Involvement of hyperprolinemia in cognitive and psychiatric features of the 22q11 deletion syndrome

Transcription

1 Human Molecular Genetics, 2007, Vol. 16, No doi: /hmg/ddl443 Advance Access published on November 29, 2006 Involvement of hyperprolinemia in cognitive and psychiatric features of the 22q11 deletion syndrome Grégory Raux 1, Emilie Bumsel 1, Bernadette Hecketsweiler 2, Therese van Amelsvoort 3, Janneke Zinkstok 3, Sylvie Manouvrier-Hanu 4, Carole Fantini 4, Georges-Marie M. Brévière 5, Gabriella Di Rosa 6, Giuseppina Pustorino 6, Annick Vogels 7, Ann Swillen 7, Solenn Legallic 1, Jacqueline Bou 1, Gaelle Opolczynski 8, Valérie Drouin-Garraud 1, Marie Lemarchand 1, Nicole Philip 9, Aude Gérard-Desplanches 10, Michèle Carlier 11, Anne Philippe 12, Marie Christine Nolen 12, Delphine Heron 13, Pierre Sarda 14, Didier Lacombe 15, Cyril Coizet 15, Yves Alembik 16, Valérie Layet 17, Alexandra Afenjar 18, Didier Hannequin 1, Caroline Demily 1,8, Michel Petit 1,8, Florence Thibaut 1,8, Thierry Frebourg 1 and Dominique Campion 1,8, * 1 Department of Genetics and Inserm U614, IFRMP, Faculty of Medicine, Rouen, France, 2 Laboratory of Biochemistry, CHRU, Rouen, France, 3 Department of Psychiatry, AMC, Amsterdam, The Netherlands, 4 Department of Genetics and 5 Department of Cardio-Pediatrics, CHRU, Lille, France, 6 Department of Medical and Surgical Pediatrics, Unit of Infantile Neuropsychiatry, University Hospital, Messina, Italy, 7 Center for Human Genetics, University Hospital, Leuven, Belgium, 8 Units of Psychiatry, CH du Rouvray and CHRU, Rouen, France, 9 Department of Genetics, CHRU Marseille, France, 10 Research Center PsyCLE (EA3273) and 11 Cognitive Psychology UMR 6146 CNRS, Aix Marseille 1 University, France, 12 INSERM U781 and Department of Genetics, Hôpital Necker-enfants malades, Paris, France, 13 Department of Genetics, Hôpital Pitié Salpétrière, Paris, France, 14 Department of Genetics, CHRU Montpellier, France, 15 Department of Genetics, CHRU Bordeaux, France, 16 Department of Genetics, CHRU Strasbourg, France, 17 Department of Genetics, CHR Le Havre, France and 18 Department of Neuropediatry, Hopital Trousseau, Paris, France Received September 28, 2006; Revised and Accepted November 16, 2006 Microdeletions of the 22q11 region, responsible for the velo-cardio-facial syndrome (VCFS), are associated with an increased risk for psychosis and mental retardation. Recently, it has been shown in a hyperprolinemic mouse model that an interaction between two genes localized in the hemideleted region, proline dehydrogenase (PRODH ) and catechol-o-methyl-transferase (COMT ), could be involved in this phenotype. Here, we further characterize in eight children the molecular basis of type I hyperprolinemia (HPI), a recessive disorder resulting from reduced activity of proline dehydrogenase (POX). We show that these patients present with mental retardation, epilepsy and, in some cases, psychiatric features. We next report that, among 92 adult or adolescent VCFS subjects, a subset of patients with severe hyperprolinemia has a phenotype distinguishable from that of other VCFS patients and reminiscent of HPI. Forward stepwise multiple regression analysis selected hyperprolinemia, psychosis and COMT genotype as independent variables influencing IQ in the whole VCFS sample. An inverse correlation between plasma proline level and IQ was found. In addition, as predicted from the mouse model, hyperprolinemic VCFS subjects bearing the Met-COMT low activity allele are at risk for psychosis (OR 5 2.8, 95% CI ). Finally, from the extensive analysis of the PRODH gene coding sequence variations, it is predicted that POX residual activity in the 0 30% range results into HPI, whereas residual activity in the 30 50% range is associated either with normal plasma proline levels or with mild-to-moderate hyperprolinemia. *To whom correspondence should be addressed at: Inserm U614, Faculté de Médecine, 22 bd Gambetta, Rouen, France. Tel: ; Fax: ; dominique.campion@univ-rouen.fr # The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 84 Human Molecular Genetics, 2007, Vol. 16, No. 1 INTRODUCTION Type I hyperprolinemia (HPI MIM ), an autosomal recessive inborn error, results from inherited deficiency of proline oxidase (POX), a mitochondrial enzyme expressed in kidney, liver and brain, which is encoded by the proline dehydrogenase (PRODH ) gene (1). Clinically, HPI is recognized by elevation of the plasma proline value above 550 mmol/l without excretion of the degradation product of proline, D1-Pyrroline-5-carboxylate (P5C). This condition was initially considered as benign (1). Nevertheless, we had previously reported evidence that HPI, resulting either from a homozygous 350 kb deletion removing the entire PRODH gene and the neighboring DGCR6 gene, or from a PRODH homozygous L441P missense mutation, was associated with seizures, mental retardation and psychiatric symptoms in children (2 3). In addition, we had shown that mild-to-moderate hyperprolinemia (i.e. plasma proline levels between mmol/l) was a weak risk factor for schizoaffective disorder (4). The PRODH gene, which is located in the 22q11 chromosomal region, is hemideleted in 22q11 deletion syndrome (22q11DS MIM ) also known as velo-cardio-facial syndrome (VCFS). It has been reported that 50% of the 22q11DS patients are hyperprolinemic (5), whereas 30 45% of them present with mental retardation (6 7) and 12 