Leech central nervous system structure

All protocols regarding the use of leeches were carried out in strict accordance with the French legislation and European Treaty, and in compliance with the Helsinki Declaration. The adult leeches Hirudo medicinalis were obtained from Biopharm (Hendy, UK). After anesthesia in 10% ethanol at 4 °C for 15 min, the CNS were dissected out in a sterile Ringer solution (115 mM NaCl, 1.8 mM CaCl2, 4 mM KCl, 10 mM Tris maleate, pH 7.4) under a laminar flow hood. After isolation of CNS, the samples were placed in 3 successive baths of antibiotics (100 UI/mL penicillin, 100 µg/mL streptomycin and 100 µg/mL gentamycin) for 15 min and later incubated in complete medium, made of Leibovitz L-15 medium (Invitrogen, Carlsbad CA, USA) complemented with 2 mM L-glutamin, 100 UI/mL penicillin, 100 µg/mL streptomycin, 100 µg/mL gentamycin, 0.6% glucose, 10 mM Hepes and 10% Exosome-depleted FBS Media Supplement (SBI System Bioscience, Palo Alto CA, USA). In situ hybridization and immunohistochemical analyses were performed on injured CNS by crushing the two connectives between the second and third ganglia, from a four ganglia long fragment.

Neurons and microglial cell preparation

The whole CNS was placed in 35 mm Petri dishes with 200 µL of complete medium. Each ganglion was carefully decapsulated by removing the collagen layer enveloping the nerve cords. The nerve cells, neurons (10–70 µm) and microglial cells (5 µm), were mechanically collected by gentle scraping and filtered through different size filters for separating the population according to size. Afterwards, the cell debris were eliminated in a 100 µm pluriStrainer filter (Dominique Dutscher, Brumath, France). Microglia were selected through a filter of 6 µm pluriStrainer and the neurons were collected in the upper part of this filter. The enriched microglial cells or neurons were centrifuged at 1,200 × g for 10 min at Room Temperature (RT). The cell pellet, corresponding to one nerve cord, was resuspended in 200 µL complete medium for the migration assays. Regarding the preparation of conditioned medium, the pellet, corresponding to 10 nerve cords, was resuspended in 500 µL complete medium (See Supplementary Method 1). The cell-free supernatant of each of the microglial cells and neurons from 10 nerve cords was used as conditioned medium (CM) in the chemotaxis experiments.

Molecular characterization

In a Hirudinea Genomics Consortium, we contributed to create a Hirudo medicinalis draft genome as previously described58. Sequences were assembled from paired short reads using Velvet and PHRAP/CONSED algorithms59,60 and given to GlimmerHMM to get predicted mRNA database61. These predicted mRNA sequences was compared in a Local BLAST program with human TGF-β type I receptor and TGF-β1 amino acid sequences62. The candidate sequences was submitted to Swiss-Prot databases using BLAST in order to specify similarities in TGF-β type I receptors and TGF-β superfamily respectively. From putative partial mRNA sequences, specific primers were designed to get the natural and complete sequences by RACE-PCR from CNS total RNAs (Supplementary Methods S2 and 3). PCR products were ligated into the pGEM T-easy vector (Promega, Madison WI, USA) and cloned into JM109 cells according to the manufacturer’s instructions. Finally, products were sequenced using BigDye Terminator v3.0 polymerization kit before detection on Genetic Analyzer (Applied Biosystems, Foster City CA, USA). The assembly of 5′ and 3′ end sequences allowed characterizing the full length mRNA of tgfbr1 and ngdf encoding respectively ALK4/5 (GenBank accession number MH346327) and nGDF (GenBank accession number MH346328) proteins.

