LANGUAGE LATERALIZATION IN MONOZYGOTIC TWIN PAIRS CONCORDANT AND DISCORDANT FOR SCHIZOPHRENIA A FMRI PILOT STUDY Filip Spaniel*, Jaroslav Tintera, Tomas Hajek, Jiri Horacek, Monika Dezortova, Pavlina Harantova, Milan Hajek, Colleen Dockery, Jiri Kozeny and Cyril Höschl From Psychiatric Center Prague, 3rd Faculty of Medicine, Charles University Prague, Center of Neuropsychiatric Studies and MR Unit, ZRIR, Institute for Clinical and Experimental Medicine, Czech Republic *corresponding author: Ústavní 191 181 03 Prague 8 Czech Republic E mail: Psychiatric Center Prague, Czech Republic spaniel@pcp.lf3.cuni.cz ABSTRACT OBJECTIVE: In this study we sought to explore possible differences in language lateralization between sporadic and familial cases of schizophrenia. fMRI was used to study cortical activation during a verbal task in Broca's area and its contralateral homologue in four pairs of monozygotic twins discordant for schizophrenia with no familial history of psychiatric illness, 4 concordant MZ twin pairs with first-degree relatives with schizophrenia, and 4 control MZ twin pairs. All subjects were right-handed. RESULTS: No statistically significant differences in task-related fMRI activation and Lateralization index (LI) were found in the direct comparison between “Sporadic” (4 affected twins from discordant MZ twin pairs) and “Familial” group (4 MZ twin pairs concordant for schizophrenia). However, pooled data from all subjects with schizophrenia (n=12) showed increased activation during verbal fluency task paradigm in the right homologue of Broca's area in contrast to healthy individuals (n=12), resulting in reduced LI in schizophrenic group. Concordant twins with high familial loading for schizophrenia showed prominent intra-pair, between co-twin differences in LI within given ROI. In this group, the intra-pair LI differences were significantly higher compared to the controls (0.69 ± 0.4 vs. 0.13 ± 0.13, p<0.03). In contrast to expectation, differences between the discordant and control group did not reach statistical significance in this variable. INTRODUCTION The aim of the present fMRI study was to determine differences in activation within Broca´s area and its contralateral homologue between affected subjects from monozygotic twin pairs discordant for schizophrenia with no familial loading of major mental illness and concordant monozygotic twins with first-degree relatives affected with schizophrenia or schizophrenia – spectrum disorder. Thus, by using this study population, we could highlight possible differences in the functional asymmetry of the brain between sporadic and familial cases of schizophrenia. METHODS Sporadic group 4 non-affected co-twins Familial group Control group. Subjects fMRI scanning procedure The study subjects consisted of 12 right-handed monozygotic (MZ) twin pairs (Tab. 1). a) 4 MZ twin pairs discordant for schizophrenia were divided into: “ ”= 4 affected twins without any family history of schizophrenia or other major psychiatric disorder. . b) 4 MZ twin pairs concordant for schizophrenia formed “ ”. Those twin pairs had family histories of schizophrenia or schizophrenia spectrum disorder in first-degree relatives. c) 4 MZ healthy MZ twin pairs were recruited as a Diagnosis of schizophrenia was made according to the Structured Clinical Interview for DSM-III-R by two experienced psychiatrists. Genetic analysis (concordance in 16 short tandem repeats markers) confirmed monozygosity of the twin pairs (PowerPlex® 16 System, Promega, Madison, USA [Thomson et al., 2001]). The following loci were used in the analysis: D3S1358, TH01, D21S11, D18S51, Penta E, vWA, AMXY, D8S1179, TPOX, FGA, D5S818, D13S317, D7S820, D16S539, CSF1PO, Penta D. · Prior to scanning, subjects completed a hand preference questionnaire, the Edinburgh Handedness Inventory (Oldfield, 1971). fMRI: l.5 T Siemens Vision. Verbal fluency task was used as an activation paradigm. EPI gradient-echo sequence: TR = 4s, TE = 54 ms. Flip angle = 90°, voxel size of 1.8 x 1.8 x 4 mm. Brain activation during the verbal fluency condition was compared to the rest period (covert counting) using SPM99, following realignment, spatial normalization and smoothing (8 mm Gaussian filter). Analysis was restricted to region of interest (ROI) encompassing pars opercularis and triangularis in the inferior frontal gyrus bilaterally (the anatomical correlates of Broca's region and its contralateral homologue). MARINA software was used as a tool for creating and manipulating masks in MNI space (Walter et al., 2003). Statistical parametric maps of Z-values were created. Voxels at a threshold of p< 0.001 uncorrected, were displayed as activated. The lateralization index (LI) was then calculated by