30% have psychosis (8 10). The Pro/Re mouse hyperprolinemic strain (11) harbors a mutation in the ortholog of the human PRODH gene that introduces a premature termination (E453X) and results in a drastic reduction in enzymatic activity. These knockdown mice have regional neurochemical alterations in the brain accompanied by a deficit in sensorimotor gating, similar to that seen in individuals with schizophrenia and other neuropsychiatric disorders (12). The catechol-o-methyl transferase (COMT ) gene, which encodes for the enzyme responsible for the degradation of dopamine (DA) in the prefrontal cortex (PFC), is also located in the 22q11 region. Recently, transcriptional and behavioral interaction between PRODH and COMT have been documented (13). Expression profiling with microarrays revealed that PRODH-deficient mice had selective upregulation of COMT in the PFC, which is likely to represent a homeostatic response to enhanced dopaminergic signaling induced by PRODH deficiency. According to this model, hyperprolinemic 22q11 individuals with low COMT activity are those who are the less efficient to compensate the dopaminergic hyperactivity and, therefore, are those who might have the higher risk to develop PFC-mediated cognitive and/or psychiatric phenotype. Interestingly, the COMT gene contains in its coding sequence a functional polymorphism (p.val158met) that determines high or low activity of this enzyme (14). The present study was undertaken to characterize the relationship between PRODH gene variations, hyperprolinemia, cognitive and psychiatric symptoms in HPI and VCFS subjects. More specifically, our goal was to test the hypothesis that the high neuropsychiatric disease risk associated with the 22q11 locus results from an interaction between the PRODH and COMT genes. Table 1. Phenotype and genotype of HPI patients Id Age (years) Phenotype Prolinemia (mmol/l) (range) PRODH Genotype a AF, seizures del/del MR, AF, seizures del/r453c þ V427M þ Q19P þ A58T MR, AF L441P/Stop c AF L441P/R431H þ G444D MR, seizures P406L þ R431H/R431H seizures R453C þ V427M þ T275N/Q19P MR, seizures R453C/T275N MR, seizures Q19P þ A58T/Q19P MR, mental retardation; AF, autistic features. Mutations with severe effect on PRODH activity appear in bold, mutations with moderate effect in normal character and mutations with unknown effect in italic. a Blood samples were also collected from parents in order to establish unambiguously the phase of DNA variations in probands. RESULTS HPI is associated with epilepsy, mental retardation and psychosis Table 1 presents the clinical features of eight children with HPI, as determined by plasma proline values above 550 mmol/l without P5C excretion. Mental retardation was a frequent feature (5/8), generalized tonic-clonic seizures were also present in 6/8 patients and autistic features including stereotypies, lack of responsiveness to people and deficits in language development were found in 4/8 patients. Three patients had psychomotor delay since birth, in other patients mental retardation appeared during early childhood. The VCFS sample In the VCFS sample, 2/92 subjects were unwilling to participate to the IQ assessment. Sixty-two subjects (69%) had mental retardation (IQ,70) including 23 subjects (25%) with an IQ, 55. Thirty-three subjects (36%) were rated as psychotic (VCFS-P) and 59 as non-psychotic (VCFS-NP). The mean full scale IQ was (VIQ ¼ , PIQ ¼ ). For each individual, significant difference between VIQ and PIQ was determined, taking age into account, at the 0.05 level. Twenty-two subjects performed significantly better to VIQ than PIQ and 14 subjects had significantly better performance to PIQ. In this sample of 92 VCFS patients, plasma proline levels ranged from 138 to 1275 mmol/l. Using the thresholds previously defined for the adult population (4) (i.e. 316 mmol/l in females and 377 mmol/l in males), we found that 34 subjects (37%) were hyperprolinemic. A subset of VCFS patients has a phenotype reminiscent of HPI subjects Plasma proline value was in the range of HPI (i.e..550 mmol/l) in seven subjects (Table 2). These subjects differed from the rest of the sample for Full Scale IQ (51+7

3 Human Molecular Genetics, 2007, Vol. 16, No Table 2. Phenotype and genotype of 92 VCFS patients according to plasma proline levels Id Age (years) Sex Prolinemia (mmol/l) VIQ PIQ IQ Psychiatric diagnosis Epilepsy COMT Genotype PRODH Genotype M A- PS Val R431H M SZ (D) þ Met R431H F SZ (P) þ Met R431H F Val R431H F SZ (P) Val Q19P-P30S M Autism þ Val Q19P-A58T M SZT þ Met T275N M Met R453C- R185W M Met Q19P- R185W M Met Q19P- R185W M Met R431H M Val R431H M Val R431H-Q19P-R185W M Met Q19P-P30S F 454 Met Q19P- R185W M Val Q19P-P8L M Val Q19P- R185W M Met wt F Val wt F SZ (P) Met R431H M Met R431H-T275N M Met Q19P-A58T F SZ (P) Val wt F SZ (P) Met R431H M Autism þ Met R431H-Q19P-P30S M SZT Met Q19P-A58T M Val wt M SZ (P) Met wt F SZ (D) Met Q19P- R185W F SZF þ Met R431H F Val Q19P-A58T M Val Q19P-P30S 8 17 M Val wt M Val wt 3 17 M A- PS Met Q19P-P30S F Met Q19P-P30S M Met R185W F SZF þ Met wt 4 15 M A- PS Met Q19P- R185W M SZ (P) Met Q19P- R185W M SZ (P) Val Q19P- R185W F Met Q19P-A58T M SZ (P) Val wt F 298 SZ (P) þ Val R43IH-T275N M Val Q19P F Met Q19P-P30S F Met Q19P- R185W M Val wt 7 37 F SZ (D) Met V427M-T275N M SZ (P) Met Q19P- R185W 1 15 F Met wt 9 20 F Met Q19P M Val Q19P- R185W M Val wt F Val Q19P 2 16 F A-PS Val Q19P- R185W-R431H M Met wt F Met wt F Val Q19P-P30S M Val Q19P M Val Q19P- R185W F Met Q19P- R185W F Val wt F Met Q19P- R185W M Val Q19P- R185W 5 23 F Met Q19P Continued

