The TNF gene predicts Hippocampus Volume in Healthy

advertisement
Multimodal imaging of a tescalcin (TESC)-regulating polymorphism
(rs7294919) – specific effects on hippocampal gray matter structure
- Supplementary material -
Udo Dannlowski, MD, MA, PhD 1,2,*,#, Hans Jörgen Grabe, MD 3,4,#, Katharina
Wittfeld 4, Johannes Klaus 1, Carsten Konrad, MD 2, Dominik Grotegerd, MSc
1,
Ronny Redlich, MA 1, Thomas Suslow, PhD 1,5, Nils Opel 1, Patricia
Ohrmann, MD 1, Jochen Bauer, PhD 1, Peter Zwanzger, MD 1, Inga Laeger,
MA 1, Christa Hohoff, PhD 1, Volker Arolt, MD 1, Walter Heindel, MD 6,
Michael Deppe, PhD 7, Katharina Domschke, MD, MA, PhD 8, David Stacey,
PhD 12, Katrin Hegenscheid, MD 9, Henry Völzke, MD 10, Henriette Meyer zu
Schwabedissen, MD 11, Harald Kugel, PhD 6, and Bernhard T. Baune, MD,
PhD, MPH, FRANZCP 12
1
Department of Psychiatry, University of Münster, Germany; 2 Department of
Psychiatry, University of Marburg, Germany; 3 Department of Psychiatry,
University Medicine Greifswald, HELIOS-Hospital Stralsund, Germany; 4
German Center for Neurodegenerative Diseases (DZNE), Site Rostock/
Greifswald, Germany; 5 Department of Psychosomatic Medicine and
Psychotherapy, University of Leipzig, Germany; 6 Department of Clinical
Radiology, University of Münster, Germany; 7 Department of Neurology,
University of Münster, Germany; 8 Department of Psychiatry, University of
Würzburg, Germany; 9 Institute of Diagnostic Radiology and Neuroradiology,
University Medicine Greifswald, Germany; 10 Institute for Community Medicine,
University Medicine Greifswald, Germany, 11 Biopharmacy, Department
Pharmaceutical Sciences, University of Basel, Switzerland; 12 Discipline of
Psychiatry, School of Medicine, University of Adelaide: North Terrace,
Adelaide, Australia
Supplementary methods
Genotyping. In the Münster sample, genotyping of 1 TESC, 10 RELN,
3 BDNF, 1 TNF-α, 4 IL6, and 5 FKBP5 SNPs was carried out following
published protocols applying the multiplex genotyping assay iPLEX™ for use
with the MassARRAY platform 1, yielding genotyping completion rates of
98.2%, 97.2%, 97.4%, 99.6%, 99.1%, and 97.8% for the above genes,
respectively. 5-HTTLPR (L/S) and rs25531 (A/G) genotypes were ascertained
by PCR-RFLP analysis. Briefly, PCR primers were forward (5’ – GGC GTT
GCC GCT CTG AAT GC – 3’) and reverse (5’ – GAG GGA CTG AGC TGG
ACA ACC AC – 3’). PCR products were digested using MspI enzyme from
New England Biolabs (NEB) (Genesearch, Adelaide, Australia), with expected
fragment lengths of 297bp for the S allele, 340bp for the LA allele, and 170bp
for the LG allele 2. The completion rate of 5-HTTLPR/rs25531 genotyping was
99.4%. All genotyping was performed by investigators blinded for the study.
In the SHIP-TREND sample, genotyping of rs7294919 was performed
using the pre-developed TaqMan® SNP Genotyping Assay C__43496979_10
(Life technologies, Applied Biosystems, Darmstadt, Germany). In detail,
reactions were carried out in a 5µl volume containing 1µl genomic DNA,
0.25µl Primer/Probe-Mix, 2.5µl Genotyping Master Mix and 1.25µl water.
Fluorescence was assessed for using the Fast Real-Time PCR system 7900
HT (Applied Biosystems) and the Sequence Detection Software SDS 2.3. All
failing samples were repeated at least twice.
VBM methods. Images were bias-corrected, tissue classified, and
normalized to MNI-space using linear (12-parameter affine) and non-linear
transformations, within a unified model
3
including high-dimensional DARTEL-
normalization. Gray matter segments were modulated only by the non-linear
components in order to preserve actual gray matter values locally (modulated
GM volumes). Using these procedures, no further correction for total brain
volume is required anymore.
Homogeneity of gray matter images was checked using the covariance
structure of each image with all other images, as implemented in the check
data quality function. The modulated gray matter images were smoothed with
a Gaussian kernel of 8 mm FWHW.
Gene x gene interactions. Gene-gene interactions of rs7294919 were
explored
with
genes
known
to
have
neurodevelopmental
(Reelin),
neurotrophic (BDNF), stress (HPA-axis; FKBP5) and neuroinflammatory (IL-6,
TNF-α) relevance as well as 5-HTTLPR/rs25531 which has repeatedly been
associated with hippocampal volumes 4. These additional genotypes were
only available for the Münster sample. For the analysis of epistasis, we have
selected a strategy focusing on the hippocampal area most robustly showing
effects of TESC rs7294919 (the overlapping area from both samples).
