An Explanation for the Monoamine Imbalance of Major Depression

ORIGINAL ARTICLE Elevated Monoamine Oxidase A Levels in the Brain An Explanation for the Monoamine Imbalance of Major Depression

Jeffrey H. Meyer, MD, PhD; Nathalie Ginovart, PhD; Anahita Boovariwala, BSc; Sandra Sagrati, BSc; Doug Hussey, BSc; Armando Garcia, BSc; Trevor Young, MD, PhD; Nicole Praschak-Rieder, MD; Alan A. Wilson, PhD; Sylvain Houle, MD, PhD

Context: The monoamine theory of depression pro- emission tomography, was used to measure MAO-A DVS poses that monoamine levels are lowered, but there is no (specific distribution volume), an index of MAO-A den- explanation for how monoamine loss occurs. Mono- sity, in different brain regions (prefrontal cortex, ante- amine oxidase A (MAO-A) is an enzyme that metabo- rior cingulate cortex, posterior cingulate cortex, cau- lizes monoamines, such as serotonin, norepinephrine, and date, putamen, thalamus, anterior temporal cortex, dopamine. midbrain, hippocampus, and parahippocampus).

Objective: To determine whether MAO-A levels in the Results: The MAO-A DVS was highly significantly el- brain are elevated during untreated depression. evated in every brain region assessed (t test; P=.001 to ϫ −7 3 10 ). The MAO-A DVS was elevated on average by 34% Setting: Tertiary care psychiatric hospital. (2 SDs) throughout the brain during major depression. Patients: Seventeen healthy and 17 depressed individu- Conclusions: The sizable magnitude of this finding als with major depressive disorder that met entry crite- and the absence of other compelling explanations for ria were recruited from the care of general practitioners monoamine loss during major depressive episodes led and psychiatrists. All study participants were otherwise healthy and nonsmoking. Depressed individuals had been to the conclusion that elevated MAO-A density is the medication free for at least 5 months. primary monoamine-lowering process during major depression. Main Outcome Measure: Harmine labeled with carbon 11, a radioligand selective for MAO-A and positron Arch Gen Psychiatry. 2006;63:1209-1216

AJOR DEPRESSIVE DISOR- ductions in monoamine transporter indi- der is an important ill- ces in depression are low compared with ness because it has a monoamine transporter loss observed in 1-year prevalence of 2% symptomatic neurodegenerative dis- to 5% and ranks fourth ease.14 Moreover, no abnormality in an in- amongM causes of death or injury.1 For more dex of serotonin transporter density was than 30 years, it has been theorized that lev- found in vivo in untreated depressed in- els of monoamines, such as serotonin, nor- dividuals.8 Decreased monoamine synthe- epinephrine, and dopamine, are generally sis is unlikely during depression because low in the brain during untreated major de- investigations of monoamine synthesis en- Author Affiliations: Vivian M. 2 Rakoff PET Imaging Centre pressive episodes (MDEs). However, no zymes in monoamine nuclei tend to find (Drs Meyer, Ginovart, convincing mechanism of monoamine loss no change or modest increases in de- 3-11 10,11,15 9 Praschak-Rieder, Wilson, and has ever been found. pressed individuals. Studies attempt- Houle, Mss Boovariwala and Previous investigations of mono- ing to determine whether monoamine pre- Sagrati, and Messrs Hussey and amine transporters and monoamine syn- cursor uptake is reduced in depression are Garcia) and Mood and Anxiety thesis enzymes have not identified a promi- inconclusive because they are typically Disorders Division (Drs Meyer nent monoamine-lowering process during confounded by recent antidepressant use. and Young), Clarke Division, untreated depressive episodes. Loss of Abnormally elevated monoamine oxi- Centre for Addiction and monoamine-releasing neurons is an un- dase B density seems less likely to occur Mental Health and Department likely mechanism of monoamine loss, since in depression, as one investigation16 of of Psychiatry, University of Toronto, Toronto, Ontario; and some investigations report no reduction monoamine oxidase B density in the amyg- Department of General in monoamine transporters and the larg- dala found no significant difference in de- Psychiatry, Medical University est reported reductions in monoamine pressed individuals. of Vienna, Vienna, Austria transporter density indices range from 14% Monoamine oxidase A (MAO-A) is a (Dr Praschak-Rieder). to 25%.3-8,12,13 Even the largest reported re- logical enzyme to investigate in depres-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1209

35 30

30 25

25 20

Activity, kBq/mL Activity, kBq/mL Activity, 15

10 10

5 5

0 0 02040 60 80 100 02040 60 80 100 Time, min Time, min

Anterior Cingulate Cortex Temporal Cortex 30 30

25 25

20 20

15 15 Activity, kBq/mL Activity, Activity, kBq/mL Activity, 10 10

5 5

0 0 02040 60 80 100 02040 60 80 100 Time, min Time, min

Figure 1. Time activity curves for harmine labeled with carbon 11 ([11C]harmine) demonstrating reversible kinetics. Time activity curves for a representative depressed individual (closed circles) and a healthy individual (open circles) are shown. This pair of study participants was chosen because each has monoamine

oxidase A (MAO-A) DVS (an index of MAO-A density) values within 10% of their group mean, and these 2 participants have near identical areas under their [11C]harmine plasma input curves (within 1%).

