PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jcinvestThe Journal of Clinical Investigation
 
J Clin Invest. 2012 September 4; 122(9): 3063–3087.
Published online 2012 August 6. doi:  10.1172/JCI62636
PMCID: PMC3428086

Modulating inflammatory monocytes with a unique microRNA gene signature ameliorates murine ALS

Abstract

Amyotrophic lateral sclerosis (ALS) is a progressive disease associated with neuronal cell death that is thought to involve aberrant immune responses. Here we investigated the role of innate immunity in a mouse model of ALS. We found that inflammatory monocytes were activated and that their progressive recruitment to the spinal cord, but not brain, correlated with neuronal loss. We also found a decrease in resident microglia in the spinal cord with disease progression. Prior to disease onset, splenic Ly6Chi monocytes expressed a polarized macrophage phenotype (M1 signature), which included increased levels of chemokine receptor CCR2. As disease onset neared, microglia expressed increased CCL2 and other chemotaxis-associated molecules, which led to the recruitment of monocytes to the CNS by spinal cord–derived microglia. Treatment with anti-Ly6C mAb modulated the Ly6Chi monocyte cytokine profile, reduced monocyte recruitment to the spinal cord, diminished neuronal loss, and extended survival. In humans with ALS, the analogous monocytes (CD14+CD16) exhibited an ALS-specific microRNA inflammatory signature similar to that observed in the ALS mouse model, linking the animal model and the human disease. Thus, the profile of monocytes in ALS patients may serve as a biomarker for disease stage or progression. Our results suggest that recruitment of inflammatory monocytes plays an important role in disease progression and that modulation of these cells is a potential therapeutic approach.

Introduction

ALS causes progressive weakness associated with degeneration of motor neurons. Mutations in more than 50 human genes cause diverse types of motor neuron pathology (1). Although ALS is not primarily considered an inflammatory or immune-mediated disease, immune mechanisms appear to play a role in pathogenesis of the disease. In both ALS patients and animal models, inflammatory responses are observed (28). Furthermore, non-neuronal cells such as microglia (9) and astrocytes (10) are activated during disease progression, and evidence suggests that they contribute to neuronal death. It has been reported that in the CNS, natural killer cells and peripheral T cells infiltrate the spinal cord (11). In addition, selective ablation of mutant SOD1 in astrocytes and microglial cells by conditional deletion (9) and neonatal wild-type bone marrow transplantation (6) increased motor neuron survival and lifespan. T cell deficiency has also been shown to lead to accelerated disease progression in SOD1 mice (3). In the peripheral nervous system, degeneration of peripheral motor axons is an early pathologic feature in ALS patients and mouse models and is preceded by the recruitment and activation of macrophages (5). Clinical trials of antiinflammatory agents in ALS have failed to show efficacy (e.g., minocycline, ref. 12; thalidomide, ref. 13; celecoxib, ref. 14; cyclophosphamide, refs. 15, 16). Nonetheless, none of these treatments were specifically targeted to the inflammatory pathways that are activated in ALS. Of note, a molecule targeting macrophage activation, NP001, is currently in phase II testing (Clinicaltrials.org NCT01091142) and may represent a more selective immunologic approach to the treatment of ALS.

It is known that the Ly6ChiCCR2+ monocyte subset participates in tissue damage and disease pathogenesis in other conditions, including the EAE model of MS (17) and both brain (18) and heart ischemia (19) and that these monocytes are recruited to the inflamed tissues by CCL2 (1921). Several groups have documented increased levels of CCL2 in the CSF and spinal cord of ALS patients (7, 22). We investigated monocytes and resident microglial cells in the G93A transgenic SOD1 mouse model of ALS and the human analog of Ly6C monocyte subsets in human subjects with ALS.

Results

Reciprocal expression of CD39 and Ly6C in CNS-resident microglia and inflammatory monocytes.

One of the major challenges confronting the study of monocytes/macrophages and their role in CNS inflammation in neurodegenerative diseases such as ALS is the ability to distinguish infiltrating peripheral monocytes from resident microglia in the CNS (23). During the course of our investigation of immune markers on monocytes and microglia, we discovered that the expression of Ly6C and CD39 distinguishes non-overlapping populations of peripheral inflammatory monocytes and resident microglia in mice (Supplemental Figures 1 and 2; supplemental material available online with this article; doi: 10.1172/JCI62636DS1). Ly6Chi is a marker for recruited CCR2+ proinflammatory monocytes (20, 24, 25), and CD39 is present on a subset of regulatory T cells (26) and has been shown to be associated with microglia (27). We analyzed CD39 (ectonucleoside triphosphate diphosphohydrolase [Entpd1]) and Ly6C (Ly6c1) expression on microglia from naive adult mouse brains (perfused for removal of non-CNS cells) compared with other immune cells (Figure (Figure1A)1A) and organ-specific CD11b-sorted macrophages (Figure (Figure1,1, B and C). As shown by real-time quantitative RT-PCR (qRT-PCR), Entpd1 and Ly6c1 identified reciprocal populations (Figure (Figure1,1, B and C). These findings were confirmed by FACS analysis of CD11b-sorted cells from naive adult CNS, bone marrow, PBMCs, and spleen (Figure (Figure1D).1D). In naive animals, specificity for microglia as measured by flow cytometry was observed both with our anti-CD39 mAb (clone 5E12) and with commercially obtained anti-CD39 antibody (eBioscience, Supplemental Figure 2E). In inflammatory conditions such as EAE and glioblastoma mouse models, microglia sepcificity was only observed with our anti-CD39 mAb.

Figure 1
Reciprocal expression of CD39 and Ly6C in CNS-resident microglia and inflammatory monocytes.

Activation of the chemotaxis pathway in CD39+ resident microglia in the spinal cord but not in the brain of SOD1 mice.

Our ability to distinguish infiltrating monocytes from resident microglia allowed us to perform expression profiling of CD11b+CD39+ microglia isolated from the spinal cord and brain of SOD1+ mice at different stages of disease. We used NanoString nCounter gene expression, which is more sensitive than microarrays, similar in accuracy to real-time PCR, and more scalable than real-time PCR or microarrays in terms of sample requirements (2830). Importantly, it does not require conversion of mRNA to cDNA by reverse transcription or the amplification of the resulting cDNA by PCR (31) and allows expression analysis of up to 800 genes from as few as 3000 cells; it is thus perfectly suited for analysis of the limited number of cells infiltrating the CNS. Of 179 inflammation-related genes measured by quantitative nCounter, 20 were upregulated (Figure (Figure2A)2A) and 38 were downregulated relative to non-transgenic (non-Tg) mice in spinal cord CD39+ resident microglia (Figure (Figure2B).2B). Microglia had prominent expression of genes related to chemotaxis (e.g., Ccl2, Ccl3, Ccl4, Ccl5, Cxcr4, and Cxcl10). Tgfb1 and Tgfbr1 were among the downregulated genes. Biological network analysis (MetaCore, GeneGo Inc.) identified key transcription factors and target genes (Table (Table1)1) and activation of inflammatory pathways, with the chemotaxis pathway being the most significantly affected (Table (Table2).2). The expression of these genes was observed 1 month prior to symptom onset and was observed in the spinal cord, but not in the brain (Figure (Figure22C).

