Reduced [3H]Cyclic AMP Binding in Postmortem Brain from Subjects with Bipolar Affective Disorder

Journal of Neurochemistry Lippincott—Raven Publishers, Philadelphia © 1997 International Society for Neurochemistry Reduced [3H]Cyclic AMP Binding in Postmortem Brain from Subjects with Bipolar Affective Disorder *shafiqur Rahman, *t~peterP. Li, §L. Trevor Young, ~OraKofman, 1~~Stephen J. Kish, and *t1~JerryJ. Warsh * Section of Biochemical Psychiatry and ¶ Human Neurochemical Pathology Laboratory, Clarke Institute of Psychiatry; Departments of ~Psvchiatry and ~:Pharmacology and 5lnstitute of Medical Sciences, University of Toronto, Toronto; §Department of Psychiatry, McMaster University, Hamilton, Ontario, Canada; and ~Department of Behavioral Sciences, Ben Gurion University of the Neget’, Beer Sheva, Israel Abstract: Findings of increased G 5cr levels and forskolinstimulated adenylyl cyclase activity in selective cerebral cortical postmortem brain regions in bipolar affective disorder (BD) implicate increased cyclic AMP (cAMP)-mediated signaling in this illness. Accumulating evidence suggests that intracellular levels of cAMP modulate the abundance and disposition of the regulatory subunits of cAMP-dependent protein kinase (cAMP-dPK). Thus, in the present study, we tested further whether hyperfunctional G5a-linked 3H]cAMPbinding, cAMP signaling a measure occursofinthe BDlevels by deof termining regulatory [subunits of cAMP-dPK, in cytosolic and membrane fractions from discrete brain regions of postmortem BD brain. Specific [3H]cAMP(5 nM) binding was determined in autopsied brain obtained from 10 patients with DSM-Ill-R diagnoses of BD compared with age- and postmortem delay-matched controls. [3H]cAMPbinding was significantly reduced across all brain regions in cytosolic fractions of BD frontal (—22%), temporal (—23%), occipital (—22%) and parietal (— 15%) cortex, cerebellum (—36%), and thalamus (—13%) compared with controls, but there were no differences in [3H]cAMPbinding in the membrane fractions from these same regions. These results suggest that changes occur in the cAMP-dPK regulatory subunits in BD brain, possibly resulting from increased cAMP signaling. The possibility that antemortem lithium and/or other mood stabilizer treatment may contribute to the above changes, however, cannot be ruled out. Key Words: Cyclic AMP-dependent protein kinase— Postmortem brain—Lithium— Bipolar affective disorder. J. Neurochem. 68, 297—304 (1997). The pathogenesis of bipolar affective disorder (BD) has been linked to disturbances in neuronal function secondary to altered action of one or more intracellular second messengers (Manji, 1992; Hudson et al., 1993; Warsh and Li, 1996). Substantial evidence implicates altered heterotrimeric guanine nucleotide binding (G) protein function as the basis for postulated signal transduction disturbances in BD (Schreiber et al., 1991; Avissar and Schreiber, 1992; Young et al., 1993). In this regard, findings of elevated cerebral cortical stimulatory G protein a subunit (Gsa) levels in postmortem BD brain, which were also associated with increased forskolin-stimulated adenylyl cyclase activity, are among the most direct evidence of postreceptor disturbance in this disorder (Young et al., 1991, 1993). Although the mechanism(s) underlying the G protein changes in brain from BD subjects is yet to be elucidated, an equally important question is that of whether such elevations are attended by increased signaling through the G 5a-mediated cyclic AMP (cAMP) signaling cascade. Aside from the above observation of increased forskolin-stimulated adenylyl cyclase activity in temporal and occipital cerebral cortical regions from BD postmortem brain (Young et al., 1993), there is a paucity of data to support this notion. Unfortunately, the rapid changes in brain cAMP levels postmortem preclude the use of measurements of this second messenger to assess the state of cAMP signaling in brain (Jones and Stavinoha, 1979). There are several important protein targets in the cAMP signaling cascade, however, the function of which is modulated by this second messenger. Thus, measurement of the concentrations and/or function of such downstream targets may provide further evidence of altered cAMP signaling in postmortem brain in neuropathological conditions (Nishino et al., 1993). cAMP-dependent protein kinase (cAMP-dPK), a tetramer composed of a dimeric regulatory (R) and two monomeric catalytic (C) subunits, is the primary Received May 23, 1996; revised manuscript received August 21, 1996; accepted August 22, 1996. Address correspondence and reprint requests to Dr. J. J. Warsh at Section of Biochemical Psychiatry, Clarke Institute of Psychiatry, 250 College Street, Toronto, Ontario, Canada MST 1R8. Abbreviotic,os used: AEBSF, 4- ( 2-aminoethyl )benzenesulfonyl fluoride; BD, bipolar affective disorder; cAMP, cyclic AMP; cAMPdPK, cyclic AMP-dependent protein kinase; cGMP, cyclic GMP; IBMX, 3-isobutyl- I -methylxanthine. 