Upload
vuongcong
View
226
Download
0
Embed Size (px)
Citation preview
The interaction of GSK3B and FXR1 genotypes may influence the
mania and depression dimensions in mood disorders
Running title : GSK3B - FXR1 genotypes and mood disorder
Alexandre Bureaua,b,*, Jean Martin Beaulieua,c,d, Thomas Paccaleta, Yvon C. Chagnona,c and
Michel Maziadea,c
a Centre de recherche de l’Institut universitaire en santé mentale de Québec du Centre
intégré universitaire en santé et services sociaux de la Capitale-Nationale, Québec, Canada; b
Département de médecine sociale et préventive, Université Laval, and c Département de
psychiatrie et neurosciences, Université Laval, Québec, Canada
d Department of Pharmacology and toxicology, University of Toronto, Toronto, Ontario,
Canada.
* Correspondence:
Alexandre Bureau, Département de médecine sociale et préventive, Université Laval,
Pavillon Ferdinand-Vandry, 1050, avenue de la Médecine, Room 2457, Quebec City, Quebec,
G1V 0A6, Canada.
Phone : 1-418-656-2131 ext. 3342
Fax : 1-418-656-5491
Email: [email protected]
1
Abstract
Background: Previous evidence in healthy subjects suggested that functional
polymorphisms GSK3B rs12630592 and FXR1 rs496250 interact in regulating mood and
emotional processing. We attempted to replicate this interaction primarily on manic and
depressive dimensions in mood disorder patients, and secondarily on schizophrenia
patients, diagnosis itself and age of onset.
Methods: Symptom dimensions were derived from the Comprehensive Assessment of
Symptoms and History 82 items rated lifetime in acute episodes and stabilized interepisode
intervals in 384 patients from the Schizophrenia and Bipolar Disorder Eastern Quebec
Kindred Study. Linear mixed effect models of symptom dimensions included rs12630592-
rs496250 main and interaction fixed effects (obtained from TaqMan genotypes), and a
polygenic random effect. The distribution of lifetime best-estimate DSM-IV diagnosis of 855
kindred members was studied versus genotype under a polytomous logistic model.
Results: In mood disorder patients, the level of mania (in both acute and stabilized periods)
and depression in stabilized periods was positively associated with GSK3B rs12630592 T
only in FXR1 rs496250 A-allele carriers (Bonferroni-corrected interaction p=0.024, 0.052
and 0.017 respectively). The two polymorphisms explained 11% of mania variance and 5%
of interepisode depression variance. The association was observed neither in schizophrenia
patients nor with the psychotic dimension in mood disorder patients. Interaction with the
diagnosis distribution (p=0.03) was driven by the decreasing prevalence of recurrent major
depression with rs12630592 T also only in carriers of rs496250 A.
2
Limitations: Sample size was limited, but power was sufficient to detect the tested
interaction effect in this replication sample.
Conclusions: We replicate in affective patients an interaction between the FXR1 rs496250
and GSK3B rs12630592 polymorphisms in regulating mood dimensions.
Word count: 250
Key words: age of onset; bipolar disorder; epistasis, genetic; Fragile X autosomal homolog
1, Glycogen Synthase Kinase 3; lithium
3
Introduction1
Genome wide association studies have identified numerous genetic risk alleles for mental
illnesses like bipolar disorder (BD) and schizophrenia (SZ) (Hou et al., 2016; Schizophrenia
Working Group of the Psychiatric Genomics, 2014). However, individual allelic variations
explain only a very small proportion of risk and established risk indicators individually
showed weak odds ratios (ORs) to predict the development of the illness (Paccalet et al.,
2016). The investigation of functional interactions between genetic factors becomes an
imperative step to further understand individual risk and increase power to develop
translational applications in the clinic (Beaulieu, 2012; Paccalet et al., 2016). Another area
of new inquiry focuses on symptom dimensions as alternative phenotypes to diagnoses that
may also accelerate our understanding of the genetic underpinning of major psychoses
(Labbe et al., 2012).
Functional interaction between the glycogen synthase kinase 3 (GSK3B) and Fragile X
mental retardation homolog 1 (FXR1) genes is suggested by previous findings (Del'Guidice
et al., 2015). Inhibition of glycogen kinase 3 (GSK3) is a potential component of the
mechanism underlying the clinical effectiveness of mood stabilizers (Beaulieu et al., 2009;
Beaulieu et al., 2008a; Klein and Melton, 1996). GSK3 is a serine threonine kinase existing
under two isoforms respectively encoded by the GSK3A and GSK3B genes (Woodgett, 2001).
Several studies in animal models also support a central role for GSK3 activity in regulating
1 Abbreviations BD: Bipolar disorder; FXR1: Fragile X autosomal homolog 1; GEE: generalized estimating equation; GSK3B: glycogen synthase kinase 3; NAAR: non-affected adult relative; OR: Odds ratio; RMDD: recurrent major depressive disorder; SAD: Schizoaffective disorder; SNP: single nucleotide polymorphism; SZ: Schizophrenia.
