Table 1.
Canonical Wnt pathway genes mutated in autism – summary of the evidence
| Gene | Gene location | Discovery strategy | Replications | Preclinical support |
|---|---|---|---|---|
|
APC |
5q21-q22 |
association study 75 unrelated patients |
single case of APC deletion |
APC’s functional role |
|
DISC1 |
1q42 |
association study in 144 families |
no |
DISC1 function |
|
EN2 |
7q36 |
association study in 3 datasets of 518 families |
yes, but opposite haplotype |
Wnt target gene |
|
MET |
7q31 |
association studies in 4 cohorts; microdeletion in 2 pts (involving >25 genes) |
yes |
post mortem expression; animal data |
|
WIF1 |
12q14.3 |
GWAS in 26 extended families; linkage peak of ≥19 genes |
no |
|
|
MARK1 |
14q41 |
GWAS in 116 families; SNPs in MARK1 |
no |
transcription of MARK1 altered by SNPs |
|
CDH10 |
5p14.1 |
GWAS in 780 families; SNPs between CDH9 and CDH10 highly significant |
replication cohort by the same authors |
|
|
WNT2 |
7q31.2 |
GWAS study in 75 families; 2 families with missense mutation in Wnt2 |
one positive, also two negative studies |
role of Wnt2 in midbrain development |
|
PTPRZ1 |
7q31.3 |
single case with deletion CNV of 20 genes |
no |
|
|
CDH15 |
16q24.3 |
genome scanning; deletion CNV of 3 genes |
no |
|
|
CDH13 |
16q23.3 |
GWAS; deletion CNV of single gene |
no |
|
|
CDH8 |
16q21 |
GWAS; detection of rare deletion CNV |
no |
data from KO mice |
|
DOCK4 |
7q31.1 |
GWAS; microdeletion CNV |
no, but dyslexia cases |
biochemical data |
|
BCL9 |
1q21 |
deletion and duplication CNVs (14 genes) |
multiple |
|
|
FZD9 |
7q11.23 |
recognized syndrome; deletion and duplication CNVs (>20 genes) |
yes, multiple |
Wnt2 receptor |
|
AHI1 |
6q23.3 |
recognized syndrome: mutation screening identified multiple disruptive mutations |
yes, multiple |
|
|
CREBBP |
16p13.3 |
recognized syndrome; microdeletion CNVs and disruptive mutations |
multiple, also cases with microduplications |
data from KO mice |
| TSC1/2 | 9q34 / 16p13.3 | recognized syndrome: mutation screening identified numerous missense mutations | yes, multiple | data from KO mice |
The genes have been ranked according to their discovery process. Association studies of candidate gene studies have frequently yielded false positive results and are considered relatively weak evidence. Genome wide association studies followed by specific investigation of genes in the ‘hot spot’ may be more reliable, but replications are crucial. Copy number variations may provide good evidence but the duplicated or deleted regions are generally large and usually contain several candidate genes. Copy number variations that involve only a few or even a single gene give a strong indication for a pathological role of those genes, but thus far such CNVs have been detected only very rarely and lack replication. The best evidence comes from CNVs that give rise to recognizable syndromes. Unfortunately, in this case the involved genes are not specific for the canonical Wnt pathway and modify other pathways as well. In some cases, there is circumstantial support for a given gene from biochemical- or whole animal studies. Abbreviations: CNV, copy number variation; GWAS, genome wide association study; SNP, single nucleotide polymorphism. For details, please refer to the individual section in the text.