testing, the differences from the sham-operated group were no longer statistically significant.

Age-dependent differences were also apparent in the effect of OBX on brain region weights (Fig. 2). In young animals, the lesion produced a small, but significant deficit in only one brain region (frontal/parietal cortex). In aged animals, OBX produced a significantly larger loss of frontal/parietal cortical weight (OBX × age, p < .02) and in addition, decrements were seen in the midbrain, which was unaffected by OBX in young animals (OBX × age, p < .02). Brain region weights in sham-operated animals were indistinguishable from those of unoperated controls (data not shown).

Behavior. In young animals, reactivity to an air puff, characterized by a sustained postural freeze, was significantly obtunded by OBX (Fig. 3). Aging alone had no effect on the response, as the sham-operated young animals and corresponding aged group had nearly identical responses. However, when aged animals were subjected to the OBX procedure, they did not show the loss of reactivity, and in fact, the tendency was toward increased reactivity. The two age groups also showed major differences in their propensity to novelty-suppressed feeding. Young animals showed no decrease in chocolate milk consumption after OBX but the same procedure caused a significant decrement in the aged animals. In neither age group did OBX change the consumption of water (range of 0-3 ml, data not shown). In the sham-operated groups, the larger, aged rats consumed more chocolate milk than did young rats.

Biochemistry. Measurements of [³H]paroxetine binding to the high-affinity, presynaptic serotonin transporter indicated significant age- and lesion-related differences in the frontal/parietal cortex (Fig. 6). At a single ligand concentration, values were increased in young OBX animals relative to sham-operated animals. In contrast, lesioning in aged animals produced a decrease, not an increase, in transporter binding. As found previously (Slotkin et al., 1997b), values for aged controls were higher than in the corresponding group of young animals. To explore the mechanisms underlying these effects on the transporter, Scatchard determinations were undertaken. Sham-operated, aged animals displayed higher binding maxima (abscissa intercept) than in young animals, without a change in transporter affinity (slope). Similarly, the increase in transporter binding seen with OBX in young animals, and the decrease with OBX seen in aged animals, reflected changes in B_max and not K_d. In the hippocampus (Fig. 7), there were no significant differences detected at a single ligand concentration but the more sensitive Scatchard determinations verified the presence of statistically significant effects. As with the other region, aging alone was associated with an increase in the number of binding sites, OBX in young animals produced an increase, and OBX in aged animals produced a decrease. Cross-region comparisons indicated significantly smaller effects in the hippocampus than in the frontal/parietal cortex (significant interactions of OBX × age × region for either the single ligand determinations or Scatchard determinations).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present results indicate that the response to OBX differs between young and aged animals at structural, behavioral, and cellular levels. In the aged group, the lesion produced a greater tissue loss in the frontal/parietal cortex than that seen with young animals and, in addition, weight deficits appeared in the midbrain, a region that, in young animals, showed no gross evidence of atrophy. Notably, there were no decrements in the hippocampus, a region susceptible to stress-induced degeneration (McEwen, 1992; Sapolsky, 1994). Our results are consistent with pathway-dependent atrophy, as the olfactory bulbs project prominently to sites in the forebrain, and through interconnections in the amygdala, to the midbrain (Kelly et al., 1997).

If aging produces differential effects on specific neural pathways, then it would be expected that similar distinctions should be detectable at the levels of behavior and neurochemistry. The young OBX group displayed the known characteristics of this depression model (Leonard and Tuite, 1981; Kelly et al., 1997), namely increased open field activity, a prominent loss of the tactile startle response and of passive avoidance behavior, but maintenance of grooming behavior and absence of novelty suppressed feeding. A different behavioral pattern emerged after OBX in aged rats. Although these animals also displayed increased locomotor activity, the effect was significantly greater than in young OBX rats. The aged OBX group did not exhibit loss of passive avoidance behavior, ordinarily one of the most prominent features of OBX, nor did they display decreased tactile startle; however, the aged OBX group showed novelty suppressed feeding and decreased grooming behavior, traits that were not elicited by OBX in young animals.

