SEARCH

SEARCH BY CITATION

Keywords:

  • stress;
  • environment;
  • psychotic disorder;
  • epigenesis;
  • epistasis

Abstract

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Peerbooms O, Rutten BPF, Collip D, Lardinois M, Lataster T, Thewissen V, Mafi Rad S, Drukker M, Kenis G, van Os J, Myin-Germeys I, van Winkel R. Evidence that interactive effects of COMT and MTHFR moderate psychotic response to environmental stress.

Objective:  A functional interaction between Catechol-O-Methyltransferase (COMT) Val158Met and methylenetetrahydrofolate reductase (MTHFR) C677T has been shown to differentially affect cognition in patients with schizophrenia and healthy controls; the effect of COMT Val158Met × MTHFR interaction on resilience to stress in patients and controls remains to be examined.

Method:  A total of 98 patients with non-affective psychotic disorder and 118 controls were genotyped for MTHFR C677T, MTHFR A1298C, and COMTVal158Met. Daily life reactivity to stress, modelled as the effect of daily life stress on psychotic experiences, was measured using the experience sampling method (ESM).

Results:  The MTHFR C677T genotype moderated the interaction between COMT Val158Met genotype and stress in patients (P < 0.0001), but not in controls (P = 0.68). Further examination of this interaction revealed that in patients with the MTHFR 677 T-allele, COMT Met/Met individuals displayed the largest increases in psychotic symptoms in reaction to ESM stress [χ2(2) = 29.51; P < 0.0001], whereas in patients with the MTHFR 677 C/C genotype no significant COMT Val158Met × ESM stress interaction was apparent [χ2(2) = 3.65; P = 0.16]. No moderating effect of MTHFR A1298C was found.

Conclusion:  Stress reactivity associated with COMT Val158Met in patients with psychosis may crucially depend on MTHFR C677T genotype.


Significant outcome

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix
  •  Complex gene-gene interactions determine vulnerability to the psychosis-inducing effects of daily life stress in patients with psychotic disorder.

Limitations

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix
  •  No data on cognitive functioning were present, therefore the trade-off effects of COMT between cognition and stress resilience were not directly tested.
  •  Dopamine or methylation levels were not tested and therefore the exact pathway by which MTHFR and COMT interact remains unclear.

Introduction

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

The Catechol-O-Methyltransferase (COMT) Val158Met polymorphism has long been considered a candidate polymorphism conferring risk of schizophrenia because of its pivotal function in dopamine breakdown in the prefrontal cortex. Recent meta-analyses, however, found absence of or minimal evidence for an association between the COMT Val158Met polymorphism and schizophrenia (1–3). Although these findings seem to refute a major role for COMT Val158Met in determining psychosis liability, replicated evidence suggests that COMT Val158Met is associated with trade-off effects in important intermediate phenotypes relevant for psychosis, such as cognition and stress sensitivity. A recent meta-analysis of 20 MRI or fMRI studies (most of which were conducted in healthy individuals) demonstrated strong and opposing effects associated with COMT Val158Met, showing greater prefrontal activation in Val carriers in cognitive tasks but lower activation in emotional paradigms, and vice versa for the Met-allele (4). Experimental evidence supporting COMT Val158Met trade-off effects comes from a study that used a genetically engineered mouse model with differential COMT function (5). Mice overexpressing the human Val/Val genotype displayed worse cognition while being more resilient to stress and pain, whereas genetic knock-out of COMT was associated with better cognitive functioning but with less resilience to stress and pain.

With regard to stress resilience in psychosis, results in agreement with trade-off effects associated with COMT Val158Met were reported by two recent studies that found that the Met/Met genotype was associated with increased reactivity to stress in patients with psychotic disorder but not in healthy controls (6, 7). This emotional paradigm was investigated using the Experience Sampling method (ESM), an innovative method to investigate reactivity to stress in a daily life prospective design, especially suited to optimize statistical power and sensitivity to detect interplay between genetic variation and environment in humans (8, 9).

