Figure 1.
(A) Schematic representation of mouse Bmcc1 gene. Exons are boxed, in black for the coding sequence and in white for the 5′ and 3′ non-coding sequences. Primers for 5′ RACE and RTPCR experiments are indicated by arrows under exons 11, 12 and 21. (B) Schematic representation of mouse Bmcc1 transcript. (C) Schematic representation of Bmcc1s cDNA and protein. The BNIP2 homology and BCH domains are indicated. Asterisks show the antigenic peptides used to generate the Bmcc1s antiserum.
Figure 2.
(A) Immunoblot of Bmcc1s in lysates of HeLa cells transfected with a plasmid expressing Bmcc1s-V5. Similar profiles were obtained using the Bmcc1 antiserum or anti-V5 antibodies. Note that the Bmcc1 antiserum recognized an endogenous protein around 50 kDa (arrow) of the same size as Bmcc1s in untransfected HeLa cells. Immunostaining of HeLa cells transfected with a plasmid expressing Bmcc1s-V5, using either the Bmcc1 antiserum or anti-V5 antibodies. The antiserum detected only the V5 positive cells, and both signals overlapped. Scale bar: 100 µm (B) Immunoblot of endogenous Bmcc1 isoforms in mouse tissue lysates using Bmcc1 antiserum. GAPDH expression is shown as a loading reference. As in HeLa cells expressing Bmcc1s-V5, the Bmcc1 antiserum detected a band around 50 kDa (arrow) in the brain lysate that appeared specific to this tissue and was the most abundant among the Bmcc1 isoforms. (C) Immunoblot of endogenous Bmcc1 in primary cultures of astrocyte and neuron lysates at DIV7, using Bmcc1 antiserum. As found in brain tissues, a major band around 50 kDa was detected (arrow).
Figure 3.
Subcellular localization of Bmcc1s in primary cultures of astrocyte.
(A–C) Confocal section images of primary astrocytes immunostained for endogenous Bmcc1s (green) and α-tubulin or GFAP (red). Merge images showed that Bmcc1s forms punctate spots mainly distributed along α-tubulin stained microtubules (A) and partially colocalized with GFAP-positive intermediate filaments (C). Boxed regions in A indicate the fields enlarged in each image. B. In nocodazole-treated primary astrocytes (10 µM, 1 h), Bmcc1s followed the disrupted α-tubulin microtubular staining. (D) Immunogold labelling and electron mircroscopy analysis of primary astrocytes showed that Bmcc1s localized on cytoskeleton-type structures compatible with microtubules (left) and intermediate filaments (right). Bars: 10 µm (A–C); 200 nm (D).
Figure 4.
Subcellular localization of Bmcc1s in primary neurons.
(A–C) Confocal section images of primary neurons after 7 days of culture immunostained for endogenous Bmcc1s (green) and α-tubulin or neurofilament subunit M (NF-M) (red). Merge images showed that Bmcc1s colocalizes with α-tubulin (A) and NF-M (C) immunoreactivity signal. Boxed regions in A and C indicate the fields enlarged in each image. B. In nocodazole-treated primary neurons (10 µM, 1 h), Bmcc1s followed the disrupted α-tubulin microtubular staining. (D) Immunogold labeling and electron microscopy analysis of primary neurons showed that Bmcc1s localized on cytoskeleton-type structures compatible with microtubules (left) and intermediate filaments (right). Bars: 10 µm (A–C); 100 nm (D).
Figure 5.
Microtubule Co-sedimentation assays.
Taxol stabilized microtubules were incubated with or without GST-Bmcc1s. The samples were then sedimented through a 60% glycerol cushion. The supernatants (S) and pellets (P) were separated by SDS-PAGE and stained with Coomassie Blue. Tubulin (50 kDa) was mostly present in the pellet fraction with or without Bmcc1s, while Bmcc1s was only detectable in the supernatant.
Figure 6.
(A) GST-Bmcc1s or GST immobilized on glutathione sepharose beads were incubated with either a lysis buffer or a mouse brain lysate. After elution, bound proteins were resolved on SDS-PAGE in parallel with the mouse brain lysate, and visualized by Coomassie staining. A unique band (square) was analyzed by MALDI-TOF, where MAP6 was identified. (B) The presence of MAP6 and the specificity of its interaction with Bmcc1s were confirmed by Western blot of the GST eluates with 23N, a polyclonal anti-MAP6 antibody. Several bands corresponding to the neuronal MAP6 isoforms N-STOP (120 kDa) and E-STOP (80 kDa), the astrocyte MAP6 isoform A-STOP (60 KDa) and a 48 kDa isoform described in total brain protein extracts were revealed. (C) MALDI-TOF analysis revealed the presence of 4 peptides (in red) corresponding to MAP6. The microtubule-stabilizing modules Mn1, Mn2 and Mc1 of MAP6 are underlined. (D) Co-immunoprecipitation of MAP6 and Bmcc1s was performed using the 175 monoclonal anti-MAP6 antibody (IP+αMAP6), or no antibody (IP-αMAP6) as control, on mouse brain lysates. Precipitates were analyzed by Western blotting with Bmcc1 antiserum, in parallel with the mouse brain lysate. Bmcc1s was co-immunoprecipitated with MAP6. (E) Pull-down experiments of purified MAP6 isoforms: neuronal, N- and E-STOP and the fibroblast F-STOP, by purified glutathione-S-transferase (GST)-Bmcc1s or GST. Bound proteins were resolved on SDS-PAGE and Coomassie stained. N- and E-STOP were specifically retained by GST-Bmcc1s.
