LRP2 is a novel SHH receptor essential for mammalian forebrain formation
Introduction. LRP2 (aka megalin) is expressed in absorptive epithelia in the yolk sac and the neural tube of the early mouse embryo as well as in kidney, brain ventricles, and urogenital tract in adults. Previously, we mainly focused on characterization of this receptor in adult tissues demonstrating its ability to act as an endocytic receptor for steroid hormones in cell-type specific delivery of vitamin D metabolites to the kidney (Nykjaer et al., Cell, 1999) and of sex steroids to reproductive organs (Hammes et al., Cell, 2005).
Fig. 2. (A)
We also showed that LRP2 mediates renal uptake of aminoglycosides, causing the severe nephrotoxic side effects of these antiinfectiva (Schmitz et al., JBC, 2002). We continued to study LRP2 in the adult kidney as part of several international collaborations (e.g., Amsellem et al., JASN, 2010; Ceccarini et al., Cell Metab, 2009; Nielsen et al., PNAS, 2007). However, our main focus in this respect has been the validation of LRP2 as drug target in the prevention of nephrotoxicity in the biotech start-up ReceptIcon.
LRP2 in mammalian forebrain development. Early on, we recognized the importance of LRP2 for brain development as mice genetically deficient for the receptor suffered from holoprosencephaly (HPE), a defect in the separation of the forebrain hemispheres and the most common forebrain anomaly in humans (Willnow et al., PNAS, 1996) (Fig. 2A-B). The significance of this receptor for humans was confirmed by others, demonstrating LRP2 defects as cause of Donnai-Barrow syndrome, a monogenic disorder characterized by renal resorption defects and brain anomalies.
To further define the role of LRP2 in forebrain formation, we characterized mouse and zebrafish models deficient for embryonic receptor expression. In Lrp2-/- mouse embryos, we showed that expression of LRP2 in the neuroepithelium but not in the yolk sac is critical for brain development. Loss of LRP2 leads to altered dorso-ventral pattering of the neural tube with a loss of ventral sonic hedgehog (Shh) signals (Fig. 2C-D). As a consequence of absent SHH activity, ventrally derived neuronal cell populations are lost in the forebrain of Lrp2-/- embryos (Spoelgen et al., Development, 2005; Willnow et al., Development, 2006).
Because expression of Lrp2 in zebrafish (zf) embryos parallels the pattern seen in the mouse, we also explored the zf as a novel model organism of receptor deficiency. Initially by morpholino knockdown (Anzenberger et al., J Cell Sci, 2006) and later in two lines with ENU-induced lrp2 mutations (Kur et al., Dev Dyn, 2011), we tested the consequence of deficiency in lrp2 and in lrp2b, a receptor homologue in zf identified by us. While receptor null zf suffered from overt renal resorption deficiency, their brain development proceeded normally. This surprising finding uncovered evolutionary conservation of receptor activity in renal clearance but not in forebrain patterning between teleosts and mammals.
Fig. 3. (A)
LRP2 acts as co-receptor in SHH signaling.
Recently, we succeeded in elucidating the mechanism of LRP2 action in forebrain development (Christ et al., Dev Cell, 2012). These studies commenced in close collaboration with the group of Annette Hammes at the MDC.
Thus, we traced the primary defect in Lrp2-/- embryos to a failure to establish SHH signaling in the rostral diencephalon ventral midline (RDVM), a major forebrain organizer. At neurulation, SHH originating from the prechordal plate (PrCP) moves to the overlying RDVM to induce its own expression and to establish its signaling domain. The exact mode of delivery is unclear, but SHH is known to signal at the apical surface of the RDVM. We demonstrated that in Lrp2 mutants, SHH fails to localize to the RDVM, suggesting a defect in morphogen sequestration (Fig. 3A). This hypothesis was confirmed in whole embryo cultures (WECs) wherein Lrp2 mutants did not bind recombinant GST-SHH despite expression of its cognate receptor Ptch1 (Fig. 3B). Using cell lines, WECs, and neural explants, we refined this pathway, ultimately demonstrating that LRP2 is an auxiliary SHH receptor that sequesters the protein in the RDVM and controls internalization and cellular trafficking of SHH/Ptch1 complexes, a step critical to SHH signaling (Fig. 4).
Intriguingly, LRP2 not only plays a role in SHH-dependent neurogenesis in the embryonic but also in the adult brain. Thus, we showed that the receptor is expressed in ependymal cells lining the lateral ventricles in the brain, a site of adult neurogenesis. Lack of LRP2 results in altered morphogen signaling in this subventricular zone, in loss of neuronal stem cells, and in impaired adult neurogenesis (Gajera et al., J Cell Sci, 2010).
It was suggested that release of the intracellular domain (ICD) of LRP2 by RIP is required for receptor function. To query this concept, we generated a knock-in mouse expressing the ICD instead of the full-length protein. The ICD did not rescue mutant mice from HPE, clarifying that it is the full-length receptor that is required for SHH signaling (Christ et al., Kidney Int., 2010).