In addition, it has been suggested that part of the phenotypic variability associated with COMT Val158Met may be explained by variations in genes that either act additively, or modify COMT function (10). One such gene may be the MTHFR gene (see Appendix), which harbours two common functional polymorphisms (C677T and A1298C) that both reduce enzymatic activity (11–13). Each copy of the 677T allele causes a 35% reduction of enzyme activity (11), whereas individuals with the 1298C allele may have more moderate reductions in enzyme activity (12–14). Both polymorphisms have been associated with a significantly increased risk of schizophrenia (15, 16). MTHFR, among others, is crucially important in the pathway leading to generation of S-adenosylmethionine (SAM). S-adenosylmethionine is the major donor in the brain for methyl groups (17), which are necessary for adequate DNA methylation. Since the methylation status of the COMT promoter influences COMT transcription, and COMT-regulated dopamine inactivation requires the transfer of a methyl group from SAM to dopamine (18), it has been suggested that COMT transcription and function may depend on the availability of SAM (19).

Of note, recent studies have suggested that functional interactions between COMT and MTHFR may be relevant for psychosis liability. One study found that the combined presence of MTHFR 677TT with COMT Met/Met was associated with an increased risk of schizophrenia (20). Furthermore, in patients with psychotic disorders, the combination of an MTHFR 677T allele and COMT Val was associated with worse cognitive performance and reduced working memory load-dependent activation of the prefrontal cortex (21, 22). In contrast, however, was the finding that reduced prefrontal activation in healthy controls, using the same cognitive paradigm, was associated with MTHFR 677C/C and COMT Met/Met suggesting that the functional relevance of the COMT × MTHFR interaction may be dependent on background factors associated with psychosis.

Aims of the study

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Thus, given i) evidence for trade-off effects between cognition and stress resilience associated with Catechol-O-Methyltransferase Val158Met, ii) the biological plausibility of functional interactions between Methylenetetrahydrofolate reductase and Catechol-O-Methyltransferase Val158Met, and iii) studies suggesting relevance of Catechol-O-Methyltransferase × Methylenetetrahydrofolate reductase interaction for schizophrenia risk and cognition, with differential effects in patients with schizophrenia and healthy controls, the current article examined the hypothesis that interaction between Methylenetetrahydrofolate reductase and Catechol-O-Methyltransferase Val158Met would be associated with reactivity to stress, with divergent effects in patients with psychosis and healthy controls.

COMT

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Catechol-O-Methyltransferase inactivates dopamine at postsynaptic sites in the human brain, especially in the prefrontal cortex. The gene encoding for COMT contains a functional polymorphism at codon 158 that results in a change from Valine (Val) to Methionine (Met) (COMTVal158Met). Val/Val carriers have a 40% higher COMT enzyme activity in the brain than the Met/Met carriers, resulting in faster breakdown of dopamine and consequently, lower baseline dopamine levels in the prefrontal cortex.

MTHFR

Methylenetetrahydrofolate reductase (MTHFR) catalyses the conversion of 5,10-methylenetetrahydrofolate (5,10-MTHF) to 5-methyltetrahydrofolate (5-MTHF), the predominant circulating form of folate. MTHFR is a crucial enzyme involved in the methylation of DNA.

A single copy of the MTHFR 677T allele causes a 35% reduction of enzyme activity, whereas the MTHFR 1298C allele causes a more moderate reduction in enzyme activity compared with the 677T allele.

Material and methods

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Sample

The current sample comprises samples of two earlier ESM studies in patients with psychosis and healthy controls (23, 24). The total sample consists of 316 participants [controls and patients; 172 pertaining to the first study (24) and 144 pertaining to the second study (23)]. In selected representative geographical areas in the South of the Netherlands and the Dutch-speaking part of Belgium, patients were identified through representative clinicians working in regional psychotic disorder services, whose caseload was screened for inclusion criteria. Subsequently, a group of patients presenting consecutively at these services either as out-patients or in-patients were recruited for the study.

The DSM-IV diagnoses (25) were confirmed with either the OPCRIT computer program (26) or the Comprehensive Assessment of Symptoms and History (27). Controls were selected through a system of random mailings to addresses in the catchment areas of the cases. Additional inclusion criteria were: i) age 16–65 years and ii) sufficient command of the Dutch language. Exclusion criteria were: i) brain disease and ii) history of head injury with loss of consciousness. For controls, additional exclusion criteria were presence of lifetime history of psychotic disorder and family history of psychotic disorder. Written informed consent, conforming to the local ethics committee’s guidelines, was obtained from all participants.