297 298 S. RAHMAN ET AL. target of intracellular cAMP signaling (Walaas and Greengard, 1991; Cho-Chung and Clair, 1993; Spaulding, 1993; Francis and Corbin, 1994). Multiple isoforms of both R (Ria, Rl1@, and RITa, RH,@) and C (Ca, C~,and Cy) subunits have been identified (Walaas and Greengard, 1991; Doskeland et al., 1993; Francis and Corbin, 1994). Each R subunit has two cAMP binding sites that show positive cooperativity in cAMP binding and activation (Ogreid and Doskeland, 1981; Ogreid et al., 1983). On binding of cAMP to each R subunit, the inactive holoenzyme dissociates into a dimeric R-subunit complex and two monomeric active C subunits (Walaas and Greengard, 1991; Spaulding, 1993). In addition to modulating catalytic activity, recent studies suggested that sustained elevations in intracellular cAMP levels cause adaptive changes in the levels of cAMP-dPK R subunits (for review, see Spaulding, 1993; Francis and Corbin, 1994). Notwithstanding the possible cell type-specific nature of such changes in RI and Rh subunit abundance (Prashad and Rosenberg, 1978; Liu et al., 1981; Budillon et al., 1995; Garrel et al., 1995), alteration in levels of one, the other, or both R subunits in response to sustained elevations in intracellular cAMP levels may afford a potential marker of modifications in cAMP signaling. We reasoned therefore that measurement of R subunit levels in membrane and/or cytosolic fractions from postmortem brainprovide samples,a 3H1cAMP binding,BDmight estimated “trace” ofbytheI increased cAMP signaling posited to occur in this disorder. In the present study, we examined specific [3H1cAMP binding in four cerebral cortical regions of postmortem BD brains in which G 5a protein levels were previously shown to be higher than in matched controls (Young et al., 1993), as well as in several regions from these same brains that showed no G5a content changes. Webinding report here significantly lower specific 3HIcAMP in cytosolic but not membrane [ fractions in postmortem brain regions from BD compared with a nonpsychiatric nonneurological control group. MATERIALS AND METHODS Materials [3H]cAMP (28.2—31.2 Ci/mmol) was obtained from New England Nuclear-Du Pont (Boston, MA, U.S.A.). cAMP, adenosine, cyclic GMP (cGMP), S ‘-AMP, ATP, EDTA, leupeptin, and 2-mercaptoethanol were purchased from Sigma (St. Louis, MO, U.S.A.). 4-(2-Aminoethyl) benzenesulfonyl fluoride (AEBSF), 3-isobutyl- 1methylxanthine (IBMX), and tris (hydroxymethyl) aminomethane were obtained from Calbiochem (San Diego, CA, U.S.A.). All other chemicals used were of analytical grade. Postmortem brain Autopsied brains were obtained from 10 subjects (five males and five females) with a verified DSMIII-R (American Psychiatric Association, 1987) diagnosis of BD and no neurological disorder as previously described (Young et al., J. Neurochein., Vol. 68, No. I, 1997 1991. 1993). A comparison group, with no history of neurologic, psychiatric, or substance abuse disorder, was matched as closely as possible on age, gender, and postmortem delay. Details of patient histories, including cause of death and antemortem drug treatment, were obtained from available medical records and are summarized in Table I. Autopsied brains were frozen (—70°C)within 24 h after death except for two subjects for which the interval from death to freezing of the brain was 33 and 39 h, respectively. Conditions for dissection and determination of brain pH were as previously described (Young et al., 1991, 1993). Brain regions assayed included frontal (middle gyrus, area 10, Brodmann’s nomenclature), temporal (middle temporal gyrus, area 21), parietal (area 7), and occipital (lips of the calcarine sulcus, area 17) cortex, thalamus (mediodorsal), and cerebellar cortex. [3HIcAMP binding assay Specific I 5H]cAMP binding was performed as previously described (Nishino et al., 1993) with minor modifications. In brief, brain samples (20—30 mg) were homogenized by ultrasonication (sonicator duty cycle setting 30%, micro tip limit at 3, 10 s; Sonics and Materials, Danbury, CT, U.S.A.) in 10 volumes of ice-cold buffer A containing 20 mM TrisHCI (pH 7.4 at 25°C),2 mMEDTA, 25 mM2-mercaptoethanol, 0.5 mM AEBSF, and 10 pg/mI leupeptin, followed by centrifugation at 48,000 g for 30 mm. The supernatant was separated and recentrifuged at 48,000 g for 30 mm. and the final supernatant (cytosolic fraction) was used for [3HJcAMP binding assay. Similarly, the retained pellets from the two centrifugations were pooled for each sample and washed twice by resuspension in 200 p1 of buffer A and centrifugation (48,000g for 15 mm). The final pellets, i.e., membrane fractions, were resuspended in ISO p1 of buffer A. Protein concentration was determined by the method of Bradford (1976) using bovine serum albumin as the standard. Aliquots of the ~.ytosolic and membrane fractions were stored at —70°Cuntil assay. [3HJcAMP binding assays were performed in triplicate in an incubation buffer containing 20 mM phosphate buffer (pH 7.4 at 25°C),2 mM EDTA, and 15 mM 2-mercaptoethanol (PEM buffer), ItHIcAMP (0.125—10 nM). membrane or cytosolic protein (~—25pg), bovine serum albumin (0.25 mg), and 1.5 mM IBMX in a total volume of 500 p1. After incubating for 60 mm at room temperature, incuhations were terminated by rapid filtration under vacuum through glass fiber receptor binding filtermat with the Skatron cell harvester (Skatron Instrument, Lier, Norway) followed by two washes with 2 ml of cold PEM buffer. The radioactivity retained on the filters was measured by liquid scintillation spectrometry (counting efficiency, 42%). Nonspecific binding was defined as the radioactivity bound in the presence of 5 p.M cAMP. For competition experiments, assays were performed with various concentrations