Plexiform and dermal neurofibromas and pigmentation are caused by Nf1 loss in desert hedgehog-expressing cells

Abstract

Neurofibromatosis type 1 (Nf1) mutation predisposes to benign peripheral nerve (glial) tumors called neurofibromas. The point(s) in development when Nf1 loss promotes neurofibroma formation are unknown. We show that inactivation of Nf1 in the glial lineage in vitro at embryonic day 12.5 + 1, but not earlier (neural crest) or later (mature Schwann cell), results in colony-forming cells capable of multilineage differentiation. In vivo, inactivation of Nf1 using a DhhCre driver beginning at E12.5 elicits plexiform neurofibromas, dermal neurofibromas, and pigmentation. Tumor Schwann cells uniquely show biallelic Nf1 inactivation. Peripheral nerve and tumors contain transiently proliferating Schwann cells that lose axonal contact, providing insight into early neurofibroma formation. We suggest that timing of Nf1 mutation is critical for neurofibroma formation.

Figure 1. Acute loss of Nf1 in dorsal root ganglion cells at E12.5+1 results in colony formation in vitro

(A). Quantification of colonies resulting from acute loss of Nf1 in E8.5 neural tube-derived crest cells, E12.5 dorsal root ganglion, (DRG) cells or differentiated Schwann cells (denoted E18.5) derived from Nf1 ^flox/flox mice cells. A few colonies were observed in the E12.5 DRG cells infected by adeno-GFP, but could not be expanded. (Mean ± SD). (B). Phase micrograph of expandable adherent colonies resulting from E12.5 DRG cells infected by adeno-Cre. (C). Adeno-Cre mediated recombination confirmed by PCR. N: Nf1 flox allele, R-Cre: recombined Nf1 allele. MOI = multiplicity of infection. Cells replated from colonies after E12.5 DRG-derived recombination (D) or adeno-GFP (E) immunostained with anti-Blbp and staining detected with a fluorescently tagged anti-rabbit antibody. (F, G). Differentiation of E12.5 DRG-derived colony-forming cells in vitro. E12.5 DRG-derived colonies were replated and cells cultured under gliogenic and smooth muscle/myofibroblast generating culture conditions for 12 days, immunostained with anti-MBP (F) or anti-SMA (G) antibodies and staining detected with appropriate second antibodies. Bars, B=50μm, D, E=26μm, F, G =15μm.

Figure 2. The DhhCre driver elicits recombination beginning at E12.5 in boundary cap cells and embryonic Schwann cells

Cryostat sections from Nf1 ^flox/flox; DhhCre; EGFP embryos dissected at E11.5 (A), E12.5 (B), and a section through spinal cord, nerve roots and peripheral nerve at postnatal day 165 (C). A – C, sections were stained with anti-EGFP antibody. B, arrows show position of EGFP+ dorsal root (BC-Dr) and ventral root (BC-Vr) boundary cap cells adjacent to the neural tube (Nt). The position of the largely unstained dorsal root ganglion (Drg) is marked, and some green cells are visible lateral to the DRG in the proximal nerve (PN). In D–J, Confocal analysis of adult DRG showing EGFP+ cells are S100β+ (D), Gfap-negative (G) and β(III)-Tubulin-negative (H). YZ view (E) and XZ view (F) of cell indicated by arrow in D. YZ view (I) and XZ view (J) of cell indicated by arrow in H. Note Gfap-positive (red) astrocytes in the spinal cord (G) and β(III)-Tubulin-positive (red) DRG neurons and spinal cord fibers (H). K – M are gross images. K, a developing spinal cord (E 12.5) with EGFP+ root cells is shown. L, a postnatal day 1 sciatic nerve is shown. M, strong expression is retained in P90 sciatic nerves. L and M, endoneurium containing Schwann cells is EGFP+ (double headed black arrow) and perineurium (adjacent to white arrows) is negative. N, DNA (50ng) isolated from designated tissues from a representative adult Nf1 ^flox/flox; DhhCre mouse was PCR amplified. An ethidium bromide stained gel demonstrates Nf1 recombination in sciatic nerve using the DhhCre driver with little recombination in other organs. FL/FL = Nfl flox allele. Bars: A–C, K–M=26μm, D–J=20μm.

Figure 3. Nf1 ^flox/flox; DhhCre mice show early lethality and develop tumors

(A) Kaplan-Meyer survival curve. Red, Nf1 ^flox/flox; DhhCre (n=28); blue, Nf1 ^flox/+; DhhCre (n=20); green, Nf1 ^flox/flox (n=9); brown, Nf1 ^flox/+ (n=8). To facilitate view of control mouse survival, blue green and brown lines are shown offset. (B–F) Gross dissections of tumors and areas of pigmentation in Nf1 ^flox/flox; DhhCre mice. In gross images B–E, a ruler showing 1mm markings is included. (B) Nf1 ^flox/flox (wild type) spinal cord from 9 month old mouse, with pairs of spinal roots attached. (C) A similar region from an Nf1 ^flox/flox; DhhCre mouse, showing three paraspinal GEM-neurofibromas (white arrows) and spinal cord compression. (D, E) Dermal GEM-neurofibromas (white arrows) in Nf1 ^flox/flox; DhhCre mice in anterior (D) and lateral views (E). (F) Pigmentation was often detected on the surface tumors or adjacent to tumor; a dramatic example is shown. White arrows point to black pigmentation. Bars=1mm.