MTHFR and COMT genotyping

Of the 316 participants, 216 agreed to provide DNA (blood: 128 participants, buccal mucosa: 88 participants). Buccal cell samples were collected with sterile swabs (Omniswab, Whatman®, Whatman, Kent, UK.). DNA was extracted using QIAamp DNA Mini Kits (Qiagen, Venlo, The Netherlands.). DNA of the blood samples was isolated either manually according to the Promega protocol or with the Autogenflex3000.

The following SNPs were genotyped using the following TaqMan® SNP Genotyping assays (Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands): rs4680 (COMT Val158Met; assay ID C_25746809_50), rs1801131 (MTHFR A1298C; assay ID C_850486_20) and rs1801133 (MTHFR C677T; assay ID C_1202883_20). Assays were run on a 7900HT Fast Real-Time PCR System (Applied Biosystems).

Experience sampling method

Experience sampling method is a structured random time-sampling self-assessment technique to assess subjects, their mental states and contexts in their daily living environments. It is a valid and reliable way to study immediate effects of the environment, thereby reducing biases in recall (6–8, 23, 24, 28–42). Participants received a pre-programmed digital wristwatch and ESM self assessment forms collated in a booklet for each day. Ten times a day on six consecutive days, the watch emitted a signal at unpredictable moments between 7.30 a.m. and 10.30 p.m. After each ‘beep’, subjects were asked to fill out the ESM self-assessment forms previously handed to them, collecting reports of thoughts, current context (activity, persons present and location), appraisals of the current situation, mood and psychotic experiences. All self-assessments were rated on 7-point Likert scales. Participants were instructed to complete their reports immediately after the beep, to minimize memory distortions and to record the time at which they completed the form. During the sampling period, research staff contacted subjects by phone to assess whether or not they were complying with the instructions. Reports are assumed valid when subjects respond to the beep within 15 min. This was ascertained by comparing the actual beep time with the reported time of completion of the reports. All reports completed more than 15 min after the signal were excluded from the analyses. Participants were only included in the analyses when they responded validly to at least one-third of the emitted beeps (29). Previous studies have demonstrated the feasibility, validity and reliability of ESM in general and patient populations (8, 37, 39). ESM is particularly well suited for the investigation of gene-environment interactions, given the detailed assessment of the environment, the prospective design and the repeated within-subject measurements enabling high statistical power (8). Because of the large number of within-person measurements, the required sample size for an ESM study cannot be equated to that of a straightforward case–control genetic association study but may be better compared with the required sample size of a neuro-imaging study of genetic association.

Measures

Stress reactivity, as described in previous work (7, 35, 37, 38) was conceptualized as psychotic reactivity to daily life events and minor disturbances in daily life. Measures of stress and psychotic symptoms were derived from the experience sampling reports as described below.

Assessment of ESM stress.  The ESM stress was conceptualized as the subjectively appraised stressfulness of distinctive events as previously described (37). To measure ESM stress, the subject was asked to report, after each beep, the most important event that had happened between the current and the previous report. This event was subsequently rated on a bipolar Likert scale (−3 = very unpleasant, 0 = neutral, 3 = very pleasant). The responses were recoded to allow high scores to reflect stress (−3 = very pleasant, 0 = neutral, 3 = very unpleasant).

Assessment of ESM psychosis.  The ESM psychosis was measured by the sum score of 6 ESM items I feel suspicious, I cannot get rid of my thoughts, I am afraid of losing control, I feel unreal, and I hear voices, I see phenomena (Cronbachs α = 0.71) as described previously (7, 35). All items were rated on 7-point Likert scales (ranging from not at all to very).

Statistical analyses

Multilevel linear regression analysis, which is suited for analyses of clustered data (43), was used since ESM data have a hierarchical structure with repeated momentary measurements (level 1) per subject (level 2). Analyses were carried out with the XTREGAR module in STATA/MP version 10.1 (44), which takes the possibility of autocorrelation into account. Outcome variables included in the analyses were expressed in units of standard deviation, dividing the variables by its standard deviation, yielding standardized effect sizes.