4 86 Human Molecular Genetics, 2007, Vol. 16, No. 1 Table 2. Continued Id Age (years) Sex Prolinemia (mmol/l) VIQ PIQ IQ Psychiatric diagnosis Epilepsy COMT Genotype PRODH Genotype M A- PS Val Q19P- R185W F Met wt F SZ (P) Met wt 6 16 F Val Q19P F Val V427M M SZT Val Q19P F Val wt M Val wt F Val Q19P- R185W F SZ (P) Met Q19P-P30S M A- PS Met Q19P-P30S F Met Q19P- R185W F Met Q19P- R185W F Met wt F Val Q19P- R185W F SZF Val wt F Val wt F Val R185W F SZ (P) Met R185W F Met Q19P- R185W F Met wt F Met wt F SZT Val Q19P F SZ (P) Met Q19P F Val Q19P F Val wt Mutations with severe effect appear in bold, mutations with moderate effect in normal character and mutations with unknown effect in italic. Wt, wild-type; IQ, intellectual quotient; VIQ, verbal intellectual quotient; PIQ, performance intellectual quotient; SZ (D), Schizophrenia (disorganized); SZ (P), schizophrenia (paranoid). versus , P ¼ 0.007, Mann Whitney U), prevalence of epileptic seizures not related to hypocalcemia (4/7 versus 4/85 P ¼ , Fisher s exact test) and frequency of psychotic symptoms (6/7 versus 27/85, P ¼ Fisher s exact test). Factors influencing IQ in VCFS patients We next conducted a series of analysis in order to assess the contribution of five variables, namely age, sex, plasma proline level, COMT genotype and psychosis to IQ value. We first examined the interrelationships between these variables. These analyses revealed an interaction between plasma proline level and gender. The mean plasma proline values were significantly higher in males ( mmol/l) than in females ( mmol/l) (P ¼ , Mann Whitney U), as classically reported (4). In subsequent analyses, this factor was controlled because the plasma proline level was used as a categorical variable (i.e. subjects were considered as hyperprolinemic or not) and different thresholds were used to define hyperprolinemia in males and females. The first factor that influenced IQ was psychosis: the mean full scale IQ in non-psychotic patients was when compared with in psychotic patients (P ¼ 0.01, Mann Whitney U). Detailed analysis of subscales was realized in 80 subjects and showed that VCFS-P patients performed significantly worse on Digit Symbol Coding, Arithmetics, Similarities, Digit Span and Picture Arrangement, when compared with VCFS-NP subjects (Table 3). The second factor that influenced IQ was hyperprolinemia: the mean full scale IQ in hyperprolinemic patients was when compared with in patients with normal plasma proline levels (P ¼ 0.01, Mann Whitney U). IQ was inversely correlated to the proline level (r ¼ 20.24, P ¼ 0.02, Spearman s test). There was also a tendency toward higher full scale IQ in Met-COMT hemizygous patients, which remained below statistical significance (P ¼ 0.07, Mann Whitney). Forward stepwise multiple regression analysis selected hyperprolinemia, psychosis and COMT genotype as independent variables influencing IQ. Sex and age did not appear as independent predictors of IQ. The regression coefficients were: hyperprolinemia; 2 7, 35 (+2, 8), P, 0.011; psychosis; 2 7, 51 (+2, 88), P, 0.01; COMT genotype; 6, 33 (+2, 7), P, When hyperprolinemia and COMT genotype were considered jointly, no significant effect on IQ was detected (Table 3). Likewise, there was no significant effect of hyperprolinemia and Met-COMT allele, either independently or conjointly, on the Digit Span and Arithmetics tasks (two subscales of WAIS relying on PFC mediated working memory performances) (Table 3). Hyperprolinemia associated with Met allele of COMT gene is a risk factor for psychosis in VCFS patients Among the 33 VCFS-P subjects, 18 satisfied DSM IIIR criteria for schizophrenia (paranoid n ¼ 15, disorganized n ¼ 3), three had schizophreniform (SZF) disorder, six had atypical

5 Table 3. Scores on cognitive measures of 90 VCFS patients VCFS-P (n ¼ 32) VCFS-NP (n ¼ 58) ANOVA (df ¼ 1, 88) VCFS HyperprolinemiaþMet-COMT allele (n ¼ 21) VCFS Others (n ¼ 69) ANOVA (df ¼ 1, 88) Measure Mean SD Mean SD F P-value Mean SD Mean SD F P-value Full scale IQ PIQ VIQ Picture completion (n ¼ 23) (n ¼ 57) ANOVA (df ¼ 1, 78) (n ¼ 16) (n ¼ 64) ANOVA (df ¼ 1, 78) Vocabulary (n ¼ 22) (n ¼ 46) ANOVA (df ¼ 1, 66) (n ¼ 14) (n ¼ 54) ANOVA (df ¼ 1, 66) Digit symbol coding (n ¼ 23) (n ¼ 57) ANOVA (df ¼ 1, 78) (n ¼ 16) (n ¼ 64) ANOVA (df ¼ 1, 78) Similarities (n ¼ 23) (n ¼ 57) ANOVA (df ¼ 1, 78) (n ¼ 16) (n ¼ 64) ANOVA (df ¼ 1, 78) Block design (n ¼ 23) (n ¼ 57) ANOVA (df ¼ 1, 78) (n ¼ 16) (n ¼ 64) ANOVA (df ¼ 1, 78) Arithmetic (n ¼ 23) (n ¼ 57) ANOVA (df ¼ 1, 78) (n ¼ 16) (n ¼ 64) ANOVA (df ¼ 1, 78) Digit span (n ¼ 20) (n ¼ 49) ANOVA (df ¼ 1, 67) (n ¼ 13) (n ¼ 56) ANOVA (df ¼ 1, 67) Information (n ¼ 23) (n ¼ 57) ANOVA (df ¼ 1, 78) (n ¼ 16) (N ¼ 64) ANOVA (df ¼ 1, 78) Picture arrangement (n ¼ 22) (n ¼ 47) ANOVA (df ¼ 1, 67) (n ¼ 14) (n ¼ 55) ANOVA (df ¼ 1, 67) Comprehension (n ¼ 22) (n ¼ 47) ANOVA (df ¼ 1, 67) (n ¼ 14) (n ¼ 55) ANOVA (df ¼ 1, 67) Human Molecular Genetics, 2007, Vol. 16, No Downloaded from at CHRU de LILLE on December 17, 2012

