Early Human Postnatal Brain Development Through the Lens of Rare Genetic Disorders

In the first years of life, the developing human brain increases in size four-fold, reaching 90% of adult volume by age 6 years (1,2). Through recent progress in human genetics, dozens of highly penetrant genetic mutations have been identified that cause neurodevelopmental syndromes, many of which are associated with specific patterns of brain undergrowth or overgrowth. Human brain development occurs over a protracted period of time, with distinct neurodevelopmental processes occurring at different stages. Therefore, the timing during which abnormalities emerge provides insight into underlying mechanisms. Of course, while important gene functions occur during prenatal development, early postnatal brain development also requires careful study. Some genetic mutations will impact postnatal events in brain development primarily, or in addition to, their prenatal effects. In childhood neurodevelopmental disorders, the postnatal developing brain is likely to be the major target for future interventions.

In the current issue of Biological Psychiatry, Levy et al. (3) carefully pinpoint the timing of abnormalities emerging in mouse postnatal cortical development caused by mutations in the gene for the dual-specificity tyrosine-phosphorylation-regulated kinase (DYRK1A). The genomic position of DYRK1A is found within the Down syndrome critical region. Most pertinent to this article, DYRK1A syndrome is an autosomal dominant neurodevelopmental disorder that arises due to de novo, loss-of-function variants in DYRK1A (4). Core features of the syndrome include intellectual disability, absent speech, and autism in a subset of patients. A prominent feature of DYRK1A syndrome is undergrowth of the brain. Importantly, patients are diagnosed with primary microcephaly and/or postnatal microcephaly. The question as to whether patients have primary microcephaly, postnatal microcephaly, or likely both, is critical background information for Levy et al. (3). Both primary and postnatal microcephaly (also called secondary or acquired microcephaly) result from reduced brain growth; however, the distinct timing of growth reductions reflects different neurodevelopmental mechanisms (2). Primary microcephaly is often associated with gene mutations and mechanisms that impede neurogenesis, a neurodevelopmental process that is completed in human brain development prior to birth. Postnatal microcephaly is generally defined as attenuated brain growth in the postnatal period, following a normal or near normal head circumference at birth. Thereby, postnatal microcephaly reflects abnormalities in mechanisms occurring postnatally, which include potential defects in neuronal growth and morphogenesis (but not cell proliferation), axonal and dendritic growth and morphogenesis, and synaptogenesis or gliogenesis. Postnatal microcephaly also may reflect an early-onset neurodegenerative mechanism. Many neurodevelopmental syndromes involve mixed components of both prenatal and postnatal microcephaly. This appears to be the case in DYRK1A syndrome.

The experiments in the mouse models presented by Levy et al. (3) are perhaps best understood within the context of the prior data on postnatal brain growth and morphology in patients with DYRK1A syndrome. In a study by Ji et al. (5), 14 individuals with DYRK1A syndrome were carefully examined with regard to head circumference. Occipital frontal circumference (OFC) was found at birth to be reduced, measuring between −1 to −4 standard deviations (SDs) below the mean. Using the conservative definition of primary microcephaly reflecting OFC less than 3 SDs below the mean, up to half of the patients studied by Ji et al. (5) did not meet the criteria for primary microcephaly. Ji et al. (5) also presented data on OFC at last follow-up on the same patients that support an ongoing attenuation of brain growth in the postnatal period. Data supporting a worsening in OFC measures in most patients were seen, measuring between −2 to −5 SDs below the mean at last follow-up. Brain magnetic resonance imaging studies of patients with DYRK1A syndrome further reflect abnormalities in neurodevelopmental processes occurring both prenatally and postnatally. Magnetic resonance imaging studies indicate general and global undergrowth of the brain, sometimes with hypoplasia of the corpus callosum, brain stem, or pituitary stalk. Also noted are indicators of likely postnatal processes, such as potential loss of tissue, reflected by enlarged ventricles, and reduced myelination (4).

In light of strong evidence that postnatal neurodevelopment is abnormal in humans with DYRK1A syndrome, the study of Levy et al. (3) is particularly interesting because it focuses on early postnatal cortical development. Levy et al. (3) employ a conditional mutagenesis strategy to isolate the Dyrk1a deletion to developing cortical cells, including a majority of neurons, as well as a subset of astrocytes and oligodendrocytes. They use the Emx1-cre driver, which deletes Dyrk1a in these cell lineages at approximately embryonic day 11 of development (6). One caveat is that Dyrk1a functions before embryonic day 11 will not be impeded. Previous studies of the germline null mutation of Dyrk1a in mice show reduced viability in homozygotes and reduced brain size in viable heterozygote mice (7). The conditional knockout (cKO) is studied by Levy et al. (3) as a homozygote and heterozygote; each shows remarkable parallels to the findings in human DYRK1A syndrome. Mutant mice have a smaller brain size, enlarged ventricles, an absent or thinning corpus callosum, and cortical thinning. Behaviorally, the heterozygote cKO shows deficits in social behavior.