To test the hypothesis that the interaction of MTHFR with COMT Val158Met genotype moderates psychotic responses to daily life stress, a three-way interaction between the COMT Val158Met genotype, MTHFR genotype and ESM stress was fitted in the model with the ESM psychosis as dependent variable. The independent variables were COMT Val158Met genotype (0 = Val/Val; 1 = Val/Met; 2 = Met/Met) and MTHFR C677T genotype (0 = C/C; 1 = C/T or T/T) or MTHFR A1298C genotype (0 = A/A, 1 = A/C or C/C) and ESM stress. Analyses were conducted separately for the patients and the control group. Main effects and interactions were assessed by Wald test. Effect sizes were calculated by applying and testing the appropriate linear combinations using the STATA LINCOM command.

Confounding is very rare in the context of a three-way interaction, as the ‘confounder’ needs to be independently associated with the three independent variables as well as with the dependent variable, which is unlikely, especially in the context of SNPs. Therefore, the analyses were not controlled for variables that were significantly different between patients and controls (see Table 1). Nevertheless, to exclude the possibility that the reported results were driven by differences in group characteristics such as age and gender, we re-ran the analyses co-varying for variables that were significantly different between groups.

Table 1.   Demographics
 Patients (N = 93)Controls (N = 115) 
  1. ESM, experience sampling method.

Mean age (SD)32.9 (11.2)40.1 (13.4)t = 4.2, df 206, < 0.0001
Sex (M/F)63/3036/79t = 5.6, df 206, < 0.0001
Education (%)
 Elementary school1.10.9t = 5.1, df 202, < 0.0001
 Secondary school77.543.5
 Higher education21.455.6
Living situation (%)
 Alone23.010.4t = −4.4, df 204, < 0.0001
 With partner22.073.9
 With parents or other family20.912.2
 Protected housing26.40.9
 Others7.72.6
Mean ESM psychosis (SD)1.6 (0.8)1.2 (0.3)t = −5.3, df 206, < 0.0001
Mean ESM stress (SD)Min 1.3 (0.9)Min 1.5 (0.6)t = −1.8, df 206, < 0.07

Results

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Subjects

Of the 316 participants (171 patients and 145 controls) who entered the study, 216 agreed to provide DNA (98 patients and 118 control participants). Two control participants were excluded ascribable to failure to complete the protocol (missing information regarding the time of the beep and completion of questionnaires) and an additional five patients and one control participant were excluded from the analyses as they had an insufficient number of valid ESM observations (<20). Genotyping of COMT and MTHFR was carried out in 208 subjects of which 93 patients (41 with schizophrenia; 10 with schizoaffective disorder; 32 with psychotic disorder NOS; one with schizophreniform disorder; three with delusional disorder; six with brief psychotic disorder) and 115 controls. A total of 13 subjects could not be analysed because of genotyping failure of COMT, another four referable to genotyping failure of MTHFR C677T and five referable to genotyping failure of MTHFR A1298C. The final sample, therefore, comprised 191 participants for MTHFR C677T and COMT Val158Met (107 controls and 84 patients) and 190 for MTHFR A1298C and COMT Val158Met (107 controls and 83 patients).

Significant differences in several socio-demographic characteristics, such as age, gender, living situation and level of education between the control group and the patient group were observed (Table 1).

The total group of subjects completed on average 43.8 valid ESM reports (SD = 9.3), control subjects reporting on average more valid ESM reports (47.1; SD = 8.0) than patients (39.7; SD = 9.2 for the patients (t = 6.2; df 206, < 0.001). The distributions of COMT Val158Met, MTHFR C677T and MTHFR A1298C genotypes were in Hardy–Weinberg equilibrium for both patients and controls.

Reactivity to stress

The ESM stress was associated with increases in ESM psychosis in both patients (P < 0.0001, β = 0.051) and controls (P < 0.0001, β = 0.020). The COMT Val158Met, MTHFR C677T or A1298C genotypes had no significant main effects on ESM psychosis, neither in patients nor controls (Table 2), but the effects of ESM stress in patients with psychosis were moderated by COMT Val158Met as reported previously, Met/Met genotypes displaying the greatest increases in ESM psychosis (6).