Gene expression analysis

The neurons were collected, as described above, from the CNS of 10 leeches for each experimental condition and incubated in complete medium. The total RNA extraction was performed as described in Supplementary Method S2. cDNA library was generated from 2 µg of total RNA using random primers and Superscript III Reverse Transcriptase kit (Invitrogen, Carlsbad CA, USA) in a final volume of 20 µL. cDNAs were treated with RNaseH (Promega, Madison WI, USA) to optimize the amplification reaction product. Real-time quantitative PCR (qPCR) were performed with the Platinum SYBR Green qPCR SuperMix (Invitrogen, Carlsbad CA, USA) by combining 2 µL of cDNA template, 2 µL of primer mix (10 mM) and 25 µL of Platinum SYBR Green qPCR SuperMix-UDG in a final volume of 50 µL. Specific primers were designed for the qPCR analyses, for ngdf gene (5′-TGCTTGTGGTTCTCGGACTC-3′, 5′-TTTCGCTCTGATCTGCTGCA-3′), hmc1q gene (5′-GTCTCGGGAGTGCAAGGAAT-3′, 5′-TGTATTGTTCCCGACTCGCC-3′) and for a leech 18 S ribosomal RNA (5′-GGAGGAGCGCGTTTATTAAG-3′, 5′-GGGCACACACTTGAAACATC-3′), used as normalizer. The qPCR reactions were conducted on CFX 96 Real-Time System (BioRad, Hercules CA, USA) with the following conditions: 2 min at 50 °C (1 cycle), 2 min at 95 °C (1 cycle), 30 s at 95 °C, 30 s at 58 °C and 30 s at 60 °C (39 cycles) followed by a final melting curve to control the amplified specificity. The expression level of ngdf gene was compared between neurons 15 minutes (T0) and 24 hours after lesion (T24 h). The expression level of hmc1q gene was compared, at first, between naive and stimulated neurons with recombinant human TGF-β (20 ng/mL, Sigma-Aldrich, Saint Louis MO, USA), as ALK5 ligand. In a second time, hmc1q gene was compared between the unstimulated neurons and the neurons in co-culture with the microglia cells, separated thanks to a Transwell® porosity 0.4 µm membrane (Corning, Corning NY, USA), with or without SB431542 (20 µM, R&D Systems, Minneapolis MN, USA), an ALK5-specific and ALK4-relative inhibitor34. Experiments were done on triplicate samples in different sets of cDNA template. The analysis of relative gene expression of hmc1q and ngdf was calculated using the 2−ΔΔCt method63. Statistical analyses were performed by Paired T-test using GraphPad Prism 6.0 software. Statistical differences were considered to be significant if p-value was <0.05.

Fluorescent in situ hybridization (FISH)

Nerve cords were incubated 24 hours post-lesion and fixed for one hour at 4 °C in 4% paraformaldehyde. Digoxigenin-UTP-labelled specific antisense and sense riboprobes (negative control) were generated. The riboprobes of tgfbr1 were generated from 688–1356 nucleotides sequence mRNA (Genbank Accession Number MH346327) (size 668nt) with specific Forward (5′-AAGTGTGGAGGGGTGTATGG-3′) and Reverse (5′-CTCTTCGTGCGTTGGATCAG-3′) primers and from the 156–592 nucleotides sequence of ngdf (Genbank Accession Number MH346328) (size 436nt) with specific primers (5′-CATCATCTTCACCGCCACCT-3′, 5′-GTTGGGATCGCTGAGTTTGC-3′). After PCR amplification and the insertion of the product in pGEM-T easy vector system (Promega, Madison WI, USA), the RNA sequence of interest was obtained by in vitro transcription using DIG RNA-labeling kit according to the manufacturer’s instructions (Roche Diagnostics, Risch-Rotkreuz, Swiss).

The hybridization protocol was performed in nerve cords as listed below64, with minor modifications (See Supplementary Method S4). After the hybridization protocol, the nerve cords were mounted on the slide with Dako Fluorescent Mounting Medium (Agilent, Santa Clara CA, USA). Slides were kept at 4 °C in the dark until observation with a Zeiss LSM700 confocal microscope connected to a Zeiss Axiovert 200 M with an EC Plan-Neofluar 40x/1.30 numerical aperture oil immersion objective (Carl Zeiss AG, Oberkochen, Germany). Processing of the images was performed using Zen software and applied on the entire images as well as on controls. The presented pictures are representative of independent triplicates.


In experiments with rabbit polyclonal anti-human TGF-β1 (1/100, ab92486, Abcam, Cambridge, UK) and rabbit polyclonal anti-human ALK5 (1/250, ab125310, Abcam, Cambridge, UK) antibodies, analyses were performed on experimentally injured nerve cords, as described above, and incubated in complete L-15 medium 15 minutes (T0), 6 hours (T6 h) or 24 hours (T24 h).

The experiment using anti-human TGF-β1 was also performed with and without ligation between ganglia 2 and 3, which were tied up with a loop of fine nylon thread to confine the lesion region.

