All animal experiments were carried out with strict adherence to National Institutes of Health (NIH) Guidelines for animal care and safety and were approved by the Mayo Clinic Institutional Animal Care and Use Committee.
Primary cell culture of Xenopus spinal neurons
Wild-type Xenopus laevis (Xenopus One) were maintained in approved animal facilities according to institutional guidelines. Experiments were conducted on spinal neurons prepared from neural tube dissections of one-day old (stage 22) X. laevis embryos . These cultures are used for experiments 10 h to 14 h after plating on fibronectin substrate at 20°C to 22°C. All cover glasses were coated with poly-D-lysine (5 mg/mL, Sigma, St. Louis, MO, USA) followed by fibronectin (20 μg/mL, Sigma). Culture medium consisted of 87.5% (v/v) Leibovitz medium (GIBCO, Grand Island, NY, USA) containing 0.4% (v/v) fetal bovine serum (HyClone, Logan, UT, USA), and 12.5% (v/v) saline solution (10 mM D-glucose, 5 sodium pyruvate, 1.26 mM calcium chloride (CaCl2) and 32 mM HEPES; pH 7.5). Experiments were performed in modified Ringers solution (120 mM sodium chloride (NaCl), 2.2 mM potassium chloride (KCl), 2 mM CaCl2, 1 mM magnesium chloride (MgCl2), 5 mM HEPES, 2 mM sodium pyruvate; pH 7.6). All animal research was performed with the approval of Mayo Clinic Institutional Animal Care and Use Committee.
Reagents, immunolabeling and microscopy
Spinal neuron cultures were treated with MAG-Fc (1 μg/mL, R&D Systems #538-M, Minneapolis, MN, USA; conjugated to Fc-specific goat anti-human immunoglobulin G, Jackson IR Labs 109-485-098, West Grove, PA, USA), BDNF (50 ng/mL, Peprotech #450-02, Rocky Hill, NJ, USA), both or a control BSA vehicle solution for predetermined times, followed by standard chemical fixation. MAG treatments for 5 min were utilized to obtain maximal integrin surface removal . Spinal neuron cultures were chemically fixed in a cytoskeleton-stabilizing buffer containing 2.5% paraformaldehyde and 0.01% glutaraldehyde for 20 min. All blocking and immunolabeling steps were performed in modified Ringers solution containing 5% goat serum. Alexa-dye-labeled secondary antibody conjugates (Invitrogen, Carlsbad, CA, USA) were used at 2 μg/mL. We immunolabeled unpermeabilized cells using a monoclonal antibody to the extracellular domain of β1-integrin (8c8-c, 0.8 μg/mL, University of Iowa Developmental Studies Hybridoma Bank) or with a β1-integrin function-blocking antibody (2999, 0.4 μg/mL, K. Yamada) along with a polyclonal anti-β-tubulin antibody (0.4 μg/mL, Abcam ab15568, Cambridge, England). Antibody staining for vinculin (2 μg/mL, Sigma V931), FAK (2 μg/mL, Santa Cruz Biotechnology sc-557, Santa Cruz, CA, USA), phosphotyrosine p-Tyr PY99 (1 μg/mL, Santa Cruz BioTechnology sc-7020), α5-integrin (1 μg/mL, D. DeSimone), TrkB (4 μg/mL, Novus NB100-92063, Littleton, CO, USA), α-actinin (1 μg/mL, Santa Cruz Biotechnology sc-59953) and talin (0.8 μg/mL, Abcam ab1188) was performed on permeabilized cells (0.1% Triton-X-100), followed by Alexa555 secondary antibody conjugates. Fluorescence microscopy was performed using a Zeiss (Jena, Germany) LSM 5LIVE confocal microscope equipped with a 63 × oil immersion objective (1.4 numerical aperture, 1.6 × optical zoom) with identical acquisition settings for control and experimental groups. The paxillin-GFP construct was provided by T.M. Gomez (University of Wisconsin, Madison). GFP-paxillin RNA was injected into four-celled Xenopus embryos. Spinal neurons were plated onto laminin (80 μg/mL) and incubated for 3 h to 4 h at 23°C to 25°C before use. High-magnification images were acquired using a Zeiss AxioCam MRm and 100 ×/1.45 NA objective on a Zeiss TIRF microscope.