Table 2.   Main effects of ESM stress and genotypes on ESM psychosis
 PatientsControls
Mean psychosis (SD)Effect size (Wald χ2, P)Mean psychosis (SD)Effect size (Wald χ2, P)
  1. COMT, Catechol-O-Methyltransferase; ESM, experience sampling method; MTHFR, Methylenetetrahydrofolate reductase.

COMT Val158Met
 Val/Val2.27 (1.2)χ2: 4.05 = 0.132.07 (0.7)χ2: 0.50 P = 0.78
 Val/Met2.83 (1.6)1.95 (0.6)
 Met/Met2.76 (1.3)2.01 (0.6)
MTHFR C677T
 C/C2.51 (1.2)χ2: 4.09 = 0.131.99 (0.6)χ2: 0.54 = 0.76
 T-allele carriers2.87 (1.6)2.00 (0.6)
MTHFR A1298C
 A/A2.64 (1.3)χ2: 1.76 = 0.411.97 (0.6)χ2: 0.07 = 0.97
 C-allele carriers2.77 (1.7)2.02 (0.7)

In addition, MTHFR C677T genotype moderated the interaction between COMT Val158Met genotype and ESM stress in patients [χ2(2) = 25.66; P < 0.0001], but not in controls [χ2(2) = 0.78, P = 0.68; Table 3]. Further exploration of the MTHFR C677T × COMT Val158Met × ESM stress three-way interaction revealed that in patients with the MTHFR 677 C/C genotype, no significant COMT Val158Met × ESM stress interaction was found [χ2(2) = 3.65; P = 0.16], whereas there was significant COMT Val158Met × ESM stress interaction in patients with the MTHFR 677 T-allele, COMT Met/Met patients displaying the largest increases in psychotic symptoms in reaction to ESM stress [χ2(2) = 29.51; P < 0.0001; Fig. 1].

Table 3.   Results of the three-way interaction of MTHFR C677T with COMT Val158Met and stress on psychosis
 PatientsControls
Wald χ2, PNWald χ2, PN
  1. COMT, Catechol-O-Methyltransferase; ESM, experience sampling method; MTHFR, Methylenetetrahydrofolate reductase.

  2. *Not applicable, i.e. genotype effects are not reported in the case that there is no significant two-way interaction.

MTHFR C677T × COMT Val158Met × ESM stressχ2: 25.66 < 0.0001 χ2: 0.78 P = 0.68 
MTHFR C/C genotypeχ2: 3.65 χ2: 0.25 
= 0.16= 0.88
 COMT Val/Val × ESM stressNA*6NA16
 COMT Val/Met × ESM stressNA13NA21
 COMT Met/Met × ESM stressNA12NA15
MTHFR T-allele carrierχ2: 29.51 χ2: 2.35 
< 0.0001= 0.31
 COMT Val/Val × ESM stressβ: 0.041 (−0.010–0.092)14NA12
 COMT Val/Met × ESM stressβ: 0.027 (0.002–0.052)31NA23
 COMT Met/Met × ESM stressβ: 0.18 (0.132–0.233)8NA20
image

Figure 1.  Psychotic symptoms in response to stress in patients with a T-allele of the MTHFR C677T polymorphism and in patients with the C/C genotype of MTHFR C677T, per COMT Val158Met genotype. In patients with the Met/Met genotype and a T allele of MTHFR C677T polymorphism, the highest increase in psychotic reaction to stress was found. No significant effect of MTHFR C677T on stress reactivity was found in patients with the C/C genotype.

Download figure to PowerPoint

Significant differences in several socio-demographic characteristics (age, gender, living situation, educational level) between patients and controls were found (Table 1). To exclude the possibility that the reported results were driven by differences in age, gender, living situation and educational level, we re-ran the analyses co-varying for these variables. This did not change the results.

In contrast to MTHFR C677T, the MTHFR A1298C genotype did not moderate the interaction between COMT Val158Met genotype and ESM stress in either patients [three-way interaction χ2(2) = 1.12; P = 0.57, Table 3] or controls [χ2(2) = 0.78, P = 0.68].

Discussion

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Findings

In the present study, the combined effect of MTHFR polymorphisms and the COMT Val158Met polymorphism on psychotic reactivity to daily life stress was investigated. It was found that COMT Val158Met × stress interaction was conditional on MTHFR C677T genotype, but only in patients. Patients with both a T-allele of MTHFR C677T and COMT Met/Met displayed the largest increase in psychotic symptoms in response to stress. In contrast, MTHFR C/C genotype ‘neutralized’ the effect of COMT Val158Met on stress reactivity. The MTHFR A1298C did not display a similar interaction with COMTVal158Met.

