Genomic imprinting is itself a relatively new finding in mammalian genetics and confers functional differences on parental genomes such that certain autosomal alleles are only expressed when they originate from the father (maternally imprinted and silenced), whereas others are only expressed on passage through the matriline (paternally imprinted and silenced). Maternal and paternal genomes are not, therefore, transcriptionally equivalent, and both sets of autosomal alleles are required for normal development and function.
An important means of investigating genomic imprinting in the brain has been achieved by the construction of chimeras. Embryos constructed from a mixture of cells that are parthenogenetic/normal (Pg) or androgenetic/normal (Ag) do survive, but survival requires the total proportion of chimeric cells not to exceed normal cells. The precise locations in the brain to which these chimeric cells, participate in development can be determined by the presence of a genetic marker (b globin, or LacZ). Using these techniques a clear and distinct patterning in brain development emerges. At birth, cells that are disomic for the paternal genome contribute substantially to those parts of the brain that are important for primary motivated behaviour (hypothalamus, pre-optic area BNST and septum) and are excluded, from the developing neocortex and striatum. At the earliest stages of brain development (days 9-10), Ag cells are present in all neural tissues and as gestation progresses they proliferate extensively in the medio-basal forebrain, but at parturition are virtually absent from telencephalic structures. By contrast parthenogenetic cells are excluded from these medio-basal forebrain areas, but selectively accumulate in those regions where Ag cells are excluded, especially neocortex and striatum.
If genomic imprinting has any impact on behaviour, these findings, along with human clinical findings of Prader-Willi syndrome, would point to paternally expressed imprinted genes influencing motivated behaviours such as sexual, feeding, aggressive and maternal behaviour. We have recently investigated two paternally expressed genes (Peg 1 and Peg 3) which have been mutated by inserting a promoterless b geo cassette into the 5’ coding exon to study both the function and expression of these genes in mice. Inheritance of the mutation from the paternal, but not the maternal germ line, causes a severe impairment in maternal behaviour, resulting in a complete loss of progeny in the first generation. In Peg 3 offspring survival improves with subsequent generations but all aspects of maternal behaviour and ability to suckle the pups are impaired, resulting in a slower growth rate of non-mutant offspring. The brain phenotype of these null mutants has a smaller nuclear areas (PVN, SON, BNST) in the hypothalamus and fewer neurons staining positive for the peptide, oxytocin. Since the magno-cellular oxytocinergic neurons control milk let-down and the parvocellular oxytocinergic neurons are important for maternal behaviour, these findings may account for the behavioural and nurturient phenotype.
It is therefore interesting that two imprinted genes, both of which are paternally expressed, map to similar areas of the developing brain as revealed from the distribution of Ag chimaeric cells. Moreover, they both have impact on primary motivated behaviour as predicted from chimaeras and interestingly mutations of these two paternally expressed genes impair maternal behaviour.