Background
Amyotrophic lateral sclerosis (ALS) is characterised by selective motoneurones degeneration in the spinal cord, brainstem and motor cortex leading to progressive muscle weakness, atrophy and paralysis. Approximately 90 % of ALS patients are sporadic whilst 10 % are familial cases with genetic mutations in SOD1 (Cu/Zn superoxide dismutase 1), FUS (fused in sarcoma), TARDBP (also known as TDP-43) and C9ORF72, among others [
1]. Transgenic mice over-expressing the human mutated gene for SOD1 develop an adult-onset paralysis that closely recapitulates human ALS [
2]. Recent studies have established that ALS is a complex multi-factorial disease that involves several cellular partners including glial cells [
3].
Microglia, the resident immune cells of the central nervous system (CNS), when activated, release pro- and anti-inflammatory cytokines and chemokines that are generally associated with M1 and M2 phenotypes [
4,
5]. Microglia have a dual role in ALS with an early protective effect on motoneurones but also a detrimental effect due to the secretion of neurotoxic factors [
6]. It is hypothesised that progressive motoneurone death results from the combination of intrinsic motoneurones vulnerability and toxicity from neighbouring cells such as microglia [
6]. In ALS patients and animal models, there is a clear microglia activation [
3], in particular we have shown an early involvement of microglia in hSOD1
G93A mice [
7]. Understanding the contribution of microglia to motoneurone degeneration is of high priority. One means of analysing the role of a cell population in a process network is to study gene expression alterations in this given population. In addition, an integrative comparison of the specific molecular signatures of several cellular partners is necessary to decipher the crosstalk between these cells. We have previously identified gene dysregulation in pure motoneurones from the lumbar spinal cord of hSOD1
G93A mice [
8] and two other mouse models of motoneurone disease [
9]. We revealed a unique motoneurone gene expression profile characterised by an absence of dysregulation of genes associated with cell death and a massive up-regulation of genes involved in cell growth [
8].
Growing evidence points toward mitochondrial dysfunction and oxidative DNA damage in ALS [
10]. Defence mechanisms, including SOD, counteract excessive accumulation of reactive oxygen species, however in ALS, cellular antioxidant defences are insufficient leading to damage of nucleic acids, proteins and lipids [
11]. Inherited mutations in breast cancer susceptibility gene 1 (
Brca1), a well-known tumour suppressor implicated in familial breast and ovarian cancers, is one of the best defined risk factor for development of breast and ovarian cancer.
Brca1 plays important roles in a broad spectrum of functions including transcription regulation, cell cycle checkpoint activation, apoptosis, chromosomal remodelling, ubiquitination and DNA repair [
12]. The role of
Brca1 in each of these processes remains to be fully understood but it is hypothesized that it act as a scaffold for the formation of complexes with a wide range of proteins [
13]. This ability of
Brca1 to interact with different proteins may underlie its involvement in a variety of cellular processes [
13].
Brca1 also exerts a protective role against oxidative stress via up-regulation of antioxidant genes and maintenance of the redox balance through up-regulating the expression of heat shock protein HSP27 [
14,
15].
In breast cancer,
Brca1 cellular localisation as well as the significance of its altered localisation, is still a matter of debate. It had been recently shown that in normal breast,
Brca1 nuclear expression is strong and uniform in parenchymal cells whereas in malignant cells its expression is reduced if not absent from the nucleus and is, in some cases, observed in the cytoplasm [
16]. Interestingly, altered expression of
Brca1 was associated with poor prognosis and shortened survival. In the adult rodent CNS, the presence of
Brca1 is detected only in neurons [
17] whereas a high
Brca1 expression is observed in embryonic [
17,
18] and adult neural stem cells and is involved in cell proliferation [
18].
Here we identify putative Brca1 involvement in ALS via hSOD1G93A microglia gene profiling and comparisons to our previous transcriptomic findings in hSOD1G93A motoneurones. We then demonstrated that Brca1 is a novel marker of human microglia and is up-regulated in ALS patients.