4
different behavioural dimensions including mood, cognition and social processes (Beaulieu
et al., 2008b; Jope, 2011). However, how GSK3 activity can contribute to clinical symptoms
and drug responsiveness in human populations with heterogeneous genetic makeups
remains poorly understood.
The Fragile X mental retardation homolog 1 (FXR1) is an RNA binding proteins that was
identified as a substrate of GSK3. A previous study has shown that FXR1 levels are
upregulated by chronic treatment with mood stabilisers, lithium, valproic acid or
lamotrigine in rodents (Del'Guidice et al., 2015). Elevated levels of FXR1 also replicate
behavioral responses to mood stabilizers in mice.
Functional polymorphisms GSK3B rs12630592 and FXR1 rs496250 have been shown to
interact in regulating mood and emotional processing in healthy humans (Del'Guidice et al.,
2015) but no study yet has approached the question in patients with mood disorders.
Furthermore, genetic variants of FXR1 have been associated to risk of SZ in a large genome
wide association study and functional evidences support an involvement for FXR1 in
regulating behavioural dimensions in this illness (Del'Guidice et al., 2015; Hauberg et al.,
2016; Schizophrenia Working Group of the Psychiatric Genomics, 2014; Stepniak et al.,
2015).
Here, we examined in mood disorder patients the interaction between GSK3B rs12630592
and FXR1 rs496250 genotypes by focusing on the manic and depressive symptom
dimensions in an Eastern Quebec kindred sample densely affected by major psychosis
disorders (Maziade et al., 2005). As secondary objectives, we examined the same
interaction in relation to the diagnosis distribution and age of onset in the kindred sample.
5
Furthermore, given that mood stabilizers inhibit GSK3 activity and increase levels of FXR1,
we explored the relationship between GSK3B and FXR1 genotypes, medication and
symptom dimensions.
Methods
Study subjects and phenotype definition
The kindred sample consisted of 855 members of 48 multigenerational families from the
Eastern Quebec population. All subjects were Caucasian of French-Canadian ancestry. A
lifetime best-estimate DSM-IV diagnosis was stringently made using personal interview,
information from relatives and the review of lifetime in- and out-patient medical records
according to a method extensively detailed previously (Maziade et al., 1992; Roy et al.,
1997). In brief, four research psychiatrists reviewed blindly all available information across
lifetime from different sources (all medical records, family informant interviews, personal
structured interview). We considered the following major psychosis diagnoses: Mood
disorders including BD I, BD II and recurrent major depressive disorder (RMDD); a broad
definition of SZ including SZ per se, schizophreniform disorder and schizotypal personality;
and schizoaffective disorder (SAD). We formed a comparison group with the genotyped
non-affected family members satisfying the following criteria: A) absence of the above
diagnoses, B) age greater or equal to 25 years and C) not the parent of a patient, and these
subjects were referred to as non-affected adult relatives (NAARs). Age of onset was defined
as the age of the first probable episode of the disorder. If unavailable, the age of the first
definitive episode was used. The number of subjects in each diagnosis category is in Table
1.
6
Ethics
The study was explained to each family member, unrelated cases and controls, and a signed
consent was obtained, as reviewed by our University Ethics Committee.
Lifetime symptoms assessment
The lifetime presence and severity of psychotic, manic and depressive symptoms was
evaluated on a six-point rating scale (each point corresponding to an operational definition
of severity adapted for each symptom - 0: None, 1: Questionable, 2: Mild, 3: Moderate, 4:
Marked, 5: Severe) on the 82 items of the Comprehensive Assessment of Symptoms and
History (CASH) instrument (Andreasen et al., 1992). Items cover 11 dimensions: Delusion,
Hallucination, Bizarre behavior, Catatonia, Thought disorder, Alogia, Anhedonia, Apathy,
Affective blunting, Mania and Depression. Item-specific ratings were averaged for each
dimension. Catatonia was excluded as it exhibited very limited variability across subjects.
The differences from the original CASH assessment scale were: 1) the rating was based not
only on interviews, but also on an estimate of the lifetime presence (across the whole
duration of illness) and severity of each symptom that was based on all available sources
used for making the best-estimate diagnosis (i.e., the SCID personal interview, family
history interviews from several family informants, and all available medical records whose
access is made possible by the universal public health system of Québec, Canada ), and 2)
symptoms were rated separately for two time frames: lifetime acute episodes and lifetime
stabilized interepisode intervals. The reliability and details of the procedure are provided in
Maziade et al. (Maziade et al., 1995).
7
Medication
Information on lifetime medication for all patients was extracted from all available medical
records. We defined indicator variables of having ever taken three types of medications:
lithium, anti-depressors and neuroleptics. Other types of medication were taken by few
subjects and were not considered.
