Susceptible stages in Schwann cells for NF1-associated plexiform neurofibroma development

Abstract

Stem cells are under strict regulation by both intrinsic factors and the microenvironment. There is increasing evidence that many cancers initiate through acquisition of genetic mutations (loss of intrinsic control) in stem cells or their progenitors, followed by alterations of the surrounding microenvironment (loss of extrinsic control). In neurofibromatosis type 1 (NF1), deregulation of Ras signaling results in development of multiple neurofibromas, complex tumors of the peripheral nerves. Neurofibromas arise from the Schwann cell lineage following loss of function at the NF1 locus, which initiates a cascade of interactions with other cell types in the microenvironment and additional cell autonomous modifications. In this study, we sought to identify whether a temporal "window of opportunity" exists during which cells of the Schwann cell lineage can give rise to neurofibromas following loss of NF1. We showed that acute loss of NF1 in both embryonic and adult Schwann cells can lead to neurofibroma formation. However, the embryonic period when Schwann cell precursors and immature Schwann cells are most abundant coincides with enhanced susceptibility to plexiform neurofibroma tumorigenesis. This model has important implications for understanding early cellular events that dictate neurofibroma development, as well as for the development of novel therapies targeting these tumors.

Figure 1. Tamoxifen-induced recombination at the Rosa26 locus in PLPCre-ERT2;R26R Mice

(A) PLPCre-ERT2;R26R double transgenic pregnant females were orally gavaged with 1 mg of tamoxifen at E12.5 and the embryos were analyzed 48 h later for expression of beta-gal (b, c). Control animals were only given sunflower oil (a). (B) Gavaged administration of tamoxifen (2 mg a day for 5 days) to lactating mothers leads to efficient recombination in sciatic nerve (a), DRG (b) and trigeminal ganglia (c) of the nourished pups. (C) Eight-week-old mice were gavaged with 4 mg of tamoxifen a day for 5 consecutive days. Animals were subsequently sacrificed for beta-gal expression analysis in sciatic nerve (a), DRG (b) and trigeminal ganglia (c).

Figure 2. Temporal induction of NF1 ablation in Schwann cells during development

Schematic representation of the experimental design for conditional inactivation of NF1 during development. Specifically, oral gavage is used to administer tamoxifen 1) to pregnant mice at E12.5 to knock out NF1 at the Schwann cell precursor stage, 2) to lactating mothers at birth to knock out NF1 at the immature Schwann cell stage, and 3) to 2–4 month-old adult mice to knock out NF1 in mature Schwann cells.

Figure 3. NF1 inactivation during the Schwann cell precursor stage leads to plexiform neurofibroma formation

(A) An example of three macroscopic plexiform neurofibromas from spinal roots (a, asterisk). H&E staining of the tumor shows features consistent with plexiform neurofibroma, including disordered, convoluted bundles of cells exhibiting spindle-cell morphology, with ovoid and spindle-shaped nuclei and adjacent fine fibrillar stroma (b). There is also massive mast cell infiltration within the tumor, as demonstrated by Leder staining, which stains mast cells red (c), (original magnification 20X). (B) Neurofibromas were also harvested for gross X-gal staining of whole tumor (a). Tumor tissues were then postfixed in formalin for paraffin sectioning. The corresponding paraffin sections also show histological evidence of neurofibroma with X-gal-positive staining (b), disordered spindle-cell morphology, ovoid and spindle-shaped nuclei and immunohistochemical stain shows spindle cells within neurofibroma are positive for S100 (c), (original magnification 20X).

Figure 4. NF1 inactivation during the immature Schwann cell stage leads to plexiform neurofibroma formation

(A) Peripheral nerve enlargement, specifically in the brachial plexus and thoracic nerves (arrows) in sick PLPCre-ERT2;NF1^flox/− mice (a) compared to control littermates (b). (B) Plexiform neurofibromas developed in close proximity to the DRG (arrow) in every sick PLPCre-ERT2;NF1^flox/− mice (a) but none were observed in control littermates (b). (C) These plexiform neurofibromas (arrows) developed exclusively near the spinal roots/DRG at the cervical (a) and thoracic (b) level. They are sometimes also seen in the cauda equina (c).