sion because it regulates levels of all 3 major monoamines tron emission tomography (PET).25,26 [11C]Harmine is a (serotonin, norepinephrine, and dopamine) in the brain.17 selective, reversible PET radiotracer for MAO-A, and 11 Whether MAO-A levels in the brain are abnormal during MAO-A DVS is an index of specifically bound [ C]har- MDEs is unknown because each previous investigation of mine.25-27 [11C]Harmine PET demonstrates the requisite MAO-A in the brain has had at least 2 critical confounders properties of a PET radiotracer for measurement of MAO- and/or limitations,18-23 including complete nonspecificity A25-27: harmine has a high affinity (Ki=2nM) and a selec- of technique for MAO-A vs monoamine oxidase B, enroll- tive affinity for the MAO-A enzyme. [11C]Harmine shows ment of study participants who had recently taken medi- high brain uptake in humans with the greatest uptake in cation, unclear diagnosis of individuals who committed sui- regions with the highest MAO-A density.25-27 [11C]Har- cide, small sample size, and lack of differentiation between mine also shows reversible kinetics in all regions with spe- early-onset depression and late-onset depression. In con- cific binding in humans25-27 (Figure 1). The MAO-A in- trast to the typical, early-onset depression before the age hibitors can fully displace specific binding of [11C]harmine of 40 years, late-onset depression probably has a different in animal models,25,27 and MAO-A inhibitors at clinically pathophysiologic mechanism attributable to lesions and/or tolerable doses can displace 80% of specific binding in hu- degenerative disease.24 mans.26 The metabolites of harmine are polar and do not 28 The MAO-A DVS (specific distribution volume), an in- cross the blood-brain barrier. The main advantage of dex of MAO-A density, is measurable in vivo in the brain [11C]harmine over clorgyline labeled with carbon 11 using harmine labeled with carbon 11 ([11C]harmine) posi- ([11C]clorgyline) is that [11C]harmine has reversible brain

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1210

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/26/2021 kinetics, whereas [11C]clorgyline shows slowly reversible brain kinetics.29 The main advantage of [11C]harmine over Table. Sample Demographics deuterium-substituted [11C]clorgyline is that deuterium- substituted [11C]clorgyline has substantial non–MAO-A Healthy Depressed 29 Group Group binding in humans in some brain regions. Demographic (n = 17) (n = 17) The hypothesis of the present study is that the MAO-A DV will be elevated throughout the brain during MDEs Age, mean ± SD, y 34 ± 8 34 ± 8 S Women, No. 7 9 in medication-free individuals with major depressive dis- Men, No. 10 8 order and typical early-onset illness. An elevation in MAO-A Education, mean ± SD, y 15 ± 2 15 ± 2 density is hypothesized because greater MAO-A could ex- Psychiatric diagnosis* None Major depressive episode; cessively lower brain monoamine levels.17 The location of major depressive disorder elevated MAO-A density was hypothesized to be through- First major depressive episode NA 8 out the brain because monoamine receptor abnormalities Second major depressive episode NA 5 Third major depressive episode NA 4 in depression consistent with lowered monoamine levels No previous antidepressant NA 11 have been reported in several brain regions, including the treatment prefrontal cortex, striatum, and midbrain.3-6,8,30 Previous antidepressant NA 6 treatment†

