Patient A is a Danish boy who carries a de novo balanced translocation t(3;8) (q21;q24). He was born in 1998 as the second child of non-consanguineous, Caucasian, healthy parents aged 37 (mother) and 38 (father) at the time of birth. Both parents have academic degrees. According to the parents the older sister is both intellectually and socially very well-functioning. The pregnancy was normal and the delivery at gestational age 40 + 2 was uncomplicated with Apgar scores 9/1 and 10/5. The birth weight was 3900 g; birth length 53 cm; and head circumference 36 cm. Growth parameters are currently still normal. No dysmorphic features were noted at birth and hearing was normal.

In the neonatal period the parents noticed an abnormal social interaction with him. Later on it was apparent that both verbal and social development was delayed; however motor milestones were achieved normally. Genetic testing for the fragile X syndrome was negative and metabolic screening showed no abnormalities. At 2 years of age he was diagnosed with childhood autism (Autism Diagnostic Observation Schedule type G (ADOS-G); Communication score: 5, Social score: 14). At 8 years of age a WISC-III test showed a very uneven profile with specific non-verbal visio-spatial difficulties (verbal IQ = 103, performance IQ = 60, global IQ = 79). Currently, he attends a class for children with special needs in a normal primary school. Verbally he is highly skilled in both Danish and English.

Periodic idiopathic trembling was noted from the age of 2 days and according to the mother it persisted for the first 5 weeks. This description is in accordance with a diagnosis of benign neonatal convulsions but this was never diagnosed. Currently, he has no epilepsy diagnosis; however, according to the parents he has brief episodes of non-responsiveness resembling absence seizures. Consequently, electroencephalographic (EEG) examination was carried out at ages 5 and 9 years during sleep, hyperventilation, photo-stimulation, and during periods of non-responsiveness but no abnormalities were observed. Cerebral magnetic resonance (MR) scanning of the brain at age 8 years was normal.

The National Ethics Committees and the Danish Data Protection Agency approved the study, and informed consent was obtained from the parents.

Mutation screening of KCNQ3 was performed in two steps. As a first step DNA from a cohort comprising 100 Portuguese and 48 Danish ASD patients were screened for KCNQ3 mutations by direct sequencing. The Portuguese ASD patients were recruited at the Hospital Pediátrico de Coimbra and all originated from mainland Portugal and the Azorean islands. The male-female ratio was 4.8:1, and the ages ranged between 2 and 18 years (mean age 6.8 years). Idiopathic subjects were included after clinical assessment and screening for known medical and genetic conditions associated with autism, including testing for Fragile X mutations (FRAXA and FRAXE), chromosomal abnormalities, neurocutaneous syndromes, endocrine (thyroid function screening), and metabolic disorders. About 35 of the 48 Danish ASD patients were recruited at child psychiatric hospitals in the western part of Denmark (Jutland) (age range 3–30 years, with mean age of 10 years and male-female ratio of 3:1). Seven autistic patients were ascertained at the Kennedy Center (Glostrup, Denmark) (age range 13–37 years, mean age 20.4 and male-female ratio of 2.5:1). These patients were unrelated and part of the IMGSAC group and accordingly some of the patients had siblings and some even additional relatives with a diagnosis of pervasive developmental disorder. These patients were screened for chromosomal abnormalities and fragile X syndrome and a physical examination included a careful search for phakamatoses to rule out Tuberous Sclerosis (TSC). Four patients diagnosed within the autism spectrum were collected at the Psychiatric Hospital in Hillerød (Frederiksborg Amt, Denmark). In addition, two DNA samples (one male, one female) from individuals diagnosed within the ASD spectrum and with chromosomal rearrangements were included in the screening. These samples were collected at the Wilhelm Johannsen Centre for Functional Genome Research, University of Copenhagen (Denmark). In all of the above ASD patients diagnosis was made in accordance with DSM-IV or ICD-10 criteria using ADI-R in addition to ADOS or the Childhood Autism Rating Scale.

The c.1720C > T variant in KCNQ3 was first identified in one Portuguese ASD patient (patient B, Table 1) by direct sequencing. As a next step 271 additional Portuguese ASD patients fulfilling the same criteria as the first cohort were specifically screened for the c.1720C > T variant by a PCR/enzyme cleavage assay whereby two additional patients (patient C and D) were identified as carriers of the c.1720C > T variant. Hence, a total of 419 ASD patients (371 Portuguese- and 48 Danish ASD patients) were screened for the c.1720C > T variant. The three male Portuguese patients (Patient B, C, D in Table 1) were diagnosed with childhood autism using the Autism Diagnostic Interview-Revised (ADI-R) and ADOS (Lord et al., 1994) and had no history of convulsions. The inheritance pattern of the c.1720C > T variant was ascertained both by direct sequencing of exon 13 of KCNQ3 and by the PCR/enzyme cleavage assay in patient B, C, and D and their parents. As controls 96 Caucasian individuals from the Human Random Control DNA panel (HRC-1, Sigma-Aldrich, St. Louis, USA), and 100 Portuguese individuals without neuropsychiatric disease (self reported) from blood donor centers throughout Portugal were included.