of cAMP (0.01 —300 nM), cGMP (0.001—30 pM), 5’-AMP or ATP (10 pM—I mM), and [3H]cAMP (2.5 nM) in a final volume of 500 pl. Animal treatment Male Sprague—Dawley rats (weighing 150—250 g: Charles River, Montreal, Quebec. Canada) were housed individually in a light- (12-h light/dark cycle) and temperature- controlled environment and fed either rat chow (control group) or lithium carbonate (0.22%) in rat chow (Bioserve Ltd.) and water ad libitum for 24 days. Animals were killed by decapitation, the brains were rapidly removed from the [ 3H]cAMP 299 BINDING IN BIPOLAR AFFECTIVE DISORDER TABLE 1. Subject characteristics Diagnosis Age (years)/sex PM (h) Control Bipolar” 48/M 40/M 14 12 Carcinoma Hepatic failure None Lithium, haloperidol Control 66/F 18 Myocardial infarction None Bipolar 70/F 17 Carcinoma with erosion into the aorta Lithium. benztropine, perphenazine Control Bipolar’ 92/F 80/M 17.5 Rectal carcinoma None 17.4 Pneumonia Thioridazine Cause of death Psychotropic drugs Temporal cortex lithium (mM)° 0.079 0.066 0.129 Control 74/F 17 Myocardial infarction 3 Bipolar 70/F 39 Pneumonia Haloperidol, lorazepam, maprotiline 0.083 0.396 Control Bipolar 71/M 71/M 10 23.5 Pneumonia Pneumonia None Lithium Control Bipolar 51/M 37/M 21) 23.5 Myocardial infarction Suicide, overdose of lithium, codeine None Lithium, amitriptyline, codeine. chlordiazepoxide Control 30/M 21 Ruptured inferior vena cava 3 Bipolar 27/F 16 Suicide, overdose of doxepin, amitriptyline Doxepin, amitriptyline 0.154 Control Bipolar 28/F 28/F 23 16 Intraabdominal hemorrhage Suicide, overdose of lithium None Lithium, thioridazine 0.166 Control Bipolar” 58/F 42/F II 33 Rectal carcinoma Suicide, hanging None Lithium, prochlorperazine, isocarboxazid, alprazolam Control Bipolar 86/M 94/F 9 8 Arrhythmia Cerebrovascular accident? None ? 0.127 ND 0.080 PM, postmortem delay; ND, not detected. “Minimal detectable level, 0.066 mM. “Coexistent alcohol abuse. ‘Differential diagnosis of schizoaffective disorder. “Bipolar type 11. cranium, and the prefrontal cortex was dissected over ice and Data analysis stored at —70°Cuntil use. Brain samples were homogenized, All data were expressed as mean ±SEM values. Analysis centrifuged, and assayed as for human postmortem brain. Plasma lithium levels in rats were determined on heparinized of binding data was performed using nonlinear reiterative trunk blood samples by a standard clinical procedure (Sampson Ct aI., 1994) using an ion-selective electrode (Nova Biomedical electrolyte analyzer). All animal procedures were performed in strict accordance with the guidelines of the Canadian Council on Animal Care and were approved by the local institutional Animal Care Committee. Determination of brain lithium levels Lithium concentrations were determined in BD temporal cortex by inductively coupled argon plasma emission spectrometry (Jones et al., 1987). In brief, —30 mg of brain curve-fitting techniques with the RADLIG software package (McPherson, 1985). Statistical analyses of the data were performed by two-way ANOVA followed by posthoc contrasts for simple effects or Tukey’s test using the SPSS 6.0 (Chicago, IL, U.S.A.) statistical software package. Simple group comparisons of age, postmortem delay, and pH were analyzed by Student’s t tests. Relationships between lithium levels, age, or postmortem delay and [‘H]cAMP binding were assessed using the Pearson Product Moment correlation analysis. Values of p < 0.05 were considered statistically significant. tissue was digested in 1 ml of concentrated ultrapure nitric acid in a sealed Teflon vessel using a pressure-programmed microwave heating cycle. After digestion was complete (22 RESULTS mm), the vessels were cooled and opened, and digestates Subject characteristics BD and comparison groups did not differ significantly (p > 0.1) in mean age (56 ± 8 vs. 60 ± 7 were rinsed out and diluted to 10.0 ml with deionized ultrafiltered water (18 MO). Samples were injected into a Jarrel Ash model 6lE, 34-channel simultaneous inductively coupled argon plasma emission spectrometer using an ultrasonic nebulizer to achieve maximum detection sensitivity. Lithium was detected and quantified at an emission wavelength of 670.784 nm. Sensitivity and coefficient of variation were 1.5 ng/ml digestate and 10%, respectively. years) or postmortem delay (21 ± 3 vs. 16 ±2 h). Six of the BD patients had a history of maintenance lithium treatment within the 6 months before death; four of these had evidence of continued lithium use within 4 weeks antemortem. The remaining four pa- .1. Neurochem., Vol. 68. No. 1, 1997 300 S. RAHMAN ET AL. otides tested, cAMP was the most potent (displacement occurred at nanomolar concentrations). In comparison, cGMP, 5 ‘-AMP, and ATP showed only marginal displacement even at micromolar concentrations (data not shown). 