Genotyping
Genotyping of GSK3B rs12630592 and FXR1 rs496250 was done on a realtime PCR
(Lightcycler 480 ; Roche) using the specific 5’_ nuclease Taq Man assays C_1029492_10 for
FXR1 and C_ 1849026_20 for GSK3B (proprietary sequences Life Technologies). For both
SNPs, PCR was made using 0.5 ul of 10X PCR MasterMix (Roche), 0.25 ul 40X TaqMan assay,
50 ng DNA in a final volume of 5 ul. PCR run included a denaturing step of 10 min at 95°C
followed by 35 PCR cycles (1 min 95°C, 1 min 55°C, 1 min 72°C). Mendelian inheritance was
checked using the computer software PedCheck (O'Connell and Weeks, 1998), and 10%
blind replicates were included for genotyping quality control.
Statistical analysis
Given the low frequency of the FXR1 rs496250 A allele, we followed Del’Guidice et al.
(Del'Guidice et al., 2015) in combining the AG and AA genotypes into the “A carrier”
category (dominant coding). For GSK3B rs12630592, we adopted an additive model of the
number of T alleles based on the phenotype-genotype relationship observed in healthy
subjects (Del'Guidice et al., 2015). Association between symptom dimensions and genotypic
variables was studied using linear mixed effect models with the genotypic variables and
8
their interaction term as fixed effects and a polygenic random effect to capture residual
familial correlation. The proportion of total variance (additive genetic plus residual
variance) explained by the genotypic variables was estimated by taking the difference of the
total variance of the null model and the total variance of the model with genotypic
variables. Association between the diagnosis distribution (RMDD, BD, SAD, SZ vs. NAARs)
and genotypic variables was studied using a polytomous logistic model estimated using
generalized estimating equations (GEEs) with an independence working correlation
structure between the subjects in the same family and an empirical standard error estimate
robust to intra-familial correlation (Zeger et al., 1988). All analyses were redone with age
and sex as covariates in the models. Since results were very similar with and without
adjusting for age and sex, we report the unadjusted results only. Association between the
prevalence of each type of medication and genotypic variables was evaluated using log
binomial models fitted using Poisson GEEs with robust standard error estimates.
Multiple testing corrections
Our primary objective was based on the apriori hypothesis that the genetic interactions
would be in relation with two symptom dimensions (mania and depression) given GSK3B
rs12630592 and FXR1 rs496250 have been shown to interact in regulating mood and
emotional processing in healthy humans. Since these two dimensions were assessed in two
states (acute episodes and stabilized interepisode intervals), we applied a Bonferroni
correction to the interaction p-values of these four analyses. The analyses of the eight other
symptom dimensions in mood disorder patients and of all ten symptom dimensions in SZ
patients, all in two states, were secondary analyses to test the specificity of the interactions
9
with the two former dimensions. A Bonferroni correction for 36 tests was applied to these
secondary analyses. Our secondary objectives of testing genetic interaction with the
diagnosis distribution and with age of onset involved a single analysis each, and our
analysis of medication was exploratory, so no multiple testing correction was applied on
the interaction tests for these additional analyses. A correction for testing association to
GSK3B rs12630592 within the two genotype categories of FXR1 rs496250 was applied
throughout.
Results
GSK3B - FXR1 interaction on mania and depression
In mood disorder patients we detected a statistically significant interaction between the T
allele of GSK3B rs12630592 and the A allele of FXR1 rs496250 in association with the
severity level of the mania symptom dimension in acute episodes, and a nearly statistically
significant one in stabilized interepisode intervals (Figure 1A and B). We also detected the
same interaction in association with the level of the depression dimension in stabilized
intervals but not in acute episodes (Figure 1C and D). The interaction pattern was the same
for the mania or the depression dimension in stabilized intervals: the severity level of the
symptom dimension increased with the number of GSK3B rs12630592 T alleles only in
carriers of the A allele of FXR1 rs496250, such that the subjects who are homozygote TT for
GSK3B rs12630592 and carriers of the A allele of FXR1 rs496250 had the highest severity of
symptoms. By contrast, subjects carrying GSK3B rs12630592 T alleles with the FXR1
rs496250 GG genotype, or carrying the FXR1 rs496250 A allele with the GSK3B rs12630592
GG genotype, had symptom severities similar to subjects with the reference GSK3B
10
rs12630592 GG - FXR1 rs496250 GG genotype. The SNPs rs12630592 and rs496250 and
their interaction explained 11% of the total variance of the manic symptom dimension (in
acute and in stabilized periods) and 5% of the total variance of the depression symptom
dimension in stabilized periods.
Specificity of the GSK3B - FXR1 interaction
To check the specificity of the interaction effect on the mania and the depression
dimensions, we examined in the mood disorder patients the relation with psychotic
symptom dimensions, and found no significant interaction between the combinations of
GSK3B rs12630592 and FXR1 rs496250 genotypes, whether in acute (Figure S1) or
stabilized periods (Figure S2). We did the same in the SZ patients and we observed no
significant interaction of the combinations of GSK3B rs12630592 and FXR1 rs496250
genotypes in relation with either the manic, depressive or psychotic symptom dimensions,
whether in acute (Figure S3) or stabilized periods (Figures S4).