Therefore, mean gray matter values of the right and left cluster were extracted
for all subjects. On each polymorphism investigated for epistasis (n=25, see
supplementary table 1 for a complete list), analysis of covariance (ANCOVA)
was conducted for each cluster (right and left). rs7294919 genotype (TT vs.
CT/TT) and the SNP in question were entered as separate factors. Each SNP
was grouped based on allele frequency to achieve a minimum cell size of
N=20 for the interaction analysis in order to ensure robustness of the results.
Age and gender were added as covariates. In addition to all main effects, the
interaction term of rs7294919 and each other polymorphism was modelled.
The results were Bonferroni-corrected for the number of investigated SNPs
(pcorrected=.05/25=.002).
Gene x environment interaction. The same two hippocampal clusters
as in the gene x gene interaction analyses were used. Gray matter values
were extracted for both samples and G x E was tested within each sample
separately. Again, rs7294919 genotype (TT vs. CT/TT) was entered as factor
into one ANCOVA model for each sample. CTQ-scores, age, and gender
were added as covariates. Besides all main effects, the interaction term
between CTQ-scores and rs7294919 was modelled.
fMRI task. The task consisted of 4 blocks of a face processing task
alternating with 5 blocks of a sensorimotor control task. During the face
processing task, participants viewed six trios of faces (all three expressing
either anger or fear). For each trio, subjects selected 1 of 2 faces (bottom)
that was identical to a target face (top). During the sensorimotor control blocks,
participants viewed six trios of geometric shapes (circles and ellipses) and
selected 1 of 2 shapes (bottom) that were identical to a target shape (top). In
the face-processing blocks, each of the 6 face trios was presented for 4s with
a variable inter-stimulus interval of 2s to 6s (mean, 4s), for a total block length
of 48s. In the sensorimotor control blocks, each of the 6 shape trios was
presented for 4s with a fixed inter-stimulus interval of 2s, for a total block
length of 36s. Total task time was 390s. Participant performance (accuracy
and reaction time) was recorded.
T2* weighted functional data were acquired at the same scanner as the
structural images, using a single shot echoplanar sequence with parameters
selected to minimize distortion in the region of central interest, while retaining
adequate signal to noise ratio (S/N) and T2* sensitivity. Volumes consisting of
34 slices were acquired (matrix 64×64, resolution 3.6mm × 3.6mm × 3.6mm;
TR=2.1s, TE=30ms, FA=90°). The slices were tilted 25° from the AC/PC line
in order to minimize drop out artefacts in the orbitofrontal and mediotemporal
region. Functional imaging data were realigned and unwarped, spatially
normalized to standard MNI space (Montreal Neurological Institute) and
smoothed (Gaussian kernel, 6mm FWHM) using Statistical Parametric
Mapping (SPM8, http://www.fil.ion.ucl.ac.uk/spm).
For each participant, one contrast image was generated in each
individual fixed-effects 1st-level analysis comparing activation in response to
faces with the shapes baseline.
Functional connectivity. We calculated functional connectivity in the
fMRI-sample using the hippocampus clusters most robustly affected by
TESC-genotype (the same overlapping area from both sites used for
morphometric GxG and GxE analyses) as one (bilateral) seed region. As in
previous studies of our group
5
the time series of the seed volume was
extracted for each subject separately and these were entered for each subject
into a new 1st-level model predicting brain-wide activity by the extracted
hippocampus signal, yielding connectivity maps of the hippocampal region of
interest. The experimental conditions were also modelled as nuisance
regressors to control for co-activation by the task.
As in the standard fMRI analyses, the resulting functional connectivity
maps were entered into a 2nd-level group analysis, again using a t-test
contrasting TT homozygotes vs. C-carrier with age and gender as nuisance
regressors.
Supplementary Results
Gene x gene interactions. Four SNPs reached nominal or marginal
significance, three in the Reelin gene and one in the IL6 gene (Please see
supplementary table 1). However, just one SNP of the Reelin gene
(rs2299403)
reached
the
corrected
significance
level
(right
cluster;
F(1,487)=12.72, p=.0004). The main effect of TESC rs7294919 was only
found in the group of rs2299403 GG homozygotes (N=346) and not in carriers
of one or two T-alleles (N=147). See supplementary figure 1.
Gene x environment interaction. In both samples, there was no
evidence for an interaction of rs7294919 and childhood maltreatment (CTQscores), Münster sample: right hippocampus, F(1,302)=1.77, p=.18; left
hippocampus: F(1,302)=.71, p=.40; SHIP-TREND: right hippocampus,
F(1,650)=1.01, p=.32; left hippocampus: F(1,650)=2.68, p=.1.