subtracting the total number of active voxels within ROI in the left from right hemisphere, and dividing the difference by the sum of the activated voxels in both hemispheres. Thus, LI of 1 correspond to complete left hemisphere lateralization and LI of -1 indicate complete right hemisphere lateralization within both subregions of the inferior frontal gyrus. For group analysis, the contrast images showing activations within the given ROI of all subjects entered into a second level analysis in SPM99, with a threshold at p<0.001, uncorrected for multiple comparisons (the voxel clusters surviving thresholds smaller than 5 voxels were excluded). Coordinates of activation in ROI were converted from MNI toTalairach andTournoux (1988) coordinates using the MATLAB software function mni2tal. ! ! ! ! ! ! ! ! ! ! ! ! ! Monozygotic twins Discordant Unaffected Mean(S.D.) Mean (S.D.) Mean (S.D.) Mean (S.D.) Subjects n=4 n=4 n=8 n=8 Gender male/female 2/2 2/2 4/4 4/4 Age 32.7(11.8) 30.3 (6.5) 28.5 (3.8) Handedness 4 right-handed 4 right-handed 8 right-handed 8 right-handed Education (years) 13.2 (2.2) 12.2 (0.4) 10.0 (1.5) 13.5 (1.5) Duration of illness (years) 8.1 (5.3) 8.1 (4.3) PANSS (P) 11.5 (1.8) 12.2 (3.9) PANSS (N) 20.0 (10.6) 25.4 (5.5) PANSS (G) 33.5 (11.9) 44.0 (13.2) PANSS (total) 65.0 (23.8) 81.6 (21.5) Current antipsychotic typicals 7.0 (1.0) 15.2 (3.3) Total antipsychotic exposure (years) 7.7 (4.4) 10.8 (5.3) Discordant Affected Concordant (“Sporadic group”) (“Familial group”) Controls exposure (years) atypicals 2.1 (0.4) 5.7 (0.2) Tab 1: Clinical description of the groups of schizophrenic patients and controls R L R L Fig. 1: Areas within pars opercularis and triangularis g. frontalis inferior bilat. showing higher BOLD fMRI signal during verbal fluency task paradigm in patients with schizophrenia (concordant, i.e. “Familial” group + affected from the discordant pairs, i.e. “Sporadic” group) compared to healthy subjects (controls + unaffected twins from the discordant MZ twin pairs). Subjects with schizophrenia showed increased activation in right BA 47 (Talairach coordinates:53, 25, -3) z score: 3.7 and right BA45 (55, 11, 18) z score: 3.24 in comparisonto healthy subjects. Two- sample t-test, p < 0.001, uncorrected, minimal number of activated voxels k=5. RESULTS No statistically significant differences in task-related fMRI activation could be found in the direct statistical comparison between “Sporadic” and “Familial” groups using a second-level analysis. Neither differences were found in the lateralization index in the ROI during the verbal fluency task paradigm between those two groups. The language lateralization index for activation in given ROI showed a significant reduction in patients with schizophrenia (concordant, i.e. “Familial” group + affected from the discordant pairs, i.e. “Sporadic” group, n=12) compared to all healthy subjects (controls + unaffected twins from the discordant MZ twin pairs, n=12): patients 0.27 ± 0.53, healthy subjects 0.86 ± 0.18 (Wilcoxon-Mann-Whitney Test, two sided, p < 0.002). According to the second-level analysis, subjects with schizophrenia showed increased activation in right BA 47 (Talairach coordinates: 53, 25, -3) z score: 3.7 and right BA 45 (55, 11, 18) z score: 3.24 compared to healthy subjects (two-sample t-test, p < 0.001, uncorrected, minimal number of activated voxels k=5, ). Fig 1 -1 -0,8 -0,6 -0,4 -0,2 0 0,2 0,4 0,6 0,8 1 Discordant Concordant Controls Lateralization index Healthy Schizophrenic Fig. 2: Language lateralization indexes in 12 pairs of right-handed monozygotic twins - discordant, concordant for schizophrenia and controls. FMRI during verbal fluency task paradigm, activations within pars opercularis and triangularis in the inferior frontal gyrus bilaterally (Broca´s area and its proximity together with contralateral homologue).Mean lateralization idnex was in discordant group, in concodrant group and in controls, respectively. 0.81±0.17 0.11±0.68 0.96±0.04 How identical is functional asymmetry of the brain in MZ twins? Apparent intra-pair, between co-twin differences in the lateralization index were observed in the discordant group, but in contrast to the expectation especially in the concordant group (Fig.2). To analyze this, the intra-pair difference in the lateralization index was obtained for each twin pair, resulting in 4 difference scores in the discordant, concordant (i.e. “Familial”) and control groups. Twins from the concordant (“Familial”) group showed significantly higher intra-pair differences compared to the controls (0.69 ± 0.4 vs. 0.13 ± 0.13, p<0.03, Mann-Whitney U rank sum test, two sided). However, differences in the variable did not reach statistical significance between the