GSK3B - FXR1 interaction on diagnosis distribution
Another way to examine the interacting effects of GSK3B and FXR1 is to consider the
prevalence of various major psychosis diagnoses as a function of GSK3B rs12630592 and
FXR1 rs496250 genotypes. A statistically significant interaction between the T allele of
GSK3B rs12630592 and the A allele of FXR1 rs496250 was then detected in relation to the
diagnosis distribution in the sample (p interaction = 0.04). The T allele of GSK3B
rs12630592 was associated with the diagnosis distribution only in FXR1 rs496250 A
carriers (p = 0.03, Figure 2 and Table 1). This association was driven mainly by the
11
decreasing prevalence of RMDD with the number of GSK3B rs12630592 T alleles, to the
point that no RMDD case was observed among the 30 GSK3B rs12630592 TT homozygotes
who are also FXR1 rs496250 A carriers. By contrast, the diagnosis distribution in carriers of
GSK3B rs12630592 T alleles with the FXR1 rs496250 GG genotype, or in carriers of the
FXR1 rs496250 A allele with the GSK3B rs12630592 GG genotype was similar to the
distribution in subjects with the reference GSK3B rs12630592 GG - FXR1 rs496250 GG
genotype.
Impact of the GSK3B - FXR1 interaction on age of onset
Given previous evidence that symptoms severity may be related to younger age of illness
onset (Propper et al., 2015; Suominen et al., 2007), we also considered a potential relation
of the interaction with onset age. In our sample, the mania dimension severity was
negatively correlated with age of onset of mood disorder (r = -0.23 in acute episodes, r = -
0.12 in stabilized interepisode intervals). In contrast, correlation between depression
severity and age of onset of mood disorder was negligible (r = -0.05 in acute episodes, r = -
0.02 in stabilized interepisode intervals). When we fitted a linear mixed model with FXR1
rs496250 and GSK3B rs12630592 genotypic variables to age of onset of mood disorder (BD
or RMDD), subjects who are homozygote TT for GSK3B rs12630592 and carriers of the A
allele of FXR1 rs496250 had an estimated mean age of onset of 21.5 years (nominal 95% CI:
[13.8, 29.1]), 13 years younger than the estimated mean age of onset of 34.4 years (nominal
95% CI: [31.4, 37.3]) for subjects who are homozygote GG for GSK3B rs12630592 and non-
carriers of the A allele of FXR1 rs496250 (Figure S5). Contrary to the mania symptom
dimensions, other genotypes had intermediate mean ages of onset between these two
12
extremes, so the interaction term was not statistically significant (nominal p = 0.19). No
association between FXR1 rs496250 and GSK3B rs12630592 genotypic variables and age of
onset of SZ was detected (results not shown).
Association between GSK3B - FXR1 genotype and lithium intake
The prevalence of lithium intake among mood disorder patients was associated with FXR1
rs496250 and GSK3B rs12630592 genotypes, the prevalence tending to decrease with the
number of GSK3B rs12630592 T alleles in FXR1 rs496250 GG subjects and to increase with
the number of GSK3B rs12630592 T alleles in FXR1 rs496250 A carriers (Table S1). To the
contrary, the intake of neuroleptic and anti-depressive medications was not associated with
FXR1 rs496250 and GSK3B rs12630592 genotypes (results not shown).
In order to assess the association between symptom dimensions and FXR1 and GSK3B in
patients with exposure to lithium, we repeated the analysis in the subset of 127 mood
disorder patients having taken lithium. Although lithium intake was highly prevalent among
BD patients, lithium intake did not equate BD diagnosis, as a sizable proportion of RMDD
patients also had lithium (Table S2). The pattern of association to the genotypes was
similar, with some attenuation of the slope of the symptoms with the number of GSK3B
rs12630592 T alleles and wider confidence intervals given reduced sample size compared
to Figure 1 (slope and 95% CI: 0. 50 [-0.18, 1.17] for mania in acute episodes; 0.28 [-0.21,
0.78] for mania in stabilized interepisode intervals and 0.35 [-0.18, 0.88] for depression in
stabilized intervals).
13
Discussion
Our findings support our a priori hypothesis of a possible interaction between GSK3B
(rs12630592 T allele) and FXR1 (rs496250 A allele) in mood disorder patients. A previous
study had indeed found this gene-gene interaction in healthy subjects on mood stability
(Del'Guidice et al., 2015).