Figure 2
Activation of the chemotaxis pathway in CD39+ resident microglia in the spinal cord but not the brain of SOD1 mice.
Table 1
Key transcription factors and target genes affected in spinal cord microglia at disease onset in SOD1 mice
Table 2
Major biological pathways affected in spinal cord microglia at disease onset in SOD1 mice

Ly6Chi monocytes in the spleen exhibit a proinflammatory profile 2 months prior to clinical disease onset and during disease progression in SOD1 mice.

We examined the gene expression profile of Ly6Chi monocytes isolated from the spleen of SOD1 mice at 1 and 2 months prior to clinical disease onset and during disease progression. We found a pronounced proinflammatory profile at all time points (Figure (Figure3A).3A). Of 179 inflammation-related genes measured by nCounter, 40 were upregulated relative to non-Tg mice. We also identified 7 genes that were downregulated in Ly6Chi cells, including the antiinflammatory cytokine Tgfb1 and Tgfbr1 (Figure (Figure3B).3B). Of note, we did not find differences in the inflammatory signature between non-Tg and SOD1WT mice in CD39+ spinal cord microglia or splenic Ly6Chi monocytes at onset and end-stage disease (data not shown).

Figure 3
Ly6Chi monocytes in the spleen exhibit a proinflammatory profile 2 months prior to clinical disease onset and during disease progression in SOD1 mice.

Biologic network (Table (Table3)3) and pathway (Table (Table4)4) analysis (MetaCore, GeneGo) demonstrated that the most significantly affected pathways related to inflammation, which included SP1, CREB1, and NF-κB (Table (Table33 and Supplemental Figure 3). These pathways have been shown to play an important role in both monocyte activation and differentiation (32, 33). Thus, the gene expression profiling demonstrates an activated proinflammatory Ly6Chi monocyte population in the peripheral immune compartment of SOD1 mice that we observed 2 months prior to disease onset.

Table 3
Key transcription factors and target genes affected in splenic Ly6Chi monocytes at disease onset in SOD1 mice
Table 4
Major biological pathways affected in splenic Ly6Chi monocytes at disease onset in SOD1 mice

The gene expression differences between non-Tg and SOD1 microglia and monocytes were determined starting from a set containing only inflammatory response–related genes. Because this may have led to a bias toward networks related to inflammation, we performed additional analysis based on a random selection of genes from the original list and found that the networks identified were indeed statistically significant relative to the network identified based on random analysis (Tables (Tables11 and and3).3). Furthermore, by interrogating gene expression found in the network analysis, but not tested by the nCounter assay set, we found that Ahr (aryl hydrocarbon receptor), which is implicated in the SP1 and NF-κB networks and was found to trigger the downstream NF-κB and CREB1 signaling (34, 35), was significantly upregulated in Ly6Chi monocytes but not in Ly6Clo monocytes in SOD1 mice (Figure (Figure3C).3C). Thus, we not only identified a gene independent of the nCounter set, we further verified the link between monocyte subsets in ALS subjects and the SOD1 mouse (see below).

Changes in microRNA profiles of spleen-derived Ly6Chi monocytes and spinal cord–derived CD39+ microglia during the course of disease.

MicroRNAs (miRNAs) belong to a family of small, non-protein-coding RNAs that regulate expression of multiple target genes and are involved in fundamental biological processes. Several hundred mammalian miRNAs have been identified, many of which are tissue specific and modulate hematopoietic lineage differentiation (36). Affected expression of specific miRNAs is associated with pathologic processes including heart disease (37), cancer (38), Alzheimer’s disease (39), and autoimmune inflammation (40). To investigate microglia and the Ly6Chi monocyte subset, we performed miRNA analysis of these populations in the spleens of SOD1 mice at the presymptomatic stage (60 days), at the onset of symptoms, and at end-stage disease. To our knowledge, no study to date has investigated miRNAs in the immune system in ALS. miRNA profiling of spleen-derived Ly6Chi monocytes during all stages of disease (Figure (Figure4A)4A) showed 34 significantly affected miRNAs in Ly6Chi cells, confirmed by Singleplex TaqMan PCR (Figure (Figure4B).4B). Inflammation-related miRNAs such as let-7a (41), let-7b (42), miR-27a (43), miR-146a (44), miR-451 (45), miR-223 (46), miR-142-5p (47), and miR-155 (48) were significantly upregulated in the Ly6Chi subset in the spleen 1 month prior to clinical onset and during disease progression in SOD1 mice (Figure (Figure4,4, A and B). These miRNAs have been shown to be upregulated during inflammatory responses. Thus, miR-146a was identified as a regulator of Ly6Chi monocyte function in inflammation (49), and miR-155 has been shown to promote tissue inflammation (48) and macrophage inflammatory responses (5052) and to be upregulated in EAE (53) and MS lesions (54). Furthermore, miR-155 has been implicated in increasing proinflammatory cytokine secretion by targeting SOCS1 mRNA (55), thereby releasing the intrinsic cellular “brake” on cytokine secretion.

Figure 4
Changes in miRNA profiles of spleen-derived Ly6Chi monocytes during the course of disease.

miRNA target scan Ingenuity pathway analysis (IPA) based on miRNA (Figure (Figure4B)4B) and inflammation-related quantitative NanoString analysis (Figure (Figure3,3, A and B) revealed that the top 10 miRNA-mRNA interactions unique in Ly6Chi splenic monocytes were related to proinflammatory genes (including Tnf, Myd88, Stat1, Ccr2, Tlr2, Myc, Jun, Cd40, H2-Eb1) that are associated with the M1 monocyte phenotype. The top 10 affected miRNAs in Ly6hi splenic monocytes (at onset) yielded 58 miRNA-mRNA interactions (Figure (Figure5)5) with 47 affected inflammatory genes. In addition to performing miRNA profiles of Ly6Chi subsets, we performed miRNA analysis of microglial cells at different stages of disease (Figure (Figure6A)6A) and found 26 miRNAs that were upregulated and 2 downregulated, as confirmed by Singleplex qRT-PCR (Figure (Figure6B).6B). For CD39+ microglia, the top 10 affected miRNAs yielded 80 miRNA-mRNA interactions (Figure (Figure7)7) with 58 affected inflammatory genes. To assess the significance of the observed interactions of the top 10 miRNA families from IPA, we mapped the miRNA families and mRNAs to the list of possible predicted miRNA-mRNA interactions in TargetScan 4.1. This resulted in 18 miRNAs showing 80 interactions in CD39+ spinal cord microglia and 17 miRNAs in Ly6Chi splenic monocytes showing 51 interactions with the affected inflammatory genes in each case. We removed the affected inflammatory genes from the database of all possible miRNA targets to assess the specificity of interaction of the observed miRNAs with non-inflammatory genes. In each case, we generated 1,000 random sets of these non-inflammatory genes (for CD39+ microglia each set contained 40 genes, and for Ly6Chi monocytes each set contained 33 genes), and we identified their predicted interactions with the affected miRNAs. The mean number of random interactions identified for CD39+ microglia was 36.09 ± 12.9 and for Ly6Chi was 25.06 ± 10.09. The number of these interactions was significantly lower than the number of affected miRNA–inflammatory mRNA interactions (for CD39, 80 vs. 36.09 and for Ly6Chi, 58 vs. 25.06). Thus, analysis of the random network showed a substantially lower number of interactions than what we observed using the inflammation-related genes and confirms that we did not bias our analysis by using the inflammatory gene set. These results demonstrate differences between resident CD39+ microglia and infiltrating Ly6Chi monocytes and identify a unique miRNA signature in spinal cord microglia and peripheral inflammatory Ly6Chi monocytes of SOD1 animals.