6 88 Human Molecular Genetics, 2007, Vol. 16, No. 1 Table 4. Distribution of hyperprolinemiaþmet-comt allele in VCFS-P and VCFS-NP patients HyperprolinemiaþMet-COMT allele VCFS-P (n ¼ 33) 12 (36%) 21 VCFS-NP (n ¼ 59) 10 (17%) 49 x 2, 1 ddl ¼ 4.38, P ¼ OR ¼ 2.8 (95% CI ¼ ). Others psychosis (A-PS), four had schizotypal (SZT) personality and two had autism. Mean age of onset for schizophrenia was years (range 12 27). Among schizophrenic patients, the most prominent symptoms were ideas of reference, paranoid ideation and auditory hallucinations. Atypical psychosis was diagnosed in six subjects who reported auditory hallucinations that did not significantly interfere with social functioning. Mean age of onset for atypical psychosis was years (range 14 16). Two subjects who received a diagnosis of autism were severely impaired and failed to acquire any language. Neither hyperprolinemia nor the Met-COMT allele was associated with psychosis, although a trend was observed in both cases. However, when these two factors were considered jointly, hyperprolinemic subjects bearing the low activity Met-COMT allele were at risk for psychosis (P ¼ 0.03, OR ¼ 2.8, 95% CI ) (Table 4). Hyperprolinemia and PRODH genotype A detailed account of phenotypic and genotypic characteristics for each HPI and VCFS subject is given in Tables 1 and 3. In all HPI and VCFS subjects, we searched for PRODH deletions, using QMPSF, and we sequenced the coding sequence of PRODH. The 350 kb deletion removing the PRODH gene was detected in two HPI patients and was present at the homozygous state in one patient. According to the in vitro functional data of Bender et al. (15), PRODH missense mutations were divided in two groups with moderate (30 70%) or severe (.70%) reduction in POX activity, respectively. For the sake of simplicity, mutations with no or mild effect, as well as silent mutations, were not included in Tables 1 and 3. Several unreported DNA variations with unknown functional consequences (P8L, P30S, A58T, T275N and G444D) were also found during this screening. In 6/8 HPI patients (Table 1), missense mutations with severe effect were found in association with a deletion, a stop codon or a mutation with moderate effect on the second allele. The second allele bore a mutation with unknown functional consequence (T275N) in one HPI patient. The HPI subject with the lowest plasma proline level was homozygous for a mutation (Q19P) with moderate effect, but also harbored in one allele a mutation (A58T) with unknown functional consequence. In the VCFS sample, in which all subjects were hemizygous for the PRODH gene, only one mutation with severe effect was detected in one hyperprolinemic patient (Table 3). Mutations with a moderate effect were detected in 66/92 VCFS subjects (72%), either with or without hyperprolinemia. These mutations were equally distributed between subjects with and without hyperprolinemia, with the exception of the R431H variation that was more frequent in hyperprolinemic subjects (P ¼ , Fisher s exact test). Four-to-seven VCFS subjects with the highest plasma proline levels and two HPI subjects bore this variation, indicating that besides its moderate effect on catalytic activity, it could have a more drastic effect on protein structure. DISCUSSION Despite the well-known association of type II hyperprolinemia with mental retardation (1), the impact of less severe increase in plasma proline levels on cognitive functions and psychiatric illness has largely escaped attention. Here, we provide further evidence that HP1 is indeed a cause of mental retardation associated with neurological and, in some cases, psychiatric features in children. To better understand the consequences of hyperprolinemia, we have constituted a sample of 92 VCFS subjects, including 33 patients with psychosis. This high proportion is not representative from that found in a randomly ascertained population, since we actively searched for psychotic patients, in order to achieve maximum statistical efficiency. Likewise, a higher proportion of subjects with mental retardation were included in this study, when compared with the proportion found in randomly ascertained populations (6 7). We show that a small subset of VCFS subjects had severe hyperprolinemia in the range of HPI, and that the phenotype of these subjects was distinguishable from that of other VCFS subjects. Notably, the mean IQ of these patients was significantly lower than that of other VCFS subjects, and the prevalence of psychosis was high. Epilepsy not related to hypocalcemia was frequent in patients with severe hyperprolinemia, although it is an uncommon feature in VCFS patients (16). Interestingly, comorbidity between seizures and psychosis has recently been reported in two VCFS subjects, one of them with an IQ of 46 (17). Overall, these findings have important potential therapeutic consequences for HPI as well as VCFS subjects with hyperprolinemia, since it is conceivable to lower plasma proline levels by reduction of proline intake (18) or by pharmacological treatment early during early childhood in these patients. Increased plasma proline level, including both severe and mild/moderate hyperprolinemia, was not significantly associated with psychosis in the whole VCFS sample. Likewise, in agreement with a previous study of adult psychotic VCFS patients (8) and in contrast with a recent report (19), no association between COMT genotype and psychosis was found in the present study. However, as predicted from the PRODH deficient mouse model, individuals with both hyperprolinemia and the Met-COMT allele were at risk for a broad spectrum of psychotic diagnoses, including schizophrenia. Consistent with a previous report (20), schizophrenic VCFS subjects experienced more often positive than negative symptoms during lifetime. Auditory hallucinations were also a key symptom in patients with SZF disorder and atypical psychosis suggesting that, despite different categorical diagnoses, most VCFS-P subjects shared prominent positive symptoms. Another notable feature in schizophrenic patients was the early age of onset, which is in part inherent to the age structure of our sample, but could also reflect the strong association