Our analysis of mood disorder patients’ manic and depressive symptom dimensions, as
assessed in lifetime acute episodes and stabilized interepisode intervals, in densely affected
kindreds in relation to GSK3B and FXR1 functional SNP genotypes revealed an interaction
pattern related to the severity of mania (in both acute and stabilized periods) and of
depression symptoms in stabilized periods. The GSK3B rs12630592 T allele was associated
with higher dimension severity only in carriers of the A allele of FXR1 rs496250. The
highest levels of these symptom dimensions were thus observed in subjects who were
homozygote TT for GSK3B rs12630592 and carriers of the A allele of FXR1 rs496250. The
sizeable proportion of variance explained by the SNPs rs12630592 and rs496250 and their
interaction (11% for mania, 5% for depression in stabilized periods) attest the value of
testing biologically driven hypotheses on relevant alternative phenotypes to identify
genetic factors for mental illness. The pattern was found specific to manic and depressive
symptom dimensions in mood disorder patients. Given the higher level of manic symptoms
in BD patients, the decrease in prevalence of RMDD with the number of GSK3B rs12630592
T alleles only in carriers of the A allele of FXR1 rs496250 is congruent with the association
of this allelic combination with higher levels of mania.
14
Noteworthily, we had reported suggestive evidence of linkage to BD at D3S3023, a marker
close to GSK3B, in a subset of the families analyzed here (Maziade et al., 2005). The need to
take into account epistasis with FXR1 and examining symptom dimensions may explain the
previous difficulty in reproducing this linkage signal and detecting one at FXR1.
The meaning of our findings is enhanced by previous studies. Our data further support the
observation of a functional interaction between these two SNPs and mood regulation in
healthy subjects (Del'Guidice et al., 2015). The FXR1 rs496250 A allele has already been
found associated to greater expression of FXR1 and GSK3B rs12630592 T associated to a
reduced expression of GSK3B in the pre-frontal cortex of healthy human subjects (Blasi et
al., 2013; Colantuoni et al., 2011; Del'Guidice et al., 2015). Furthermore, control subjects
carrying the FXR1 rs496250 A allele display greater emotional stability, a potential
protective effect against mood disorder only observed in GSK3B rs12630592 T carriers
(Del'Guidice et al., 2015). Our present observations in medicated BD patients may appear
contradictory since the combination of FXR1 rs496250 A and GSK3B rs12630592 T alleles
would be associated with greater severity of mania and earlier age of onset. A similar
paradox has been described for other risk alleles such as DRD2 rs2514218 T which exhibit a
level of protection against SZ (Schizophrenia Working Group of the Psychiatric Genomics,
2014) and yet, in first episode psychosis, would be associated with more severe symptoms,
poorer responsiveness to treatment and greater side effects (Zhang et al., 2015). In light of
this, it might then be possible that, while being protective in controls, the FXR1 rs496250 A
- GSK3B rs12630592 T allele combination would foster treatment resistance for some
dimensions of symptoms in subjects with mood disorder. Further studies need to address
treatment responsiveness for mania in testing this gene-gene interaction.
15
In line with previous studies showing more severe mania in patients with younger age of
illness onset, subjects of our sample who were homozygote TT for GSK3B rs12630592 and
carriers of the A allele of FXR1 rs496250 were diagnosed with BD at a substantially younger
age on average (21.5 years vs 34.4 years) than subjects with the reference GSK3B
rs12630592 GG - FXR1 rs496250 GG genotype were diagnosed with mood disorder (BD or
RMDD). Our findings in patients need replication but our data nonetheless suggest that the
BD patients who carry the variant allele in the two genes would suffer a greater impact of
illness on their life course because they would develop the disorder earlier. Congruent with
our findings regarding the mania and depression dimensions, the interaction effect with age
of onset was observed only in mood disorder patients, not in SZ patients. When replicated,
this finding could have future implications in the clinic such as the possibility of testing
young patients at entry into illness and inducing particular forms of management and
surveillance depending on genotype.
Our study has strengths and limitations. A first strength resides in our symptom ratings
based on a lifetime assessment of the patients’ records and our measurements of manic,
depression and psychotic symptoms on all patients in the study regardless of their
diagnosis. Second, the kindred sample comes from a population where we previously
demonstrated that population stratification confounding is negligible (Bureau et al., 2013),
so adjustment for such stratification was unnecessary. Lastly, the relatedness of the kindred
members was taken into account by polygenic random effect in linear mixed models for
quantitative outcomes and by the use of generalized estimating equations for categorical
outcomes. Limitations reside first in the limited sample size resulting in a small number of
subjects who were homozygote TT for GSK3B rs12630592 and carriers of the A allele of
16
FXR1 rs496250, although we had sufficient power to detect the interaction effect assuming
effects of the GSK3B rs12630592 T allele are additive, as suggested in healthy individuals
(Del'Guidice et al., 2015). Second, generalization of our findings to all BD patients may be
limited by the fact that the present sample was made of familial patients, although we have
already shown that most phenotypic, endophenotypic and genetic characteristics we
previously reported in this familial sample were very much alike those observed in studies
of sporadic patients (Maziade and Paccalet, 2013).
In conclusion, our study confirms previous observations of an interaction between FXR1
rs496250 and GSK3B rs12630592 genetic variants in regulating mood related behavioural
dimensions in humans (Del'Guidice et al., 2015). It also suggests that this allelic
combination may impact disease severity and age of onset in people with mood disorder.