The experiments using rabbit polyclonal anti-human C1qBP (1/500, HPA026483, Sigma-Aldrich, Saint Louis MO, USA) were performed on injured nerve cords, as described above, 24 h after incubation in complete L-15 medium and with injection of SB431542 (20 µM, R&D Systems, Minneapolis MN, USA) as ALK4/5-specific inhibitor, or its vehicle DMSO (100 mM). The injection of the inhibitor was performed in decapsulated ganglia adjacent to the lesioned connective so that it reaches the injury site. For injections, patch pipettes were pulled from borosilicate glass capillaries (outer diameter 1.5 mm, Clark GC 150 F-10) using a two-stage horizontal micropipette puller (model P-97, Sutter Instrument, Novato, CA, USA) (pipette resistance 3 to 5 MΩ).

The experiments using rabbit polyclonal anti-leech Iba1 antibodies (1/5000) were performed on intact or injured nerve cords 6 h after incubation in complete L-15 medium as previously described8.

After the different incubation times, the nerve cords were fixed with 4% paraformaldehyde at RT for 1 h. After fixation, tissues were washed 3 times in PBS, permeabilized by a 24 h incubation at 4 °C in permeabilization solution (1% Triton X100 in PBS) and pre-incubated in blocking buffer (in 1% Triton, 3% Normal Donkey Serum (NDS) and 1% ovalbumin in PBS/glycine 0.1 M) for 8 h at 4 °C. Then the samples were incubated overnight at 4 °C with the appropriate primary antibody diluted in blocking buffer. After 3 washes with PBS, samples were incubated 1 h at 37 °C with secondary donkey anti-rabbit antibody conjugated to Alexa Fluor 488 (1:2000, Invitrogen, Carlsbad CA, USA) in blocking buffer. They were rinsed with PBS and the cell nuclei were counterstained by Hoechst 33342 fluorescent dye (1/10000, Invitrogen, Carlsbad CA, USA) for 20 min at 4 °C. Finally the nerve cords were mounted on the slide with Dako Fluorescent Mounting Medium (Agilent, Santa Clara CA, USA). Samples without the addition of primary antibody were used as negative control. Slides were maintained, observed and processed in the same manner as described above. The presented pictures are representative of independent triplicates.

Chemotaxis assays

In vitro chemotaxis assays were performed by using the double-P assay as described by Köhidai, with minor modifications65 (Supplementary Method S5). Experiments were performed in triplicates. The results were expressed as the mean percentage of cells migrated, taking into account the starting amount as 100 percent ± S.E.D. Comparisons between means were made using the Ordinary one-way Anova using GraphPad Prism 6.0 software. Statistical differences were considered significant if p was <0.05.

Either recombinant form of human TGF-β (H8541, Sigma-Aldrich, Saint Louis MO, USA) (0, 0.5, 1, 5 and 10 ng/ml) or conditioned media (CM) containing their respective secretome (microglial cells and neurons collected for 15 min primary cultures, see below) were used as chemotactic factors. CM-dependent chemotaxis assays were performed also with neutralizing antibodies. Microglial cells were preincubated for 1 hour at RT with rabbit polyclonal anti-ALK5 antibody (1/100, ab125310, Abcam Cambridge, UK) to inhibit the receptor prior to assays. In another condition, neuron-conditioned medium was preincubated for 1 hour at RT with rabbit polyclonal anti-human TGF-β1 antibody (1/100, ab92486, Abcam, Cambridge, UK) to neutralize nGDF prior to assays. In every experiment, negative controls were carried out with complete L-15 medium alone.

Ex vivo microglial cell recruitment assays

The experiment was performed in injured leech connectives (as described above) at different times post-injury (T6 h, T16 h and T24 h) perfused with a rabbit polyclonal anti-ALK5 (1/50, ab125310, Abcam, Cambridge, UK) or with a control rabbit IgG (1/50, SC2027, Santa Cruz TX, USA) at the same dilution. The antibodies were respectively injected (8 μL) inside the 2nd and 3rd ganglia in a 4 ganglia fragment. The connectives were crushed immediately after injection with fine forceps in the middle side of the two ganglia injected and the tissues were fixed in 4% paraformaldehyde, pH 7.4 for 1 h after T6 h, T16 h and T24 h. Microglial cells recruitment was followed by using a nuclear fluorescent dye Hoechst 33342 (1:10000, Invitrogen, Carlsbad CA, USA) for 20 min at 4 °C. Microglial cell movement in response to these different injections was then observed and processed in the same manner as described above. The presented pictures are representative of independent triplicates.