3H]cAMPin cytosolic fracFIG. of 1. postmortem Saturation experiment with [ cortex. Cytosolic fractions tion control temporal (—-25 pg of protein) were incubated with varying concentrations of[3H]cAMP in the absence (0; total binding) and presence (LI; nonspecific binding) of 5 pM cAMP as described in Materials and Methods. Specifically bound [3H]cAMPis also indicated (.). Inset: Scatchard plot of specific [3H]cAMPbinding data. Each point represents the mean of triplicate determinations. The data were best fit for a one-site model. B/F, bound/free. Distribution of 113H]cAMP binding in postmortem human brain Cytosolic [3H]cAMP binding varied across the brain regions from 409 to 629 fmol/mg of protein in control and from 298 to 488 fmol/mg of protein in BD subjects (Table 2). [3H]cAMP binding in membrane fractions varied more widely, ranging from 185 to 51 8 fmol/mg of protein in control and from 167 to 430 fmol/mg of protein in BD subjects. For both cytosolic and membrane fractions, there was a significant main effect of brain region on [3H I cAMP binding identified by two-factor ANOVA [cytosolic fraction, F = 3.08, df(5,80); p = 0.013; membrane fraction, F = 8.39, df(5,80); p < 0.001] without a significant interaction with subject group. Pairwise post hoc analysis (Tukey’s test) of the mean differences for the brain regions tients had no documented history of lithium use in the 6-month interval before death, although previous lithium therapy could not be completely ruled out based on available medical records (Table 1). Brain pH, which affords an index of agonal status, was slightly but significantly higher (4%) in BD (6.56 ±0.05) compared with control (6.30 ± 0.08; p <0.05) subjects. Characterization of I 3H] cAMP binding In preliminary experiments, specific binding of [3H1cAMP (5 nM) was measured in cytosolic and membrane fractions of postmortem frontal cortex from a control subject. Maximum [3HJcAMP binding was reached at 60 mm, remained stable up to 120 mm, and was linear with respect to protein concentration between 0.01 and 0.08 mg of protein (data not shown). As demonstrated in Fig. I, specific [3H1 cAMP binding in cytosolic fractions of control postmortem temporal cortex was saturable and exhibited a single high-affinity binding site with a dissociation constant (Ks) of 0.73 ±0.04 nM and maximal binding density (Br,,ax) of 232 ±7 fmol/mg of protein (n = 3) as estimated by nonlinear reiterative curve fitting and Scatchard analyses. Nonspecific binding determined in the presence of 5 fLM cAMP was 8—10% of the total binding. In parallel experiments with autopsied temporal cortical membranes, nonlinear regression analysis of the data was best fit by a two-site model: a high-affinity site with KD of 0.9 :t 0.1 nM and B,,,, 5 of 154 ±28 fmol/mg of protein and a low-affinity site with KD of 3.6 ± 0.4 nM and Bma. of 283 ±36 fmol/mg of protein (n = 3). To characterize further the binding sitesnucleotides labeled by 3H]cAMP, the ability of several different [ displace specific [3H]cAMP binding was analyzed to in competition experiments. Among the various nudej. Neurochein., Vol. 68, No. 1, 1997 across the control and BD groups demonstrated significantly (p = 0.05) higher cytosolic [3H]cAMP binding only in frontal and occipital cortex compared with temporal cortex. In contrast, membrane [3H]cAMP binding was significantly (p < 0.05) higher in occipital and parietal cortical areas (390—518 fmol/ mg of protein) compared with temporal cortex, thalamus, and cerebellum (170—27 1 fmol/mg of protein). For both control and BD groups, the ratio of cytosolic to membrane [3H]cAMP binding varied from 1.4 to 2.6 across the brain regions examined except for panetal cortex, for which the ratio of cytosohic to membrane [3H]cAMP binding was —-~0.8. Specific [3H]cAMP binding in BD brain Mean specific [3HI cAMP binding in cytosolic fractions was significantly lower [F = 6.55, df(l,80); p = 0.012] across all brain regions in BD compared with matched control subjects, with reductions of 15—23% in the cerebral cortical regions, 36% in cerebellar cortex, and 13% in thahamus (Fig. 2A). As noted above, there was no interaction, however, between diagnostic group and brain region. Post hoc analysis of simple effects of diagnosis revealed that only the largest reduction in [3H]cAMP binding in BD cerebellum approached statistical significance (p = 0.06) compared with controls, whereas the 23% decrement in the temporal cortex did not reach statistical significance (p = 0.132). In contrast, there were no significant differences [F = 1.41, df(l,80); p > 0.1] in [3HJcAMP binding in membrane fractions between BD and matched controls for the six brain regions examined (Fig. 2B). In both control and BD groups, there were no significant correlations between age, postmortem delay, or brain pH and [3H]cAMP binding in cytosolic or membrane fractions from any of the brain regions examined. [3HJcAMP BINDING IN BIPOLAR AFFECTIVE DISORDER 30] TABLE 2. Regional distribution of specific [3HJcAMP binding in cytosolic and membrane fractions from control and BD postmortem brain Specific [3H]cAMPbind ing (fmol/mg of protein) Membrane fraction Cytosolic fraction Brain region Control Frontal cortex 613 ± 62 Temporal cortex Parietal cortex Occipital cortex Cerebellum 417 409 629 435 Thalamus 476 ± 48 ± 46 ± 110 ±60 ±92 BD Control 478 ± 60 439 ±79 276 ± 38 (8)” 323 ± 46 (10) 340 ± 51(5) 488 ± 59 (10)” 298 ±45 (7) 415 ±25 (6) 518 ± 15 430 ± 5t) 240 ± 59 185 ± 30 BD 326 272 390 430 167 200 ±54 (8) ±51(10) ±34 (5)” ±60 (10)” ±25 (7) ±20 (6) Data are mean ± SEM values for the numbers of subjects indicated in parentheses. [‘HIcAMP(5 nM) was incubated in triplicate with cytosolic or membrane fractions of various brain regions as described in Materials and Methods, using 5 pM cAMP to define nonspecific binding. Statistical differences between brain regions were assessed across BD and controls using pairwise