discordant and control groups. DISCUSSION Pooled data from all schizophrenic patients revealed a significant decrease in the language lateralization index in the ROI compared to healthy individuals. This finding corresponds to a previous fMRI study showing an increase in language- related activity within the right hemisphere in patients with schizophrenia (Sommer et al., 2001). Absence of any significant differences in task-related fMRI activation between “Sporadic” and “Familial” groups could result from an extreme intra-group variability in the “Familials” (i.e. concordant twins with high familial loading of schizophrenia) with considerable intra-pair differences in activation patterns between the co-twins. The intra-pair differences in the lateralization index were significantly higher in the “Familial” group compared to controls, whereas no significant differences in the variable were observed between the discordant and control groups (Fig. 2). This particular issue warrants further consideration. There is an evidence, that under normal conditions, handedness- concordant MZ twins show high intra-pair resemblance in language lateralization (Sommer et al., 2002). Prominent intra-pair variability in the lateralization indexes in the monozygotic “Familial” group could be explained by a right-shift genetic model of handedness and brain laterality proposed by Annett (Annett, 1985). According to the predictions of the model, identical handedness accompanied by discordance in cerebral asymmetry could be theoretically found in MZ twin pairs homozygous for putative inactive form of allele that determines hemispheric dominance (RS-/RS-). Cerebral dominance is influenced by stochastic factors in those gene carriers (Annett, 2003). RS- homozygozity may theoretically account for an extreme intra-pair and thus intra-group variability in our “Familial” group of right-handed MZ twins concordant for schizophrenia. Whether cases with extremely high genetic risk for schizophrenia are also prevalent carriers of the RS-/RS- homozygosity resulting in deviations in both functional and structural asymmetry of the brain, remains elusive and warrants further investigation. Findings of structural and functional brain asymmetry in familial patients and genetically vulnerable relatives (Falkai et al., 2002, Sharma et al., 1999, Malaspina et al., 1998, Honer et al., 1995) and absence of such reports in sporadic cases provide partial support for this conceptual framework. SUMMARY The current study, as well as several other available studies, provides evidence of reduced cerebral dominance for language processing in patients with schizophrenia. The preliminary findings of this pilot study further suggest the link between deviations of functional asymmetry in schizophrenia and genetic susceptibility resulting in random specification of cerebral lateralization. Acknowledgments: This work was supported by project MSMT 1M0002375201, Czech Republic. REFERENCES Annett M (1985) Left, Right, Hand and Brain: The Right Shift Theory Lawrence Erlbaum, London, pp. 291–292 Falkai P, Honer WG,Alfter D, Schneider-Axmann T, Bussfeld P, Cordes J, Blank B, Schonell H, Steinmetz H, Maier W, Tepest R. (2002) The temporal lobe in schizophrenia from uni- and multiply affected familiesNeurosci Lett. 31;325(1):25–8. Honer WG, Bassett AS, Squires-Wheeler E, Falkai P, Smith GN, Lapointe JS, Canero C, Lang DJ. (1995) The temporal lobes, reversed asymmetry and the genetics of schizophrenia. Neuroreport. 29;7(1):221–4. Malaspina D, Goetz RR, Yale S, Berman A, Friedman JH, Tremeau F, Printz D, Amador X, Johnson J, Brown A, Gorman JM.(2000). Relation of familial schizophrenia to negative symptoms but not to the deficit syndrome. Am J Psychiatry. 157(6):994–1003. Oldfield RC (1971) The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9: 97–113. Sharma T, Lancaster E, Sigmundsson T, Lewis S, Takei N, Gurling H, Barta P, Pearlson G, Murray R. (1999) Lack of normal pattern of cerebral asymmetry in familial schizophrenic patients and their relatives The Maudsley Family Study. Schizophr Res.40:111–20. Sommer I , Ramsey N, Kahn R (2001) Language lateralization in schizophrenia, an fMRI study. Schizophr Res 52: 57–67. Sommer IE, Ramsey NF, Mandl RC, Kahn RS. (2002).Language lateralization in monozygotic twin pairs concordant and discordant for handedness. Brain 125:2710–8. Thomson JA, Ayres KL, PilottiV, Barrett MN, Walker JI, Debenham P (2001) Analysis of disputed single-parent/child and sibling relationships using 16 STR loci. Int J Legal Med 115: 128–134. Walter B, Blecker C, Kirsch P, Sammer G, Schienle A , Stark R, Vaitl, D (2003) An easy to use tool for the creation of masks for region of interest analyses. NeuroImage 19: 2.