Discussion
Non-cell autonomous toxicity plays a major role in ALS [
20] but microglia participation is dual and complex. Microglia reactivity over the course of the disease may be characterised by a continuum of activation states from a M2 neuroprotective state to a deleterious M1 state. In culture, microglia have, at disease onset, a M2 phenotype whereas they are typified by a M1 phenotype at disease end-stage [
21]. Comparison of our data to previous studies [
21,
22] reveals an up-regulation of five M2 priming genes (
Clec7a,
Igf1,
Mmp12,
Spp1 and
Lgals3) and a down-regulation of
Retnla and
F13a1. Interestingly, four M1 priming genes are up-regulated
CD86,
Tnfα,
Bcl2a1a and
Cxcl10, whilst growth arrest and DNA-damage-inducible alpha gene (
Gadd45gip1) is down-regulated. A previous study has reported the gene-expression profile of isolated microglia in hSOD1
G93A mice and shown that potentially neuroprotective and neurotoxic factors are induced concurrently during disease progression [
23]. The authors have analysed microglia from the entire spinal cord whereas we have restricted our investigations to the lumbar segment where onset of degeneration occurs. We report that hSOD1
G93A microglia from the lumbar region of the spinal cord over-express
progranulin,
Igf1 and
osteopontin, all potential neurotrophic factors, and thus confirmed findings from a previous study [
23]. Interestingly, we had previously identified in pure motoneurones of two mouse models of motoneurone disease (hSOD1
G93A and
pmn) an increase in
IGFBP. Also, an
IGFBP that binds to
IGF-
1 and
IGF-
2 (nephroblastoma over-expressed gene) was up-regulated at all disease stages in hSOD1
G93A mice [
8] and
IGFBP4 mRNA was induced at pre-symptomatic age in
pmn mice [
9]. We also confirmed the up-regulation of potential neurotoxic factors (including
Mmp12,
tnf-
α and interferon-induced protein with tetratricopeptide repeats) [
23]. However, we did not confirm the dysregulation of the genes coding for
IL-
1β,
IL-
α,
IL-
10,
Ifnar 1 and
Ifnar 2 as well as
Nox2 at P90. It had been shown that delayed forelimb motor impairment in ALS mice may be partially explained by augmented protective responses in the cervical spinal cords [
24], thus gene expression profile of lumbar hSOD1
G93A microglia is potentially more homogenous and is more likely to reflect a pathological gene profile than microglia taken from the entire spinal cord. Together, these data confirm that microglia activation states are best characterised as a continuum of M2 and M1 states [
21] with a M2 phenotype at early stage of the disease that evolves into a M1 phenotype at disease end-stage.
An unexpected finding was the up-regulation in hSOD1
G93A microglia of
Brca1 with a 1.76 fold. Using
in silico comparison with data from Chiu et al. [
23], we found that
Brca1 was also deregulated in their study and presented a steady increased with 2.78 and 3.08 fold changes at P100 and P130, respectively. In our study,
Brca1 involvement was substantiated by the concomitant dysregulation of a number of other genes. As previously stated,
Igf1 was robustly up-regulated in hSOD1
G93A microglia; a complex interplay between
Brca1 and
IGF signalling pathways had been reported in familial cancer, in particular through the convergence of Brca1-mediated tumour protective pathways and IGF1 receptors-mediated cell survival [
25,
26]. This simultaneous up-regulation may represent a potential neuroprotective phenotype of microglia in ALS at early stage. Converging elements toward the involvement of Brca1 was also pointed through the dysregulation of genes linked to
Brca1 and belonging to the DNA damage pathway (Fig.
3). Indeed,
GADD45 was down-regulated in hSOD1
G93A microglia (P90) and up-regulated in motoneurones (P120) and it had been demonstrated that
Brca1 can modulate
GADD45 that in turn mediates DNA repair mechanisms and regulates growth arrest [
15]. Importantly, we found that the gene coding for cyclin-dependent kinase inhibitor 1A (
p21) was up-regulated both in hSOD1
G93A microglia (P90) and motoneurones (P120). Indeed,
p21 is a downstream target of
p53 and regulates several processes such as DNA repair, cell cycle arrest, cell differentiation and apoptosis. Through its antioxidant effects,
p21 also protects cells from oxidative damage
in vitro and
in vivo [
15]. Activation of microglial cells and acquisition of deleterious M1 state is associated with an increased generation of reactive oxygen species (ROS) [
27] that is likely to participate in motoneurone demise. Polarisation of microglia/macrophages to pro- and anti-inflammatory states is driven by cytokines and other factors such as ROS within the tissue microenvironment [
28]. While the functional role of deregulated
Brca1 pathway in microglia remains to be determined, one hypothesis is that it may represent an attempt to counteract the detrimental effects of ROS and reflect an antioxidative defence mechanism through modulation of microglia polarisation.