Figure 5
Top miRNA-mRNA interactions in splenic Ly6Chi monocytes in SOD1 mice.
Figure 6
Changes in miRNA profiles of spinal cord–derived CD39+ microglia during the course of disease.
Figure 7
Top miRNA-mRNA interactions in spinal cord CD39+ microglia in SOD1 mice.

Ly6Chi inflammatory monocytes infiltrate the spinal cord with disease progression in SOD1 mice.

We measured CD11b+Ly6C+ monocytes and CD11b+CD39+ microglia in the CNS of SOD1 mice during disease progression. As shown in Figure Figure8A,8A, in wild-type mice, 98% of CD11b+ cells were CD39+ and 1%–2% were Ly6C+ in both spinal cord and brain. In SOD1 mice with end-stage disease (135 days), 31% of CD11b cells in the spinal cord were Ly6C+, and there was a decrease in the number of CD39+ cells (22%). No change in Ly6C+ cells or in CD39+ cells was observed in the brains of SOD1 mice (Figure (Figure8B).8B). Our observation of changes in the spinal cord and not the brain is consistent with our observation of changes in inflammatory gene expression in microglia from spinal cord but not brain (Figure (Figure8C).8C). Furthermore, these findings suggest a relationship between the recruitment of Ly6C+ cells and the areas of CNS damage in SOD1 mice. We then quantified the percentage of CD11b+Ly6C+ monocytes and CD11b+CD39+ microglia in the CNS over time. We found an increase in Ly6C+ monocytes that began at 60 days of age (1 month before disease onset) and which continued as the disease progressed (Figure (Figure8,8, C and D). At 120 and 135 days, respectively, the proportion of Ly6C monocytes and myeloid cells significantly increased compared with that in wild-type mice at age 135 days (P < 0.01 and P < 0.001, respectively), whereas CD39+ microglia significantly decreased (P < 0. 01 and P < 0.001, respectively). No contribution of myeloid subsets was detected in brains of SOD1 mice (Figure (Figure8C).8C). No Ly6C+ monocytes were detected in the spinal cord at 30 days of age, even though they had increased expression of inflammatory genes at this time (Figure (Figure3A).3A). As shown in Figure Figure8E,8E, Ly6C expression was upregulated with disease progression, and CD39+ microglia remained negative for Ly6C, which is consistent with our observation that CD39 and Ly6C represent non-overlapping CD11b populations (Figure (Figure1,1, B–D).

Figure 8
Ly6Chi monocytes are recruited to the spinal cord with disease progression in SOD1 mice.

SOD1 spinal cord microglia induce recruitment of Ly6C+ monocytes.

It is known that CCL2 interacts with CCR2 receptors on the surface of Ly6Chi monocytes and is required for the recruitment of Ly6Chi monocytes to areas of inflammation (1921). As shown above, gene profiling revealed an increase in the expression of Ccl2 on microglia (Figure (Figure2A)2A) and Ccr2 on Ly6Chi monocytes (Figure (Figure3A).3A). We validated this observation using qRT-PCR to measure the kinetics of Ccr2 expression in Ly6Chi splenic monocytes and Ccl2 expression in CD39+ microglia in the spinal cord over the course of disease. We found that Ccr2 was upregulated in splenic Ly6Chi monocytes both at disease onset and at end-stage disease. This was paralleled by an upregulation of Ccl2 in CD39+ microglia at disease onset. In addition, we observed no expression of Ccr2 in CD39+ microglia or of Ccl2 in Ly6Chi monocytes at any time during the disease course (Figure (Figure8F).8F). This suggests that expression of Ccl2 and other chemokines (Figure (Figure2A)2A) on microglia plays a role in the recruitment of Ly6Chi monocytes to the CNS.

To address whether CD39+ spinal cord microglia in SOD1 mice facilitate recruitment of peripheral Ly6C+ monocytes, we isolated microglia from the spinal cord of both wild-type and SOD1G93A mice at disease onset and transplanted them intracranially in WT and SOD1 recipient mice. Forty-eight hours after transplantation, we isolated brain mononuclear cells and analyzed them by FACS for Ly6C and CD11b expression (Figure (Figure9A).9A). We found a significant increase in recruited Ly6C+ monocytes when we transferred microglia from SOD1 mice versus WT or PBS (sham)–injected mice (Figure (Figure9B).9B). These experiments demonstrate that spinal cord–derived CD39+ microglia directly recruit the Ly6C+ monocytes. Furthermore, we observed a greater increase in Ly6C+ monocytes when microglia from SOD1 mice were transplanted into recipient SOD1 brain as opposed to WT brain (Figure (Figure9B).9B). This is consistent with our observation that splenic Ly6C+ monocytes in SOD1 mice are activated in the periphery and thus there is enhanced recruitment of these cells to the spinal cord in SOD1 mice. Of note, we did not observe recruitment of T, B, or NK cells following intracranial transplantation of microglia (data not shown).

Figure 9
SOD1 spinal cord microglia induce recruitment of Ly6C+ monocytes.

Although we observed changes in Ly6Chi monocytes and CD39+ microglia, during the course of disease, it is theoretically possible that these markers are affected by inflammation that occurs in the spinal cord of SOD1 animals. To rule out this possibility, we generated chimeras in which donor peripheral monocytes expressing GFP under the Cx3cr1 promoter were transplanted into recipient irradiated SOD1 or non-Tg mice. In these chimeras, peripheral monocytes are distinguishable from microglia by FACS. Consistent with our observations above, we found progressive recruitment of Cx3cr1GFP/+ monocytes (Figure (Figure10,10, A–C), which were Ly6C+ and CD39 negative (Figure (Figure10,10, B and D). Of note Ly6Chi recruited monocytes express low levels of CX3CR1 during all disease stages (Figure (Figure10E).10E). This inflammatory monocyte subset (CD11b+CCR2+Ly6ChiCX3CR1lo) (56) has been shown to infiltrate into inflamed tissues and contribute to inflammation (18, 20, 24, 25).