METHODS Abbreviation: NA, not applicable. *Study participants did not have comorbid Axis I disorders, borderline personality disorder, or antisocial personality disorder. PARTICIPANTS †No study participant with depression had received antidepressant treatment within the past 5 months. Twenty individuals with an MDE and major depressive disorder were recruited, and 17 depressed individuals completed the protocol (mean±SD age, 34±8 years; 8 men and 9 women). IMAGE ACQUISITION AND ANALYSIS Seventeen age-matched healthy individuals were recruited (mean±SD age, 34±8 years; 10 men and 7 women). Each un- A dose of 370 MBq of intravenous [11C]harmine was adminis- derwent an [11C]harmine PET scan. Participants were be- tered as a bolus for each PET scan. An automatic blood sampling tween 20 and 49 years of age. Healthy participants were age system was used to measure arterial blood radioactivity continu- matched within 4 years to depressed patients (Table). ously for the first 10 minutes. Manual samples were obtained at Allstudyparticipants(MDEandhealthy)werephysicallyhealthy 5, 10, 15, 20, 30, 45, 60, and 90 minutes. The radioactivity in whole and nonsmoking and had no history of neurotoxin use. Partici- blood and plasma was measured as described previously.26 Frames pants were nonsmoking because it is reported that smoking can were acquired as follows: 15 frames of 1 minute, then 15 frames lower MAO-A levels, which could create greater variance in mea- of 5 minutes. [11C]Harmine was of high radiochemical purity surement.31 Women in perimenopause or menopause were ex- (Ͼ96%; mean±SD, 98.4%±0.8%; n=34) and high specific activ- cluded. Healthy participants were screened to rule out any Axis ity (mean±SD, 43±18 terabecquerels/mmol at the time of injec- I disorders, and depressed participants were screened to rule out tion). The PET images were obtained using a GEMS 2048-15B any comorbid Axis I disorders using the Structured Clinical In- camera (intrinsic in-plane resolution; full width at half maxi- terview for DSM-IV.32 All participants were screened to rule out mum, 5.5 mm; Scanditronix Medical, General Electric, Uppsala, borderline and antisocial personality disorder using the Structured Sweden). All images were corrected for attenuation using a ger- Clinical Interview for DSM-IV for Axis II disorders.33 All partici- manium 68/gallium 68 transmission scan and reconstructed by pants underwent a urine drug screen on the day of the [11C]har- filtered back projection using a Hanning filter. mine PET scan. All depressed patients underwent common blood For the region of interest (ROI) method, each participant tests to rule out medical causes of disturbed mood (thyroid func- underwent magnetic resonance imaging (GE Signa 1.5-T scan- tion, electrolyte levels, and complete blood cell count). ner; spin-echo sequence, T1-weighted image; x, y, z voxel di- For depressed patients, the mean±SD age at onset of ill- mensions, 0.78, 0.78, and 3 mm, respectively; GE Medical Sys- ness was 23±10 years. Patients were in their first (n=8), sec- tems, Milwaukee, Wis). The ROIs were drawn on magnetic ond (n=5), or third (n=4) MDE. No patient with depression resonance images that were coregistered to each summed had received antidepressant treatment within the past 5 months, [11C]harmine PET image using a mutual information algo- and 11 depressed patients had never received antidepressant rithm.35 The location of the ROI was verified by visual assess- treatment. For depressed patients, a diagnosis of MDE second- ment of the ROI on the summated [11C]harmine PET image. ary to major depressive disorder was based on the Structured The ROIs were drawn to sample the prefrontal cortex, ante- Clinical Interview for DSM-IV for Axis I disorders32 and a con- rior cingulate cortex, posterior cingulate cortex, caudate, pu- sultation with a psychiatrist (J.H.M.). For patients with MDE, tamen, thalamus, anterior temporal cortex, midbrain, and a hip- the minimum severity for enrollment was based on a cutoff score pocampal and parahippocampal region. The definitions of the of 17 on the 17-item Hamilton Depression Rating Scale.34 The ROIs were similar to our previous investigations.8,36 The pre- mean±SD Hamilton Depression Rating Scale score for partici- frontal cortex regions (left and right) were drawn in trans- pants with MDE was 22±3. Additional exclusion criteria in- verse planes extending 32.5 mm in the z-axis and included Brod- cluded MDE with psychotic symptoms, bipolar disorder (type mann areas 9, 10, 46, and part of 8 and 47. The anterior cingulate I or II), history of self-harm or suicidality outside episodes of cortex (Brodmann areas 24 and part of 32) was sampled from depression, and history of alcohol or other drug abuse. adjacent transverse planes extending 26 mm in the z-axis. The For each study participant, written consent was obtained putamen and thalamus were drawn within adjacent transverse after the procedures had been fully explained. The study and planes to maximally sample the individual structures. These recruitment procedures were approved by the Research Ethics planes extended 13 mm in the z-axis. The remaining regions Board for Human Subjects at the Centre for Addiction and Men- were sampled from adjacent transverse planes that extended tal Health, University of Toronto. 19.5 mm in the z-axis. For the temporal cortex, the anterior

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1211

30 s

20 MAO-A DV

0 Prefrontal Temporal Anterior Posterior Thalamus∗ Caudate† Putamen† Hippocampus† Midbrain‡ Cortex∗ Cortex∗ Cingulate Cortex† Cingulate Cortex∗

Figure 2. Comparison of monoamine oxidase A (MAO-A) DVS (an index of MAO-A density) between depressed and healthy study participants. On average, MAO-A DVS was elevated by 34% or 2 SDs in depressed individuals. The hippocampal region also samples the parahippocampus. Differences between groups were highly statistically significant in each region. *PϽ1ϫ10−5;†PϽ1ϫ10−4;‡P=.001.