Figure 7.
Bmcc1s inhibits the MAP6-induced microtubule cold stability.
(A) Inhibition of N-STOP-induced microtubule cold stability by Bmcc1s in vitro. Microtubules polymerized at 37°C and subjected to cold were recovered by sedimentation and analyzed by SDS-PAGE and coomassie staining. The observed 50 kDa band corresponds to polymerized tubulin. At 4°C, almost no microtubules could be recovered. In contrast, they were preserved at 4°C in presence of N-STOP or F-STOP. Adding increasing concentrations of GST-Bmcc1s progressively decreased the level of microtubules in presence of N-STOP, but not of F-STOP. In contrast, GST alone had no effect. Numbers indicate the final concentration of the proteins in micromolar in the depolymerization reaction mix. Concentration of tubulin was 30 µM. (B,C,D) Confocal microscopy image projections of cells transiently transfected with a plasmid expressing Bmcc1s-V5. Twenty-four hours after transfection, cells were exposed to 0°C for 45 minutes. Following free tubulin extraction by cell permeabilization, cells were fixed and double-stained for α-tubulin antibody (red), and V5 (green). Nuclei were stained with DAPI (blue). (B) HeLa cells stably transfected with GFP-N-STOP; (C) Primary culture of astrocytes; (D) Primary culture of neurons. In Bmcc1s-V5 transfected cells (green), α-tubulin staining was almost gone and V5 staining either retracted in a ball shape in the case of GFP-N-STOP HeLa cells and astrocytes, or filled the cell body in neurons. Bars: 10 µm.
Figure 8.
Bmcc1s overexpression displaces MAP6 away from the microtubules and induces the formation of membrane protrusions.
(A) HeLa cells stably transfected with GFP-N-STOP (GFP-N-STOP HeLa) transiently transfected with an expression plasmid for Bmcc1s-V5. Twenty-four hours after transfection, cells were fixed and double-stained for N-STOP using the 23N polyclonal MAP6 antibody (green) and for Bmcc1s-V5 using a monoclonal anti V5 antibody (red). In untransfected GFP-N-STOP HeLa, N-STOP staining showed a microtubule-like pattern. The Golgi apparatus was also labeled (asterisks). Insert (a) is an enlargement of the squared region showing N-STOP staining in more detail. In the Bmcc1s-V5 GFP-N-STOP HeLa transfected cell (white arrows), N-STOP labeling became brighter, no longer featuring its typical microtubule-type distribution, and numerous membrane protrusions (white arrowheads) labeled for both V5 and N-STOP were seen. Insert (b) is an enlargement of the Bmcc1s-V5 GFP-N-STOP HeLa transfected cell. (B) Confocal microscopy images of HeLa cells transiently transfected with an expression plasmid for Bmcc1s-V5. Twenty-four hours after transfection, cells were fixed and double-stained for Bmcc1s-V5 using a monoclonal anti V5 antibody (green) and for F-actin using TRITC-conjugated phalloidin (red). (C) Confocal microscopy images of a Bmcc1s-V5 stably transfected HeLa cell (Bmcc1s-V5 HeLa) transiently transfected with an expression plasmid for GFP-N-STOP. Twenty-four hours after transfection, cells were fixed and stained for N-STOP using the 23N polyclonal anti-MAP6 antibody (green), for microtubules using a α-tubulin antibody (blue), and for F-actin using TRITC-conjugated phalloidin (red). Merge images show that N-STOP partially loses its microtubular staining, being more diffuse in the cell, and located in actin-rich membrane protrusions (white arrowheads). Bars: 10 µm.
Figure 9.
Morphological changes induced by Bmcc1s overexpression requires MAP6.
Confocal microscopy image projections of cells transfected with a Bmcc1s-V5 or GFP expressing plasmid, and stained for V5 (green) and F-actin (detected with TRITC-conjugated phalloidin in red). Cells were fixed 24 h after transfection. (A) primary astrocytes; (B) primary neurons. The morphology of GFP-expressing cells (green) was unchanged compared to untransfected cells. In contrast, Bmcc1s-V5-expressing astrocytes and neurons developed numerous membrane protrusions (white arrowheads). Images in B illustrate representative confocal projections of the effect of Bmcc1s-V5 on neuritic growth and number in wild-type neurons. The whole Bmcc1s-V5 transfected neuron is shown in the insert. Histograms present means ± sd of the length of the longest neurite and of the number of neurites. *** p-value<0.0001 ** p-value<0.001. ns, not significant for 3 independent experiments using the two sample independent t-test. In neurons, length of the longest neurite, and number of neurites (or cell extensions starting from the soma) were significantly increased by Bmcc1s-V5 transfection, but not in MAP6-deleted neurons. Bars: 10 µm.