7 Human Molecular Genetics, 2007, Vol. 16, No between childhood-onset schizophrenia and VCFS (21). Recently, two studies have reported that a high proportion of VCFS adolescents experienced transient psychotic episodes that could be predictive of schizophrenia in adulthood (22 23). The symptoms of these patients are similar to those found in subjects diagnosed as atypical psychosis or SZT personality in the present study. Stepwise multiple regression analysis indicated that three factors, psychosis, hyperprolinemia and COMT genotype had an independent effect on IQ in the VCFS sample. A similar trend toward a relationship between psychosis and lower IQ in VCFS subjects has previously been reported in three studies (8,22,24). Although relying on a less complete battery of tests, our data are also consistent with the results obtained by Chow et al. (24) and van Amelsvoort et al. (25), showing that adult VCFS-P subjects performed significantly worse on a large number of cognitive tasks, when compared with other VCFS subjects. In the present study, increased plasma proline level, including both severe and mild/moderate hyperprolinemia, was the second factor associated with lower mean full scale IQ in VCFS subjects. Moreover, an inverse correlation between plasma proline level and IQ was found. Although it is recognized that a high level of proline has a detrimental effect on neuronal integrity (26) and is associated with a pattern of global cognitive impairment in HPI and HPII, there is, to our knowledge, no previous study addressing the involvement of mild-to-moderate hyperprolinemia in cognitive impairment in humans. The third factor influencing IQ was the COMT genotype. In contrast to the lack of proline studies, the impact of COMT genotype on IQ has previously been investigated in several studies, that yielded conflicting results. In VCFS children, the Met-COMT low activity allele has been associated either with higher verbal and full scale IQ (27), as in the present study, or inversely with a tendency toward a decline in verbal (19) and full scale IQ (28). The PRODH deficient mouse model predicts a specific impairment of PFC-mediated cognitive functions, related to an excess of dopaminergic transmission, in VCFS patients with both hyperprolinemia and Met-COMT genotype. The neuropsychological evaluation performed in this study does not allow to address properly this question. Indeed, IQ accounts for a very modest proportion of the variance in executive functions (29) and the subtests of WAIS are lacking specificity to adequately capture working memory deficits. Similarly, literature data do not deal efficiently with this issue. It should be stressed that, although the impact of COMT genotype on PFC cognition has been investigated in numerous studies, hyperprolinemia and COMT genotype have never been studied together, thus limiting the value of these reports. In the general population or in schizophrenic patients, most studies (30 34) have concluded that the Met-COMT low activity allele was associated with better prefrontal functioning, although a large study failed to detect any effect (35) and another concluded that the main effect was on processing speed and attention (36). In VCFS subjects, two studies (28 29) have reported better PFC cognition in Met-COMT hemizygotes, whereas one study only found a trend (37) and another one reported no effect (38). In contrast, in a recent study, it was reported that the Met-COMT allele was a risk factor for decline in PFC volume and cognition (19). Thus, although most studies have concluded that low COMT activity (i.e. increased level of DA transmission) was associated with fewer perseverative errors on frontal tests, some discrepancies persist and the problem is probably more complex than initially thought. There is now evidence that either a reduction (39) or an excess (such as in PRODH-deficient mice) of DA signalling have detrimental effects on PFC cognition and that different neuronal process (i.e. phasic and tonic DA release) affect the capacity to built efficiently working memory representations (40 42). Therefore, future studies aimed to delineate the pattern of cognitive impairment in VCFS subjects with both hyperprolinemia and the Met-COMT allele, should rely on more sensitive tasks (43 44) allowing to discriminate between these different aspects of dopaminergic transmission in the PFC. In the present study, we have progressed in our attempt to establish a link between PRODH genotypes and plasma proline levels, but the picture remains incomplete, since (i) functional data are still lacking for some individual variations or some haplotypes with multiple mutations in cis and (ii) regulatory regions have not been examined. Nevertheless, on the basis of available data, our working hypothesis is that predicted POX residual activity in the 0 30% range results into HPI, whereas residual activity