third of the temporal cortex was sampled, and this included RESULTS Brodmann area 38 and part of 20, 21, and 22. The anterior cingulate cortex and the posterior cingulate cortex (part of Brod- mann areas 23 and 30) were drawn in transverse planes rela- As expected, given the previous report of no relationship tive to the corpus callosum. between age or sex with MAO-A density,37 there was no The kinetics of [11C]harmine can be described with an un- relationship between age or sex and regional MAO-A DVS constrained 2-tissue compartment model (described as method 26 in our sample (analysis of covariance; effect of age, F1,32=0.3 B in our previous publication). Highly identifiable fits with the to 0.001; P=.50 to .98; analysis of variance; effect of sex, unconstrained 2-tissue compartment model are obtainable for the 26 F1,32=0.4 to 0.001; P=.50 to .98). DVS. The DVS is an index of specific binding and represents the concentration of the specifically bound radiotracer in tissue rela- There was a highly significant elevation in MAO-A DVS tive to plasma concentration at equilibrium. (In previous publi- in all regions in the depressed group compared with the cations, DV was referred to as DV .26) The DV can be expressed healthy group (independent-sample t test, P=.001 to S B S Ͻ in terms of kinetic rate parameters as: .001; mean difference in MAO-A DVS between groups, 34%; mean effect size, 2) (Figure 2). Because this was not the situation of a single significant finding among a = K1 × k3 DVs number of nonsignificant findings, a correction for mul- k k 2 4 tiple comparisons was not performed. A multiple analysis of variance was also performed, with regional MAO-A where K1 and k2 are influx and efflux rates for radiotracer pas- DVS as the dependent variable and diagnosis as a predic- Ͻ sage across the blood-brain barrier and k3 and k4 describe the ra- tor variable (effect of diagnosis: F9,24=5.8; P .001). dioligand transfer between the free and nonspecific compart- To examine whether MAO-A DVS is related to par- ment and the specific binding compartment. K1/k2 is similar among ticular clinical characteristics in addition to diagnosis, different individuals (for further details see Ginovart et al26). 11 secondary post hoc analyses were performed using the The [ C]harmine PET measure of the MAO-A DVS was previously found to be reliable. Under test-retest conditions, for Pearson correlation coefficient, correlating regional the regions evaluated in this study, the mean absolute differ- MAO-A DVS with the following clinical characteristics: ence in MAO-A DVS, expressed as a percentage of MAO-A DVS, duration of illness, episode number, duration of epi- ranged from 5% to 17% (n=6 individuals) (J.H.M., A.A.W., S.H., sode, illness severity based on the 17-item Hamilton et al, unpublished data, 2005). Depression Rating Scale Score,34 and lifetime history of antidepressant treatment. None of the correlations reached the trend level (PϽ.10). STATISTICAL ANALYSIS

The primary analysis was an independent-sample t test compar- COMMENT ingMAO-ADVS betweendepressedandhealthyindividualsforeach brain region, since it was expected that MAO-A DVS would be significantly elevated in each brain region in the depressed group. We The main finding was that MAO-A DVS, the index of chose to examine each individual region because it would also be MAO-A density, was elevated throughout the brain on of interest whether there were no differences in some brain regions. average by 34% (2 SDs). Monoamine oxidase A metabo-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1212

Vesicle Containing Monoamine

Monoamine Monoamine Presynaptic Postsynaptic Transporter Receptor

C D MAO-A MAO-A

Monoamine Monoamine Presynaptic Postsynaptic Transporter Receptor

Figure 3. Modernization of monoamine theory of depression. A, Description of monoamine release in a synapse in a healthy person. B, During a major depressive episode, monoamine oxidase A (MAO-A) density is elevated, resulting in greater metabolism of monoamines, such as serotonin, norepinephrine, and dopamine, in the brain. C and D, Range of outcomes. If the monoamine transporter density for a particular monoamine is low during a major depressive episode (C), the effect of an elevated MAO-A level on reducing that particular monoamine in the extracellular space is somewhat attenuated, resulting in a moderate loss of monoamine. This eventually results in a moderate severity of symptoms associated with long-term loss of that particular monoamine. If the monoamine transporter density for a particular monoamine is not low during a major depressive episode (D), then there is no protection against the effect of elevated MAO-A levels. The extracellular concentration of the monoamine is severely reduced, and symptoms associated with long-term loss of that particular monoamine eventually become severe. Some postsynaptic receptors increase in density when their endogenous monoamine level is low in the long term. Mostly, MAO-A is found in norepinephrine- releasing neurons but is reported to be detectable in other cells, such as serotonin-releasing neurons and glia. Even so, MAO-A metabolizes serotonin, norepinephrine, and dopamine in vivo.

lizes all 3 major monoamines (serotonin, norepineph- elevated MAO-A levels can be viewed as a general mono- rine, and dopamine)17 in the brain, and no previous study amine-lowering process (with no relationship to particu- has convincingly explained why monoamine levels may lar symptoms), whereas the regional density of mono- be low during MDEs; therefore, it is plausible that an el- amine transporters has a selective influence on particular evation in brain MAO-A density is the primary mono- monoamines (with a strong relationship with particular amine-lowering process during MDEs. symptoms). Elevated brain MAO-A density during MDEs has im- Data to support this advancement of the monoamine portant implications for the monoamine theory of de- model are found in investigations of drug-free (ie, no pression when combined with previous neuroimaging re- medication for Ն3 months), depressed individuals with sults in medication-free depressed patients4,8,36,38 (ie, no early-onset depression who do not have comorbid psy- medication for Ն3 months). An advanced monoamine chiatric or medical illnesses.4,8,36,38 This includes 4 earlier theory (Figure 3A-D) can be conceptualized: During PET brain imaging studies4,8,36,38 in addition to the pres- an MDE, elevated MAO-A levels increase the metabo- ent study. The binding potential (BP) is often measured lism of monoamines such as serotonin, norepinephrine, in PET studies and is an index of receptor density. Dur- and dopamine. Thereafter, individual monoamine trans- ing MDEs, greater regional serotonin transporter BP is porter densities have a secondary influence on specific associated with more severely pessimistic thinking (dys- extracellular monoamine levels. If the monoamine trans- functional attitudes),8 and greater striatal dopamine porter density for a particular monoamine is low, the effect transporter BP is associated with more severe motor of greater monoamine metabolism on extracellular mono- retardation.4 In addition, during MDEs, increased cortex amine levels is somewhat attenuated, resulting in a mod- serotonin 2 BP occurs when severe pessimism is erate monoamine loss. Long-term, moderate loss of a par- present,36 and increased striatal dopamine 2 BP occurs ticular monoamine in specific brain regions eventually when severe motor retardation is present.38 Increased results in moderate severity of particular symptoms. If serotonin 2 receptor density can occur when serotonin the monoamine transporter density for a particular mono- is lowered in the long term,39-42 and increased dopamine amine is not low during an MDE, then the extracellular 2 BP (as measured with PET with raclopride labeled concentration of the monoamine is severely reduced and with carbon 11) can occur when the extracellular dopa- symptoms associated with long-term regional loss of that mine level is low.38 These studies4,8,36,38 argue that particular monoamine eventually become severe. In short, patients with MDEs and greater monoamine transporter