Figure 10
Reciprocal expression of CD39 and Ly6C in CNS-resident microglia and bone marrow–derived monocytes in SOD1 chimeric mice.

Ly6Chi monocytes proliferate and CD39+ microglia undergo apoptosis and are lost in the spinal cord during disease progression in SOD1 mice.

To further investigate Ly6Chi monocytes and CD39+ microglia in the spinal cord during the course of disease, we assessed cellular proliferation as measured by BrdU and apoptosis as measured by Annexin V and 7-AAD staining for apoptotic and necrotic cells, respectively. As shown in Figure Figure11,11, A and B, CD39+ microglia in the spinal cord underwent apoptosis at all disease stages . Concomitant with this, we observed that Ly6Chi monocytes were recruited to the spinal cord and proliferated at all stages of disease (Figure (Figure11,11, C and D). In addition, immunohistochemical analysis of spinal cord sections at different stages of disease demonstrated that microglia were also located in the ventral horn in close proximity to motor neurons, and with disease progression there was a loss of resident microglia (Supplemental Figure 4) in association with recruited monocytes that were in close proximity to dying motor neurons in the ventral horn (Supplemental Figure 5, A and B). These results demonstrate reciprocal changes in these two cell populations during the course of disease.

Figure 11
Ly6Chi monocytes proliferate and CD39+ microglia undergo apoptosis during disease progression in the spinal cord of SOD1 mice.

Anti-Ly6C mAb treatment delays disease onset and extends survival in SOD1 mice.

Based on our finding that Ly6Chi monocytes infiltrate the spinal cord and appear to participate in disease progression, we treated SOD1 mice with anti-Ly6C mAb to determine whether modulation of Ly6Chi monocytes affected disease progression. Animals were treated i.p. every other day beginning at disease onset (defined by body weight loss) until end-stage disease. We monitored body weight (every 3–4 days starting at day 80), clinical neurologic score (daily), and rotarod performance (3 times/week). Treatment with 100 μg anti-Ly6C antibody prolonged survival by 16 days (P = 0.0097) (Figure (Figure12A),12A), extended time to reach a neurologic score of 2 by 9 days (P = 0.0127) (Figure (Figure12B),12B), enhanced rotarod performance (Figure (Figure12C),12C), reduced weight loss (P = 0.001 at day 137) (Figure (Figure12D),12D), and delayed early (P < 0.001) and late (P < 0.05) disease onset (Figure (Figure12E).12E). These results are representative of 3 independent experiments in female animals. In the other 2 experiments (n = 7–9/group), treatment with 100 μg anti-LyC6 mAb extended survival by 18 days (P < 0.01) and 8 days (P < 0.05). No significant effects were observed in animals treated with 10-μg or 1-μg doses (data not shown). Cumulative results of statistical analysis of isotype-treated and anti-Ly6C–treated groups are shown in Table Table5.5. In one experiment, males treated with anti-Ly6C had an increased survival of 6 days versus isotype control (n = 10/group).

Figure 12
Anti-Ly6C mAb treatment delays disease onset and extends survival in SOD1 mice.
Table 5
Cumulative results of statistical analysis of isotype-treated and anti-Ly6C–treated groups

Anti-Ly6C treatment affects the phenotype of splenic Ly6Chi monocytes in SOD1 mice.

To examine the effects of anti-Ly6C treatment on Ly6C+ cells, we sorted CD11b+Ly6C+ monocytes from the spinal cord and spleen of treated animals after 1 month of treatment (age 120 days). We performed quantitative nCounter analysis of splenic Ly6Chi cells for the 179 inflammation-related genes we examined in untreated animals above (Figure (Figure3,3, A and B). As shown in Figure Figure13A,13A, genes related to inflammation, including Tnf, Il1b, and Il6 were downregulated, whereas Tgfb1 and Tgfbr1 were upregulated. We then validated these results by qRT-PCR (Figure (Figure13B).13B). Consistent with our nCounter results, anti-Ly6C mAb suppressed Il1b and Tnf and increased expression of Tgfb1 in Ly6Chi splenic monocytes. In the spinal cord, we observed downregulation of Tnf and upregulation of Tgfb1. No changes in Il1b or Il6 were observed in the spinal cord.

Figure 13
Anti-Ly6C treatment affects the phenotype of splenic Ly6Chi monocytes in SOD1 mice.

In addition, to determine the degree to which recruitment to the spinal cord affected the expression profile of Ly6Chi recruited monocytes, we measured Il1b, Il6, Tnf, and Tgfb1 expression in spleen-derived and spinal cord–recruited Ly6Chi monocytes in both WT and SOD1 untreated mice at 120 days of age (Figure (Figure13C).13C). In the spleen, we found increased expression of Il1b and Tnf and decreased expression of Tgfb1 in SOD1 versus WT mice. We found similar patterns in the spinal cord, viz., increased expression of Il1b and Tnf and decreased expression of Tgfb1 in SOD1 versus WT mice. In addition, we found increased expression of Il6 in the spinal cord in SOD1 versus WT mice. Thus, in untreated animals, we observed an increase in expression of Il1b and Il6 when cells were recruited to the spinal cord, but no difference in Tnf or Tgfb1 expression. This may explain why anti-Ly6C treatment did not affect Il1b and Il6 expression in the spinal cord, whereas it did affect Tnf and Tgfb1.

Anti-Ly6C mAb treatment decreases infiltration of Ly6Chi monocytes into the spinal cord and attenuates neuronal loss.

To further investigate the effect of anti-Ly6C mAb treatment, we performed FACS analysis of Ly6Chi monocytes in the spinal cord. At 30 days after treatment, we found a marked decrease in the percentage of Ly6C+ cells as shown both by FACS (Figure (Figure14A)14A) and quantitative analysis (Figure (Figure14B),14B), indicating depletion of these cells. It is not known whether there is a particular activation/signaling event; however, based on the effects shown in Figure Figure13,13, A and B, anti-Ly6C mAb treatment modulates the M1 Ly6Chi phenotype. Concomitant with the decrease in Ly6C+ cells, we found an increase in CD39+ microglial cells. Because we treated animals with anti-Ly6C antibody and also used anti-Ly6C antibody to measure Ly6C cells in the spinal cord, it is possible that the decrease we observed is related to the antibody and not to actual loss of cells. To address this issue, we used CD169 (Siglec1), which is a well-characterized marker of inflammatory monocytes and differentiated dendritic cells (57). Using SOD1-chimeric mice, we found that CD169 was expressed by all recruited GFP+ bone marrow–derived monocytes, but not by microglia (Supplemental Figure 5B). Consistent with our results above, we found a decrease in CD169+ monocytes in the spinal cord following anti-Ly6C treatment both by FACS (Figure (Figure14C)14C) and by immunohistochemical analysis (Figure (Figure14D).14D). We then asked whether anti-Ly6C treatment affected neurons in the spinal cord. As shown in Figure Figure14,14, D and E, we found an increase in the numbers of neurons both in the dorsal and ventral horns of anti-Ly6C–treated animals. Concomitant with this, there was a decrease in the number of inflammatory monocytes (CD169+ cells). Because CD169, like Ly6C is expressed on the surface of inflammatory monocytes, we treated SOD1 mice with anti-CD169 mAb, but unlike with anti-Ly6C, we found no effect on disease (data not shown).

