Molecular Psychiatry (2008) 13, 1093–1101

& 2008 Nature Publishing Group

All rights reserved 1359-4184/08 $30.00

www.nature.com/mp

ORIGINAL ARTICLE

Reduced gray matter brain volumes are associated

with variants of the serotonin transporter gene in

major depression

T Frodl1, N Koutsouleris1, R Bottlender1, C Born2, M Ja¨ger1, M Mo¨rgenthaler1, J Scheuerecker1, P Zill1, T Baghai1, C Schu¨le1, R Rupprecht1, B Bondy1, M Reiser2, H-J Mo¨ller1 and EM Meisenzahl1

1Department of Psychiatry and Psychotherapy, Ludwig-Maximilians-University of Munich, Munich, Germany and 2Department of Radiology, Ludwig-Maximilians-University of Munich, Munich, Germany The serotonergic system is involved in the pathophysiology of major depression as well as in the early central nervous system development and adult neuroplasticity. The aim of the study was to examine in 77 patients with major depression and 77 healthy controls the association between the triallelic polymorphism in the promoter region of the serotonin transporter gene (5-HTTLPR) and gray matter (GM) brain volumes measured with 1.5 T magnetic resonance imaging. Voxel-based morphometry were estimated on magnetic resonance images and genotyping was performed. We found that healthy controls have a strong association between the 5-HTTLPR and GM volumes of the dorsolateral prefrontal cortex, left anterior gyrus cinguli, left amygdala as well as right hippocampus, whereas there is no such association in patients with major depression. Healthy subjects carrying the S- or LG-allele have smaller GM volumes than those with the LA-allele, indicating that 5-HTTLPR contributes to the development of brain structures. Patients with depression show reduced GM volumes, particularly when they are homozygous for the LA-allele, suggesting that these patients are more vulnerable for morphological changes during depressive episodes.

Molecular Psychiatry (2008) 13, 1093–1101; doi:10.1038/mp.2008.62; published online 1 July 2008

Keywords: major depression; VBM; serotonin; imaging genetics; MRI; morphometry Introduction

hippocampus and the frontal cortex,4 has neurotrophic effects5 and increases hippocampal neurogenesis.6 More-Dysfunction of neuronal plasticity could contribute to

over, serotonergic function is modulated by BDNF. 7

the pathophysiology of mood disorders.1 The under-This hypothesis is also supported by preclinical

standing of these neuroplastic processes may offer the

studies demonstrating that stress and depression lead

possibility to define new pathology-related markers.

to reductions in total hippocampal volume, and

Such markers might be helpful to improve the

atrophy and loss of neurons in the adult hippocam-

understanding of the neurobiology of major depres-

pus. 8,9 Many in vivo neuroimaging studies have sion. Moreover, they might offer the opportunity to

detected reduced hippocampal volumes in elderly

obtain surrogate markers and to find new antidepres-

and also in younger patients with major depression.

sants, as well as new substances, which may improve

The results for other brain regions are rather incon-

the clinical outcome.

sistent, perhaps because of problems with the exact

The neuroplasticity hypothesis of major depression

manual measurement of anatomical difficult brain

is supported by studies demonstrating that serotoner-

regions: enlarged amygdala volumes, as well as

gic signaling is an important regulator of early central

reduced volumes of the anterior cingulum, prefrontal

nervous system development2 and of adult neurogen-cortex and basal ganglia, have been reported in only

esis.3 This is supported by findings of reciprocal some of the studies with regions of interest analysis. 10

regulations between the serotonergic system and

Voxel-based morphometry (VBM) is extremely useful

brain-derived neurotrophic factor (BDNF). Anti-

because brain regions whose boundaries are difficult

depressant treatment, which increases serotonergic

to define can be analyzed. VBM has detected smaller

neurotransmission, for example, selective serotonin

volumes of the medial frontal lobes in patients with

reuptake inhibitors, and elevates BDNF in the

subthreshold depression,11 as well as of the right hippocampus and the middle frontal gyrus in patients

Correspondence:

Dr

T

Frodl,

Department

of

Psychiatry

with major depression. 12

and Psychotherapy Ludwig-Maximilians-University of Munich,

Neuroimaging genetics offer a good possibility to

Nussbaumstr. 7, Munich 80336, Germany.

combine the morphological and neurochemical as-

E-mail: [e-mail]

pects, for example, structural and serotonergic altera-

Received 19 December 2007; revised 20 March 2008; accepted 12

May 2008; published online 1 July 2008

tions. Substantial relative decreases in hippocampal





Brain volumes and serotonin transporter in depression

T Frodl et al

1094

volume have been found in Met-carriers of the BDNF

diagnoses based on DSM-IV criteria and on the

polymorphism, in either healthy controls13,14 or structured clinical interview for DSM-IV (SCID)

patients with major depression. 15 Furthermore, the were determined by a consensus of at least two