Brca1 is implicated in a broad spectrum of functions; it regulates transcription and cell cycle progression, it is also involved in function that preserve genomic stability such as DNA repair pathways [
29] and protection against oxidative damage to DNA. Many of these functions have been associated with CNS development but also with neurodegenerative diseases and in particular with ALS. Brca1 is required for normal cerebral cortex size development [
30] by preventing apoptosis [
31]. Using a neural progenitor-specific driver to delete
Brca1, Pao et al. demonstrated an important role of
Brca1 in apoptotic and centrosomal functions in neuronal progenitors that may underlie DNA damage and brain size during development [
31].
Brca1 is also associated with lack of spinal cord neural tube closure in
spina bifida meningomyelocele [
32,
33]. Moreover,
Brca1-deficient embryos presented disorganised neuroepithelium associated with rapid proliferation and enhanced cell death [
32].
De-regulation in
Brca1 expression had been reported in Alzheimer’s [
34,
35] and Huntington’s diseases [
36]. Even though motor neuron diseases are not typical paraneoplastic syndromes, association with breast cancer had been regularly reported [
37‐
41]. Moreover, there are occasional reports on improvement of motor neuron syndrome after cancer treatment [
42‐
44].
Methods
Animals
Transgenic mice carrying the G93A human SOD1 mutation, B6SJL-Tg (SOD1-G93A)1Gur/J (ALS mice, high copy number) were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and bred on a B6SJL background. Transgenic mice were housed in controlled conditions (hygrometry, temperature and 12 h light/dark cycle). Ninety days old (P90, symptomatic) males were used for transcriptomic analysis and immunohistochemistry. Litter-matching between groups were done. We carried out all animal experiments in accordance with the guidelines approved by the French Ministry of Agriculture and following the European Council directive (2010/63/UE). We minimised the number and suffering of animals.
Flow cytometry sorting of spinal cord microglia from SODG93A and control littermate mice
Mice were deeply anesthetised with tribromoethanol (500 mg/kg) and intracardially perfused with cold RNAse-free 0.1 M phosphate base saline (PBS, Invitrogen, Carlsbad, USA); spinal cords were dissected. Only the lumbar (L1 - L5) segment was used and dissociated in 750 μl PBS, 100 μl trypsin 13 mg/ml, 100 μl hyaluronidase 7 mg/ml, 50 μl kinurenic acid 4 mg/ml (Sigma Aldrich, Saint Louis, USA) and 20 μl DNAseI 10 mg/ml (Roche, Rotkreuz, Switzerland) for 30 min at 37 °C. Finally, gentle mechanic dissociation was carried out by pipetting. Cell suspension was sieved on a 40 μm cell strainer (BD Biosciences, Franklin Lakes, USA). To eliminate myelin, cells were re-suspended in PBS-25 % sucrose and centrifuged for 20 min at 750 g. Cells were incubated for 20 min on ice in the primary antibody CD11b-APC 1/100 in PBS (BD biosciences, Franklin Lakes, USA) that specifically labels microglia. Cells were washed with cold PBS and re-suspended in PBS 7-AAD 2 μg/ml (Sigma Aldrich). Cells were sorted with a FACS ARIA (BD Biosciences, Franklin Lakes, USA), equipped with a 488 nm Laser Sapphire 488–20. Size threshold, morphology and 7-AAD were used to eliminate cellular debris and dead cells.
Microarray analysis of gene transcripts
Our data comply with the “Minimal Information About Microarray Experiment (MIAME)” guidelines. Total RNA was isolated using RNeasy Mini Kit, (Qiagen, Maryland, USA) including DNAse treatment to remove potential genomic DNA contamination. We tested the quality of the starting RNA and of the amplified cRNA (Agilent 2100 bioanalyzer, RNA 6000 Pico LabChip, Palo Alto, USA) and proceeded only if the RNA quality was satisfactory. A criterion was a cut point for RNA integrity number (RIN) at 7 [
45]. Fifty nanograms of RNA per chip were hybridized (three chips per condition) following a T7-based double amplification procedure.