Acknowledgements
We thank Chantal Mérette (Université Laval) for her comments to improve the study and
manuscript. We are grateful to our professional research assistants: Louise Bélanger, Marie-
Claude Boisvert, Linda René, Lisette Gagnon, Claudie Poirier, Nicole Leclerc, Julie Lamarche,
Pierrette Boutin, Lise St-Germain, Mélanie Mercier, Jordie Croteau, Alain Fournier, Claudia
Émond and Isabel Moreau, and to the family members, adults and children, who
participated in this study.
Funding
This work was supported by the Canadian Institutes of Health research (CIHR, grants MOP-
74430, MOP-119408 and MOP-114988) and by a Canada Research Chair (# 950-200810) in
the genetics of neuropsychiatric disorders of which M. Maziade is the Chair. The data
17
management system was supported by the Canada Foundation for Innovation Leadership
Opportunity Fund (grant 27592). J.-M. Beaulieu holds a Canada Research Chair in Molecular
Psychiatry and is an International Mental Health Organization Rising Star Awardee.
18
References
Andreasen, N.C., Flaum, M., Arndt, S., 1992. The Comprehensive Assessment of Symptoms and History (CASH). An instrument for assessing diagnosis and psychopathology. Archives of general psychiatry 49, 615-623.
Beaulieu, J.M., 2012. A role for Akt and glycogen synthase kinase-3 as integrators of dopamine and serotonin neurotransmission in mental health. J Psychiatry Neurosci 37, 7-16.
Beaulieu, J.M., Gainetdinov, R.R., Caron, M.G., 2009. Akt/GSK3 signaling in the action of psychotropic drugs. Annual review of pharmacology and toxicology 49, 327-347.
Beaulieu, J.M., Marion, S., Rodriguiz, R.M., Medvedev, I.O., Sotnikova, T.D., Ghisi, V., Wetsel, W.C., Lefkowitz, R.J., Gainetdinov, R.R., Caron, M.G., 2008a. A beta-arrestin 2 signaling complex mediates lithium action on behavior. Cell 132, 125-136.
Beaulieu, J.M., Zhang, X., Rodriguiz, R.M., Sotnikova, T.D., Cools, M.J., Wetsel, W.C., Gainetdinov, R.R., Caron, M.G., 2008b. Role of GSK3beta in behavioral abnormalities induced by serotonin deficiency. Proceedings of the National Academy of Sciences of the United States of America 105, 1333-1338.
Blasi, G., Napolitano, F., Ursini, G., Di Giorgio, A., Caforio, G., Taurisano, P., Fazio, L., Gelao, B., Attrotto, M.T., Colagiorgio, L., Todarello, G., Piva, F., Papazacharias, A., Masellis, R., Mancini, M., Porcelli, A., Romano, R., Rampino, A., Quarto, T., Giulietti, M., Lipska, B.K., Kleinman, J.E., Popolizio, T., Weinberger, D.R., Usiello, A., Bertolino, A., 2013. Association of GSK-3beta genetic variation with GSK-3beta expression, prefrontal cortical thickness, prefrontal physiology, and schizophrenia. The American journal of psychiatry 170, 868-876.
Bureau, A., Chagnon, Y.C., Croteau, J., Fournier, A., Roy, M.A., Paccalet, T., Merette, C., Maziade, M., 2013. Follow-up of a major psychosis linkage site in 13q13-q14 reveals significant association in both case-control and family samples. Biological psychiatry 74, 444-450.
Colantuoni, C., Lipska, B.K., Ye, T., Hyde, T.M., Tao, R., Leek, J.T., Colantuoni, E.A., Elkahloun, A.G., Herman, M.M., Weinberger, D.R., Kleinman, J.E., 2011. Temporal dynamics and genetic control of transcription in the human prefrontal cortex. Nature 478, 519-523.
Del'Guidice, T., Latapy, C., Rampino, A., Khlghatyan, J., Lemasson, M., Gelao, B., Quarto, T., Rizzo, G., Barbeau, A., Lamarre, C., Bertolino, A., Blasi, G., Beaulieu, J.M., 2015. FXR1P is a
19
GSK3beta substrate regulating mood and emotion processing. Proceedings of the National Academy of Sciences of the United States of America 112, E4610-4619.
Hauberg, M.E., Roussos, P., Grove, J., Borglum, A.D., Mattheisen, M., Schizophrenia Working Group of the Psychiatric Genomics, C., 2016. Analyzing the Role of MicroRNAs in Schizophrenia in the Context of Common Genetic Risk Variants. JAMA Psychiatry 73, 369-377.