post hoc analyses (Tukey’s test): °p< 0.05 compared with the temporal cortex; 5p < 0.05 compared with temporal cortex, thalamus, and cerebellum. Relationship of brain lithium levels to [3H IcAMP binding In rats receiving chronic lithium treatment, {3H]cAMP binding showed a trend toward a significant reduction (84 ± 5% of control values, n = 6;p = 0.06) in cytosolic (controls, 910 ± 42 fmol/mg of protein; lithium-treated, 762 ± 47 fmol/mg of protein) but not membrane (control, 1,018 ± 68 fmol/mg of protein; lithium-treated, 861 ± 63 fmol/mg of protein) fractions from prefrontal cortex. Although the plasma lithium concentrations in these animals (0.85 ± 0.09 mmol/L) did not correlate significantly with either cytosolic or membrane [3H]cAMP binding, the reduced cytosohic [1H]cAMP binding in rat prefrontal cortex suggested that antemortem lithium exposure may have altered the levels of cAMP-dPK R subunits. As this effect might have contributed to the lower cytosolic [3H]cAMP binding in BD brain, we determined whether a relationship existed between lithium concentrations and specific [3H]cAMP binding in cytosohic and membrane fractions from BD brain. Lithium levels measured in the BD temporal cortex varied from 0.066 to 0.396 mM (Table 1). Furthermore, no significant correlation was observed between lithium concentration and membrane or cytosolic (Fig. 3) [3H]cAMP binding in the temporal cortex. DISCUSSION The present study demonstrates, for the first time, significantly reduced [3HI cAMP binding in the cytosolic but not membrane fractions from postmortem brain of BD compared with control subjects matched on age and postmortem delay. As [3HIcAMP binds predominantly to the R subunits of cAMP-dPK in brain, the observed decrements likely reflect changes involving one or more cAMP-dPK R-subunit isoforms in BD brain compared with controls. The differences in [3H]cAMP binding between BD and control subjects could not be explained by differences in age. postmortem delay in removal and freezing of brain tissue, or agonal status as estimated by brain pH, nor was there any relationship with history of lithium treatment or tissue lithium concentrations. Accordingly, these observations raise the possibility that reduced [3H]cAMP binding in postmortem BD brain may reflect changes in intracellular signaling linked either directly or indirectly to the pathophysiology of this disorder. The decrease in [3H]cAMP binding in the cytosolic fractions, although small, was evident across all BD brain regions as supported by the significant main effect of diagnostic group without an interaction between the latter factor and brain region on the ANOVA. Despite the significant main effect, post hoc analysis performed to identify specific regions that accounted for this effect only revealed marginally statistically significant reductions in [3H]cAMP binding for cerebellum and temporal cortex, regions that showed the largest differences in BD compared with control subjects. The inability to detect statistically significant reductions across a broader range of brain regions, given the significant main effect and lack of interaction on ANOVA, is likely attributable to the small sample size and relatively large variance of the estimates of [3H1cAMP binding obtained. Of note, however, is that the manifestation of the trend to diminished [3H]cAMP binding also in the cerebellum suggests the process(es) that accounts for this reduction occurs widely in BD brain. It is unlikely that the reduced [3H]cAMP binding found in BD brain is attributable to differences in age or postmortem delay, as BD and control subjects were matched on these variables, and [3H]cAMP binding did not correlate significantly with these factors. There were also no apparent differences in [3H]cAMP binding between BD subjects who died by suicide versus .1. “l,’uroch,’,n. , t’ol. 68. No. 1, /997 302 S. RAHMAN ET AL. not affected by neuroleptics (Nishino et al., 1993) or lithium (up to 1 mM) in vitro (S. Rahrnan et al.. unpublished data), it is unlikely that residual drug in postmortem brain accounts for the observed changes. but the contribution of long-term drug treatment to the reductions in [3HIcAMP binding in BD brain cannot be completely ruled out. However, chronic antidepres- sant administration in rats actually increased membrane cerebral cortex cAMP binding (Perez et al.. 1989, 1991). Although [3H1cAMP binding, in this study, was moderately reduced in the cytosolic but not membrane fractions from prefrontal cortex in rats that received lithium chronically, the decreases did not correlate significantly with plasma lithium levels. Furthermore, we did not observe a significant relationship between cytosolic [3HI cAMP binding and lithium concentration in BD temporal cortex, and the lithium lev- els were subtherapeutic based on in vivo estimates in brain (0.3—0.9 mM) of BD patients on maintenance lithium therapy (Kato et al., 1992; Sachs et al.. 1995). Taken together, these findings suggest reduced [3H]cAMP binding in BD brain is not likely related to 3H]cAMP binding in (A) cytoFIG. 