L-allele of the 5-HTTLPR, which shows increased

psychiatrists. Clinical variables were documented

5-HTT expression

and

increased

5-HT-reuptake

using the 21-item Hamilton Depression Rating

in vitro and in vivo,16–18 is associated with hippocampal Scale. 22

volume reduction in major depression. 19,20

On the day of the magnetic resonance imaging,

The L-allele of the 5-HTTLPR can be subtyped into

patients were receiving the following medication: 17

La and Lg alleles, the latter of which is thought to be

patients serotonin reuptake inhibitors (7 sertraline,

similar to the S-allele in terms of reuptake efficiency

7 citalopram, 2 paroxetine, 1 fluvoxamine), 8 tricyclic

perhaps explaining some of these discrepancies.21

antidepressants (1 amitriptyline, 2 amitriptylinoxide, 4

This polymorphism can then be handled as a triallelic

doxepine, 1 trimipramine), 34 other new antidepres-

polymorphism. To avoid inconsistencies, we use the

sants (11 venlafaxine, 10 reboxetine, 13 mirtazapine),

term triallelic for this new polymorphism and

2 maprotiline and 16 patients no antidepressant.

diallelic for the older polymorphism.

For comparison, 77 healthy control subjects were

The aim of this study was to investigate the

matched with respect to age (mean age: 43.6ÂÄ…11.3

influence of this functional triallelic 5-HTTLPR/

years), gender and handedness.

rs25531 on gray matter (GM) volumes of the brain in

A structured interview was used to assess medical

patients with major depression and healthy controls

history, trauma and other exclusion criteria for all

using voxel-based morphology. We hypothesized that,

subjects. Exclusion criteria for patients and controls

compared to healthy controls, GM volume is particu-

were previous head injury with loss of consciousness,

larly reduced in the hippocampus, amygdala, anterior

cortisol medication in the medical history, previous

gyrus cinguli (ACC) and the dorsolateral (DLPFC) and

alcohol or substance abuse, neurological diseases or

dorsomedial prefrontal cortex (DMPFC) in patients

age over 65 years. Comorbidity with other mental

with major depression who are homozygous for the

illnesses and personality disorders were also ex-

L-allele of the 5-HTTLPR.

cluded, as were patients with bipolar disorders.

Healthy controls had no history of neurological or

Materials and methods

mental illnesses. No subject received an electrocon-

vulsive therapy before investigation. Handedness was

Participants

determined by the Edinburgh inventory.23

Seventy-seven inpatients with major depression

Written informed consent was obtained from all

(mean

age:

46.1ÂÄ…11.3

years,

Table

1)

were

subjects subsequent to a detailed description of

recruited from the Department of Psychiatry of the

the study. The study design was approved by

Ludwig-Maximilians-University, Munich. Psychiatric

the local ethics committee and was prepared in

Table 1

Demographic and clinical data of patients with an episode of major depression and healthy controls Patients (n = 77)

Controls (n = 77)

T-test

(meanÂÄ…s.d.)

(meanÂÄ…s.d.)

(P-value)

Age (years)

42.6ÂÄ…12.4

40.5ÂÄ…11.6

0.29

Female/male*

35/42

35/42

1.0

Handedness (R/L)*

65/12

72/5

0.07

Height (cm)

172.7ÂÄ…8.8

175.1ÂÄ…10.3

0.12

Weight (kg)

73.2ÂÄ…15.2

75.1ÂÄ…14.2

0.43

Alcohol (g day 1)

9.8ÂÄ…18.1

8.4ÂÄ…11.3

0.57

Age of onset (years)

37.1ÂÄ…12.1

Duration of illness (years)

5.4ÂÄ…8.2

Number of depressive episodes

2.05ÂÄ…2.1

HDRSâ€"baseline

22.8ÂÄ…6.1

Triallelic 5-HTTLPR*

LALA/LALG/LGLG/LAS/LGS/SS

14/18/11/23/1/10

10/10/1/34/8/14

0.002

Allele frequency

LA/LG/S

0.45/0.27/0.28

0.42/0.13/0.45

0.001

Functional consequence* (LA = L, LG = S)

LL/LS/SS

14/41/22

10/44/23

0.6

Abbreviation: HDRS, Hamilton Depression Rating Scale.

Mean and s.d.(ÂÄ…) are given.

No significant differences were found between patients and controls as measured with t-test or w2 test.

*P < 0,05, only w2 test.

Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1095

accordance with the ethical standards laid down in

Hilden,

Germany)

according

to

the

supplier’s

the Declaration of Helsinki.

instructions. We used the triallelic polymorphism

SNP-rs25531, because the L-allele can be subtyped

Magnetic resonance imaging procedures

into La and Lg alleles, the latter of which is thought to

be similar to the S-allele in terms of reuptake

Data acquisition.