Functional relevance of MTHFR × COMT interaction

Earlier studies have found evidence for functional interaction between MTHFR C677T and COMT Val158Met, with differential effects on prefrontal recruitment in a cognitive task in patients with schizophrenia and healthy controls (21, 22). Roffman and colleagues speculate that different hypothesized positions on the prefrontal ‘inverted U dopamine curve’ may explain their findings. The inverted U curve refers to findings suggesting an optimal level of dopamine in the prefrontal cortex, which corresponds with optimal cognitive performance (45). Dopamine levels below or above the optimum result in poorer cognitive performance. It is assumed that patients may have low prefrontal dopamine availability, in contrast to healthy controls. Theoretically, optimal prefrontal dopamine levels would correspond with the COMT Val/Met genotype. The COMT Val/Val genotype are hypothesized to have sub-optimal levels of prefrontal dopamine, whereas Met/Met individuals are hypothesized to have too high levels of dopamine.

In patients, the co-presence of a MTHFR T-allele in addition to the COMT Val/Val genotype may further decrease availability of prefrontal dopamine, given that the MTHFR T allele may lead to hypomethylation i.e. increased expression of COMT gene, enhancing dopamine inactivation. In controls, however, the combination of COMT Met/Met and MTHFR C/C would lead to increased availability of prefrontal dopamine, as the combination of normal methylation potential by MTHFR C/C genotype and lower activity of COMT Met/Met will cause prefrontal dopamine to be higher. Since both allele combinations are at the extremes of the ‘inverted U curve’, this could potentially explain the fact that poorest cognitive performance is observed in these groups (Fig. 2).

image

Figure 2.  Inverted U curve. Dopamine availability is associated with efficiency of cortical functioning. Either too much or too little dopamine is associated with reduced efficiency. In patients, the addition of a Methylenetetrahydrofolate reductase (MTHFR) T-allele to the Catechol-O-Methyltransferase (COMT) Val/Val genotype may further decrease the availability of prefrontal dopamine, as a MTHFR T allele may lead to a hypomethylated, i.e. increased expression of COMT gene thereby enhancing inactivation of dopamine, whereas in controls the combination of COMT Met/Met and MTHFR C/C would lead to increased availability of prefrontal dopamine. Since both allele combinations are at the extremes of the ‘inverted U curve’, this could potentially explain that poorest cognitive performance is observed in these groups. The finding that psychotic patients with MTHFR 677T and COMT Met/Met are most reactive to stress, however, is difficult to reconcile with this theory, as patients with this combination are not at one of the extremes of the inverted U curve (a).

Download figure to PowerPoint

The current results suggesting that psychotic patients with MTHFR 677T and COMT Met/Met are less resilience to stress, however, are difficult to reconcile with this theory, as patients with this combination are not at one of the extremes of the inverted U curve (Fig. 2). Instead, when interpreting the current results in the context of previous findings (21, 22) it could be hypothesized that COMT Val158Met trade-off effects between cognition and resilience to stress are dependent on the co-presence of a MTHFR C677T T-allele (4, 5).

The present study found no interaction between COMT Val158Met and MTHFR A1298C. As far as we are aware this is the first study examining the possibility of interaction between COMT Val158Met and MTHFR A1298C. Although the reported results need independent replication, a possible explanation for the lack of detectable behavioural consequences of MTHFR A1298C on stress reactivity may be situated in the fact that the effects of this SNP on protein function are smaller and may therefore be more dependent on other factors, such as folate intake. Studies with more statistical power are needed to examine whether MTHFR A1298C does not interact with COMT Val158Met, or does so in a more subtle way than the C677T SNP (13).