in the 30 50% range is associated either with normal plasma proline level or with mild-to-moderate hyperprolinemia. Finally, two limitations of our study have to be underlined. First, few studies have addressed the question of the reproducibility of plasma proline determination. We had previously assessed this issue in 25 controls or mildly hyperprolinemic subjects. The maximum variation between two independent measures was 12% and the mean variation was 8% (unpublished data). In contrast, for an unknown reason, there are often considerable variations in plasma proline levels in HPI patients over time, but the mean value remains generally above 550 mmol/l. Secondly, it should be stressed that 40 VCFS-NP subjects had less than 30 years of age. Therefore, it is conceivable that some of these subjects may yet develop psychotic symptoms. To avoid the risk of such misclassification, future studies should rely on older VCFS-NP populations that have escaped to a larger part of the risk period for schizophrenia. Owing to the scarcity of such subjects, this will require larger collaborative studies. METHOD Subjects Eight children with HP1, ascertained in France and Italy from medical genetics centers, and 92 VCFS participants, aged 15 and older (46% males, mean age years, range 15 58) were included in this study. Eighty-one VCFS subjects were recruited from nine medical centers from France, Belgium and in The Netherlands, five were ascertained from psychiatric sources and six via the French association Generation 22. In all patients, the diagnosis of VCFS was confirmed by fluorescence in situ hybridization and Quantitative Multiplex PCR of Short Fragments (2). Information about behavioral, somatic, neurological, pharmacological and psy-

8 90 Human Molecular Genetics, 2007, Vol. 16, No. 1 chiatric history were collected. Consensus diagnoses for psychiatric cases were established by two raters according to DSM IIIR criteria from all available information following examination of case notes and in most cases, direct interview of patients with the appropriate sections of the Schedule for Affective Disorder and Schizophrenia (45). In addition, the Schedule for Schizotypal Personality (SSP) (46) was administered when necessary. Clinical data were compiled into case vignettes and diagnoses were made blind of proline data. Intellectual functioning was assessed by trained psychologists using the Wechsler Intelligence Scale for Children 3rd edition (WISC III) for subjects younger than age 17 and either Wechsler Adult intelligence Scale-Revised (WAIS-R) or Wechsler Adult intelligence Scale 3rd edition (WAIS III) for older subjects. The proportion of subjects assessed with WAIS-R was not significantly different between VCFS-P and VCFS-NP subjects. All VCFS-P subjects were assessed during a stable or remitted phase of their psychotic illness. At the time of assessment, 12 VCFS-P subjects received atypical neuroleptic drugs, 11 received conventional neuroleptics and six had anticholinergic medications. Seven epileptic subjects were treated by valproate. Informed consent was obtained from all participants and the Local Research Ethics Committee approved the study protocol. Determination of plasma proline levels In VCFS patients, plasma proline levels were determined in the morning after overnight fasting. All samples were analyzed by the same laboratory using ion exchange chromatography on a BIOTRONIK LC 3000 system. As previously reported (4), subjects whose plasma proline levels were two standard deviations above the controls mean values (i.e. 316 mmol/l in females and 377 mmol/l in males) were considered to be hyperprolinemic. Quantitative multiplex PCR of short fluorescent fragments Ten exons (1, 2, 4, 5, 9, 10, 11, 12, 14, 15) of the PRODH gene were analyzed by quantitative multiplex PCR of short fluorescent fragments (QMPSF) in all subjects. When a rearrangement involving the PRODH gene was detected, four different QMPSF assays covering the entire 22q11.2 region were performed, as previously described (2). Sequence analysis Exons 2 15 of PRODH were PCR-amplified from peripheral blood lymphocytes. PCR conditions and primers were identical to those previously described (4), except the sense primer of exon 12 (5 0 -TTGGGGACAGAGGTTGGAGGC-3 0 ) and exon 11 sense (5 0 -GCTCGGGCAGAGGGTACC-3 0 ) and antisense (5 0 -CCGCTGGGTGAGCTGTGCTG-3 0 ) primers. After purification by low-melt agarose gel electrophoresis, PCR products were directly sequenced on both strands using the ABI PRISM BigDye Terminator v3.1 Cycle Sequencing kit (PE Applied Biosystems, Foster city, CA, USA) and an Applied Biosystems model 3100 automated sequencer (PE Applied Biosystems). Screening for specific DNA variation The Val158Met polymorphism in the COMT gene was detected using the ABI Prism SNaPshot multiplex kit. Statistical analyses Quantitative data were expressed as mean + SD. Comparisons for continuously distributed variables were performed using the Mann Whitney non-parametric test. Statistical comparisons for categorical variables were performed using x 2 test or Fisher s exact test, when the cells size was small. Correlations between quantitative variables were assessed by the Spearman s test. All tests reported were two-tailed and significance level was set to 5%. OR are given with their 95% CI. One-way analysis of variance (ANOVA) was used to detect group differences for neuropsychological measures. To determine the relative contribution of five variables (age, sex, hyperprolinemia, psychosis and COMT genotype) to the full scale IQ, we submitted the data to a forward stepwise multiple regression analysis, using a selection criterion of P, ACKNOWLEDGEMENTS Supported by a grant from Programme Hospitalier de Recherche Clinique. A.G.