Magnetic resonance images were

efficiency. 27

obtained with a Magnetom Vision scanner (Siemens,

Genotyping was carried out by applying the PCR

Erlangen, Germany) operating at 1.5 T. All subjects

amplification in a final volume of 25 ml consisting of

were scanned with a T1-weighted 3D-MPRAGE

50-ng DNA, l mmol l 1 of each primer, 200-mM deoxy-

sequence

(repetition

time,

11.6 ms;

echo

time,

nucleotide triphosphate, 100-mM 7-deaza-guanosine

4.9 ms; total acquisition time, 9 min; number of

triphosphate,

5%

dimethyl

sulfoxide,

10-mM

acquisitions, 1; field of view, 230 mm; matrix,

Tris-hydrochloride

(pH

8.3),

50-mM

potassium

512 512; section thickness, 1.5 mm) yielding 126

chloride, 1.5-mM magnesium chloride and 2.5 U of

contiguous axial slices with a defined voxel size of

DNA polymerase (AmpliTaq Gold; PerkinElmer,

0.45 0.45 1.5 mm. After manually reorienting and

Langen, Germany). The PCR products were separated

centering the images on the anterior commissure,

on a 3% agarose gel (FMC NuSieve 3:1; Biozym

data preprocessing were performed with the SPM5

Diagnostic

GmbH,

Oldendorf,

Germany)

and

software package (Wellcome Department of Cognitive

visualized by ethidium bromide staining.

Neurology, London, UK) running under MATLAB 6.5

(The MathWorks, Natick, MA, USA).

Statistical analysis

An analysis of covariance was designed to investigate

VBM preprocessing

focal GM volume (GMV) differences between the

This study employed the VBM5 toolbox, which

patients with major depression and healthy controls

utilizes and extends the new unified segmentation

as well as between genotypes. Age and gender were

approach implemented in SPM5.24 Unified segmenta-entered as covariates of no interest in the statistical

tion provides a generative model of VBM preproces-

design. First, GMV differences (increases/decreases)

sing that integrates tissue classification, image

between patients with major depression and healthy

registration and magnetic resonance imaging inho-

controls were assessed at the whole-brain level using

mogeneous bias correction. Thus, the model avoids

T contrasts (P < 0.05, family wise error correction

the â€Åšcircularity problem’ of the optimized VBM

(FWE) corrected). Then, the overall genotype effects

procedure, as the initial image registration does not

were calculated again at the whole-brain level using T

require an initial tissue segmentation and vice versa. 25

contrasts (P < 0.05, FWE corrected).

The VBM5 toolbox extends the unified segmentation

Additionally, VBM analysis was performed to

model, as it increases the quality of segmentation by

examine the interactions between diagnosis and

applying a Hidden Markov Field model on the

genotype, with additional masking for higher GMV

segmented tissue maps. 26 The Hidden Markov Field in the S-allele within the patient group thresholded at

algorithm provides spatial constraints based on

P < 0.05, uncorrected, as the exclusive mask.

neighboring voxel intensities within a 3 3 3 voxel

The VBM results were also analyzed to test

cube. It removes isolated voxels, which are unlikely

differences between patients and controls within

to be a member of a certain tissue class and also closes

each genotype and to test differences between

holes in a cluster of connected voxels of a certain

genotypes within the patients and within the controls

class, resulting in a higher signal-to-noise ratio of the

only. Additionally, small-volume-corrected VBM ana-

final tissue probability maps. The VBM5 toolbox

lyses were performed for the following regions of

offers the possibility to write the estimated tissue

interest: hippocampus, amygdala, ACC, DMPFC and

probability maps without making use of the respec-

DLPFC at P < 0.01 for comparison of cross-sectional

tive ICBM tissue priors (International Consortium for

GMV differences between patients and controls. For

Brain Mapping) from SPM5.

these five regions of interest patient subgroups,

This study used this option, as it improved the

analyses were performed at P < 0.01, small volume

delineation of the subcortical structures and the sulci

corrected: cross-sectional GMV differences between

in the final tissue maps. The final tissue maps of gray

the patients compared to controls, between patients

matter (GM), white matter and cerebro-spinal fluid

with each genotype and controls with the same

were modulated with the deformation fields obtained

genotype, as well as in the patients and control group

by normalization to standard space to analyze volume

between each genotype. Coordinates of peak signifi-

differences between study populations. Finally, the

cant voxels were assigned to anatomical regions by

modulated GM partitions were smoothed with a

means of automated anatomical labeling.28

12 mm FWHM Gaussian kernel and used for statis-

tical analysis.

Results

Genetics

Age, gender, handedness, height, weight and alcohol

DNA was extracted from a 5 ml blood sample using

consumption were similar in patients and controls

the QIAamp Blood Isolation Kit (QIAGEN GmbH,

(Table 1). The 5-HTTLPR distributions for the patients Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1096

and controls were in Hardy–Weinberg equilibrium.

observed between heterozygous patients and patients

For each genotype, age, weight and height did not

homozygous for either the S-(LG or S) or LA-allele

differ between patients and controls. There was

(Table 4).

no

significant

difference

in

illness

duration

(F(2/74) = 0.6;

P = 0.55)

or

depression

severity

Genotype effects within healthy controls

(Hamilton Depression Rating Scale, F(2/74) = 1.5;

Healthy controls homozygous for the LA-allele had

P = 2.3) between patients with a specific genotype.

significantly larger GMVs in the ACC, DLPFC right

Interestingly, with respect to the allele frequencies

pronounced, right DMPFC, left amygdala and left

of the 5-HTTLPR-rs25531 the L

hippocampus than those homozygous for the S-allele

G-allele

was more

frequent in our patients sample, whereas the S-allele

(Table 4).

was less frequent compared to the healthy controls

(X2 = 13.5; P = 0.001). As the LG-allele is considered

Discussion

to be functionally comparable to the S-allele,

functional triallelic LL/LS/SS genotypes can be used.