2.andScatter plots of specific [ from postmortem BD brain solic (B) membrane fractions (.) and matched controls (0): temporal cortex (TCX), frontal (FCX), parietal (PCX), and occipital (OCX) cortices, cerebellar cortex (CBX), and thalamus (THAL). Horizontal bars indicate the mean values. Two-way ANOVA of cytosolic [3H]cAMP binding data revealed significant main effects of group [F = 6.55, df(1,80); p = 0.012] and region [F = 3.10, df(5,80); p = 0.013] but no interaction [F = 0.31, df(5,80); p > 0.1]. ANOVA of membrane [3HJcAMP binding revealed a significant effect of region [F = 8.39, df(5,80); p < 0.0011 but not group [F = 1.41, df(1 80); p > 0.01] and no interaction [F = 0.30, df(5,80); p >0.1]. those who did not. Thus, it is unlikely that suicidality contributed to the observed differences. Although brain pH, an index of the severity of the agonal state (Harrison et al., 1995), was slightly higher (4%) in BD compared with control postmortem brain, this parameter did not correlate with either cytosolic or membrane I ~H1cAMP binding, thus making it unlikely that differences in agonal status account for the observed reduction in [3H]cAMP binding in BD brain. Moreover, the decrease does not appear to be related to a general loss of nerve cells because [‘H] cAMP binding in the membrane fractions was not different between BD and control subjects. In the present study, saturation experiments were not performed owing to limited availability of postmortem BD brain samples. Thus, it was not possible to discern whether the lower [31-I]cAMP binding in BD temporal cortex is due solely to changes in B,,,,,, or also involves alterations in K[). The BD subjects in the present study had taken drugs such as neuroleptics and antidepressants, in addition to lithium (see Table 1). As [3H1cAMP binding is J. N,’urochem., Vol. 68, No. I, 1997 antemortern lithium or antidepressant treatment, although effects of other maintenance medications cannot be ruled out. The extent to which [‘H]cAMP binding reflects specific labeling of the R subunits of cAMP-dPK in postmortem human brain fractions is a crucial question underpinning the attribution of the observed differences in binding to changes in the levels of specific cAMP-dPK R-subunit isoforms. In line with previous observations (Nishino et al., 1993), [3H]cAMP labeled a homogeneous population of sites in cytosolic fractions at least in the representative temporal cortex region in which the binding kinetics were examined. Furthermore, the results of competition experiments with nucleotide analogues support the specificity of [3H]cAMP binding to a cAMP binding protein(s). That this binding occurs primarily to cAMP-dPK R subunits in brain is supported by the finding that cAMP FIG. 3. Scatter plot ofcytosolic [3H]cAMP binding versus lithium levels in BD temporal cortex. The hatched line indicates the calculated regression line. No significant relationship was found between cytosolic [3H]cAMP binding and lithium concentrations in this region (r = 0.01 ‘p > 0.1; Pearson Product Moment correlation analysis). [ 3HJcAMP BINDING IN BIPOLAR AFFECTIVE DISORDER reduced the photoactivated incorporation of 8-N 332PIcAMP into RI and RH subunits in rat brain by [90 and 94%, respectively (Walter et al., 1978). Furthermore, Stein et al. (1987) also showed a high percentage of cAMP binding (56—84%) to RIl subunits in membrane and cytosol from neurons, astrocytes, and oligodendroglia fractionated from rat brain. In the membrane fractions from temporal cortex, [3H]cAMP binding was best described by a two-site model with estimated KD values of 0.9 and 3.6 nM, respectively. Other investigators have also reported heterogeneous [1H]cAMP binding sites in membrane fractions of bovine myocardium and skeletal muscle (Doskeland and Ogreid, 1981; Ogreid and Doskeland, 1981; Ogreid et al., 1983; Doskeland et al., 1993). Whereas cAMP binds nonselectively to type I or type II cAMP-dPK, it is unlikely that differential localization of R-subunit subtypes in cellular fractions accounts for the observed difference in binding kinetics between the two fractions. Both RI and Rh subtypes are expressed in cytosolic fractions of rat brain (Walter et al., 1978), yet cytosolic [3HIcAMP binding in human temporal cortex is characterized by single-site binding kinetics as shown in this study and reported by others (Nishino et al., 1993). A more plausible explanation for the two classes of [3HIcAMP binding site characterized in the membrane fractions, however, is that they may represent stepwise binding of cAMP to the R subunits in a positive cooperative manner (Doskeland and Ogreid, 1981; Ogreid et al., 1989). The distribution of [3H 1 cAMP binding in cytosolic and membrane fractions varied across the brain areas examined, although the regional differences were more marked in the latter fractions. Cytosolic [3H]cAMP binding was highest in occipital and frontal cortices, 303 sensitization, may result in increased kinase activity and enhanced protein phosphorylation at subsaturating cAMP concentrations (Greenberg et al., 1987) consequent to a decreased R to C subunit ratio. Consonant with this notion, Perez et al. (1995) recently reported increased cAMP-dependent endogenous phosphorylation in platelets from euthymic bipolar patients. Further studies are required to determine whether such a relative reduction in R subunits, accompanied by altered cAMP-dPK activity and cAMP-dependent endogenous phosphorylation, occurs in BD brain. Moreover, from a pathophysiological perspective, it remains to be demonstrated whether changes in cAMP-dPK function are primary disease-related changes specific to BD or reflect