Vascularization in mTOR Mouse Mutants: An Effort Not in Vein

The mechanistic target of rapamycin (mTOR) signaling pathway is activated by cellular inputs from growth factors, ATP, amino acids, and oxygen levels. Mutations in mTOR pathway genes (MPG) such as TSC2 and PTEN, both negative regulators of mTOR, are linked to developmental brain malformations, intellectual disability, autism, and epilepsy (Mirzaa et al., 2016; Marsan and Baulac, 2018). TSC2 and PTEN mutations lead to mTOR pathway hyperactivation during fetal brain development causing morphologic and physiological abnormalities in the developing brain, i.e., neuronal hypertrophy, aberrant dendritic architecture, altered cortical lamination, brain overgrowth, and network hyperexcitability. Curiously, the effects of MPG variants on brain vascular development have not been deeply investigated although the vascular network results from the concomitant development of both neural and vascular components via angiogenesis (generation of new blood vessels from preexisting ones) and vasculogenesis (generation of blood vessels de novo) starting early in embryonic development. Thus, it seems likely that MPG might have effects on blood vessel structure and function. For example, MPG variants causing mTOR activation may alter expression of vascular remodeling genes such as VEGF, Ang-1/2, HIF-1α, and matrix metalloproteinases (Xue et al., 2018; Broekaart et al., 2020) and indeed, vascular remodeling has been reported in heterozygous Tsc2 Eker rats (Kútna et al., 2020). Seizures associated with MPG variants may disrupt the blood-brain-barrier integrity (Mendes et al., 2019).

Recently, Dusing et al. (2023) have investigated the effects of mTOR hyperactivation on brain vasculature because of loss of either Pten or Tsc2, in three mouse models comparing a focal Pten knock-out (KO) in dentate granule cells (Gli1-DGC-P10 KO) following tamoxifen treatment at postnatal day 14 (P14), a conditional Pten knock-out in forebrain excitatory neurons (CamK2α-FB-P10 KO), and an AAV9-mediated Tsc2 knock-out in cortical excitatory neurons (targeted injection of AAV9-CaMK2αshort>mCherry:T2A:Cre:WPRE vector; f-Tsc2 KO). Blood vessel area was significantly increased (51.9%) in DGC-P10 KO mice at 10 weeks of age relative to both earlier timepoints. To determine whether the observed increase in vessel area also produced an increase in vessel density, dentate volume for each brain slice was examined. Quantification of dentate vessel volume revealed significant increases across groups with DGC-P10 KO vessel volume greater at 10 weeks (51.8%) compared with previous timepoints and controls. However, when dentate vessel area was normalized to dentate volume (to calculate vessel density), no significant differences were seen across age groups. Using confocal image stacks and Neurolucida 360 software analysis, the authors were able to create 3-D-reconstructed images of vasculature in each dentate, and gain in-depth quantitative analysis of vessel length, volume, diameter, and tortuosity (how “twisted” a vessel is). They noted significantly increased vessel length and volume in 10-week DGC-P10 KO mice; however, when normalized for dentate volume increase (because of Pten KO), there was no significant difference in vessel density. Vessel diameter and tortuosity were also not different among groups. It appeared that increasing vessel area was driven by lengthening, with no marked changes in diameter or vessel path. In parallel experiments, blood-brain barrier leakage was assessed in the DGC-P10 KO group using AF488-conjugated bovine serum albumin before perfusion but showed no statistically significant alterations (Fig. 4).

The investigators hypothesized that the lack of change in vascular density was because of the comparatively smaller population of affected neurons in the DGC-P10 KO group, thus, they next assayed the FB-P10 KO mice where vessel area in cortex and hippocampi was larger than controls, however, when normalized to brain volume, there was again, no significant changes in vessel density. Two mice expressed unexpected germline loss of one Pten allele (vs conditional), and those values seemed to overlap with the values of FB-P10 KO group, indicating that Pten heterozygosity does not change the phenotype and that even loss of Pten from neurons is not enough to drive persistent hypervascularization. In parallel experiments, the expression of the angiogenic factor VegfA was increased significantly (65%) in the FB-P10 KO mice relative to controls.

Finally, in the f-Tsc2 KO strain, there was increased vascularity in areas of focal Tsc2 loss. Post hoc analyses confirmed that there was significantly increased vascular density in the mCherry-expressing hemisphere relative to the nonexpressing hemisphere within f-Tsc2 KO strains compared with controls. This is contrary to the effect seen in the previous FB-P10 KO group, suggesting that Pten and Tsc2 KOs create differing vascular characteristics.

Overall, Dusing et al. (2023) performed a thorough and rigorous analysis. Indeed, by using varying analysis techniques (3-D reconstructive analysis vs 2-D analysis), defined regional expression of Pten deletion (DGC vs FB), and distinct mTOR regulators (Tsc2 vs Pten), the group provided solid evidence that mTOR hyperactivation results in distinct profiles of altered vascularization depending on genotype and brain region. In PTEN-associated hemimegalencephaly or megalencephaly and in TSC, altered vascular structure could have important implications for understanding disease pathogenesis or treatment and indeed, there is an increase in microvasculature in white matter of TSC cortical tubers (Veersema et al., 2019). Vascular anomalies are identified in patients with PTEN germline mutation syndromes (Tan et al., 2007). The significance of the current findings to patients with either PTEN or TSC2 mutation syndromes should be defined in further studies.

There are some limitations to the study. Focal loss of either PTEN or TSC2 is uncommon in the hippocampus in human patients and thus analysis of the hippocampus may not yield insights into human disease. The choice of a P14 tamoxifen induction to Pten KO does not model when PTEN mutations usually exert an effect, i.e., during embryonic development. Since all three mouse model platforms target a neuron-specific KO, consideration of how noncell autonomous mechanisms might alter brain vasculature would be interesting. It would have been helpful to discuss how loss of either Pten or Tsc2 in neurons might lead to altered vascular structure, an effect that likely requires signaling to endothelial cells. The DGC-P10 KO also hits small numbers of astrocytes (∼1%) in cortex and midbrain but not endothelial cells. Assay of VegfA levels in the DGC-P10 KO mice or f-Tsc2 KO strains would have been intriguing since Vegf levels are elevated in human TSC and might shed mechanistic light into how loss of Pten or Tsc2 might alter vasculature in various brain regions. Further work could aim to reconcile the elevation in VegfA, a known angiogenesis factor, with the paucity of overall changes in vascularization in either Pten or Tsc2 KO.