Figure legends
Figure 1. Analysis of NF1 pre-mRNA splicing.
(A) Schematic overview of NF1 exon trap constructs derived as part of this study, showing the approximate location of the tested variants. Exonic sequences are shown as boxes, intronic sequences as horizontal lines; intronic sequence representing a pseudo-exon is indicated by the shaded box. § indicates a variant also subjected to functional assessment (see Supplementary Table S2). Variant nomenclature is according to reference transcript NM_000267.3, except ¶: reference transcript NM_001042492.2.
(B) Approximate location of the NF1 variants investigated with RT-PCR and Sanger sequencing in patient RNA in relation to the NF1 reference transcript (NM_000267.3).
(C) Exon trap and RT-PCR analysis of the NM_000267.3(NF1):c.2710T>A variant. Agarose gel electrophoresis of the NF1 exon 21 wild-type (WT) and c.2710T>A exon trap RT-PCR products (left) and RT-PCR products from the index patient and 3 control individuals (center). A indicates the canonical exon 21 product; B indicates the product obtained due to the cryptic splice site at c.2706 created by the c.2710T>A substitution. Electropherogram of Sanger sequence analysis of the RT-PCR products, confirming the abnormal r.2707_2850del transcript.
(D) Exon trap and RT-PCR analysis of the NM_000267.3(NF1):c.3871-3T>G and c.3974+5G>C variants. Agarose gel electrophoresis of the NF1 exon 29 wild-type (WT) and c.3871-3T>G, c.3974+5G>C and c.3871-3T>G + c.3974+5G>C exon trap RT-PCR products (indicated with the red box). Note the complete loss of the wild-type product from the constructs containing both substitutions.
Figure 2. In vitro functional assessment of RAS GAP activity of the NF GRD.
(A) Schematic overview of the pull-down assay for NF RAS GAP activity. Variants were first introduced into the wild-type NF expression construct by site-directed mutagenesis (SDM), co-transfected into mammalian cells in culture, together with a RAS reporter expression construct and, after 5 minutes stimulation with EGF, the cells were lysed and GTP-bound RAS subjected to GST-RAF-RBD pull-down with glutathione-agarose beads. Lysate and pull-down fractions were subsequently analyzed by immunoblotting.
(B) Schematic overview of NF (above), showing the different proposed functional domains (CSRD, orange; TBD, green; SPRED1 interaction, violet; SEC14-PH, cyan; HEAT, pink; SBD, blue) and the NF GRD expression constructs used in this study (below). Amino acid changes are given according to reference transcript NM_000267.3. Variants derived in the NF V5-p.1180_1504 expression construct [Thomas et al., 2012] are indicated in cyan; other variants were derived from the wild-type NF p.V5-1180_1504-V5 expression construct.
(C) Relative RAS GAP activity of the NF p.V5-1180_1504 variants. NF V5-p.1180_1504 variants were co-expressed with HA-H-RAS and RAS GAP activity was estimated in a pull-down assay using recombinant GST-RAF-RBD to specifically bind RAS-GTP. The signals for HA-H-RAS in the pull-down fractions were determined in 3 independent experiments. The mean estimated RAS GAP activity is shown relative to the wild-type (V5-GRD); = 1.0). Error bars represent the standard error of the mean. Variants showing significantly reduced RAS GAP activity are indicated in red (see main text for details).
(D) Relative RAS GAP activity of the NF p.V5-1180_1504 variants. Pull-down assay was performed as in (C); RAS GAP activity was estimated relative to the wild-type control (V5-GRD-V5). Variants showing no evidence for RAS GAP activity are indicated in red; variants with reduced RAS GAP activity are shown in orange; active variants are indicated in black (see main text for details). Quantification data of at least 3 independent experiments is shown.
(E) Representative immunoblot showing the GST-RAF-RBD pull-down fractions.
(F) Cell lysate fractions corresponding to the samples shown in (E).
Figure 3. In vitro functional assessment of NF RAS GAP activity.
(A) Schematic overview of neurofibromin (NF) (top) and the expressed truncated NF proteins used for the functional experiments (below). The C-terminal region absent from the expressed NF p.1_2069-myc and p.420ins10-myc proteins is indicated and the approximate positions of the different variants are shown, including the 10 amino acid insertion (420ins10) that differentiates between the two expressed wild-type proteins. Variants derived in the NF p.1_2069ins10-myc construct are indicated in cyan; variants derived in the NF p.1_2069-myc construct are shown in black. To determine the RAS GAP activity of the variants, the scheme shown in Figure 2A was used. Briefly, NF p.1_2069-myc and NF p.1_2069ins10-myc variants were co-expressed with HA-H-RAS. RAS GAP activity was estimated in a pull-down assay using recombinant GST-RAF-RBD in at least 4 independent experiments, relative to the wild-type NF p.1_2069-myc or NF p.1_2069ins10-myc (indicated in cyan) proteins.
(B) Quantification of NF1 variant RAS GAP activity. Variants showing evidence of disruption of RAS GAP activity (> 50% reduction in activity; P < 0.05) are indicated in red; variants with reduced RAS GAP activity are shown in orange (< 50% reduction; P < 0.05); active variants (no evidence for reduced RAS GAP activity)(P > 0.05) are indicated in black (see main text for details). IT/RC: NF1 p.Ile1799Thr/Arg1809Cys (double cis variant).
(C) Representative immunoblot showing the lysate (above) and GST-RAF-RBD pull-down (below) fractions.
Figure 4. Functional assessment of the effects of NF1and SPRED1 variants on NF-SPRED1 interaction and NF expression.
(A) Schematic overview of the in vitro functional assessment of the NF1-SPRED1 interaction. Variants were introduced into the wild-type (WT) expression construct by site-directed mutagenesis (SDM) and the NF p.1_2069-myc or p.1_2069ins10-myc and FLAG-SPRED1 expression constructs co-transfected into mammalian cells. NF-SPRED1 complexes were isolated by immunoprecipitation (IP) using anti-FLAG agarose beads. Lysate and IP fractions were subsequently analyzed by immunoblotting. Signals for the variants relative to the WT proteins (NF p.1_2069-myc, NF p.1_2069ins10-myc or FLAG-SPRED1) were determined in at least 3 independent experiments.
(B) Quantification of the IP signals for the NF1 (left) andSPRED1 (right) variants. Variants showing a reduced interaction (> 50% reduction in signal in the coIP fraction; P< 0.05) are indicated in red; variants showing < 50% reduction (P < 0.05) are shown in orange; variants showing a comparable interaction to wild-type NF-SPRED1 (P> 0.05) are indicated in black (see main text for details). IT/RC: NF1 p.Ile1799Thr/Arg1809Cys (double cis variant).
(C) Quantification of the signals for the expressed NF1 (left) and SPRED1 (right) variants. Variants showing > 50% reduction (P < 0.05) are indicated in red; variants showing < 50% reduction (P < 0.05) are shown in orange; variants expressed at levels comparable to the wild-type control (P > 0.05) are indicated in black (see main text for details). IT/RC: NF1 p.Ile1799Thr/Arg1809Cys (double,cis variant).
(D) Representative immunoblot showing the anti-FLAG IP (above) and lysate (below) fractions for NF1 variants.