Figure 14
Anti-Ly6C mAb treatment decreases infiltration of Ly6Chi monocytes into the spinal cord and attenuates neuronal loss.

Our results thus far demonstrate that peripheral Ly6Chi cells play an important role in disease progression. It is known that the spleen is a major reservoir of Ly6C+ cells (58), and it has been reported that the spleen atrophies during the course of disease in the SOD1 mouse (4). Thus, we examined spleen size to determine whether splenic atrophy had occurred and whether this was affected by treatment with anti-Ly6C mAb. We found that the splenic atrophy that occurred in SOD1 mice was reversed by anti-Ly6C mAb treatment (Supplemental Figure 6). It has been shown that splenectomy is of benefit in an animal model of stroke (59) and cardiac ischemia (58). Thus, we asked whether splenectomy would have a beneficial effect on the disease course in the SOD1 mouse. We performed splenectomy during the presymptomatic stage (60 days of age). However, rather than ameliorating disease, splenectomy worsened disease (Supplemental Figure 7).

miRNA and gene profiling of peripheral monocytes (CD14+CD16) in ALS patients demonstrate an inflammatory phenotype analogous to that observed in SOD1 mice.

To investigate whether the inflammatory phenotype we observed in splenic Ly6Chi monocytes in SOD1 mice was also present in patients with ALS, we sorted CD14+CD16 cells (Supplemental Figure 8A), which are the analog of Ly6Chi cells in mice (56) and which we found to be decreased compared with those from healthy controls (Supplemental Figure 8B). We then profiled blood-derived CD14+CD16 and CD14+CD16+ monocyte subsets for 664 miRNAs using quantitative nCounter analysis. We also profiled untreated relapsing-remitting MS patients to determine whether any inflammatory signatures we observed in ALS subjects were unique to ALS or were also present in a classic inflammatory CNS disease such as MS. In ALS, we found 56 miRNAs that were significantly affected in CD14+CD16 monocytes (Figure (Figure15,15, A and B). We identified 19 miRNAs similarly affected in SOD1 mice and human ALS (Supplemental Table 3). Principal component analysis (PCA) showed that both ALS and MS were distinct from healthy controls and that we could distinguish ALS from MS, although there was some overlap (Figure (Figure16A).16A). To validate these findings, we performed quantitative Singleplex qRT-PCR on 6 miRNAs (Figure (Figure16B),16B), including 5 (miR-27a, miR-155, miR-142-5p, miR-223, and miR-532-3p) that we found upregulated in Ly6Chi monocytes in the SOD1 mouse (Figure (Figure4A).4A). In our validation, these miRNAs showed the least overlap between MS and healthy control subjects. Of note, miR-27a was highly expressed in ALS subjects, with no expression in healthy control or MS subjects. To further validate these findings, we investigated a separate independent cohort of ALS patients and healthy controls and found identical results (data not shown). When we analyzed CD14+CD16+ monocytes, we found relatively small numbers (17 miRNAs) of significantly affected miRNAs (Supplemental Figure 9). However, 12 miRNAs were similarly affected in CD14+CD16 monocytes. These results are consistent with our findings in the SOD1 mice of prominent changes in Ly6Chi monocytes, the mouse analog of CD14+CD16 monocytes (56). We also tested cerebrospinal fluid (CSF) from ALS subjects for 43 selected miRNAs that we found upregulated in SOD1 spinal cord CD39+ microglia and splenic Ly6Chi monocytes, and found 3 miRNAs (miR-27b, miR-146a, and miR-532-3p) in common elevated in ALS CSF that were elevated both in monocytes and microglia from SOD1 mice and from human ALS subjects (Supplemental Figure 10). To further evaluate monocytes in subjects with ALS, we performed gene expression, in which we examined 184 inflammatory-related genes as we had in SOD1 mice (Figure (Figure3,3, A and B). We found upregulation of inflammatory genes in ALS that were not observed in healthy controls. Nonetheless, although there were some differences, unlike in our experiments above with miRNAs, the inflammatory gene expression pattern in ALS was similar to that observed in MS (Supplemental Figure 11).

Figure 15
miRNA profiling of peripheral monocytes (CD14+CD16) in ALS patients demonstrates an inflammatory phenotype analogous to that observed in SOD1 mice.
Figure 16
Identification of a unique microRNA signature in CD14+CD16 blood monocytes from ALS subjects.

To further investigate immune-related gene expression in ALS, we examined 511 immune-related genes in monocytes from ALS subjects. With this panel, we were able to distinguish ALS subjects from healthy controls and sporadic ALS (sALS) from familial ALS (fALS) with SOD1 mutation in the small number of samples tested (Figure (Figure17,17, A and B, and Figure Figure18A).18A). We validated these findings using Singleplex qPCR in an independent cohort of ALS patients and healthy controls in which we chose the genes that were the most significantly upregulated (AHR, CCL2, PTAFR, NFKB1, TRAF3, and FCER1A) or downregulated (CXCR4, SOCS1) in ALS (Figure (Figure18B).18B). Of note, we found differences between fALS and sALS in 3 of 8 genes (NFKB1, TRAF3, and FCER1A). Although only 4 fALS subjects were studied, it is possible that these genes may be related to the SOD1 mutation. Furthermore, Ingenuity miRNA-mRNA target filter analysis revealed that the top 10 miRNA-mRNA interactions in ALS were linked to the genes we found to be the most significantly affected in ALS subjects (Figure (Figure19).19). When we analyzed the human peripheral monocyte gene data for degree of overlap with the mouse, we did not find different or more extensive overlap of fALS than sALS with data from the mouse mutant model (see Supplemental Table 3 for a list of similar miRNAs and mRNAs between splenic Ly6Chi monocytes and CD14+CD16 blood monocytes in ALS).

Figure 17
Immune gene profiling of peripheral monocytes (CD14+CD16) in ALS patients.
Figure 18
Identification of a unique gene signature in CD14+CD16 blood monocytes from ALS subjects.
Figure 19
Top miRNA-mRNA interactions in CD14+CD16 blood monocytes from ALS subjects.