Hybridization targets were obtained following a double amplification procedure according to the protocol developed by Affymetrix (GeneChip® Eukaryotic Small Sample Target Labeling Assay Version II, Affymetrix, Santa Clara, USA) and previously used [
8,
9]. A hybridization mixture containing 5.5 μg of biotinylated cRNA was generated. The biotinylated cRNA was hybridized to Affymetrix GeneChip® MOE 430 2.0. Three chips per group (wild type and hSOD1
G93A) were hybridized, each corresponding to microglia from at least six pooled mice. Chips were visualised on a 3000 gene scanner (Affymetrix, Santa Clara, USA). We selected the differentially expressed transcripts using the Affymetrix software MAS 5.0 and carried out pair-wise comparison analyses where each of the mutant samples was compared to each of their respective control samples. This analysis is based on the Mann–Whitney pair-wise comparison test and allows the ranking of the results by concordance as well as the calculation of significance (
p value) of each identified gene expression [
46,
47]. A gene must exhibit 50 % or more of the “present” calls in all samples to be considered “expressed” and has two or more “present” calls among the three sets of samples. Fold differences were calculated as the ratio between the average values within each condition. Signal values and detection calls (present or absent) for all samples were determined using Affymetrix MAS5.0. Based on power analysis, we had selected a cut off threshold of 1.75 (p (α) 0.05, β 0.80) to identify transcripts that are differentially expressed between the controls and hSOD1
G93A mutant mice. Statistics: t-test with un-equal variance. Pathway analysis was done with MetaCore (Thomson Reuters).
Quantitative real-time polymerase chain reaction
Candidate genes involved in Brca1 pathway were validated using qPCR. Similar to microarray, total RNA was extracted as described above from CD11-positive microglia isolated using FACS and used as a template in real time PCR. At least two animals were used for each analysis. To assess the involvement of microglia
Brca1 at initial stages of the disease progression, we carried out qPCR at 60 and 90 days of age in hSOD1
G93A and wild type mice. One round of amplification was done following the first cycle (first cDNA and cRNA synthesis) of the Affymetrix double amplification procedure before undertaking reverse transcription with random hexamers (Superscript II, Invitrogen, Carlsbad, CA). Real time PCR using Syber Green PCR Master Mix and Abi Prism SDS 7900 HT (Applied Biosystems, Foster City, CA) was done according to the manufacturer’s protocol. All amplicons were designed within the 3′ end of the cDNA using Primer Express Software 2.0 (Applied Biosystems, Foster City, CA) and when possible, overlapped exon-exon junctions. For the sequences of the primers, see Additional file
7: Table S5. All samples were analysed in triplicate and the values were normalised to four reference genes mitochondrial ribosomal protein S9 (
RPS9), TATA box binding protein (
TBP), actin β and eukaryotic translation elongation factor 1 (
EEF1).
Human spinal cord samples
Human low thoracic and lumbar (T11-L5) spinal cords were obtained from 14 controls (males and females; 23 to 74 years of age; mean age: 52.4 years) and five ALS patients (males and females; 66 to 79 years of age; mean age: 71 years) from the Kantonsspital St. Gallen Fachbereichsleiter Muskelzentrum/ALS clinic under the approval of the Swiss legislation and from the New York Brain Bank–Taub Institute, Columbia University (NYBB), New York, USA. All donors had given their written consent for the autopsy and we followed the Declaration of Helsinki.
Immunohistochemistry
Mice were anesthetised with tribromoethanol (500 mg/kg) and perfused intracardially with cold PBS followed by cold 4 % paraformaldehyde (PFA, Sigma Aldrich). Spinal cords were removed and post fixed for 2 h in 4 % PFA. Samples were cryoprotected in sucrose 30 %, included in Tissue Teck (Sakura, Alphen aan den Rijn, The Netherlands), frozen and kept at −80 °C until processing.