secondary adaptive responses consequent to up- stream alterations in processes modulating cAMP levels and, if so, whether they afford any protective advantage in this disorder. Finally, in light of the results showing moderate reduction of cytosolic [3H]cAMP binding in prefrontal cortex of rats receiving lithium chronically, the possibility that antemortem lithium treatment may, in part, contribute to the above changes also cannot be excluded. Regardless of the mechanism underlying the loss of R subunit of cAMP-dPK in BD brain, the present results, along with the growing body of evidence of altered second messenger and signal transduction processes in brain and peripheral tissues of BD patients (Young et al., 1991, 1993, 1994; Perez et al., 1995; Mathews et al., 1996; Jope et al., 1996; Warsh and Li, 1996), support the hypothesis that abnormalities in signal transduction mechanisms play an important role in the pathophysiology of this major psychiatric disorder. in agreement with the findings of Nishino et al. (1993), Acknowledgment: This work was supported by a grant whereas in the membrane fractions binding was highest in occipital and parietal cortices, moderate in temporal from the Medical Research Council of Canada (to J.J.W. and L.T.Y.). S.R. is a postdoctoral fellow supported by the Clarke Institute of Psychiatry Research Foundation. We are cortex and cerebellum, and lowest in thalamus. Greater [3H]cAMP binding in the cytosolic fractions (1.4— 2.6-fold) than in corresponding membrane fractions also concurs with earlier findings of enrichment of cAMP-dPK R subunits in the cytosolic fractions of rat, guinea pig, and bovine cerebral cortex (Walter et al., 1978). Among possible mechanisms that might account for the reduced cytosolic [3H]cAMP binding in BD brain, altered synthesis or degradation of cAMP-dPK R subunits merits further consideration. It is unlikely, however, that differential redistribution of cAMP-dPK R subunits occurs between cytosolic and membrane pools in BD in contrast to control brain, because specific [3H]cAMP binding was not significantly different in the membrane fractions of BD compared with control brain. Some evidence suggests that the loss of R subunits may impact on regulatory functions in the cAMP signaling cascade. In Aplysia, reduced levels of R subunits, found after treatments that produce long-term grateful to Dr. J. Forrester of Novamann International for brain lithium content determination, Ms. Kathleen Shanak and Ms. Kin Po Siu for autopsied brain preparation, and Ms. Cathy Spegg for her advice on statistical analyses. REFERENCES American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorder,,, 3rd edit., revised. American Psychiatric Press, Washington, D.C. Avissar S. and Schreiber G. (1992) The involvement of guanine nucleotide binding proteins in the pathogenesis and treatment of affective disorders. Biol. Psychiatry 31, 435—459. Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein—dye binding. Anal. Biochem. 72, 248—252. Budillon A., Creseto A., Kondrashin A., Nesterova M., Merlo G., Clair T., and Cho-Chung Y. S. (1995) Point mutation of the autophosphorylation site or in the nuclear location signal causes protein kinase A Rll 3 regulatory subunit to lose its ability to revert transformed fibroblasts. Proc. NatI. A cad. Sd. USA 92, 10634—10638. Cho-Chung Y. S. and Clair T. (1993) The regulatory subunit of cAMP-dependent protein kinase as a target for chemotherapy J. Neurochem., Vol. 68, No. 1, 1997 S. RAHMAN ET AL. 304 of cancer and other cellular dysfunctional-related disease. Pharmacal. Ther. 60, 265—288. Dtskeland S. 0. and Ogreid D. (1981) Binding proteins for cyclic AMP in mammalian tissues. ml. J. Biochem. 13, 1— 19. Doskcland 5. 0., Maronde E., and Gjertsen B. T. (1993) The genetic subtypes of cAMP-dependent protein kinase—functionally different or redundant. Biochim. Biophy.c. Acta 1178, 249—258. Francis S. H. and Corbin J. D. (1994) Structure and function of cyclic nucleotide-dependent protein kinases. Annu. Rev. Physiol. 56, 237—272. Garrel G., Delahaye R.. Hemmings B. A.. and Counis R. (1995) Modulation of regulatory and catalytic subunit levels of cAMPdependent protein kinase A in anterior pituitary cells in response to direct activation of protein kinases A and C or after GnRH stimulation. Neuroendocrinology 62, 514—522. Greenberg S. M., Castellucci V. F., Bayley H., and Schwartz J. H. I 987) A molecular mechanism for long—term sensitization in Aplvsia. Nature 329, 62—65. Harrison P. J., Heath P. R., Eastwood S. L., Burnet P. W. J., McDonald B., and Pearson R. C. A. (1995) The relative importance of premortem acidosis and postmortem interval for human brain gene expression studies: selective mRNA vulnerability and comparison with their encoded proteins. Neuro,sci. Let!. 200, 151— 154. Hudson C. J., Young L. T., Li P. P., and Warsh J. J. (1993) CNS signal transduction in the pathophysiology and pharmacotherapy of affective disorders and schizophrenia. Synapse 13, 278— 293. Jones C. L., Hodge V. F., Schoengold D. M., Biesiada H., and Starks T. H. ( I 987) An Inierlahorators’ Study of Inductively Coupled Plasma Atomic Emission Spectro.cc’ops’ Method 6010 and Dige.vlion Method 3050. Publication no. EPA-600/4-87-032, U.S. Environmental Protection Agency, Las Vegas. Nevada. Jones D. J. and Stavinoha W. B. (1979) Microwave inactivation as a tool for studying the neuropharmacology ofcyclic nucleotides, in Neuropharmadology of Cyclic Nucleotides (Palmer G. C., ed), pp. 253—281. Urban and Schwarzenbcrg, Munich. Jope R. S., Song L., Li P. P., Young L. T., Kish S. J., Pacheco M. A., and Warsh J. J. (1996) The phosphoinositide signal transduction system is impaired in bipolar affective disorder brain. J. Neurochem. 