We then analyzed both the miRNA and gene expression profile in monocytes from ALS subjects. We found that the abnormalities related to miRNA and gene expression in monocytes from ALS subjects were linked to inflammatory and immune-related genes (Supplemental Table 6 and Supplemental Figure 12). When these miRNA-mRNA interactions in CD14+CD16 blood monocytes in ALS were analyzed, the interactions were shown to be statistically significant using TargetScan 4.1 prediction analysis both in SOD1 mice and in ALS subjects (Supplemental Figure 13, A and B). Furthermore, GeneGo pathway analysis identified 9 inflammation-related networks (Supplemental Figure 14) that were identical to those that we observed in splenic Ly6Chi monocytes in SOD1 mice (Table (Table33).

Discussion

We have identified an important role for Ly6Chi monocytes in a mouse model of ALS. One month prior to disease onset, Ly6Chi monocytes in the spleen had a pronounced proinflammatory profile, as detected by both miRNA and gene expression profiling. The inflammatory gene profile in the spleen was also observed at 2 months prior to disease onset (age 30 days). Ly6Chi monocytes were recruited to the spinal cord, but not brain, and proliferated in the spinal cord during disease progression. This occurred in association with increased Ccr2 expression in Ly6Chi peripheral monocytes and increased Ccl2 expression on CD39+ resident microglia. Even before disease onset and neuronal loss, CD39+ resident microglia expressed Ccl2 and other chemotaxis-associated molecules in the spinal cord, but not brain. Furthermore, we observed increased apoptosis of CD39+ microglia in association with Ly6Chi infiltration and neuronal loss.

Given these findings, we treated SOD1 mice with anti-Ly6C mAb at the time of disease onset. We found that targeting of peripheral Ly6Chi monocytes with anti-Ly6C mAb slowed disease progression and attenuated neuronal damage by decreasing infiltration of Ly6Chi cells to the spinal cord. In addition, anti-Ly6C mAb treatment reduced microglial loss. Anti-Ly6C treatment modulated the cytokine profile of Ly6Chi monocytes in both the spinal cord and spleen by decreasing Tnf and increasing Tgfb1. Thus, it appears that we were both decreasing the infiltration of disease-promoting monocytes and inducing an alternatively (M2) monocyte population.

We then investigated subjects with ALS to determine whether the SOD1-like inflammatory monocyte profile occurred in people with the disease. In humans, CD14+CD16 monocytes express CCR2 and low levels of CX3CR1 and are functionally equivalent to Ly6Chi monocytes (56). We tested CD14+CD16 monocytes from the peripheral blood of ALS subjects for expression of the miRNAs we found upregulated in Ly6Chi cells in SOD1 mice and found analogous patterns. In addition, using gene expression profiling, we found an analogous inflammatory profile in ALS subjects, consistent with classically activated (M1 type) monocytes, the same pattern we observed in the SOD1 mouse.

The chemokine receptor CCR2 and its ligand CCL2 participate in regulating monocyte Ly6ChiCCR2+ subset infiltration (19, 20). The CCL2 protein is expressed in astrocytes in ALS and is increased in the CSF of ALS patients (7). These results support the involvement of immune/inflammatory responses in motor neuron degeneration in ALS. CCL2 is produced at 3-fold-higher levels by SOD1-mutant microglia compared with wild-type human SOD1 microglia activated with LPS in vitro (60). Henkel et al. reported increased CD14+ monocytes in the spinal cord in close proximity to motor neurons, and this was associated with disease progression in ALS patients (7). Consistent with this, the authors reported increased expression of CCL2 in ALS glial cells. Furthermore, Mantovani et al. reported a decrease in CD14+ cells in the blood of ALS patients and postulated that this was related to their early recruitment to CNS areas of primary neurodegeneration (61). We also observed a modest but significant decrease in CD14+CD16 peripheral blood monocytes in the ALS subjects we studied. These findings in humans are consistent with our findings that Ly6Chi monocytes are recruited to the CNS in the SOD1 mice in association with increased Ccl2 expression in resident microglia. Of note, this occurs both prior to disease onset and during disease.

An important question is why Ly6Chi cells are recruited to the spinal cord but not to the brain in SOD1 mice. These findings are consistent with the observation that neurons in the spinal cord, but not the brain, are damaged in the SOD1 mouse. They are also consistent with the fact that we only found changes in microglia in the spinal cord of SOD1 mice. It appears that expression of CCL2 and other chemokines in microglia plays a role in the recruitment of Ly6Chi monocytes to the CNS. We found increased recruitment of Ly6C+ monocytes to the CNS when we transferred microglia from SOD1 mice to the CNS of both WT and SOD1 mice. There was greater recruitment of Ly6C+ monocytes to the CNS when SOD1 microglia were transplanted into SOD1 as opposed to WT CNS. This is consistent with our observation that SOD1 splenic Ly6C+ monocytes are activated and this results in enhanced recruitment of these cells to the spinal cord in SOD1 mice. This process plays an important role in damage to spinal cord neurons, since modulation of Ly6Chi monocytes prolongs survival and decreases motor neuron loss. Thus, based on our results in the SOD1 mouse, we postulate that treatment to modulate inflammatory monocytes in ALS may slow disease progression.

Of note, we found that the activation of splenic inflammatory Ly6Chi monocytes occurs at 30 days of age in SOD1 mice, which is 2 months prior to disease onset. Furthermore, we observed no changes in CNS microglia at this age. This raises the possibility that the SOD1 mutation triggers autonomous activation of peripheral monocytes independent of its effects in the nervous system. Of note, in Huntington’s disease peripheral innate immune activation is detectable before clinical onset (62). Our identification of an analogous SOD1-like miRNA profile in monocytes of subjects with ALS provides a direct link between the animal model and the human disease. Furthermore, it is now possible to test whether the SOD1-like profile of monocytes in the blood of ALS patients is linked to disease stage or progression and whether it has prognostic significance.

One of the important insights from our studies was that CD39+ was able to distinguish resident microglial cells from infiltrating monocytes. This led to the discovery that there was resident microglial loss, not “microgliosis,” in the spinal cord of these animals. Our findings support previous work by Fendrick et al., who demonstrated the occurrence of severe abnormalities in microglia and hypothesized that degeneration of microglia is part of the pathogenesis in ALS (63). It has been proposed that microglia in the adult brain arise from Ly6ChiCCR2+ monocytes under defined host conditions (20). However, in line with our data, recent studies suggest that this is not the case and that resident microglia represent a unique, indigenous cell population in the brain (25, 64, 65).

Numerous studies demonstrate that miRNAs play an essential role in the regulation of immune function (40). Based on miRNA profiling, we found that the peripheral monocytes in the SOD1 mouse and in humans with ALS exhibit a unique inflammatory phenotype. In the SOD1 mouse, we observed upregulation of let-7, miR-15b, miR-16, miR-27a, miR-34a, miR-132, miR-146a, miR-155, miR-223, and miR-451, all of which have been associated with inflammatory responses (4144, 48). Specifically, the expression of miR-15b, miR-16, miR-132, miR-146a, miR-155, and miR-223 is induced in response to LPS stimulation and by proinflammatory stimuli such as IL-1β and TNF-α and mediated by the TLR ligands via the NF-κB pathway and signaling through the JNK pathway (50, 66, 67). Interestingly, a recent study demonstrated that extracellular let-7b induces neurodegeneration through TLR7 (42). We found let-7b to be upregulated in inflammatory monocytes in both SOD1 mice and ALS subjects, which may contribute to the neuronal loss observed in ALS.