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1213

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/26/2021 density are additionally vulnerable to monoamine loss, to years an MDE is typically present before treatment.1 Sec- show up-regulation (or increase in BP) of receptors sen- ond, long-term increasing of monoamine levels reduces sitive to monoamine loss, and have greater severity of symptoms substantially more than short-term increasing particular symptoms. of monoamine levels.30 Third, monoamine levels in- This advancement in monoamine theory relies on the creased in the long term, consequent to antidepressant ad- principle that a single explanation for the findings that ministration, influence secondary and tertiary messenger directly support it is likely to be the correct explana- system targets, such as cyclic adenosine monophosphate tion. Although one could conceive of different explana- response element binding protein and brain-derived neu- tions for individual findings,4,8,36,38 it would be ex- rotrophic factor, particularly in animal models.43 Avail- tremely difficult to generate an alternative, single, cohesive able evidence demonstrates that these targets are abnor- explanation for the entire study series. The statistical sig- mal during MDEs.43 An advanced model of long-term nificance of the comparison of mean regional MAO-A DVS monoamine loss has the advantage of potentially being con- for the current study was P=3ϫ10−6, and the statistical sistent with secondary and tertiary messenger theories of significances of the previous 4 studies were high (P=.003 depression if long-term monoamine loss influences the for elevations in prefrontal serotonin 2 BP in depression same secondary and tertiary messenger targets as mono- with severe pessimistic thinking, P=.005 for elevation in amine levels increased in the long term. striatal dopamine 2 BP for depressed individuals with mo- An important facet of this advanced monoamine model tor retardation, PϽ.001 for correlation between prefron- is that, depending on the combinations of long-term sero- tal serotonin transporter BP and pessimistic thinking, and tonin, norepinephrine, or dopamine loss, one would ex- P=.006 for correlation between striatal dopamine trans- pect different combinations of the associated symptoms porter BP and motor retardation). Each study sample was during depressive episodes. This is entirely consistent with gathered separately. In general, the probability of mul- the particular definition of an MDE that requires 5 of 9 tiple independent chance events occurring together can symptoms.44 In addition, this advanced monoamine model be estimated by multiplying the probability of each sepa- proposes that variable levels of long-term monoamine loss rate event occurring. For each study, the significance re- result in variable levels of symptom severity across in- flects the likelihood of the finding occurring by chance dividuals, as is commonly observed. alone. If one took the specific viewpoint that the find- A significant problem in depression research is a lack ings of these studies are unrelated, independent chance of valid animal models.43 An important implication of this events (ie, there is no common model to explain them), study is that new animal models of major depression can then the probability of all of these findings occurring to- be created that have chronic monoaminergic abnormali- gether results in a P value of 2.7ϫ10−16. Given the re- ties similar to what is found in untreated depressive epi- sulting statistical significance, it is exceedingly unlikely sodes. For example, dexamethasone administration to that these findings represent independent chance events. older Sprague-Dawley rats can increase MAO-A density In addition, the entire sample consisted of 89 depressed in the brain by 300%,45 and Wistar rats can be geneti- and 103 healthy individuals. Depressed patients were drug cally selected to breed high platelet serotonin trans- free for at least 3 months plus 5 half-lives of medication. porter density and low platelet serotonin density sub- None had comorbid medical illnesses or Axis I psychi- lines.46 Breeding of rat sublines with high brain atric illness,4,8,36,38 and 94% of the sample was nonsmok- monoamine transporter density and exposing these ani- ing. The collective set of highly significant findings in mals to an MAO-A–increasing paradigm could generate unbiased samples across a series of studies to support a a new animal model of depression with long-term mono- common theory must be considered strongly. amine loss. Our monoamine model of depression is consistent with There are some limitations in this work. Our study has many other findings of monoamine research in depres- the major advantage of measuring MAO-A DVS in vivo. On sion. This modern model views elevated MAO-A levels the other hand, there are some disadvantages with PET. as a pathologic process that results in excessive loss of 3 The resolution of PET does not permit detailed localiza- major monoamines, since increasing monoamine levels tion of MAO-A at the cellular level. For example, an in- through antidepressant inhibition of MAO-A or inhibi- crease in MAO-A density could be attributed to a greater tion of monoamine transporter function are long- density of MAO-A per mitochondrion or an increase in mi- standing therapeutic treatments.30 This model also pre- tochondrial density within MAO-A–containing neurons. dicts that indices of monoamine loss should be best A second issue is that, although MAO-A DVS is an index detected in subgroups with high severity of particular of MAO-A density, it also reflects 2 other parameters, namely symptoms. Therefore, investigations that compare mono- free and nonspecific binding and MAO-A affinity for the amine abnormalities between depressed and healthy in- radioligand. These other parameters are unlikely to influ- dividuals and do not selectively sample individuals with ence the interpretations of this study. The estimates of free severe symptom clusters should find abnormalities in de- and nonspecific binding in the present study (found by ob- pression at a rate exceeding chance but not at a fre- taining the difference between total distribution volume26 quency near 100%, as is observed.3-8,30,38 and specific distribution volume) were similar in patients In this advanced monoamine model, it is proposed that and healthy controls, and an elevation in MAO-A affinity long-term rather than short-term monoamine loss is im- would be expected to have a similar functional effect of in- portant for several reasons. First, the period for which creasing monoamine removal through increased binding monoamine receptor abnormalities are found in the stud- to monoamines. We also acknowledge in the interpreta- ies directly supporting this model4,8,36,38 spans the months tion of this study that there are other neurochemical and