Functional connectivity. While across all subjects, strong connectivity
patterns with prefrontal and occipital / inferior temporal areas emerged, no
significant genotype effects could be discerned at a corrected level of
significance. Using the same lenient threshold as used for the other fMRI and
DTI data (p<0.005, k=30), only one cluster survived showing any genotype
effects on functional connectivity, mapping to the left middle and inferior
occipital gyrus (x=-32, y=-90, z=0, Z=3.67, k=76, p=.0001). TT homozygotes
showed higher functional connectivity compared to C-allele carriers. However,
this cluster would not survive any kind of correction (e.g., pFWE-corrected=0.689).
Supplementary Table 1
Selection of genes/ SNPs, genotype distribution, and p-values of the
interaction respective terms with TESC rs7294919 genotype.
SNP
Groups
Grouping
Min. cell
Results
size
Interaction 2
5HTTLPR
171LL, 232LS, 99SS
LL vs. S
32
None (p>.5)
5HTTLPR/ rs255311
132LL, 241LS, 129SS
3 Groups
20
None (p>.5)
RELN_rs12705136
48AA, 198GA, 255GG
A vs. GG
39
None (p>.5)
RELN_rs2249372
56AA, 194AG, 251GG
A vs. GG
42
None (p>.36)
RELN_rs2299403
13TT,134GT, 346GG
T vs. GG
30
p=.0004R,
p=.003L
RELN_rs2528856
48CC, 209CT, 242TT
C vs. TT
40
None (p>.26)
RELN_rs2711844
31AA, 156 CA, 314CC
A vs. CC
33
p=.056R
RELN_rs362691
5CC,109CG,389GG
C vs. GG
20
None (p>.14)
RELN_rs39367
79TT,227CT,194CC
T vs. CC
34
p=.014L
RELN_rs4460306
121AA, 251CA,131CC
C vs. AA
25
None (p>.5)
RELN_rs4621738
63GG, 198AG, 201AA
G vs. AA
35
None (p>.28)
RELN_rs528528
114TT,224CT,149CC
3 Groups
25
None (p>.36)
BDNF_rs6265
12AA,151GA,318 GG
GG vs. A
29
None (p>.5)
BDNF_rs7103411
15CC,169CT, 306TT
TT vs. C
35
None (p>.5)
BDNF_rs7124442
52CC,214CT,226TT
TT vs. C
33
None (p>.43)
TNF-α_rs1800629
15AA, 154GA,333GG
A vs. GG
21
None (p>.5)
IL6_rs1800795
76CC, 253GC,174GG
GG vs. C
37
None (p>.29)
IL6_rs2069833
77CC, 250TC, 175TT
C vs TT
37
None (p>.38)
IL6_rs2069840
69GG,226GC,207CC
G vs CC
33
None (p>.26)
IL6_rs7801617
5A,89AG,406GG
GG vs. A
16
p=.013R
FKBP5_rs1360780
40TT, 260CC, 200TC
T vs CC
43
None (p>.5)
FKBP5_rs3800373
35GG, 195GT, 272TT
TT vs G
44
None (p>.28)
FKBP5_rs4713916
40AA,195AG, 265GG
GG vs. A
43
None (p>.5)
FKBP5_rs9296158
39AA, 203AG, 259GG
GG vs. A
43
None (p>.5)
FKBP5_rs9470080
55TT,208CT, 239CC
CC vs T
38
None (p>.14)
Bold: nominal or marginal significance.
1
The tri-allelic model was used.
Groups were built as carriers of 0,1, or 2 LALA alleles.
cluster; L, left hippocampus cluster
2
R, right hippocampus
Supplementary Figure 1
Interaction of TESC (rs7294919) and RELN (rs2299403) genotype on right
hippocampal gray matter volume. Error bar, S.E.M.
References
1.
Oeth P, Beaulieu M, Park C, Kosman D. iPLEXTM Assay: Increased
Plexing Efficiency and Flexibility for MassARRAY System Through
Single Base Primer Extension with Mass-Modified Terminators
[Internet]. 2007;Available from:
http://www.agrf.org.au/docstore/snp/iPlex.pdf
2.
Stacey D, Cohen-Woods S, Toben C, Arolt V, Dannlowski U, Baune BT.
Evidence of increased risk for major depressive disorder in individuals
homozygous for the high-expressing 5-HTTLPR/rs25531 (L(A)) allele of
the serotonin transporter promoter. Psychiatr Genet [Internet] 2013
[cited 2014]; 23: 222–223. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/23969989
3.
Ashburner J, Friston KJ. Unified segmentation. Neuroimage [Internet]
2005 [cited 2011]; 26: 839–851. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/15955494
4.
MacQueen GM, Frodl T. The hippocampus in major depression:
evidence for the convergence of the bench and bedside in psychiatric
research? Mol Psychiatry [Internet] 2010 [cited 2010]; 16: 252–264.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/20661246
5.
Dannlowski U, Ohrmann P, Konrad C, Domschke K, Bauer J, Kugel H
et al. Reduced amygdala-prefrontal coupling in major depression:
association with MAOA genotype and illness severity. Int J
Neuropsychopharmacol 2009; 12: 11–22.
Download