The main finding from this study is that healthy

Frequencies of these were not significantly different

controls

show

a

strong

association

between

between patients and controls (Table 1).

5-HTTLPR

polymorphism

and

GMVs

in

all

Using w2 test, there was no significant difference

investigated brain regions, with the smallest volumes

between genotypes with respect to different medica-

being found in healthy subjects homozygous for the

tion (no medication, selective serotonin reuptake

S-allele (LGLG, LGS, SS), whereas in patients with

inhibitor, tricyclic, newer antidepressants).

major depression this association is only seen in

In the overall VBM analyses, patients with major

subregions of the hippocampus. The explanation for

depression had significantly smaller GMVs in the left

this discrepancy may be that patients with depression

DLPFC, bilaterally in the supplementary motor area,

who are homozygous for the LA-allele have prominent

in the thalamus and the precentral area, and showed

GM reductions compared to controls with the same

trends bilaterally in the DMPFC, the right orbitofron-

genotype. Patients have the largest volume reductions

tal cortex as well as the left insular cortex compared

compared to controls in cortical brain regions such as

to healthy controls (Figure 1).

the DLPFC, DMPFC, ACC as well as in the amygdala

Small volume correction for regions of interest

and anterior hippocampus when they are homo-

revealed that patients had significantly reduced

zygous for the LA-allele. Patients homozygous for the

GMVs bilaterally in the ACC, DLPFC and in

S-allele (LGLG, LGS, SS) or heterozygous patients

the DMPFC compared to controls (Table 2). The (LAS, LALG) only differ in small regions from controls

hippocampus and amygdala GMVs were not smaller

with the same genotype. For example, only the

in patients.

middle hippocampal volume was found to be reduced

in patients carrying the S-allele compared to controls

Interaction between genotype and diagnosis

with the same genotype.

The interaction between diagnosis and genotype was

This finding is very interesting because the normal

significant in the DLPFC right pronounced, left ACC,

influence of 5-HTTLPR on the brain may be dimin-

left amygdala as well as right hippocampus: the

ished in patients with major depression, indicating

difference between controls homozygous for the

that other factors may have a greater influence on

LA-allele and those homozygous for the S-allele

these processes, for example psychosocial factors or

(LGLG, LGS, SS) was larger than the difference between

early life events. 29

patients homozygous for the LA-allele and those

Interestingly, patients had higher LG-allele frequen-

homozygous for the S-allele (Table 3).

cies and less S-alleles (LGLG, LGS, SS) compared to

healthy controls. Unfortunately, our sample is too

Effects of diagnosis within genotypes

small to conclude that the LG-allele may be a risk

Patients homozygous for the LA-allele had signifi-

factor for major depression or to analyze the different

cantly smaller GMVs in the hippocampus, amygdala,

influences between LG-allele and S-allele (LGLG, LGS,

ACC, DLPFC and DMPFC than healthy controls with

SS) on the brains of patients with major depression

the same genotype. Heterozygous patients differed

and healthy controls.

significantly from controls with the same genotype

These data suggest that the main GM alterations are

with respect to the DLPFC, and patients homozygous

seen in patients homozygous for the functional

for the S-allele (LGLG, LGS, SS) had significantly

LA-allele of the 5-HTTLPR-rs25331. In our earlier

smaller subregions within the hippocampus DLPFC

study, the homozygous L-allele of the 5-HTTLPR was

and DMPFC compared to controls with the same

found to be associated with reduced hippocampal

genotype (Figure 2).

volumes in patients with major depression and not in

controls.19 Moreover, patients with a late-onset Genotype effects within the depressive group

geriatric depression who were homozygous for the

Patients homozygous for the LA-allele had signifi-

L-allele of the 5-HTTLPR had smaller hippocampal

cantly

reduced

bilateral

hippocampal

volumes

volumes than other groups. 30

than those patients being homozygous for S-allele

One possible explanation may be that the higher

(LGLG, LGS, SS). No significant differences were

reuptake of serotonin in subjects homozygous for the

Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1097

Figure 1

Overall brain gray matter (GM) volume reduction in patients compared to controls. Regions where GM density is reduced in patients with major depression compared to healthy controls. P-values are corrected for multiple comparisons (FDR, P < 0.01). Numbers below slices represent the vertical distance in millimeter to the anterior commissure.

Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1098

Table 2

Gray matter volume reduction in patients with major depression compared to healthy controls with SVC for ACC, hippocampus, amygdala, DLPFC, DMPFC (P < 0.01, uncorrected)

k

k (%)

FWE

FDR

T

P

x

y

z

Left ACC

219

1.96

0.057

0.028

3.21

0.001

2

36

28

Right ACC

1200

11.42

0.031

0.028

3.44

< 0.001

4

42

28

Left DLPFC

12667

32.56

0.006

0.004

4.34

< 0.001

25

5

61

Right DLPFC

13570

33.23

0.038

0.004

3.78

< 0.001

43

4

51

Left DMPFC

6302

26.33

0.004

0.002

4.26

< 0.001

0

32

46

Right DMPFC

2886

16.9

0.004

0.002

4.24

< 0.001

1

32

46

Left amygdala

9

0.51

0.017

0.126

3.27

0.001

16

2

17

Right amygdala

NS

Left hippocampus

7

0.09

0.041

0.166

3.33

0.001

15

2

16

Right hippocampus

1

0.01

0.053

0.166

3.24

0.001

26

40

3

Abbreviations: ACC, anterior gyrus cinguli; DLPFC, dorsolateral prefrontal cortex; DMPFC, dorsomedial prefrontal cortex; FDR, false discovery rate; FWE, family wise error correction; k, number of suprathreshold voxels in anatomical region; k (%), percentage of suprathreshold voxels in anatomical region, T-test; NS, not significant; SVC, small volume correction.

Table 3

Difference in GM volume between LA/LA and S/S (LGLG, LGS, SS) is greater for controls compared to the difference between LA/LA and S/S (LGLG, LGS, SS) for patients with SVC for ACC, hippocampus, amygdala, DLPFC, DMPFC (P < 0.01, uncorrected)

k

k (%)

FWE

FDR

T

P

x

y

z

Left ACC

49

0.44

0.153

0.347

2.79

0.003

12

31

26

Right ACC

NS

Left DLPFC

11

0.03

0.591

0.266

2.54

0.006

34

1

50

Right DLPFC

926

2.27

0.175

0.266

3.21

0.001

29

35

29

Left DMPFC

NS

Right DMPFC

NS

Left amygdala

21

1.19

0.117

0.128

2.47

0.007

29

0

26

Right amygdala

NS

Left hippocampus

NS

Right hippocampus

17

0.22

0.269

0.253

2.52

0.006

35

12

27

Abbreviations: ACC, anterior gyrus cinguli; DLPFC, dorsolateral prefrontal cortex; DMPFC, dorsomedial prefrontal cortex; FDR, false discovery rate; FWE, family wise error correction; GM, gray matter; k, number of suprathreshold voxels in anatomical region, k (%), percentage of suprathreshold voxels in anatomical region, T-test; NS, not significant; SVC, small volume correction.

LA-allele is an additional factor that aggravates stress-

during major depression. Importantly, this may not be

induced neurotoxic effects during depression. In

a common phenomenon because healthy control

addition to its role as a neurotransmitter, serotonin

subjects with the S-allele (LGLG, LGS, SS) have the

acts as a trophic factor modulating developmental

smallest hippocampal volumes. Therefore, the pre-

processes such as neuronal division, differentiation,

sence of major depression or other factors that play a

migration, synaptogenesis31,32 and adult neurogen-part in the pathophysiology of major depression

esis. 33 Moreover, several studies found interactions seems to be necessary for the changes seen in patients

between the serotonergic system and BDNF, as the

homozygous for the LA-allele.

5-HTT function is modulated by BDNF, 34 which, in Another possibility for the association of brain

turn, was found to be elevated in the hippocampus

volume and the 5-HTTLPR polymorphism comes

and the frontal cortex after antidepressant treat-

from the studies reporting that the L-allele of the

ment. 35 Upregulation of the cAMP response ele-5-HTTLPR may be related to Alzheimer disease. 37–39

ment-binding

protein

due

to

effective

therapy

The larger reduction in GMVs in patients homozy-

increases the expression of BDNF, which seems to

gous for the L-allele may then be explained by a larger

have neurotrophic effects.36

susceptibility to neurodegenerative changes. How-

One might therefore expect that the high-activity

ever, this seems to be unlikely, although not exclu-

LA-allele with its increased number of 5-HTT trans-

sive, because our patients were young.

porter proteins, concomitant decrease in serotonin

The effect of stress on the brain is relevant for the

levels and reduced effects on neuroplastic processes

discussion of these findings. It is well known from

would be more likely to cause structural changes

clinical and experimental studies that stress is one of

Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1099

Figure 2

Small-volume-corrected (SVC) gray matter volume (GMV) reduction in patients compared to controls for each genotype pair. Regions where GMV is reduced in patients with the LALA genotype compared to healthy controls with the LALA genotype (red), LAS, LALG genotypes (green), (LGLG, LGS, SS genotypes (yellow), respectively with SVC (P < 0.01, uncorrected). Numbers below slices represent the vertical distance in millimeter to the anterior commissure. Significantly reduced GMVs in patients compared to controls are detected bilaterally in the hippocampus, bilaterally in the anterior gyrus cinguli (ACC), in the left amygdala and the right dorsomedial prefrontal cortex (DMPFC).

Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1100

Table 4

Difference in GM volume between patients with LA/LA and those with S/S genotype (LGLG, LGS, SS) as well as difference in GM volume in healthy controls with LA/LA compared to those with S/S genotype (LGLG, LGS, SS) with SVC for ACC, hippocampus, amygdala, DLPFC, DMPFC (P < 0.01, uncorrected)

k

k (%)

FWE

FDR

T

P

x

y

z

Major depression

Left hippocampus

185

2.48

0.02

0.097

3.59

< 0.001

26

38

4

Right hippocampus

281

3.71

0.012

0.097

3.76

< 0.001

30

38

5

Healthy controls

Left ACC

40

0.36

0.24

0.144

2.55

0.006

11

29

25

Right ACC

94

0.89

0.218

0.144

2.61

0.005

11

36

28

Left DLPFC

20

0.05

0.62

0.248

2.5

0.007

27

7

62

Right DLPFC

867

2.12

0.063

0.248

3.6

< 0.001

28

34

29

Left DMPFC

NS

Right DMPFC

54

0.32

0.122

0.16

3.08

0.001

12

22

41

Left amygdala

35

1.99

0.078

0.114

2.66

0.004

17

6

18

Right amygdala

NS

Left hippocampus

65

0.87

0.212

0.22

2.64

0.005

17

6

20

Right hippocampus

NS

Abbreviations: ACC, anterior gyrus cinguli; DLPFC, dorsolateral prefrontal cortex; DMPFC, dorsomedial prefrontal cortex; FDR, false discovery rate; FWE, family wise error correction; GM, gray matter; k, number of suprathreshold voxels in anatomical region; k (%), percentage of suprathreshold voxels in anatomical region, T-test; NS, not significant; SVC, small volume correction.

the major factors that contribute to the development

certain 5-HTTLPR genotype did not differ with

of depression. Several studies have found that the

respect to age, gender, illness duration, age of onset,

S-allele of 5-HTTLPR increases the vulnerability to

medication and origin, it does not seem likely that the

depression only in the presence of significant adverse

population stratification had any negative effects.

life events. Relative to homozygous individuals for

Another important factor is medication. There was no

the L-allele, young adult carriers of the S-allele were

difference in medication between genotypes, how-

more vulnerable to major depression when exposed to

ever, we cannot exclude from our study an interaction

adverse events.40 Young adults homozygous for the S-between medication, genotype and brain volumes,

allele also had significantly greater sensitivity to

because our sample is too heterogeneous with respect

depressogenic effects of stressful life events than

to medication and too small, to analyze this effect.

carriers of the L-allele. 41 An increased risk for major Further studies are therefore necessary.

depression was detected in maltreated children

In summary, healthy subjects with the LGLG, LGS or

homozygous for the S-allele.42 Moreover, S-alleles SS genotype have smaller GMVs than those being

independently predicted greater depression severity

homozygous for the LA-allele. A hypothetical expla-

and greater severity of major depression with moder-

nation may be that the S-allele may have impact

ate-to-severe life events compared with the higher

on neurodevelopmental processes, perhaps due to

expressing L-allele. 43 These studies indicate a role of increased sensitivity to childhood and adolescent

the S-allele in predisposing to major depression.

stressful life events as suggested from genetic studies.

However, one recent study found that the high-

Patients with depression show the most reduced

activity L-allele was associated with a higher risk of

GMVs compared to controls with the same genotype

depression in the presence of adverse life events in a

when they are homozygous for the LA-allele, suggest-

large sample of 247 young adult female twins from

ing that these patients are more vulnerable for

Missouri.44

morphological or even neuroplastic changes during

Also, controversial is the neurobiological interpre-

depressive episodes.

tation. It seems puzzling that lower serotonin uptake

resulting from a genetic predisposition would be

associated with a higher risk of depression because

Acknowledgments

lowering serotonin uptake by treatment with seroto-

nin selective reuptake inhibitors relieves depression.

This study was supported by the German Federal

Therefore, more studies are necessary to explore the

Research Ministry within the promotion â€ÅšGerman

role of the serotonin transporter in the neurobiology

Research Networks in Medicine’ as part of the project

of major depression.

â€ÅšGerman Research Network on Depression’ and Maike

The primary limitation of this study is its

Moergenthaler carried out her doctoral thesis within

case–control design, which is sensitive to population

this study. Moreover, we thank Mrs Jaquie Klesing for

stratification. However, as cases and controls with a

native english editing.

Molecular Psychiatry





Brain volumes and serotonin transporter in depression

T Frodl et al

1101

References

22 Hamilton M. A rating scale for depression. J Neurol Neurosurg

Psychiatry 1960; 23: 56–62.

1 Duman RS. Pathophysiology of depression: the concept of

23 Oldfield RC. The assessment and analysis of handedness: the

synaptic plasticity. Eur Psychiatry 2002; 17(Suppl 3): 306–310.

Edinburgh inventory. Neuropsychologia 1971; 9: 97–113.

2 Lauder JM. Neurotransmitters as growth regulatory signals: role of

24 Ashburner J, Friston KJ. Unified segmentation. Neuroimage 2005;

receptors and second messengers. Trends Neurosci 1993; 16: 233–240.

26: 839–851.

3 Gould E. Serotonin and hippocampal neurogenesis. Neuropsycho-

25 Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ,

pharmacology 1999; 21(Suppl 2): 46S–51S.

Frackowiak RS. A voxel-based morphometric study of ageing in 465

4 Nibuya M, Morinobu S, Duman RS. Regulation of BDNF and trkB

normal adult human brains. Neuroimage 2001; 14(1 Part 1): 21–36.

mRNA in rat brain by chronic electroconvulsive seizure and

26 Cuadra MB, Cammoun L, Butz T, Cuisenaire O, Thiran JP.

antidepressant drug treatments. J Neurosci 1995; 15: 7539–7547.