66, 2402—2409. Kato T., Takahashi S., and Inubushi T. (1992) Brain lithium concentration by ‘Li-and- ‘H-magnetic resonance spectroscopy in bipolar disorder. Psychiatry Re,,. 45, 53—63. Licameli V., Mattiace L. A., Erlichman J., Davies P., Dickson D., and Shalit-Zagardo B. (1992) Regional localization of the regulatory subunit (R%8) of the type II cAMP-dependent protein kinase in human brain. Brain Res. 578, 61—68. Liu A. Y.-C., Chan T., and Chen K. Y. (1981) Induction of the regulatory subunit of type I adenosine cyclic 3 ‘,S ‘-monophosphate-dependent protein kinase in differentiated N- 18 mouse neuroblastoma cells. Cancer Res. 41, 4579—4587. Manji H. K. (1992) G proteins: implications for psychiatry. Am. J. Psychiatry 149, 746-760. Mathews R., Li P. P., Young L. T., Kish S. J., and Warsh J. J. (1996) Increased Ga,,,, 1 immunoreactivity in postmortem occipital cortex from patients with bipolar affective disorder. Biol. P.svchiatry (in press). McPherson G. A. (1985) Analysis of radioligand binding experiments: a collection of computer programs for the IBM PC. I. Pharomacol. Methods 14, 213—218. Nishino N., Kitamura N., Hashimoto T., Kajimoto Y., Shirai Y., Murakami N., Nakai T., Osamu K., Shirakawa 0., Mita T., and Nakai H. (1993) Increase in [‘HI cAMP binding sites and decrease in Ga and G,,a immunoreactivities in left temporal J, N,’,,,’o,’hcn,., Vol. 68, No. I, /997 cortices from patients with schizophrenia. Brain Rca. 615, 41 — 49. Ogreid D. and Doskeland S. 0. (1981) The kinetics of association of cyclic AMP to the two types of binding sites associated with protein kinase 11 from bovine myocardium. FEBS Lelt. 129, 287—292. Ogreid D., Doskeland S. 0., and Miller J. P. (1983) Evidence that cyclic nucleotides activating rabbit muscle protein kinase I interact with both types of cAMP binding sites associated with the enzyme. J. Biol. Chem. 258, 1041—1049. Ogreid D., Ekanger R., Suva R. H.. Miller J. P., and Doskeland S. 0. (1989) Comparison of the two classes of binding sites (A and B) of type I and type II cyclic-AMP dependent protein kinascs by using cyclic nucleotide analogues. Furl. Biocheoi. 181, 19—31. Prashad N. and Rosenberg R. N. (1978) Induction of cyclic AMPbinding proteins by dibutyryl cyclic AMP in mouse neuroblastoma cells. Biochim. Biophy.s. Ac/a 539, 459—469. Perez J., Tinelli D., Brunello N.. and Racagni G. (1989) cAMP dependent phosphorylation of soluble and microtubule fraction after prolonged DM1 treatment in rat cerebral cortex. Liii. I. Phar,nacol. 172, 305—3 16. Perez J., Tinelli D., Bianchi E.. Brunello N., and Racagni G. (1991 1 cAMP binding proteins in the rat cerebral cortex after administration of selective 5-HT and NE reuptake blockers with antidepressant activity. Neurop.vvchophormacology 4, 57—64. Perez J., Zanardi R., Mon S., Gasperini M., Smeraldi E., and Racagni G. (1995) Abnormalities of cAMP-dependent endogenous phosphorylation in platelets from patients with bipolar disorder. Am. J. Psychiatry 152, 1204— 1206. Sachs G. S., Renshaw P. F., Lafer B ., .Stoll A. L., Guimaraes A. R.. Rosenbaum J. F., and Gonzalez R. G. (1995) Variability of’ brain lithium levels during maintenance treatment: a magnetic resonance spectroscopy study. Biol. P.sychiatr,’ 38, 422—428. Sampson M., Ruddel M., and Elm R. J. (1994) Lithium determinations evaluated in eight analyzers. Clin. Chem. 40, 869—872. Schreiber G., Avissar S., Danon A., and Belmaker R. H. (1991 Hyperfunctional G proteins in mononuclear leukocytes of patients with mania. Biol. Psychiatry 29, 273—280. Spaulding S. W. (1993) The ways in which hormones change cyclic adenosine 3 ‘,S ‘-monophosphate-dependent protein kinase subunits, and how such changes affect cell behavior. Endocr. Re,’. 14, 632—650. Stein J. C., Farooq M., Norton W. T., and Rubin C. S. (1987) Differential expression of isoforms of the regulatory subunit of type II cAMP-dependent protein kinase in rat neurons, astrocytes and oligodendrocytes. J. Biol. Chemn. 262, 3002—3006. Walaas S. I. and Greengard P. (1991) Protein phosphorylation and neuronal function. Pharmacol. Rev. 43, 299—349. Walter U., Kanof P., Schulman I-I., and Greengard P. (1978) Adenosine 3.5 ‘-monophosphate receptor proteins in mammalian brain. J. Biol. Chem. 253, 6275—6280. Warsh J. J. and Li P. P. (1996) Second messenger systems and mood disorders. Curr. Opin. Psychiatry 9, 23—29. Young L. T.. Li P. P., Kish S. J., Siu K. P., and Warsh J. J. (1991) Postmortem cerebral cortex G, a subunit levels are elevated in bipolar affective disorder. Brain Re,,. 553, 323—326. Young L. T., Li P. P., Kish S. J., Siu K. P., Kamble A., Hornykiewicz 0.. and Warsh J. J. (1993) Cerebral cortex G,a protein levels and forskolin-stimulated cyclic AMP formation are increased in bipolar affective disorder. J. Neurochem. 61, 890— 898. Young L. T., Li P. P., Kamble A., Siu K. P., and Warsh J. J. (1994) Mononuclear leukocyte levels of G proteins in depressed patients with bipolar or major depressive disorder. Am. J. P,s’s’chiatry 151, 594—596.