Differential effects in patients and controls

Previous research also identified MTHFR × COMT interaction in healthy controls; however, the implicated alleles were different for patients and controls (21, 22). The current study found no evidence for MTHFR × COMT interaction in healthy controls. It can be argued that this discrepancy has arisen ascribable to the low variability in the phenotype investigated in the control group; subtle psychosis-like symptoms were examined in reaction to daily life stress in individuals not suffering from a psychotic disorder. In agreement with this explanation, healthy controls reported low levels of psychosis-like experiences. Nevertheless, a strong main effect of ESM stress on the ESM psychosis outcome measure was found in the healthy control group. This supports the use of the ESM psychosis phenotype in healthy control groups as applied in previous studies (35, 41) and suggests that the current findings may index biological differences between the patient and control group that have yet to be determined.

Robustness of the reported results

The present results were not corrected for multiple testing, although it may be argued that this is necessary given the test of not only the three-way interaction but also the underlying main effects and two-way interactions. However, testing the three-way interaction is non-informative if the underlying main effects and two-way interactions are not reported. Therefore, in our opinion, two hypotheses of three-way interaction in (COMT × MTHFR A1298C × stress and COMT × MTHFR C677T × stress) in two different samples (controls and patients) could be controlled for. In support of their robustness, the reported findings would have remained highly significant when applying Bonferroni correction (0.05/4 = 0.0125) for this number of independent hypotheses. In addition, even if the most stringent correction would be applied, i.e. correction for all main effects, two-way and three-way interactions (22 tests), the results would also have remained significant at this P-value (0.05/22 = 0.002).

Strengths and limitations

A major strength of the current study is the use of the ESM for the assessment of gene-gene-environment interaction (8, 9). This method allows gathering insight in the course of symptoms and the influence of genes and environmental stimuli in a momentary, ‘real-world’ design by providing a prospective collection of cumulative, repeated measures of proximal environmental risk factors (8, 46).

As cognitive functioning was not assessed in the current sample, it was not possible to directly test for possible trade-off effects between cognition and stress resilience. However, when placing current results in context with data by Roffman et al. on MTHFR × COMT interaction in cognitive functioning (21, 22), the current study provides additional support for COMT trade-off effects in patients with a psychotic disorder, and for functional relevance of MTHFR C677T in this trade-off effect. Future studies are warranted to replicate these findings, to test MTHFR × COMT effects on both cognition and resilience to stress in individuals pertaining to a single sample, and to further explore the mechanism underlying differential effects of the MTHFR × COMT in patients and healthy individuals.

Declaration of interest

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

The research leading to these results has received funding from the European Community’s Seventh Framework Programme under grant agreement No. HEALTH-F2-2009-241909 (Project EU-GEI). Inez Myin-Germeys was supported by a 2006 NARSAD Young investigator award and by the Dutch Medical Council (VENI and VIDI grant). Bart Rutten is the recipient of a Kootstra fellowship of Maastricht University.

Odette Peerbooms, Shahed Mafirad and Marjan Drukker have received no compensation for professional services in any of the previous 2 years, and do not anticipate receiving such compensation in the near future. Dina Collip has received honoraria from Lundbeck BV.

Marielle Lardinois received honoraria from Lundbeck BV. Tineke Lataster has received honoraria from Lundbeck BV, AstraZeneca BV, Eli Lilly, and Servier Farma. Viviane Thewissen received honoraria from Bristol-Myers Squibb B.V., AstraZeneca BV, Eli Lilly B.V., Organon Nederland B.V., Wyeth Pharmaceuticals B.V. and Servier Nederland Farma B.V.

Gunter Kenis has received financial compensation as an independent symposium speaker from Eli Lilly. Jim van Os is an unrestricted research grant holder with, or has received financial compensation as an independent symposium speaker from Eli Lilly, BMS, Lundbeck, Organon, Janssen-Cilag, GSK, AstraZeneca, Pfizer and Servier, companies that have an interest in the treatment of psychosis.

Ruud van Winkel has been a consultant for Eli Lilly and Astra Zeneca, has received honoraria from Eli Lilly, AstraZeneca, Bristol Myers-Squibb and Janssen-Cilag and unrestricted grants from Eli Lilly and AstraZeneca.

References

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Appendix

  1. Top of page
  2. Abstract
  3. Significant outcome
  4. Limitations
  5. Introduction
  6. Aims of the study
  7. COMT
  8. Material and methods
  9. Results
  10. Discussion
  11. Declaration of interest
  12. References
  13. Appendix

Function of the COMT and MTHFR genes and the biological significance of the functional polymorphisms examined.