Figure 5. Histological and immunohistological analysis of the Genetic Engineered Mouse (GEM)-neurofibromas

The mutant PLPCre-ERT2;NF1^flox/− mice robustly developed paraspinal tumors near the DRG when NF1 is ablated at the immature Schwann cell stage. (A) Representative H&E staining of these tumors showing classic features of plexiform neurofibroma (original magnification: 10X in the upper panel and 40X in the lower panel). (B) Leder staining shows heavy infiltration of mast cells (red) within the plexiform neurofibroma (original magnification: 10X in the upper panel and 40X in the lower panel). (C) Anti-S100β and (D) Anti-GAP43 immunohistochemistry shows that spindle cells in the plexiform neurofibroma are positive for the Schwann cell markers S100β and GAP43 (original magnification: 10X in the upper panel and 40X in the lower panel).

Figure 6. NF1 ablation in adulthood rarely leads to plexiform neurofibroma formation

Adult PLPCre-ERT2;NF1^flox/−Rosa26 and control mice were orally gavaged with 4 mg of tamoxifen a day for five days to activate Cre activity. We then harvested DRG, sciatic nerve (SN) and trigeminal ganglia (TG) for X-gal staining to confirm recombination (A; a-c). We also observe the co-localization of X-gal stain with S100 marker (brown color) in a cross section of nerve fiber (A; d, arrow) and horizontal section of the DRG (A; e, arrow). The majority of the PLPCre-ERT2;NF1^flox/− mice (17 out of 19 mice) did not develop macroscopic or microscopic plexiform neurofibroma into their second year of life when NF1 is ablated at the mature Schwann cell stage (B). Only about 10% (2 out of 19 mice) of PLPCre-ERT2;NF1^flox/− mice develop tumors, including spinal neurofibroma (C, arrow head) and a classic plexiform neurofibroma at the left hindlimb causing left leg gigantism (D, arrow).

Figure 7. Diagram outlining the stages of Schwann cell development, and the periods of Cre-mediated recombination in various neurofibroma models

Neural crest cells generate mature Schwann cells in a process that parallels embryonic development: migrating neural crest cells move through immature connective tissue before the time of nerve formation at E9-E11, and then differentiate into Schwann cell precursors (SCPs) between E12-E13. These Schwann cell precursors then become immature Schwann cells, which are generated from E14 until right after birth. The immature Schwann cells eventually differentiate into mature Schwann cells in the postnatal period. Cre-mediated recombination to ablate NF1 expression in the NCSCs using the Wnt-1 promoter does not result in neurofibroma formation (12). However, NF1 inactivation with Cre recombinase expression driven by the Krox20, P0A, or Dhh promoter does result in neurofibroma formation (–11). The Krox20Cre;NF1^flox/− mice develop only paraspinal neurofibroma, and require sacrificing at 10–12 months of age. On the other hand, the P0ACre;NF1^flox/− mice develop signs of sickness including lethargy, ruffled hair and hindlimb paralysis in their second year of life. These sick mutant mice exhibit neurofibroma formation throughout the PNS. The DhhCre;NF1^flox/flox also develop plexiform neurofibromas near the dorsal root ganglia, predominately at lower cervical or upper thoracic levels leading to paralysis and requiring sacrifice by 13 months of age. A portion of these mice also has neurofibromas located under the dermal muscle layer in the back at the thoracic and lumbar areas (–10). While PLPCre-ERT2;NF1^flox/− mutant mice robustly develop paraspinal neurofibroma near the DRG when NF1 is ablated at, or prior to, the neonatal period, the inactivation of NF1 specifically in adulthood rarely leads to plexiform neurofibroma formation in this model.