Comparison and validation of tissue modelization and statistical

5 Duman RS. Pathophysiology of depression: the concept of

classification methods in T1-weighted MR brain images. IEEE

synaptic plasticity. Eur Psychiatry 2002; 17(Suppl 3): 306–310.

Trans Med Imaging 2005; 24: 1548–1565.

6 Santarelli L, Saxe M, Gross C, Surget A, Battaglia F, Dulawa S et al.

27 Hu X, Oroszi G, Chun J, Smith TL, Goldman D, Schuckit MA. An

Requirement of hippocampal neurogenesis for the behavioral

expanded evaluation of the relationship of four alleles to the level

effects of antidepressants. Science 2003; 301: 805–809.

of response to alcohol and the alcoholism risk. Alcohol Clin Exp

7 Mossner R, Daniel S, Albert D, Heils A, Okladnova O, Schmitt A et

Res 2005; 29: 8–16.

al. Serotonin transporter function is modulated by brain-derived

28 Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F,

neurotrophic factor (BDNF) but not nerve growth factor (NGF).

Etard O, Delcroix N et al. Automated anatomical labeling of

Neurochem Int 2000; 36: 197–202.

activations in SPM using a macroscopic anatomical parcellation of

8 Reagan LP, McEwen BS. Controversies surrounding glucocorti-

the MNI MRI single-subject brain. Neuroimage 2002; 15: 273–289.

coid-mediated cell death in the hippocampus. J Chem Neuroanat

29 Heim C, Nemeroff CB. The role of childhood trauma in the

1997; 13: 149–167.

neurobiology of mood and anxiety disorders: preclinical and

9 Woolley CS, Gould E, McEwen BS. Exposure to excess glucocorti-

clinical studies. Biol Psychiatry 2001; 49: 1023–1039.

coids alters dendritic morphology of adult hippocampal pyramidal

30 Taylor WD, Steffens DC, Payne ME, MacFall JR, Marchuk DA,

neurons. Brain Res 1990; 531: 225–231.

Svenson IK et al. Influence of serotonin transporter promoter

10 Campbell S, Marriott M, Nahmias C, MacQueen GM. Lower

region polymorphisms on hippocampal volumes in late-life

hippocampal volume in patients suffering from depression: a

depression. Arch Gen Psychiatry 2005; 62: 537–544.

meta-analysis. Am J Psychiatry 2004; 161: 598–607.

31 Gaspar P, Cases O, Maroteaux L. The developmental role of

11 Taki Y, Kinomura S, Awata S, Inoue K, Sato K, Ito H et al. Male

serotonin: news from mouse molecular genetics. Nat Rev Neurosci

elderly subthreshold depression patients have smaller volume of

2003; 4: 1002–1012.

medial part of prefrontal cortex and precentral gyrus compared

32 Sikich L, Hickok JM, Todd RD. 5-HT1A receptors control neurite

with age-matched normal subjects: a voxel-based morphometry.

branching during development. Brain Res Dev Brain Res 1990; 56:

J Affect Disord 2005; 88: 313–320.

269–274.

12 Bell-McGinty S, Butters MA, Meltzer CC, Greer PJ, Reynolds III CF,

33 Gould E. Serotonin and hippocampal neurogenesis. Neuropsycho-

Becker

JT.

Brain

morphometric

abnormalities

in

geriatric

pharmacology 1999; 21(Suppl 2): 46S–51S.

depression: long-term neurobiological effects of illness duration.

34 Mossner R, Daniel S, Albert D, Heils A, Okladnova O, Schmitt A

Am J Psychiatry 2002; 159: 1424–1427.

et al. Serotonin transporter function is modulated by brain-derived

13 Bueller JA, Aftab M, Sen S, Gomez-Hassan D, Burmeister M,

neurotrophic factor (BDNF) but not nerve growth factor (NGF).

Zubieta JK. BDNF Val66Met allele is associated with reduced

Neurochem Int 2000; 36: 197–202.

hippocampal volume in healthy subjects. Biol Psychiatry 2006; 59:

35 Nibuya M, Morinobu S, Duman RS. Regulation of BDNF and trkB

812–815.

mRNA in rat brain by chronic electroconvulsive seizure and

14 Pezawas L, Verchinski BA, Mattay VS, Callicott JH, Kolachana BS,

antidepressant drug treatments. J Neurosci 1995; 15: 7539–7547.

Straub RE et al. The brain-derived neurotrophic factor val66met

36 Duman RS. Pathophysiology of depression: the concept of

polymorphism and variation in human cortical morphology.

synaptic plasticity. Eur Psychiatry 2002; 17(Suppl 3): 306–310.

J Neurosci 2004; 24: 10099–100102.