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1214

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/26/2021 neuropathologic models of depression (which may be in- port for other projects from Eli Lilly, GlaxoSmithKline, terrelated). In addition, this article is not intended to ar- and Lundbeck. gue that other neurochemical and neuropathologic pro- Funding/Support: This research received project sup- cesses do not play a role in antidepressant treatment. Finally, port from the Canadian Institutes of Health Research we acknowledge that scientific models of disease are typi- (FRN-69052). cally replaced with more complicated scientific models of Additional Information: To be presented in part at a sym- disease, and it is possible that this new theory of mono- posia at the Society of Biological Psychiatry; May 19, 2006; amine dysregulation in untreated MDEs will be replaced Toronto, Ontario. by a more complicated theory at a future date. Acknowledgment: We thank technicians Alvina Ng, BSc, Future study should investigate why MAO-A levels are and Ken Singh, BSc, research assistant Irina Vitcu, BSc, elevated during MDEs. It is well known that the plasma chemist Alexandra Chestakova, MSc, and engineers Terry cortisol concentration is elevated during depression,30,43 Bell, BTech, and Ted Brandts-Harris, BASc, for their as- and it was demonstrated that dexamethasone adminis- sistance with this project. We thank Stephen Kish, PhD, tration can substantially increase MAO-A activity in for his suggestions regarding the manuscript. 19-month-old Sprague-Dawley rats.45 Since dexamethasone and cortisol both have agonist effects on glucocorticoid receptors, it is possible (should this animal model REFERENCES be representative of adult humans) that greater cortisol agonist effects during depressive episodes contribute to 1. Ustun TB, Ayuso-Mateos JL, Chatterji S, Mathers C, Murray CJ. Global burden of depressive disorders in the year 2000. Br J Psychiatry. 2004;184: an elevation in MAO-A levels. It would also be interest- 386-392. ing to explore a relationship between genotypes and 2. Schildkraut JJ, Kety SS. Biogenic amines and emotion. Science. 1967;156: MAO-A levels through complex models such as gene- 21-37. environment interaction designs in large samples.47 It 3. Mann JJ, Huang YY, Underwood MD, Kassir SA, Oppenheim S, Kelly TM, Dwork AJ, Arango V. A serotonin transporter gene promoter polymorphism (5- seems unlikely that a simple, single genotype to MAO-A HTTLPR) and prefrontal cortical binding in major depression and suicide. Arch activity relationship will be explanatory, since Balciu- Gen Psychiatry. 2000;57:729-738. niene et al48 did not find any simple, single-genotype re- 4. Meyer JH, Kruger S, Wilson AA, Christensen BK, Goulding VS, Schaffer A, Mini- lationship to MAO-A activity in a postmortem study of fie C, Houle S, Hussey D, Kennedy SH. Lower dopamine transporter binding po- humans. tential in striatum during depression. Neuroreport. 2001;12:4121-4125. 5. Klimek V, Schenck JE, Han H, Stockmeier CA, Ordway GA. Dopaminergic ab- This is the first study, to our knowledge, to measure normalities in amygdaloid nuclei in major depression: a postmortem study. Biol MAO-A levels in the brain in medication-free individu- Psychiatry. 2002;52:740-748. als during MDEs secondary to major depressive disor- 6. Klimek V, Stockmeier C, Overholser J, Meltzer HY, Kalka S, Dilley G, Ordway GA. Reduced levels of norepinephrine transporters in the locus coeruleus in major der. An index of MAO-A density, the MAO-A DVS, was depression. J Neurosci. 