Cytogenetic analysis and fluorescence in situ hybridization (FISH) was performed according to standard protocols, and array-based comparative genome hybridization was performed as previously described (Erdogan et al., 2006).

When necessary, genomic DNA was uniformly amplified using GenomiPhi™ DNA Amplification Kit (GE Healthcare, Buckinghamshire, UK).

All 15 coding exons and intron-exon boundaries of KCNQ3 (NM_004519.3) were amplified by PCR. Sequencing reactions were carried out using BigDye^® Terminator v 1.1 Cycle Sequencing Kit (Life Technologies, California, USA) and analyzed by an ABI 3100 AVANT Genetic Analyzer (Life Technologies, California, USA). ChromasPro version 1.33 (Technelysium Pty Ltd, Australia) was used to visualize the data. Nucleotide changes were verified by a second PCR amplification of non-genome amplified patient DNA, sequencing and restriction cleavage.

A PCR product of 461 base pairs (bp) encompassing exon 13 of KCNQ3 was amplified using primers KCNQ3_13a: TATTCCAAACCCTTATCTCAT and KCNQ3_13b: AAACAGGTGGGG CTATTA. PCR fragments amplified from the WT allele were digested into two fragments with lengths 438 and 23 bp by the restriction enzyme Hpy188III, whereas PCR fragments amplified from the c.1720C > T allele were digested into three fragments with lengths 337, 101, and 23 bp.

The plasmids hK_V7.2-hK_V7.5 in pXOOM or pXOON, hK_V7.3-flag in pNS2z, and hK_V7.2-cmyc in pNS2z used in this study have been described previously (Bentzen et al., 2006; Rasmussen et al., 2007). K_V7.4 cDNA was amplified with PCR and inserted into pNS2z to generate C-terminally myc-tagged K_V7.4. To generate the extracellularly tagged expression plasmid hK_V7.5-3xHA in pXOOM, 3 HA-tags were inserted into the TM3-TM4 linker of hK_V7.5 by PCR using the primers 5′-CCAGATTACGCGTACCC TTACGACGTTCCAGATTACGCTGGTAATATTTTTGCCAC-3′ and 5′-GACATCGTAT GGGTAAGCGTAGTCTGGGACGTCGTATGGGTACTGAGTTTTTGCAGAAAC-3′. Human CD4-WT in pcDNA3.1 was a kind gift from James Trimmer (University of California Davis, CA, USA) and has been described earlier (Gu et al., 2003). The chimera hCD4-hK_V7.3CT in pcDNA3.1 was generated using standard PCR and in-frame insertion of cDNA corresponding to K_V7.3 amino acids 358–873 into NotI and XhoI sites of the wild-type (WT) construct. The point mutation c.1720C > T leading to the amino acid exchange P574S was introduced using mutated oligonucleotide extension (PfuTurbo Polymerase, Stratagene, La Jolla, CA, USA) from the plasmid template harboring the cDNA of interest, digested with DpnI (Fermentas, St. Leon, Germany) and transformed into E. coli XL1 Blue cells. All plasmids were verified by complete DNA sequencing of the cDNA insert (Macrogen Inc., Seoul, Rep. of Korea). The Gene Bank Accession numbers of the human cDNAs are: NM_004519 (K_V7.3), NM_004518 (K_V7.3, isoform c), NM_004700 (K_V7.4), and NM_019842 (K_V7.5). Protein accession number for K_V7.3 is NP_004510.

HEK 293 cells were grown in DMEM (Invitrogen, Glostrup, Denmark) supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin and 10% FCS (Sigma-Aldrich, Copenhagen, Denmark) at 37°C in a humidified atmosphere with 5% CO₂. Transfections were carried out using the Lipofectamine-Plus Reagent system (Invitrogen) according to the manufacturer’s instructions. Hippocampal cultures were prepared as previously described (Rasmussen et al., 2007) and transfected at 7–8 days in vitro (DIV) using the lipofectamine 2000 method (Invitrogen) with a total of 0.9 μg of DNA per cover-slip. Transfection was carried out for 1 h at 37°C, 5% CO₂ after which the neurons were transferred back to the dishes containing the glial cell layer. Neurons were left for expression for 48 h.