Innate immune effector cells are armed with an extensive repertoire of pattern recognition receptors including CD14 and TLRs that are necessary for detection and rapid elimination of invading microorganisms. Upon ligand binding, the CD14/TLR signaling cascade via a common adaptor, MyD88, induces activation of transcription factor 2 (ATF2) and NF-κB transcription factors, both of which subsequently regulate the expression of numerous immune response genes including inflammatory cytokines and also have been recognized as responding to endogenous ligands (68). Several reports indicate the involvement of TLR2 and TLR4 in oxidative stress–related cell activation, a hypothesized mechanism of ALS pathogenesis (69). Moreover, in ALS, mSOD1 protein aggregates (70) and is secreted into the extracellular space (71). Thus, having observed a marked upregulation of Myd88 and Nfkb1 in Ly6Chi splenic monocytes, we hypothesize that in ALS, recruited monocytes could be activated autonomously in the periphery prior to disease onset by aggregated extracellular mSOD1 protein via innate immune receptors, which in turn modulate proinflammatory activation of Ly6Chi monocytes associated with neuronal death. Interestingly, miR-132, miR-146a, and miR-155 were shown to be upregulated by LPS via the adaptor MyD88 and NF-κB (50, 66) and are all upregulated in Ly6Chi monocytes in SOD1 mice. Consistent with this, in our gene expression profiling, we found activation of SP-1, NF-κB, and CREB1, which are pathways related to inflammatory responses. Furthermore, we found that Ahr, which is implicated in the SP1 and NF-κB networks, is upregulated in splenic Ly6Chi monocytes in SOD1 mice and CD14+CD16 blood monocytes in ALS subjects. AhR is reported to trigger the downstream CREB1 (34) and NF-κB (35) signaling, leading to transactivation of the c-Myc oncogene, which we also found to be upregulated in SOD1 mice and ALS. AhR expression is induced during monocytic differentiation (72), and AhR-dependent upregulation of MMP-12, CCL2, and IL-8 in vascular macrophages correlates with recruitment and activation of inflammatory cells in a mouse model of atherosclerosis (73). We found high expression of Jun and Fos in Ly6Chi monocytes in SOD1 mice, indicating that the JNK pathway is activated, and recent studies showed that the NF-κB pathway is activated in ALS (74). Thus, AhR activation in monocytes may be related to misfolded SOD1 and serve a “sensor of toxin exposure” that is linked to inflammation in ALS.

The miRNA profile we found in SOD1 mice was also observed in ALS and consisted of miR-27a, miR-155, miR-146a, and miR-532-3p, with miR-27a being highly expressed in ALS subjects and not expressed in healthy control or MS subjects. Concomitant with the upregulation of these miRNAs, we observed downregulation of TGF-β1 in monocytes from both SOD1 mice and ALS patients. It has been shown that overexpression of miR-155 reduced TGF-β1 production by repressing SMAD2 (52) and SMAD5 (75). In ALS patients, miR-21 and miR-106b were upregulated, and these miRNAs have been shown to target TGF-β1 (76). Indeed, in mice, treatment with anti-Ly6C antibody upregulated Tgfb1 and decreased Tnf in Ly6C+ monocytes in both the spleen and CNS.

Recently, the spleen has been identified as a main reservoir for Ly6C monocytes. This has been shown most dramatically in a myocardial ischemia model in which Ly6Chi monocytes accumulate in the myocardium, which is associated with en masse exit of monocytes from the spleen (58). Consistent with these findings, Offner et al. report that splenic atrophy occurs in a model of experimental stroke (77). This phenomenon has also been reported in ALS models. Banerjee et al. showed that spleens of SOD1 mice were reduced in size and weight at the end stage of disease (4), and in our experiments we found a decrease in spleen size in untreated SOD1 mice. Furthermore, we found that in association with disease amelioration in the CNS following treatment with anti-Ly6C mAb, the spleen returned to normal size. We hypothesize that splenic atrophy was linked to the deployment of inflammatory monocytes during disease progression. Nonetheless, we found that splenectomy worsened, rather than ameliorated disease, suggesting that the spleen may harbor disease-protective monocytes or other protective elements.

Although immune abnormalities have been reported before in ALS, it is not generally considered an inflammatory disease. MS is a classic neuroinflammatory disease with well-documented changes in the peripheral immune compartment and proven response to immune-modulatory treatment. Although we found broad similarities between the miRNA and inflammatory gene patterns in monocytes from MS and ALS patients, we identified unique miRNA and gene profiles in the blood of ALS patients. General immunosuppressive treatments including cyclophosphamide (15, 16) and total lymphoid irradiation (78) have been tested in ALS without positive results. Furthermore, as described above, affecting the entire monocyte pool by splenectomy worsened disease in the mouse model. Thus, specific modulation of inflammatory monocytes may be required. The identification of common inflammatory pathways between ALS and MS further supports a role for inflammation in ALS and raises the possibility that immunomodulatory therapies that target monocytes may be beneficial in both MS and ALS.

Most subjects with ALS have the sporadic form (sALS), though some have a familial form (fALS) linked to specific gene mutations. The clinical characteristics of the sporadic and familial forms are similar. We studied monocytes from 4 fALS subjects with SOD1 mutations and found, by miRNA and gene profiling, that they had profiles similar to those we had observed in SOD1 mice. Although only 4 fALS (mSOD1) subjects were studied, the similar immune abnormalities we identified may indicate that these immune abnormalities represent common pathways that are affected in different forms of ALS. It is now known that there are other forms of familial ALS-related mutations such as TDP-43, FUS, and the newly discovered mutation in a noncoding region of C9ORF72 (79, 80), which is the most common known genetic cause of fALS (46%) and also accounts for approximately 21% of sALS cases in Finland. Based on our findings, it will now be possible to determine the degree to which the changes in inflammatory monocytes we identified are linked to the various genetic forms of the disease.

In conclusion, we demonstrate that recruitment of inflammatory monocytes into the CNS plays an important role in ALS disease progression. Furthermore, we found that the inflammatory miRNA signature in peripheral monocytes from ALS subjects to be analogous to that observed in the spleen of SOD1 mice, providing a direct link between the mouse model and the human disease. Thus, the SOD1-like profile of monocytes in the blood of ALS patients we identified may serve as a biomarker linked to disease stage or progression and modulation of these cells is a potential therapeutic approach.

Methods

See Supplemental Methods for detailed methodology.

Animals.