1997;17:8451-8458. elevated by 34% (or 2 SDs) in depressed individuals. The 7. Stockmeier CA. Involvement of serotonin in depression: evidence from post- sizable magnitude of this finding and the absence of other mortem and imaging studies of serotonin receptors and the serotonin transporter. compelling explanations for a primary monoamine- J Psychiatr Res. 2003;37:357-373. lowering process during MDEs led to the conclusion that 8. Meyer JH, Houle S, Sagrati S, Carella A, Hussey DF, Ginovart N, Goulding V, Kennedy elevated MAO-A density should be viewed as the pri- J, Wilson AA. Brain serotonin transporter binding potential measured with carbon 11-labeled DASB positron emission tomography: effects of major depres- mary monoamine-lowering process during untreated ma- sive episodes and severity of dysfunctional attitudes. Arch Gen Psychiatry. 2004; jor depression. This finding has major implications for 61:1271-1279. modernizing the monoamine theory of depression. Taken 9. Rosa-Neto P, Diksic M, Okazawa H, Leyton M, Ghadirian N, Mzengeza S, Nakai in conjunction with other studies of serotonin and do- A, Debonnell G, Blier P, Benkelfat C. Measurement of brain regional alpha- [11C]methyl-L-tryptophan trapping as a measure of serotonin synthesis in medi- pamine receptors in medication-free depressed individu- cation-free patients with major depression. Arch Gen Psychiatry. 2004;61: 4,8,36,38 als, it can be argued that monoamine transporter 556-563. density has a secondary role to influence long-term loss 10. Zhu MY, Klimek V, Dilley GE, Haycock JW, Stockmeier C, Overholser JC, Melt- of specific monoamines, resulting in a model of variable zer HY, Ordway GA. Elevated levels of tyrosine hydroxylase in the locus coer- severity of long-term monoamine loss during untreated uleus in major depression. Biol Psychiatry. 1999;46:1275-1286. 11. Boldrini M, Underwood MD, Mann JJ, Arango V. More tryptophan hydroxylase major depression. Important implications of the new in the brainstem dorsal raphe nucleus in depressed suicides. Brain Res. 2005; monoamine theory of depression are that different com- 1041:19-28. binations of severe long-term monoamine loss can ex- 12. Austin MC, Whitehead RE, Edgar CL, Janosky JE, Lewis DA. Localized decrease plain the variety of symptom cluster presentations found in serotonin transporter-immunoreactive axons in the prefrontal cortex of depressed subjects committing suicide. Neuroscience. 2002;114:807-815. in MDEs and that great potential exists to generate more 13. Neumeister A, Willeit M, Praschak-Rieder N, Asenbaum S, Stastny J, Hilger E, complex animal models of persistent monoamine loss that Pirker W, Konstantinidis A, Kasper S. Dopamine transporter availability in symp- may closely resemble the monoaminergic abnormalities tomatic depressed patients with seasonal affective disorder and healthy controls. of untreated depression. Psychol Med. 2001;31:1467-1473. 14. Hornykiewicz O. Biochemical aspects of Parkinson’s disease. Neurology. 1998; 51(suppl 2):S2-S9. Submitted for Publication: January 1, 2006; final revision 15. Bonkale WL, Murdock S, Janosky JE, Austin MC. Normal levels of tryptophan received February 22, 2006; accepted February 23, 2006. hydroxylase immunoreactivity in the dorsal raphe of depressed suicide victims. Correspondence: Jeffrey H. Meyer, MD, PhD, College J Neurochem. 2004;88:958-964. Street Site, Centre for Addiction and Mental Health, PET 16. Karolewicz B, Klimek V, Zhu H, Szebeni K, Nail E, Stockmeier CA, Johnson L, Ordway GA. Effects of depression, cigarette smoking, and age on monoamine Centre, 250 College St, Toronto, Ontario, Canada M5T oxidase B in amygdaloid nuclei. Brain Res. 2005;1043:57-64. 1R8 ([email protected]). 17. Shih JC, Chen K, Ridd MJ. Monoamine oxidase: from genes to behaviour. Annu Financial Disclosure: Dr Meyer has received grant sup- Rev Neurosci. 1999;22:197-217.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1215