HEK 293 cells or primary hippocampal neurons were fixed in 3% paraformaldehyde in PBS for 15–20 min at room temperature. Blocking and permeabilization was performed by a 30 min incubation with 0.2% fish skin gelatin in phosphate buffered saline supplemented with 0.1% Triton X-100 (PBST). The cells were then incubated for 1 h in primary antibodies diluted in PBST. Primary antibodies employed were: rabbit anti-c-myc (A-14, 1:50 dilution, Santa Cruz Biotechnology, Heidelberg, Germany), mouse anti-FLAG (M2, 1:250 dilution, Sigma-Aldrich, Copenhagen, Denmark), and rat anti-HA (3F10, 1:50 dilution, Santa Cruz Biotechnology). For immunofluorescent detection, Alexa Fluor^®–coupled secondary antibodies were diluted in PBST and applied for 45 min. The cover-slips were mounted in Prolong Gold (Invitrogen, Glostrup, Denmark).

Laser scanning confocal microscopy was performed using the Leica TCS SP2 system equipped with argon and helium-neon lasers. Images were acquired using a 63× water immersion objective, NA 1.2 with a pinhole size of 0.8-1 and a pixel format of 1024 × 1024. Line averaging was used to reduce noise. For double- and triple-labeling experiments sequential scanning was employed to allow the separation of signals from the individual channels. Acquired images were treated using Adobe Photoshop CS4 and Adobe Illustrator CS4.

By cytogenetic analysis a de novo t(3;8) (q21;q24) translocation was identified in Patient A. The breakpoints of this translocation were mapped using FISH. The breakpoint on chromosome 3 was delineated to a 26 kb region (chr3:131.189.991-131.216.096; NCBI36/hg18) at 3q21.3 located 10.5 kb downstream of the thyrotropin-releasing hormone (THR) gene (Figure 1). No annotated human genes or mRNAs are located within this breakpoint region. On chromosome 8q24.22 the breakpoint was localized in a 25 kb region (chr8:133.318.034-133.343.450; NCBI36/hg18) within exon 1 of the KCNQ3 gene (Figure 1). No disease-related copy number variations were identified by array comparative genomic hybridization (CGH) in this patient.

A paternally inherited c.1720C > T missense mutation in exon 13 of KCNQ3 was identified in patient B by direct sequencing (Figure 2A). There is no history of psychiatric- or neurological disorders in this family. The mutation results in an amino acid change at position 574 replacing proline by serine (p.P574S). By restriction enzyme assay the c.1720C > T (p.P574S) variant in KCNQ3 was identified in two additional ASD patients (patient C and D) (Figure 2B) and was confirmed in patient B. Patient C inherited the variant from the mother who suffers from major depression and patient D inherited the variant from the father who does not suffer from any psychiatric- or neurological disorders (Figure 2B). The c.1720C > T mutation in patients C and D was confirmed by direct sequencing of a second PCR product from non-amplified DNA. No c.1720C > T mutations were identified in 96 UK Caucasian and 100 Portuguese controls.

To address effects of the mutation P574S on ion channel function, we heterologously expressed mutant channels in Xenopus laevis oocytes. Since K_V7.3 does not form functional channels on its own (Schwake et al., 2000), we investigated whether the K_V7.3_{_}P574S mutation could affect the function of other neuronal members of the K_V7 family. K_V7.3_P574S or K_V7.3 WT was co-expressed with K_V7.2, K_V7.4, or K_V7.5 in Xenopus laevis oocytes and currents were recorded by TEVC. In agreement with Neubauer et al. (2008), we found that current levels for K_V7.2/K_V7.3_P574S were similar to those of K_V7.2/K_V7.3 (Figure 3A). Similarly, the function of K_V7.4 channels did not appear to be affected by the mutation, as oocytes expressing K_V7.3_{_}P574S/K_V7.4 had similar current levels as K_V7.3/K_V7.4 (Figure 3B). In a final set of experiments, we tested the effect of K_V7.3_{_}P574S on K_V7.5 currents. In line with previous work by Lerche et al. (2000), co-expression of K_V7.5 with K_V7.3 dramatically increased current levels compared to K_V7.5 alone (Figure 4). Expression of K_V7.3_{_}P574S also enhanced K_V7.5 current levels but to a significantly lesser extent than WT K_V7.3. These results suggest that K_V7.3_{_}P574S has not lost its ability to interact with K_V7.5. Since both patients B and C were heterozygous for the K_V7.3_{_}P574S mutation, we mimicked the heterozygous state by co-expressing K_V7.5 with K_V7.3 and K_V7.3_{_}P574S in a 2:1:1 ratio. The resulting current levels were intermediate of that of K_V7.3/K_V7.5 and K_V7.3_{_}P574S/K_V7.5, indicating that (1) K_V7.3_{_}P574S is not dominant-negative, and (2) co-expression of WT does not rescue the K_V7.3_{_}P574S phenotype. The difference in current levels between the heterozygote and WT K_V7.3/K_V7.5 is statistically significant as indicated by the asterisk in Figure 4B. Meticulous inspection of the curves in Figure 4B reveals that for K_V7.5 expressed alone, there is a tendency for the current-voltage relationship to flatten out at higher voltages and this tendency appears to be removed by co-expression with K_V7.3, making the current-voltage curve more linear.