For mice, free floating spinal cord transverse sections (20 μm) were washed twice in PBS (5 min), treated for 30 min in PBS containing lysine (20 mM, pH 7.2) and for 15 min in 1 % H2O2. Sections were blocked for 1 h with PBS containing bovine serum albumin (BSA, 1 %, Sigma Aldrich) and Triton X-100 (0.1 %, Fisher Scientific, Illkirch, France) and then incubated 48 h at 4 °C with CD11b (1/200, Developmental Studies Hybridoma Bank, Iowa, USA) primary antibody. Alexa-conjugated 594 secondary antibody was used (1/1000; Molecular Probes, Eugene, OR, USA).
For human spinal cord 22 μm-thick cryosections of lumbar and lower thoracic segments were collected on super frost plus slides and were processed as described above. For dual fluorescence labelling, sections were placed in a cocktail of rat anti CD11b (1/100, Hybridoma Bank, University of Iowa, USA) and rabbit anti-Brca1 (1/100, Santa Cruz Biotechnology, Dallas, USA) primary antibodies for 48 h at 4 °C. Sections were washed in 0.1 M PBS followed by incubation in corresponding secondary antibodies conjugated to Alexa 488 and 594 (1/1000; Molecular Probes, Eugene, OR, USA). For peroxidase labelling, sections were placed for 48 h at 4 °C in either rabbit anti Iba1 (macrophage/microglia-specific calcium-binding protein) (1/1000, Wako Pure Chemical Industries, Osaka, Japan) or rabbit anti-Brca1 (1/100, Santa Cruz Biotechnology, Dallas, USA) primary antibodies. Spinal cord sections were then incubated in donkey anti-rabbit (1/500, Jackson Immunoresearch, Carlsbad, USA) antibody for 2 h at 4 °C. Sections were then washed in TRIS buffer and enzymatic revelation was done with nickel enhanced DAB and H2O2 0.1 % as a substrate. Sections were then dehydrated in ascending concentration of ethanol and finally xylene. Coverslips were applied using Entellan (Merck KGaA, Darmstadt, Germany).
Morphometric bright field photographs had been obtained and analysed using NanoZoomer RS slide scanner (NanoZoomer Digital Pathology System and NDP view software, Hamamatsu, Japan). For immunofluorescence images, we used laser scanning inverted confocal microscopy (Leica SP5, Mannheim, Germany). Laser intensity and detector sensitivity settings were kept constant for all image acquisitions within a given experiment. Brca1 staining intensity measurement was done by measuring their optical density (OD) using ImageJ (National Institutes of Health, USA), as described previously [
48]. For each given sample we analysed at least three 22-μm-thick section with 330 μm distance from each other. Statistics: un-paired t-test done with GraphPad Prism version 5.03 (GraphPad software, CA, USA). Significance was accepted at
p ≤ 0.05. Results are expressed as mean ± S.E.M.
Acknowledgments
We are grateful to ALS patients and their relatives that donate their tissues. We acknowledge the New York Brain Bank–The Taub Institute, Columbia University (NYBB). The hybridoma CD11b antibody developed by Timothy A. Springer was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the University of Iowa, Department of Biology, Iowa city, IA 52242. We thank the iGE3Genomics Platform, University of Geneva Switzerland for their assistance in transcriptomic and qPCR analysis.
This work was supported by the Spanish Government, Plan Nacional de I+D+I 2008–2011 and ISCIII-Subdirección General de Evaluación y Fomento de la investigación (PI10/00709) [to FEP], the Government of the Basque Country grant (Proyectos de Investigacion Sanitaria and Fondo Comun de Cooperacion Aquitania-Euskadi) [to FEP], the “Fondation pour la Recherche Médicale” [to FEP] and the French Government, ANR-FNS grant, GliALS (N° ANR-14-CE36-0009-01) [to FEP], the patient organisations “Demain Debout Aquitaine” [to YNG and HNN] and “Verticale” [to FEP and HNN].
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HNN: participated in the design of the study, performed immunohistology, participated to the analysis and interpretation of data and helped to draft the manuscript. JCS: participated in the design of the study, carried out FACS and participated to immunohistology. YNG: participated to acquisition of FACS data. MT: participated to immunohistology. AS: performed autopsy. MdmV: participated to acquisition of FACS data. MW: performed patient selection and obtained patients consent. FEP: conception, design of the work; analysis and interpretation of data, drafting the work and final approval. All authors read and approved the final manuscript.