-D. held a Jerôme Lejeune Foundation fellowship. We are indebted to our colleagues S. Lebreton, O. Boespflug-Tanguy, C. Francannet, M. Lacambre, M.O. Krebs, N. Bazin, who referred patients for this study and to J. Benichou and J.F. Ménard for statistical advice. We would like to thank the patients who took part in this study and the support of the French association of parents Génération 22. Conflict of Interest statement. None declared. REFERENCES 1. Phang, J., Chien-an, A. and Valle, D. (2001) Disorders of Proline and Hydroxyproline Metabolism, 8th edn. McGraw-Hill, New York. 2. Jacquet, H., Raux, G., Thibaut, F., Hecketsweiler, B., Houy, E., Demilly, C., Haouzir, S., Allio, G., Fouldrin, G., Drouin, V. et al. (2002) PRODH mutations and hyperprolinemia in a subset of schizophrenic patients. Hum. Mol. Genet., 11, Jacquet, H., Berthelot, J., Bonnemains, C., Simard, G., Saugier-Veber, P., Raux, G., Campion, D., Bonneau, D. and Frebourg, T. (2003) The severe form of type I hyperprolinaemia results from homozygous inactivation of the PRODH gene. J. Med. Genet., 40, e7. 4. Jacquet, H., Demily, C., Houy, E., Hecketsweiler, B., Bou, J., Raux, G., Lerond, J., Allio, G., Haouzir, S., Tillaux, A. et al. (2005) Hyperprolinemia is a risk factor for schizoaffective disorder. Mol. Psychiatry, 10, Goodman, B.K., Rutberg, J., Lin, W.W., Pulver, A.E. and Thomas, G.H. (2000) Hyperprolinaemia in patients with deletion (22)(q11.2) syndrome. J. Inherit. Metab. Dis., 23, Moss, E.M., Batshaw, M.L., Solot, C.B., Gerdes, M., McDonald-McGinn, D.M., Driscoll, D.A., Emanuel, B.S., Zackai, E.H. and Wang, P.P. (1999) Psychoeducational profile of the 22q11.2 microdeletion: a complex pattern. J. Pediatr., 134, Swillen, A., Devriendt, K., Legius, E., Eyskens, B., Dumoulin, M., Gewillig, M. and Fryns, J.P. (1997) Intelligence and psychosocial adjustment in velocardiofacial syndrome: a study of 37 children and adolescents with VCFS. J. Med. Genet., 34, Murphy, K.C., Jones, L.A. and Owen, M.J. (1999) High rates of schizophrenia in adults with velo-cardio-facial syndrome. Arch. Gen. Psychiatry, 56,

9 Human Molecular Genetics, 2007, Vol. 16, No Pulver, A.E., Nestadt, G., Goldberg, R., Shprintzen, R.J., Lamacz, M., Wolyniec, P.S., Morrow, B., Karayiorgou, M., Antonarakis, S.E., Housman, D. et al. (1994) Psychotic illness in patients diagnosed with velo-cardio-facial syndrome and their relatives. J. Nerv. Ment. Dis., 182, Shprintzen, R.J., Goldberg, R., Golding-Kushner, K.J. and Marion, R.W. (1992) Late-onset psychosis in the velo-cardio-facial syndrome. Am. J. Med. Genet., 42, Blake, R.L. and Russell, E.S. (1972) Hyperprolinemia and prolinuria in a new inbred strain of mice, PRO-Re. Science, 176, Gogos, J.A., Santha, M., Takacs, Z., Beck, K.D., Luine, V., Lucas, L.R., Nadler, J.V. and Karayiorgou, M. (1999) The gene encoding proline dehydrogenase modulates sensorimotor gating in mice. Nat. Genet., 21, Paterlini, M., Zakharenko, S.S., Lai, W.S., Qin, J., Zhang, H., Mukai, J., Westphal, K.G., Olivier, B., Sulzer, D., Pavlidis, P. et al. (2005) Transcriptional and behavioral interaction between 22q11.2 orthologs modulates schizophrenia-related phenotypes in mice. Nat. Neurosci., 8, Lachman, H.M., Papolos, D.F., Saito, T., Yu, Y.M., Szumlanski, C.L. and Weinshilboum, R.M. (1996) Human catechol-o-methyltransferase pharmacogenetics: description of a functional polymorphism and its potential application to neuropsychiatric disorders. Pharmacogenetics, 6, Bender, H.U., Almashanu, S., Steel, G., Hu, C.A., Lin, W.W., Willis, A., Pulver, A. and Valle, D. (2005) Functional consequences of PRODH missense mutations. Am. J. Hum. Genet., 76, Ryan, A.K., Goodship, J.A., Wilson, D.I., Philip, N., Levy, A., Seidel, H., Schuffenhauer, S., Oechsler, H., Belohradsky, B., Prieur, M. et al. (1997) Spectrum of clinical features associated with interstitial chromosome 22q11 deletions: a European collaborative study. J. Med. Genet., 34, El Tahir, M.O., Kerr, M. and Jones, R.G. (2004) Two cases of generalized seizures and the Velocardiofacial syndrome a clinically significant association? J. Intellect. Disabil. Res., 48, Jaksic, T., Wagner, D.A. and Young, V.R. (1990) Plasma proline kinetics and concentrations in young men in response to dietary proline deprivation. Am. J. Clin. Nutr., 52, Gothelf, D., Eliez, S., Thompson, T., Hinard, C., Penniman, L., Feinstein, C., Kwon, H., Jin, S., Jo, B., Antonarakis, S.E. et al. (2005) COMT genotype predicts longitudinal cognitive decline and psychosis in 22q11.2 deletion syndrome. Nat. Neurosci., 8, Bassett, A.S., Chow, E.W., AbdelMalik, P., Gheorghiu, M., Husted, J. and Weksberg, R. (2003) The schizophrenia phenotype in 22q11 deletion syndrome. Am. J. Psychiatry, 160, Nicolson, R., Giedd, J.N., Lenane, M., Hamburger, S., Singaracharlu, S., Bedwell, J., Fernandez, T., Thaker, G.K., Malaspina, D. and Rapoport, J.L. (1999) Clinical and neurobiological correlates of cytogenetic abnormalities in childhood-onset schizophrenia. Am. J. Psychiatry, 156, Debbane, M., Glaser, B., David, M.K., Feinstein, C. and Eliez, S. (2006) Psychotic symptoms in