Figure 4. Nf1 ^flox/flox; DhhCre mice: histological and electron microscopic analysis

Paraffin embedded tissue sections from plexiform and dermal GEM-neurofibromas stained with hematoxylin and eosin (H&E) (A and C). Adjacent sections were stained with anti-S100β antibody visualized with DAB [brown] (B and D). (E) Plexiform neurofibroma sections stained with Masson’s trichrome showing abundant collagen (blue) in the tumor. (F) Toluidine blue staining showing metachromatic mast cell infiltration (black arrows). Electron micrographs of plexiform (G) and dermal GEM-neurofibromas (H); small black arrowheads point to non-myelinating Schwann cells that are apparently not wrapping multiple axons. White arrows in G and H indicate intact myelinated fibers. In I, a higher magnification electron micrograph shows abundant collagen deposition (C) and prominent Schwann cell cytoplasmic processes identified by continuous basal lamina (black arrows). Fibroblasts are identified by prominent intracellular vesicles, prominent endoplasmic reticulum and patchy basal lamina. (Bar: A–F, 20μm; G, H=10μm; I=1μm).

Figure 5. Only p75+ cells recombine Nf1 in Nf1 ^flox/flox; DhhCre mouse nerves and tumors

EGFP expression (B, D) in sections from adult Nf1 ^flox/flox; DhhCre; EGFP sciatic nerves, cryo-sectioned and stained with anti-von Willebrand factor to mark endothelial cells (red, A), or anti IgE Fc receptor to mark mast cells (red, C). In B and D show merged images of EGFP fluorescence and immunostains. FACS sorted EGFP+ (E, F) and EGFP- (G, H) cytospun cells were immunostained with anti-p75 antibodies (E, G). All EGFP+ cells are p75+, but some EGFP-are p75+ while other are p75-negative cells (H, white arrows). FACS sorted EGFP+ cells from P1/P2 sciatic nerves (I–L) or tumor were stained with Blbp ( I, M, red) and S100β (J, N, green). Bars, A–P: 13 μm. FACS sorted cells from P1 sciatic nerve or tumor were analyzed for Cre mediated recombination of Nf1 by PCR (Q, R). Q, Endothelial cells (CD31+/CD45−; EC) or mast cells (c-kit+/IgE-FcR+; MC) from dissociated tumor cells did not recombine Nf1. DNA from cells from two Nf1 ^flox/flox; DhhCre; EGFP tumors (T1, T2) show recombination. Nf1fl allele (FL/FL), recombined Nf1 allele (R-Cre). R, postnatal day 1 EGFP+ cells (EGFP+) showed Nf1 recombination, as did tumor DNA (T1). Water is shown as a negative control.

Figure 6. Nf1 ^flox/flox; DhhCre mouse nerves show disrupted morphology and transiently increased Schwann cell proliferation

Paraffin sections from 6.5 month old wild type (A) and Nf1 ^flox/flox; DhhCre (B) mouse sciatic nerves. Bars in A, B = 50μm. Electron micrograph from wild type (C) and Nf1 ^flox/flox; DhhCre (D) saphenous nerve. Black arrows in C indicate well organized wild type non-myelinated axons; those in D show disorganized non-myelinated axons. Bars in C, D = 10 microns. (E) Quantification of BrdU positive cells in Nf1 ^flox/flox; DhhCre P1, P15, and P60 sciatic nerves (white bars) and wild type littermate controls (black bars). Proliferation in the Nf1 ^flox/flox; DhhCre mouse nerve compared to wild type controls differs significantly in P1 sciatic nerves (p<0.01, n=3) and P15 sciatic nerves (p<0.001, n=3), but not in P60 sciatic nerves (p=0.17, n=3). Data analyzed by Student’s t-test are presented as mean values ± standard deviation. (F) Quantification of apoptotic cells in Nf1 ^flox/flox; DhhCre P15 sciatic nerves (white bars) and wild type littermate controls (black bars). There is no significant difference in TUNEL staining between Nf1 ^flox/flox; DhhCre mice and their littermates (p= 0.42, n=4, data analyzed by Student’s t-test are presented as mean values ± standard deviation. A representative image of double staining is shown in I (BrdU: red; P75: green). (G) FACS analysis of BrdU labeled P1/P2 Nf1 ^flox/flox; DhhCre sciatic nerve cells indicating BrdU+/EGFP+ and BrdU+/EGFP− cells (G, H). (J) Sorted BrdU+/EGFP+ cells are S100β positive (red).

Figure 7. Proposed model of neurofibroma initiation in the neural crest-derived glial lineage

In the paraspinal region of the trunk, migrating neural crest (NC) cells (E8.5–9) give rise to boundary cap (BC) cells (E10.5–11) in the spinal roots and Schwann cell precursors (SCP, E12–13) in the peripheral nerve. The diagram shows a crest cell and general appearance of the nerve at subsequent developmental stages. Markers used in this report and expression over time are designated by horizontal boxes (P75, Dhh, Blpb, S100β, Gfap). We place the neurofibroma-initiating cell at the Schwann cell precursor/immature Schwann cell boundary.