Given the role of Dyrk1a in embryonic neurogenesis, the authors predicted that smaller brain size would reflect a decreased number of cells; however, unexpectedly, using an isotropic fractionator to count absolute cell numbers, the authors found no differences in cell number at postnatal day 0 (P0) between mutant and control brains, attributing this to equivalent neuronal and glial cell counts. The authors nicely show that at birth in the mutant brains the cell size is smaller and the packing density of cells is increased. The packing density is further shown to be attributed to decreased neuronal arbors. Importantly, Levy et al. (3) then studied neuronal cell numbers across the first postnatal week. Therein, they observed a reduction of neuronal numbers at P7 but not P0. Through careful attention to developmental timing, the authors have pinpointed defects involving neuronal growth (soma size), morphogenesis (size and complexity of arbors), and cell survival in the early postnatal period, as opposed to neurogenesis in their Dyrk1a model.

DYRK1A is one of hundreds of different genes that, when mutated individually, cause neurodevelopmental syndromes. Fortunately, there is a smaller range of neurodevelopmental mechanisms these gene functions converge onto that perturb brain development. Thereby, the hopeful possibility is that treatments targeting a given neurodevelopmental mechanism may be helpful for large subgroups of genetic syndromes. Moving forward, the field will require more clinical biomarkers of subgroupings. Physical and neurological examinations and magnetic resonance imaging findings may be helpful, but additional molecular biomarkers are needed. In this way, the proteomic studies in Levy et al. (3) define the underlying signaling events that are abnormal in Dyrk1a mutant mice and are hypothesis generating in terms of the development of future biomarkers for this syndrome. Using a combination of proteomic methods, Levy et al. (3) discover signatures of potential defects in growth factor signaling cascades, including pathways previously implicated in a range of other neurodevelopmental disorders, such as RhoA, TrkB-BDNF (tyrosine receptor kinase B–brain-derived neurotrophic factor), ERK/MAPK (extracellular signal-regulated kinase/mitogen-activated protein kinase), and mTOR (mechanistic target of rapamycin) signaling. For example, one target suggested by the proteomic studies was TrkB, and decreased phosphorylation of TrkB (Y816) was nicely validated by Western blotting. Similarly, Levy et al. (3) validated additional downstream growth factor signaling events and found severe reductions in phosphorylated S6, S6K1, and ERK1/2. To further validate their findings in situ, Levy et al. (3) report decreased phosphorylation of S6, a reporter of growth signaling such as through mTOR, in immunohistochemical studies of layer V cortical pyramidal neurons.

Levy et al. (3) present two distinct experimental rescues of Dyrk1a mutant defects in their mouse models, including both a genetic and a pharmacological rescue. Importantly, these rescues exemplify the theme of convergence of genetic mutations onto specific cellular mechanisms. Drawing from their evidence of diminished TrkB-BDNF and mTOR signaling, Levy et al. (3) rescue their Dyrk1a cKO heterozygous mice by mating them to a conditional heterozygous Pten mutant line. Heterozygous Pten mutations cause a neurodevelopmental syndrome with brain overgrowth (8). The double heterozygous conditional mutants demonstrate a rescue of cortical phenotypes. In addition, to stimulate growth signaling pathways, Levy et al. (3) treat their Dyrk1a conditional heterozygous mutants with (1–3)IGF-1 GPE (glycine-proline-glutamate), an amino terminal fragment of IGF-1. This experiment had an interesting design, wherein treatment was studied from daily IGF-1 treatment in early postnatal brain development between P0 and P7. IGF-1 treatment rescued soma size defects and the p-S6 defect in vivo by P7. Preclinical evidence for IGF-1 in the treatment of neurodevelopmental disorders has provided support for pilot clinical trials (9,10). Initial clinical evidence in humans with Rett syndrome in a placebo-controlled crossover trial has not yielded favorable results (9). Studies such as Levy et al. (3) highlight the potential challenges inherent in these trials: neurodevelopmental mechanisms that are targets of these treatments may require specific treatment windows. In conclusion, Levy et al. (3) nicely exemplify the careful study of early postnatal development. The developing brain during this early postnatal period may well be the target of future treatments in childhood neurodevelopmental disorders.