children and adolescents with 22q11.2 deletion syndrome: neuropsychological and behavioral implications. Schizophr. Res., 84, Baker, K.D. and Skuse, D.H. (2005) Adolescents and young adults with 22q11 deletion syndrome: psychopathology in an at-risk group. Br. J. Psychiatry, 186, Chow, E.W., Watson, M., Young, D.A. and Bassett, A.S. (2006) Neurocognitive profile in 22q11 deletion syndrome and schizophrenia. Schizophr. Res., 87, van Amelsvoort, T., Henry, J., Morris, R., Owen, M., Linszen, D., Murphy, K. and Murphy, D. (2004) Cognitive deficits associated with schizophrenia in velo-cardio-facial syndrome. Schizophr. Res., 70, Delwing, D., Bavaresco, C.S., Wannmacher, C.M., Wajner, M., Dutra-Filho, C.S. and Wyse, A.T. (2003) Proline induces oxidative stress in cerebral cortex of rats. Int. J. Dev. Neurosci., 21, Shashi, V., Keshavan, M.S., Howard, T.D., Berry, M.N., Basehore, M.J., Lewandowski, E. and Kwapil, T.R. (2006) Cognitive correlates of a functional COMT polymorphism in children with 22q11.2 deletion syndrome. Clin. Genet., 69, Bearden, C.E., Jawad, A.F., Lynch, D.R., Sokol, S., Kanes, S.J., McDonald-McGinn, D.M., Saitta, S.C., Harris, S.E., Moss, E., Wang, P.P. et al. (2004) Effects of a functional COMT polymorphism on prefrontal cognitive function in patients with 22q11.2 deletion syndrome. Am. J. Psychiatry, 161, Crinella, F.M. and Yu, J. (1999) Brain mechanisms and intelligence: psychometric g and executive function. Intelligence, 27, Egan, M.F., Goldberg, T.E., Kolachana, B.S., Callicott, J.H., Mazzanti, C.M., Straub, R.E., Goldman, D. and Weinberger, D.R. (2001) Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia. Proc. Natl Acad. Sci. USA, 98, Goldberg, T.E., Egan, M.F., Gscheidle, T., Coppola, R., Weickert, T., Kolachana, B.S., Goldman, D. and Weinberger, D.R. (2003) Executive subprocesses in working memory: relationship to catechol-o-methyltransferase Val158Met genotype and schizophrenia. Arch. Gen. Psychiatry, 60, Rosa, A., Peralta, V., Cuesta, M.J., Zarzuela, A., Serrano, F., Martinez-Larrea, A. and Fananas, L. (2004) New evidence of association between COMT gene and prefrontal neurocognitive function in healthy individuals from sibling pairs discordant for psychosis. Am. J. Psychiatry, 161, Joober, R., Gauthier, J., Lal, S., Bloom, D., Lalonde, P., Rouleau, G., Benkelfat, C. and Labelle, A. (2002) Catechol-O-methyltransferase Val-108/158-Met gene variants associated with performance on the Wisconsin Card Sorting Test. Arch. Gen. Psychiatry, 59, Malhotra, A.K., Kestler, L.J., Mazzanti, C., Bates, J.A., Goldberg, T. and Goldman, D. (2002) A functional polymorphism in the COMT gene and performance on a test of prefrontal cognition. Am. J. Psychiatry, 159, Stefanis, N.C., Van Os, J., Avramopoulos, D., Smyrnis, N., Evdokimidis, I., Hantoumi, I. and Stefanis, C.N. (2004) Variation in catechol-o-methyltransferase val158 met genotype associated with schizotypy but not cognition: a population study in 543 young men. Biol. Psychiatry, 56, Bilder, R.M., Volavka, J., Czobor, P., Malhotra, A.K., Kennedy, J.L., Ni, X., Goldman, R.S., Hoptman, M.J., Sheitman, B., Lindenmayer, J.P. et al. (2002) Neurocognitive correlates of the COMT Val(158)Met polymorphism in chronic schizophrenia. Biol. Psychiatry, 52, Kates, W.R., Antshel, K.M., Abdulsabur, N., Colgan, D., Funke, B., Fremont, W., Higgins, A.M., Kucherlapati, R. and Shprintzen, R.J. (2006) A gender-moderated effect of a functional COMT polymorphism on prefrontal brain morphology and function in velo-cardio-facial syndrome (22q11.2 deletion syndrome). Am. J. Med. Genet. B Neuropsychiatr. Genet., 141, Glaser, B., Debbane, M., Hinard, C., Morris, M.A., Dahoun, S.P., Antonarakis, S.E. and Eliez, S. (2006) No evidence for an effect of COMT Val158Met genotype on executive function in patients with 22q11 deletion syndrome. Am. J. Psychiatry, 163, Goldman-Rakic, P. (1991) Prefrontal Cortical Dysfunction in Schizophrenia: the Relevance of Working Memory. Raven Press, New York. 40. Mattay, V.S., Goldberg, T.E., Fera, F., Hariri, A.R., Tessitore, A., Egan, M.F., Kolachana, B., Callicott, J.H. and Weinberger, D.R. (2003) Catechol O-methyltransferase val158-met genotype and individual variation in the brain response to amphetamine. Proc. Natl Acad. Sci. USA, 100, Seamans, J.K. and Yang, C.R. (2004) The principal features and mechanisms of dopamine modulation in the prefrontal cortex. Prog. Neurobiol., 74, Winterer, G. and Weinberger, D.R. (2004) Genes, dopamine and cortical signal-to-noise ratio in schizophrenia. Trends Neurosci., 27, Stefanis, N.C., van Os, J., Avramopoulos, D., Smyrnis, N., Evdokimidis, I. and Stefanis, C.N. (2005) Effect of COMT Val158Met polymorphism on the Continuous Performance Test, Identical Pairs Version: tuning rather than improving performance. Am. J. Psychiatry, 162, Nolan, K.A., Bilder, R.M., Lachman, H.M. and Volavka, J. (2004) Catechol O-methyltransferase Val158Met polymorphism in schizophrenia: differential effects of Val and Met alleles on cognitive stability and flexibility. Am. J. Psychiatry, 161, Fyer, A.J., Endicott, J., Manuzza, S. and Klein, D.F. (1985) Schedule for Affective Disorders and Schizophrenia, Lifetime Version Modified for the Study of Anxiety Disorders, French Version, INSERM Ed, 1989, Paris. 46. Baron, M., Asnis, L. and Gruen, R. (1981) The schedule for Schizotypal Personalities (SSP): a diagnostic interview for schizotypal features. Psychiatry Res., 4,