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/26/2021 18. Grote SS, Moses SG, Robins E, Hudgens RW, Croninger AB. A study of selected 33. Blais MA, Norman DK. A psychometric evaluation of the DSM-IV personality dis- catecholamine metabolizing enzymes: a comparison of depressive suicides and order criteria. J Personal Disord. 1997;11:168-176. alcoholic suicides with controls. J Neurochem. 1974;23:791-802. 34. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 19. Gottfries CG, Oreland L, Wiberg A, Winblad B. Lowered monoamine oxidase ac- 23:56-62. tivity in brains from alcoholic suicides. J Neurochem. 1975;25:667-673. 35. Studholme C, Hill D, Hawkes D. An overlap invariant entropy measure of 3D medi- 20. Mann JJ, Stanley M. Postmortem monoamine oxidase enzyme kinetics in the cal image alignment. Pattern Recognit. 1999;32:71-86. frontal cortex of suicide victims and controls. Acta Psychiatr Scand. 1984; 36. Meyer JH, McMain S, Kennedy SH, Korman L, Brown GM, DaSilva JN, Wilson 69:135-139. AA, Blak T, Eynan-Harvey R, Goulding VS, Houle S, Links P. Dysfunctional atti- 21. Sherif F, Marcusson J, Oreland L. Brain gamma-aminobutyrate transaminase and tudes and 5-HT(2) receptors during depression and self-harm. Am J Psychiatry. monoamine oxidase activities in suicide victims. Eur Arch Psychiatry Clin Neurosci. 2003;160:90-99. 1991;241:139-144. 37. Saura J, Andres N, Andrade C, Ojuel J, Eriksson K, Mahy N. Biphasic and region- 22. Ordway GA, Farley JT, Dilley GE, Overholser JC, Meltzer HY, Balraj EK, Stockmeier specific MAO-B response to aging in normal human brain. Neurobiol Aging. 1997; CA, Klimek V. Quantitative distribution of monoamine oxidase A in brainstem mono- 18:497-507. amine nuclei is normal in major depression. Brain Res. 1999;847:71-79. 38. Meyer JH, McNeeley HE, Sagrati S, Boovariwala A, Martin K, Verhoeff NPLG, Wil- 23. Galva MD, Bondiolotti GP, Olasmaa M, Picotti GB. Effect of aging on lazabemide son AA, Houle S. Elevated putamen D2 receptor binding potential in major de- binding, monoamine oxidase activity and monoamine metabolites in human fron- pression with motor retardation: an [11C]raclopride positron emission tomogra- tal cortex. J Neural Transm Gen Sect. 1995;101:83-94. phy study. Am J Psychiatry. 2006;163:1594-1602. 24. Krishnan KR. Biological risk factors in late life depression. Biol Psychiatry. 2002; 39. Stockmeier CA, Kellar KJ. In vivo regulation of the serotonin-2 receptor in rat 52:185-192. brain. Life Sci. 1986;38:117-127. 25. Bergstrom M, Westerberg G, Langstrom B. 11C-harmine as a tracer for mono- 40. Roth BL, McLean S, Zhu X, Chuang D. Characterization of two [3H]ketanserin amine oxidase A (MAO-A): in vitro and in vivo studies. Nucl Med Biol. 1997; recognition sites in rat striatum. J Neurochem. 1987;49:1833-1838. 24:287-293. 41. O’Regan D, Kwok RP, Yu PH, Bailey BA, Greenshaw AJ, Boulton AA. A behav- 26. Ginovart N, Meyer JH, Boovariwala A, Hussey D, Rabiner EA, Houle S, Wilson ioural and neurochemical analysis of chronic and selective monoamine oxidase AA. Positron emission tomography quantification of [11C]-harmine binding to inhibition. Psychopharmacology (Berl). 1987;92:42-47. monoamine oxidase-A in the human brain. J Cereb Blood Flow Metab. 2005; 42. Todd KG, McManus DJ, Baker GB. Chronic administration of the antidepres- 26:330-344. sants phenelzine, desipramine, clomipramine, or maprotiline decreases binding 27. Bergstrom M, Westerberg G, Kihlberg T, Langstrom B. Synthesis of some 11C- labelled MAO-A inhibitors and their in vivo uptake kinetics in rhesus monkey brain. to 5-hydroxytryptamine2A receptors without affecting benzodiazepine binding Nucl Med Biol. 1997;24:381-388. sites in rat brain. Cell Mol Neurobiol. 1995;15:361-370. 28. Tweedie DJ, Burke MD. Metabolism of the beta-carbolines, harmine and har- 43. Nestler EJ, Barrot M, DiLeone RJ, Eisch AJ, Gold SJ, Monteggia LM. Neurobi- mol, by liver microsomes from phenobarbitone- or 3-methylcholanthrene- ology of depression. Neuron. 2002;34:13-25. treated mice: identification and quantitation of two novel harmine metabolites. 44. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Drug Metab Dispos. 1987;15:74-81. Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994. 29. Fowler JS, Logan J, Ding YS, Franceschi D, Wang GJ, Voklow ND, Pappas N, 45. Slotkin TA, Seidler FJ, Ritchie JC. Effects of aging and glucocorticoid treatment Schlyer D, Gatley SF, Alexoff D, Felder C, Biegon A, Zhu W. Non-MAO A binding on monoamine oxidase subtypes in rat cerebral cortex: therapeutic implications. of clorgyline in white matter in human brain. J Neurochem. 2001;79: Brain Res Bull. 1998;47:345-348. 1039-1046. 46. Jernej B, Cicin-Sain L. Platelet serotonin level in rats is under genetic control. 30. Nemeroff CB, Owens MJ. Treatment of mood disorders. Nat Neurosci. 2002; Psychiatry Res. 1990;32:167-174. 5(suppl):1068-1070. 47. Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington H, McClay J, Mill 31. Fowler JS, Volkow ND, Wang GJ, Pappas N, Logan J, Shea C, Alexoff D, MacGregor J, Martin J, Braithwaite A, Poulton R. Influence of life stress on depression: RR, Schler DJ, Zezulkova I, Wolf AP. Brain monoamine oxidase A inhibition in moderation by a polymorphism in the 5-HTT gene. Science. 2003;301: cigarette smokers. Proc Natl Acad Sci U S A. 1996;93:14065-14069. 386-389. 32. First M, Spitzer R, Williams J, Gibbon M. Structured Clinical Interview for DSM-IV 48. Balciuniene J, Emilsson L, Oreland L, Pettersson U, Jazin E. Investigation of the Axis I Disorders, Patient Edition (SCID-P), version 2. New York, NY: Biometrics functional effect of monoamine oxidase polymorphisms in human brain. Hum Research; 1995. Genet. 2002;110:1-7.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 63, NOV 2006 WWW.ARCHGENPSYCHIATRY.COM 1216