Hou, L., Bergen, S.E., Akula, N., Song, J., Hultman, C.M., Landen, M., Adli, M., Alda, M., Ardau, R., Arias, B., Aubry, J.M., Backlund, L., Badner, J.A., Barrett, T.B., Bauer, M., Baune, B.T., Bellivier, F., Benabarre, A., Bengesser, S., Berrettini, W.H., Bhattacharjee, A.K., Biernacka, J.M., Birner, A., Bloss, C.S., Brichant-Petitjean, C., Bui, E.T., Byerley, W., Cervantes, P., Chillotti, C., Cichon, S., Colom, F., Coryell, W., Craig, D.W., Cruceanu, C., Czerski, P.M., Davis, T., Dayer, A., Degenhardt, F., Del Zompo, M., DePaulo, J.R., Edenberg, H.J., Etain, B., Falkai, P., Foroud, T., Forstner, A.J., Frisen, L., Frye, M.A., Fullerton, J.M., Gard, S., Garnham, J.S., Gershon, E.S., Goes, F.S., Greenwood, T.A., Grigoroiu-Serbanescu, M., Hauser, J., Heilbronner, U., Heilmann-Heimbach, S., Herms, S., Hipolito, M., Hitturlingappa, S., Hoffmann, P., Hofmann, A., Jamain, S., Jimenez, E., Kahn, J.P., Kassem, L., Kelsoe, J.R., Kittel-Schneider, S., Kliwicki, S., Koller, D.L., Konig, B., Lackner, N., Laje, G., Lang, M., Lavebratt, C., Lawson, W.B., Leboyer, M., Leckband, S.G., Liu, C., Maaser, A., Mahon, P.B., Maier, W., Maj, M., Manchia, M., Martinsson, L., McCarthy, M.J., McElroy, S.L., McInnis, M.G., McKinney, R., Mitchell, P.B., Mitjans, M., Mondimore, F.M., Monteleone, P., Muhleisen, T.W., Nievergelt, C.M., Nothen, M.M., Novak, T., Nurnberger, J.I., Jr., Nwulia, E.A., Osby, U., Pfennig, A., Potash, J.B., Propping, P., Reif, A., Reininghaus, E., Rice, J., Rietschel, M., Rouleau, G.A., Rybakowski, J.K., Schalling, M., Scheftner, W.A., Schofield, P.R., Schork, N.J., Schulze, T.G., Schumacher, J., Schweizer, B.W., Severino, G., Shekhtman, T., Shilling, P.D., Simhandl, C., Slaney, C.M., Smith, E.N., Squassina, A., Stamm, T., Stopkova, P., Streit, F., Strohmaier, J., Szelinger, S., Tighe, S.K., Tortorella, A., Turecki, G., Vieta, E., Volkert, J., Witt, S.H., Wright, A., Zandi, P.P., Zhang, P., Zollner, S., McMahon, F.J., 2016. Genome-wide association study of 40,000 individuals identifies two novel loci associated with bipolar disorder. Human molecular genetics.
Jope, R.S., 2011. Glycogen synthase kinase-3 in the etiology and treatment of mood disorders. Front Mol Neurosci 4, 16.
Klein, P.S., Melton, D.A., 1996. A molecular mechanism for the effect of lithium on development. Proceedings of the National Academy of Sciences of the United States of America 93, 8455-8459.
Labbe, A., Bureau, A., Moreau, I., Roy, M.A., Chagnon, Y., Maziade, M., Merette, C., 2012. Symptom dimensions as alternative phenotypes to address genetic heterogeneity in schizophrenia and bipolar disorder. European journal of human genetics : EJHG 20, 1182-1188.
20
Maziade, M., Paccalet, T., 2013. A protective-compensatory model may reconcile the genetic and the developmental findings in schizophrenia. Schizophr Res 144, 9-15.
Maziade, M., Roy, M.A., Chagnon, Y.C., Cliche, D., Fournier, J.P., Montgrain, N., Dion, C., Lavallee, J.C., Garneau, Y., Gingras, N., Nicole, L., Pires, A., Ponton, A.M., Potvin, A., Wallot, H., Merette, C., 2005. Shared and specific susceptibility loci for schizophrenia and bipolar disorder: a dense genome scan in Eastern Quebec families. Mol Psychiatry 10, 486-499.
Maziade, M., Roy, M.A., Fournier, J.P., Cliche, D., Merette, C., Caron, C., Garneau, Y., Montgrain, N., Shriqui, C., Dion, C., et al., 1992. Reliability of best-estimate diagnosis in genetic linkage studies of major psychoses: results from the Quebec pedigree studies. Am J Psychiatry 149, 1674-1686.
Maziade, M., Roy, M.A., Martinez, M., Cliche, D., Fournier, J.P., Garneau, Y., Nicole, L., Montgrain, N., Dion, C., Ponton, A.M., 1995. Negative, psychoticism, and disorganized dimensions in patients with familial schizophrenia or bipolar disorder: continuity and discontinuity between the major psychoses. Am J Psychiatry 152, 1458-1463.
O'Connell, J.R., Weeks, D.E., 1998. PedCheck: A Program for Identification of Genotype Incompatibilities in Linkage Analysis. American Journal of Human Genetics 63, 259-266.
Paccalet, T., Gilbert, E., Berthelot, N., Marquet, P., Jomphe, V., Lussier, D., Bouchard, R.H., Cliche, D., Gingras, N., Maziade, M., 2016. Liability indicators aggregate many years before transition to illness in offspring descending from kindreds affected by schizophrenia or bipolar disorder. Schizophrenia research.