Serotonergic and catecholaminergic pathways represent prominent features of projections to and from the olfactory bulbs and pharmacological targeting of these transmitter systems reverses the behavioral effects of OBX (Leonard and Tuite, 1981; Kelly et al., 1997). Accordingly, we determined whether the differential age effects on behavior were associated with corresponding alterations in monoaminergic systems. In the frontal/parietal cortex, OBX in young animals produced an increase in the number of presynaptic serotonin transporter sites, as evidenced by an enhanced capacity to bind [³H]paroxetine. Although aging, by itself, also increased the number of sites, the OBX lesion in aged animals produced a change in the opposite direction, namely a decrease. A similar pattern emerged for the interaction of aging with OBX in the hippocampus, albeit with a smaller magnitude of effect when compared to the frontal/parietal cortex: OBX increased transporter expression in young animals but decreased it in aged animals. Although the two regions share the same effect on serotonin transporter sites, the hippocampus did not show any weight loss after OBX, and it is therefore unlikely that the age dependence of the effects of lesioning on serotonergic function is secondary to atrophy. Indeed, degenerative changes have been shown to produce an increase in transporter gene transcription (Meister et al., 1995), whereas we saw no change in transporter mRNA. The lack of change of mRNA also indicates that changes in transporter number probably reflect post-transcriptional alterations related to cellular function. In support of the concept of age-specific CNS effects, blood platelets from young and aged OBX animals showed only small changes that were in the same direction (decrease), regardless of age, the same type of change reported previously for young animals (Leonard and Tuite, 1981; Kelly et al., 1997). Thus, in parallel with behavioral differences in the effect of OBX, actions directed toward the presynaptic serotonin transporter show opposite changes in brain regions of aged versus young animals.

Earlier work in human geriatric depression indicates a dichotomy between platelet serotonin transporter expression and function (Nemeroff et al., 1988; Slotkin et al., 1989, 1997a). Specifically, elderly depressed patients with normal dexamethasone suppression tests show decreased serotonin uptake and reduced imipramine effect, whereas those with non-suppression do not (Slotkin et al., 1997a). Both groups, however, show decreased transporter expression as monitored by [³H]imipramine binding (Nemeroff et al., 1988; Slotkin et al., 1997a). Accordingly, it would be useful to examine whether similar divergence of transporter expression and function occur with the OBX model and specifically whether the dichotomy is selective for age or suppression status. This was not done in the current study because the preparation of platelet membranes for binding studies is not compatible with the intact platelets needed for serotonin uptake measurements. However, previous work on platelet transporter function in aged rats with or without dexamethasone treatment indicates poor homology between effects on platelet transporter expression or function and those on the CNS transporter (Slotkin et al., 1997b). Accordingly, platelet studies in the OBX model of geriatric depression may be an inappropriate representation of corresponding events in the CNS and might be specifically different from platelet effects seen in human depression.

Among the most prominent neurotransmitter-related effects of aging are the loss of adrenergic receptors and of cell-signaling properties required by those receptors (Makman et al., 1978, 1980; Heinsimer and Lefkowitz, 1985; Slotkin et al., 1996). In the current study, aged sham-operated animals displayed lower basal adenylyl cyclase activity and reduced -adrenergic receptor numbers in temporal/occipital cortex and cerebellum. However, in each case, the receptor linkages for catecholamines (assessed with isoproterenol in the cortex and cerebellum, or with dopamine in the striatum) to stimulate adenylyl cyclase showed no specific decrement: activities in the presence of stimulant were affected only to the same extent as was activity without the neurotransmitter stimulant (i.e., no change in the percent stimulation evoked by neurotransmitter agonists). However, upon OBX lesioning, major differences appeared in the neurotransmitter response of aged animals as compared to the young cohort. OBX in young animals elicited up-regulation of adenylyl cyclase total activity (increased forskolin-Mn²⁺ response) and of the linkage of cyclase to activation of G-proteins (fluoride response); accordingly, there was an increase in the G protein-linked response to activation of -adrenergic (isoproterenol) or dopaminergic receptors. In aged animals, the same lesion produced either no change or an actual loss of responsiveness. The age-dependent differences were not secondary to altered receptor expression, which was unaffected by OBX at either age. Thus, as was true for the serotonergic system, catecholaminergic pathways show major, even opposite, agerelated differences in their response to OBX.

Our findings thus substantiate the idea that, in animals, lesions that elicit many of the features of major depressive disorder in human, show distinctly different and even opposite response profiles in aged brain, effects that are expressed at both behavioral and neurochemical levels. Equally important, the neurotransmitter systems displaying age-specific effects are those involved in the mechanisms thought to underlie depression in human, and that represent the systems targeted by the major antidepressants. Accordingly, the OBX model may provide a useful paradigm with which to identify age-dependent differences in synaptic adaptability that may underlie the specificity of biological markers for depression or the effectiveness of antidepressants. This model may then prove useful in uncovering additional biological markers that can distinguish subpopulations that do or do not respond to drug therapies (Kin et al., 1997; Slotkin et al., 1997a).