Western blotting

CNS, microglial cells and neurons protein extract analysis were performed from 5 and 10 nerve cords respectively T24 h post-injury with RIPA buffer. For each experimental condition, SDS-PAGE was conducted (4–12% polyacrylamide gel) using 30 µg of protein extract homogenized (v/v) in 2X Laemmli sample buffer and loaded in the gel wells, further details are found in Supplementary Method S6. After migration of proteins the gel is transferred in a membrane and was incubated for 1 hour at RT in blocking buffer (0.05% Tween 20 w/v, 5% milk powder w/v in 0.1 M PBS, pH 7.4) and then overnight at 4 °C in rabbit polyclonal anti-TGF-β1 (1/200, Abcam, Cambridge, UK) in blocking buffer. After rinsing three times with PBS-0.05% Tween 20 for 15 minutes, the membrane was incubated for 1 hour at RT in secondary goat anti-rabbit polyclonal antibody conjugated with horseradish peroxidase (1:20,000, Jackson Immunoresearch, Cambridgeshire, UK) in PBS-0.05% Tween 20. Finally, after another rinsing with PBS, immunoreactive bands were revealed using the ECL Kit SuperSignal West Dura ChemoLuminescent Substrate (Thermo Fisher Scientific, Waltham MA, USA). Chemiluminescence analyses were performed by ImageQuant LAS-4000 mini system (Fujifilm, Tokyo, Japan).

In situ micro-extraction of proteins

A large-scale proteomic analysis was developed on the injured connectives from the leech nerve cord. This spatially- and temporally-resolved proteomic study was performed on organotypic culture of isolated fragments of nerve chain respecting the integrity of several ganglia joined by connective tissues. Protein micro-extraction experiments were performed using the TriVersa Nanomate platform (Advion BioSciences, Ithaca NY, USA), with the Liquid Extraction Surface Analysis (LESA) feature66 with several modifications as previously described67. For our study, a CNS fragment composed by 4 leech ganglia is dissected. Then the leech nerve cord is injured by cutting one of the two connectives between each pair of ganglia (3 extraction points per fragment). The fragments were incubated in complete medium with a specific ALK4/5 inhibitor, SB431542 (20 µM, R&D Systems, Minneapolis MN, USA) or with its vehicle (100 mM DMSO). The fragments were mounted on the Poly-D-lysine slide (Dominique Dutscher, Brumath, France) at different times post-injury 15 minutes (T0), 6 hours (T6 h) and 24 hours (T24 h). The complete details of the LESA method are found in Supplementary Method S7. Proteins coming from cells and intercellular spaces are then extracted and stored in a collection tube. For proteomics analysis, the sample was migrated on an acrylamide gel. Then, the gels band were cut and subjected in gel digestion were the proteins underwent reduction, alkylation and enzymatic digestion (See Supplementary Method S8).


The samples were separated by online reversed-phase chromatography using a Proxeon EasynLC1000 system (Thermo Fisher Scientific, Waltham MA, USA) equipped with a Proxeon trap column (100 µm ID 2 cm) and a C18 packed-tip column (Acclaim PepMap, 75 µm ID 50 cm). Peptides were separated using an increasing amount of acetonitrile (5–30% over 120 min) at a flow rate of 300 nL/min. The chromatography system was coupled to a Q-exactive mass spectrometer (Thermo Fisher Scientific, Waltham MA, USA) programmed to acquire the top 10 MSMS in data-dependent mode. All the details of the NanoLC-HR-MS/MS can be found in Supplementary Method S9.

Data analyses

All the MS data were processed with MaxQuant (version software68 using the Andromeda search engine69. Proteins were identified by searching MS and MS/MS data against a homemade database of Hirudo medicinalis All the predicted protein sequences are annotated by homology with the human database. For identification, the FDR at the peptide spectrum matches (PSMs) and protein level was set to 0.01. Label-free quantification of proteins was performed using the MaxLFQ algorithm integrated into MaxQuant with the default parameters. Analysis of the proteins identified were performed using Perseus (version software70,71. Multiple-samples tests were performed using ANOVA test with a p-value of 5% and preserving grouping in randomization. Visual heatmap representations of significant proteins were obtained using hierarchical clustering analysis. Functional annotation and characterization of identified proteins were obtained using PANTHER (version 13.0) software72 and STRING (version 9.1)73. The analysis of gene ontology, cellular components and biological processes, were performed with FunRich 3.0 analysis tool74. The details regarding the data analysis are found in Supplementary Method S10.