Since the P574S mutation reportedly is without effect upon the current characteristics of the K_V7.2/K_V7.3 complex (Miceli et al., 2009), we decided to investigate whether the P574S mutation could affect the localization of the heteromeric K_V7.2/K_V7.3 complex. We first analyzed the localization of K_V7.2 and K_V7.3 upon co-expression in HEK 293 cells. As illustrated in Figure 5, both channel subunits displayed a primarily intracellular staining pattern. The subunits appeared to co-localize to a large degree in the intracellular structures and only weak staining could be detected in association with the cell surface. Importantly, co-expression of K_V7.2 and K_V7.3_{_}P574S resulted in a staining pattern that was indistinguishable from the co-expression of the WT channels. Thus, K_V7.3_{_}P574S does not appear to have an impact on the localization of the K_V7.2/K_V7.3 heteromeric complex in HEK 293 cells.

In neurons, the K_V7.2/K_V7.3 complex is localized to the AIS (Devaux et al., 2004; Chung et al., 2006; Rasmussen et al., 2007). We therefore speculated that the specific localization of the complex to the AIS could be disturbed by the P574S mutation. To address this question, we first examined the localization of K_V7.3 and K_V7.3_{_}P574S upon exogenous expression in cultured rat hippocampal neurons. As previously reported, singly expressed K_V7.3 was primarily observed intracellularly with no significant enrichment in the AIS (Rasmussen et al., 2007). Likewise, K_V7.3_{_}P574S demonstrated a primarily intracellular staining pattern similar to the WT subunit.

Upon co-expression of K_V7.2 and K_V7.3, the channel complex appears in the AIS (Rasmussen et al., 2007). To investigate whether the K_V7.3_{_}P574S mutation perturbed the localization of the complex to the AIS, we transiently expressed K_V7.2 with either WT K_V7.3 or K_V7.3_{_}P574S in cultured hippocampal neurons. As illustrated in Figure 6B, the P574S mutation did not impair the localization of the K_V7.2/K_V7.3_{_}P574S complex as it localized to the AIS similar to the WT complex. These results were further emphasized by experiments using chimeric constructs of the transmembrane protein CD4 and K_V7.3/K_V7.3_{_}P574S. We have previously demonstrated that the ability of K_V7.3 to direct the K_V7.2/K_V7.3 complex to the AIS critically depends on an ankyrin-G binding sequence in the C-terminal tail of K_V7.3 (Rasmussen et al., 2007). We therefore attached the C-terminal tail of K_V7.3 to a truncated version of the CD4 receptor to examine whether this part of K_V7.3 would be sufficient to redirect the otherwise non-polarized protein CD4 to the AIS (Figure 7, left panel). As expected, the truncated version of CD4 displayed a non-polarized localization pattern upon expression in cultured hippocampal neurons (Figure 7, top panel). Attachment of the K_V7.3 C-terminus was sufficient to drive an AIS localization of the chimera (Figure 7, middle panel). As illustrated, introduction of the P574S mutation into the chimera was without effect as the mutated chimera was still able to target efficiently to the AIS (Figure 7, lower panels).

Since the P574S mutation did not affect the localization of the classical K_V7.2/K_V7.3 complex, we investigated the impact of the mutation on the localization of heteromeric channels including the K_V7.4 or K_V7.5 subunits. We transiently co-expressed WT K_V7.3 or the mutant P574S with either the K_V7.4 or the K_V7.5 subunit in HEK 293 cells and analyzed the localization of the subunits by confocal microscopy. As illustrated in Figure 8, K_V7.3/K_V7.4 and K_V7.3/K_V7.5 complexes demonstrated a mixed surface and intracellular staining pattern demonstrating that both K_V7.4 and K_V7.5 subunits can pull a fraction of K_V7.3 subunits to the cell surface. However, the same localization pattern was observed when analyzing the mutant complexes. Thus, the P574S mutation was without significant effect on the localization of K_V7.4 and K_V7.5 containing complexes.