Image analysis and processing
To measure only receptors at the plasma membrane, permeabilized growth cones were excluded inadvertently from the analysis, as identified by tubulin immunofluorescence with polyclonal anti-β-tubulin. The original 14-bit images were analyzed using ImageJ (Bio-Formats ZVI plug-in, Madison, WI, USA). A region of interest encompassing the entire growth cone (defined as the distal 20 μm) was used to determine the mean fluorescence intensity of thresholded images (identical for experimental and control conditions). Data were background subtracted and normalized to the appropriate control images. Growth cones with a mean β1-integrin fluorescence level less than double the background fluorescence were excluded from analysis. For quantification of integrin clustering, we used a three-fold fluorescence inclusion criterion of β1-integrin cluster intensity over the mean background fluorescence in the growth cone central domain. Only clusters within the distal 20 μm of the axon and within 5 μm of the lateral borders of the growth cone (growth cone peripheral domain) were included. Quantification of β1-integrin clustering for quality control analysis was performed by randomly selecting growth cones (n = 50) for reevaluation with a range of fluorescence threshold values (two-, two and a half-, three-, and four-fold above the background fluorescence) . To quantify co-labeling with β1-integrin puncta, a two-fold fluorescence increase relative to the mean fluorescence in the growth cone central domain was utilized for identifying cytoplasmic protein clusters. The growth cone diameter was determined using ImageJ by measuring from the lateral edges of lamellipodia at the widest point of the growth cone, excluding filopodia. The GFP-paxillin TIRF microscopy movie was made using ImageJ software by exporting time-lapse stacks to a QuickTime format (MOV, MPEG4 compression, 3 frames per second).
Manipulation of β1-integrin clustering
We disrupted BDNF-induced integrin clustering using a 30-min pretreatment with the glycosphingolipid L-t-LacCer: β-D-lactosyl-N-octanoyl-L-threo-sphingosine (20 μM, Avanti Polar Lipids, Alabaster, AL, USA). The natural stereoisomer D-e-LacCer: D-lactosyl-β1-1'-N-octanoyl-D-erythro-sphingosine (20 μM, Avanti Polar Lipids) was used as a control lipid. Both L-t-LacCer and D-e-LacCer were complexed to defatted BSA and incubated with cells at a final concentration of 20 μM . Integrin function was inhibited by a 20-min pretreatment with the function-blocking antibody 2999 (5 μg/mL). To buffer intracellular Ca2+, neuron cultures were incubated for 30 min in low-Ca2+ (30 nM) solution consisting of 50% culture medium and 50% ethylene glycol tetraacetic acid (EGTA)-buffered saline (120 NaCl mM, 4.9 KCl mM, 1.55 mM MgCl2, 1.25 mM glucose, 5 mM sodium pyruvate, 4 mM HEPES, 0.65 mM EGTA, pH 7.6) and BAPTA-AM (1,2-bis-(o-aminophenoxy)-ethane-N, N, N', N'-tetraacetic acid, tetraacetoxymethyl ester, 1 μM; Calbiochem, Gibbstown, NJ, USA) or dimethyl sulfoxide vehicle for 30 min, followed by consecutive washes in low-Ca2+ saline. To broadly disrupt Ca2+ influx via voltage-dependent Ca2+ channels in the plasmalemma, neurons were pretreated for 20 min with CdCl2 (50 μM, Sigma) as reported previously .
Live-cell Ca2+ imaging
X. laevis spinal neurons were loaded with the fluorescence Ca2+ sensor Fluo-8H (2 μM, 30-min loading in culture medium containing 0.01% pluronic acid, AAT Bioquest, Sunnyvale, CA, USA). Growth cones were subjected to Ca2+ imaging within 45 min of dye loading, using a Zeiss 200 M inverted microscope equipped with a 100X/1.45 NA objective and EM-CCD camera (Hamamatsu, Bridgewater, NJ, USA). Image acquisition was every 15 s and was started at least 2 min prior to application of BDNF (50 ng/mL). To measure fluorescence intensity, each series of images was thresholded to eliminate background noise, and the mean fluorescence intensity within a region of interest drawn around the growth cone was measured using ImageJ software (NIH). For each growth cone, the mean fluorescence intensity of each image after BDNF treatment (F) was then compared to the mean baseline fluorescence (pre-BDNF treatment; F0) to obtain the displayed value of normalized fluorescence (F/F0).
Functional axon outgrowth assay
For measurements of neurite outgrowth in response to the various treatment groups, a series of time-lapse images were taken to record growth over a 60-min period (ProgRes CapturePro 2.7, Jenoptik Inc., Jupiter, FL, USA). Neurons were pretreated with L-t-LacCer or D-e-LacCer for 30 min in appropriate assays. A 20-min pretreatment was used for outgrowth assays with the function blocking antibody 2999 and a control antibody. All outgrowth assays were performed on a Zeiss 40 compact fluorescent lamp microscope equipped with a Ludl Electronic Products (Hawthorne, NY, USA) BioPoint 2 motorized stage, cooled charged-coupled device camera and a 20 × objective. Only axons > 50 μm in length were included in the analysis. Analysis was conducted using ImageJ.
All statistical analyses were performed using GraphPad Prism software (v5, La Jolla, CA, USA). The figure legends state the statistical tests used. Data with a normal distribution (D'Agostino and Pearson omnibus normality test) were assessed using repeated-measures one-way analysis of variance with a Tukey post hoc analysis.