Propper, L., Ortiz, A., Slaney, C., Garnham, J., Ruzickova, M., Calkin, C.V., O'Donovan, C., Hajek, T., Alda, M., 2015. Early-onset and very-early-onset bipolar disorder: distinct or similar clinical conditions? Bipolar Disord 17, 814-820.
Roy, M.A., Lanctot, G., Merette, C., Cliche, D., Fournier, J.P., Boutin, P., Rodrigue, C., Charron, L., Turgeon, M., Hamel, M., Montgrain, N., Nicole, L., Pires, A., Wallot, H., Ponton, A.M., Garneau, Y., Dion, C., Lavallee, J.C., Potvin, A., Szatmari, P., Maziade, M., 1997. Clinical and methodological factors related to reliability of the best-estimate diagnostic procedure. The American journal of psychiatry 154, 1726-1733.
Schizophrenia Working Group of the Psychiatric Genomics, C., 2014. Biological insights from 108 schizophrenia-associated genetic loci. Nature 511, 421-427.
Stepniak, B., Kastner, A., Poggi, G., Mitjans, M., Begemann, M., Hartmann, A., Van der Auwera, S., Sananbenesi, F., Krueger-Burg, D., Matuszko, G., Brosi, C., Homuth, G., Volzke, H., Benseler, F., Bagni, C., Fischer, U., Dityatev, A., Grabe, H.J., Rujescu, D., Fischer, A., Ehrenreich, H., 2015.
21
Accumulated common variants in the broader fragile X gene family modulate autistic phenotypes. EMBO Mol Med 7, 1565-1579.
Suominen, K., Mantere, O., Valtonen, H., Arvilommi, P., Leppamaki, S., Paunio, T., Isometsa, E., 2007. Early age at onset of bipolar disorder is associated with more severe clinical features but delayed treatment seeking. Bipolar Disord 9, 698-705.
Woodgett, J.R., 2001. Judging a protein by more than its name: GSK-3. Science's STKE : signal transduction knowledge environment 2001, RE12.
Zeger, S.L., Liang, K.Y., Albert, P.S., 1988. Models for longitudinal data: a generalized estimating equation approach. Biometrics 44, 1049-1060.
Zhang, J.P., Robinson, D.G., Gallego, J.A., John, M., Yu, J., Addington, J., Tohen, M., Kane, J.M., Malhotra, A.K., Lencz, T., 2015. Association of a Schizophrenia Risk Variant at the DRD2 Locus With Antipsychotic Treatment Response in First-Episode Psychosis. Schizophrenia bulletin 41, 1248-1255.
22
Figure legends
Figure 1 Distribution of manic and depressive symptom dimensions in mood disorder
patients of the Eastern Quebec kindred study by GSK3B rs12630592 and FXR1 rs496250
genotypes. Interaction p-values are Bonferroni-corrected for testing 4 dimensions. The
slopes of the relationships between mean symptom levels and the number of GSK3B
rs12630592 T alleles, in FXR1 rs496250 GG subjects and in A carriers (A*), are shown
below the interaction p-values, with simultaneous 95% confidence intervals for two
genotypes and four symptom dimensions, obtained using the Bonferroni method. In box
plots, center lines are the median, bottom and top lines are the 1st and 3rd quartiles and the
whiskers extend to 1.5x the interquartile range. N : number of subjects with given genotype.
Figure 2 Distribution of major psychosis diagnoses by GSK3B rs12630592 and FXR1
rs496250 genotypes in the Eastern Quebec kindred study.
23
Table
Table 1 : Association of the GSK3B rs12630592 T allele with major psychosis diagnosis
distribution stratified by FXR1 rs496250 genotype in the Eastern Quebec kindred study.
FXR1 rs496250 GG FXR1 rs496250 A carrier
GSK3B rs12630592 T
allele
GSK3B rs12630592 T allele
Diagnosis N OR (95% CI)* P† OR (95% CI)* P†
RMDD 86 0.94 (0.58, 1.53) 0.13 (0.02, 0.89)
BD 128 0.92 (0.58, 1.45) 0.81 (0.34, 1.96)
SAD 48 1.26 (0.65, 2.44) 1.76 (0.58, 5.31)
SZ 122 0.84 (0.50, 1.40) 1.39 (0.77, 2.50)
NAAR 471 1 (reference) 1.0 1 (reference) 0.03
RMDD: recurrent major depressive disorder; BD: bipolar disorder; SAD: schizoaffective
disorder; SZ: schizophrenia; NAAR: non-affected adult relative
* Odds ratio of diagnosis compared to non-affected category for an additional copy of the
GSK3B rs12630592 T allele, with simultaneous 95% confidence interval for two genotypes
and four diagnoses.
† P-value for global test of association of GSK3B rs12630592 T with diagnosis, Bonferroni
corrected for two genotypes.
24