37 Zill P, Padberg F, de JS, Hampel H, Burger K, Stubner S et al.

15 Frodl T, Schule C, Schmitt G, Born C, Baghai T, Zill P et al.

Serotonin transporter (5-HTT) gene polymorphism in psychoger-

Association of the brain-derived neurotrophic factor Val66Met

iatric patients. Neurosci Lett 2000; 284: 113–115.

polymorphism with reduced hippocampal volumes in major

38 Li T, Holmes C, Sham PC, Vallada H, Birkett J, Kirov G et al. Allelic depression. Arch Gen Psychiatry 2007; 64: 410–416.

functional variation of serotonin transporter expression is a

16 Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S et al.

susceptibility factor for late onset Alzheimer’s disease. Neurore-

Association of anxiety-related traits with a polymorphism in the

port 1997; 8: 683–686.

serotonin transporter gene regulatory region. Science 1996; 274:

39 Hu M, Retz W, Baader M, Pesold B, Adler G, Henn FA et al.

1527–1531.

Promoter polymorphism of the 5-HT transporter and Alzheimer’s

17 Bengel D, Murphy DL, Andrews AM, Wichems CH, Feltner D, Heils A

disease. Neurosci Lett 2000; 294: 63–65.

et al. Altered brain serotonin homeostasis and locomotor insensitivity 40 Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington H et

to 3, 4-methylenedioxymethamphetamine (â€ÅšEcstasy’) in serotonin

al. Influence of life stress on depression: moderation by a

transporter-deficient mice. Mol Pharmacol 1998; 53: 649–655.

polymorphism in the 5-HTT gene. Science 2003; 301: 386–389.

18 Heinz A, Jones DW, Mazzanti C, Goldman D, Ragan P, Hommer D

41 Kendler KS, Kuhn JW, Vittum J, Prescott CA, Riley B. The

et al. A relationship between serotonin transporter genotype and

interaction of stressful life events and a serotonin transporter

in vivo protein expression and alcohol neurotoxicity. Biol

polymorphism in the prediction of episodes of major depression: a

Psychiatry 2000; 47: 643–649.

replication. Arch Gen Psychiatry 2005; 62: 529–535.

19 Frodl T, Meisenzahl EM, Zill P, Baghai T, Rujescu D, Leinsinger G

42 Kaufman J, Yang BZ, Douglas-Palumberi H, Grasso D, Lipschitz D,

et al. Reduced hippocampal volumes associated with the long

Houshyar S et al. Brain-derived neurotrophic factor-5-HTTLPR

variant of the serotonin transporter polymorphism in major

gene interactions and environmental modifiers of depression in

depression. Arch Gen Psychiatry 2004; 61: 177–183.

children. Biol Psychiatry 2006; 59: 673–680.

20 Taylor WD, Steffens DC, Payne ME, MacFall JR, Marchuk DA,

43 Zalsman G, Huang YY, Oquendo MA, Burke AK, Hu XZ, Brent DA et

Svenson IK et al. Influence of serotonin transporter promoter

al. Association of a triallelic serotonin transporter gene promoter

region polymorphisms on hippocampal volumes in late-life

region (5-HTTLPR) polymorphism with stressful life events and

depression. Arch Gen Psychiatry 2005; 62: 537–544.

severity of depression. Am J Psychiatry 2006; 163: 1588–1593.

21 Hu X, Oroszi G, Chun J, Smith TL, Goldman D, Schuckit MA. An

44 Chorbov VM, Lobos EA, Todorov AA, Heath AC, Botteron KN,

expanded evaluation of the relationship of four alleles to the level

Todd RD. Relationship of 5-HTTLPR genotypes and depression

of response to alcohol and the alcoholism risk. Alcohol Clin Exp

risk in the presence of trauma in a female twin sample. Am J Med

Res 2005; 29: 8–16.

Genet B Neuropsychiatr Genet 2007; 144: 830–833.

Molecular Psychiatry





Document Outline


Reduced gray matter brain volumes are associated with variants of the serotonin transporter gene in major depression Introduction

Materials and methods Participants





Table 1 Demographic and clinical data of patients with an episode of major depression and healthy controls Magnetic resonance imaging procedures Data acquisition





VBM preprocessing

Genetics

Statistical analysis





Results Interaction between genotype and diagnosis

Effects of diagnosis within genotypes

Genotype effects within the depressive group

Genotype effects within healthy controls





Discussion

Figure 1 Overall brain gray matter (GM) volume reduction in patients compared to controls.

Table 2 Gray matter volume reduction in patients with major depression compared to healthy controls with SVC for ACC, hippocampus, amygdala, DLPFC, DMPFC (Plt0.01, uncorrected)

Table 3 Difference in GM volume between LA/LA and S/S (LGLG, LGS, SS) is greater for controls compared to the difference between LA/LA and S/S (LGLG, LGS, SS) for patients with SVC for ACC, hippocampus, amygdala, DLPFC, DMPFC (Plt0.01, uncorrected)

Figure 2 Small-volume-corrected (SVC) gray matter volume (GMV) reduction in patients compared to controls for each genotype pair.

Acknowledgments

Table 4 Difference in GM volume between patients with LA/LA and those with S/S genotype (LGLG, LGS, SS) as well as difference in GM volume in healthy controls with LA/LA compared to those with S/S genotype (LGLG, LGS, SS) with SVC for ACC, hippocampus, am

References