Lewis rats were used to generate hybridoma cells for antibody pro­duction and were purchased from Charles River Laboratories. B6/SJL-SOD1G93A Tg, SOD1 wild-type, CX3CR1-GFP (B6.129P-Cx3cr1tm1Litt/J), and non-Tg mice were provided by Prize4Life or purchased from the Jackson Laboratory.

Behavioral analysis.

See Supplemental Methods.

Generation of chimeric mice.

See Supplemental Methods.

Generation of mAbs specific for resident microglia and peripheral inflammatory monocytes.

See Supplemental Methods.

Cell transfection.

See Supplemental Methods.

Flow cytometry.

See Supplemental Methods.

Patients.

See Supplemental Methods.

Blood samples.

See Supplemental Methods.

CSF samples.

See Supplemental Methods.

Isolation of human blood monocytes.

Fresh peripheral blood mononuclear cells were obtained by Ficoll density gradient centrifugation. CD14+CD16 and CD14+CD16+ monocyte subsets stained with mouse anti–human CD14-PE and CD16-PeCy7 (BD Biosciences — Pharmingen) were sorted with a FACSAria (BD Biosciences). The sorted cells were further prepared for the RNA isolation protocol indicated below.

Analysis of miRNA and mRNA expression.

To assess miRNA and mRNA expression in human CD14+CD16 and CD14+CD16+ monocyte subsets and brain and spinal cord CD11b+CD39+ microglia and CD11b+Ly6C+ spinal cord and splenic monocytes from WT and SOD1 mice, total RNA was isolated and analyzed by qRT-PCR using specific primers for selected mRNAs and miRNAs all purchased from Applied Biosystems. For analysis of miRNA expression, qRT-PCR analyses were carried out using TaqMan miRNA assays (Applied Biosystems), and relative expression was calculated using the ΔCt method as described elsewhere (81) and normalized to uniformly expressed U6 (Applied Biosystems) for miRNAs and to housekeeping genes for relative gene expression indicated in each corresponding figure legend. All qRT-PCRs were performed in duplicate or triplicate, and the data are presented as mean ± SD.

Administration of BrdU and tissue preparation.

See Supplemental Methods.

Immunohistochemistry.

See Supplemental Methods.

RNA isolation and miRNA profiling in CSF.

See Supplemental Methods.

TaqMan miRNA low-density array (TLDA) miRNA analysis and TLDA data analysis.

See Supplemental Methods.

Statistical analysis of TLDA data.

See Supplemental Methods.

Quantitative NanoString nCounter miRNA/gene expression analysis.

See Supplemental Methods.

GeneGo biological pathway analysis.

See Supplemental Methods.

IPA miRNA-mRNA target filter analysis.

See Supplemental Methods.

GEO dataset.

The gene expression data set has been deposited in the GEO database (GSE39644).

Statistics.

Prism software (GraphPad Software) was used for 1-way ANOVA with Dunnett’s or Bonferroni post hoc test. Differences between 2 groups were examined using Student’s t test (2-tailed). Unsupervised hierarchical clustering of significantly affected miRNA and gene expression profiles was analyzed by nonparametric Kruskal-Wallis test, with significance based on Benjamini-Hochberg false discovery rate (FDR). The results were visualized by heatmap and PCA using MeV (Multi Experiment Viewer software version 4.7.4). miRNA and mRNA expression between WT and SOD1 mice was compared by ANOVA with Dunnett’s post-hoc test. A P value less than 0.05 was considered statistically significant.

Study approval.

Animal protocols were approved by the IACUC at Harvard Medical School and were performed according to NIH guidelines. Concerning the use of human materials, the study was approved by the MGH and BWH institutional review boards, and all subjects provided written informed consent prior to being enrolled in the study. ALS patients were at least 18 years old and met El Escorial criteria for possible, probable, or definite ALS.

Supplementary Material

Supplemental data:

Acknowledgments

This work was supported in part by the Amyotrophic Lateral Sclerosis Association; NIH grant AG027437; Tufts Center for Neuroscience Research P30 NS047243. The Harvard NeuroDiscovery Center provided CSF specimens (Biomarker Study). We thank Prize­4Life for providing SOD1 mice; V. Kuchroo and D. Selkoe for critical discussions; N. Kassam for technical support.

Footnotes

Conflict of interest: The authors have declared that no conflict of interest exists.

Citation for this article: J Clin Invest. 2012;122(9):3063–3087. doi:10.1172/JCI62636.

References

1. Pasinelli P, Brown RH. Molecular biology of amyotrophic lateral sclerosis: insights from genetics. Nat Rev Neurosci. 2006;7(9):710–723. [PubMed]
2. McGeer PL, McGeer EG. Inflammatory processes in amyotrophic lateral sclerosis. Muscle Nerve. 2002;26(4):459–470. doi: 10.1002/mus.10191. [PubMed] [Cross Ref]
3. Beers DR, Henkel JS, Zhao W, Wang J, Appel SH. CD4+ T cells support glial neuroprotection, slow disease progression, and modify glial morphology in an animal model of inherited ALS. Proc Natl Acad Sci U S A. 2008;105(40):15558–15563. doi: 10.1073/pnas.0807419105. [PubMed] [Cross Ref]
4. Banerjee R, et al. Adaptive immune neuroprotection in G93A-SOD1 amyotrophic lateral sclerosis mice. PLoS One. 2008;3(7):e2740. doi: 10.1371/journal.pone.0002740. [PMC free article] [PubMed] [Cross Ref]
5. Chiu IM, et al. Activation of innate and humoral immunity in the peripheral nervous system of ALS transgenic mice. Proc Natl Acad Sci U S A. 2009;106(49):20960–20965. doi: 10.1073/pnas.0911405106. [PubMed] [Cross Ref]
6. Beers DR, et al. Wild-type microglia extend survival in PU.1 knockout mice with familial amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A. 2006;103(43):16021–16026. doi: 10.1073/pnas.0607423103. [PubMed] [Cross Ref]
7. Henkel JS, et al. Presence of dendritic cells, MCP-1, and activated microglia/macrophages in amyotrophic lateral sclerosis spinal cord tissue. Ann Neurol. 2004;55(2):221–235. doi: 10.1002/ana.10805. [PubMed] [Cross Ref]
8. Lincecum JM, et al. From transcriptome analysis to therapeutic anti-CD40L treatment in the SOD1 model of amyotrophic lateral sclerosis. Nat Genet. 2010;42(5):392–399. doi: 10.1038/ng.557. [PubMed] [Cross Ref]
9. Boillee S, et al. Onset and progression in inherited ALS determined by motor neurons and microglia. Science. 2006;312(5778):1389–1392. doi: 10.1126/science.1123511. [PubMed] [Cross Ref]
10. Nagai M, et al. Astrocytes expressing ALS-linked mutated SOD1 release factors selectively toxic to motor neurons. Nat Neurosci. 2007;10(5):615–622. doi: 10.1038/nn1876. [PMC free article] [PubMed] [Cross Ref]
11. Chiu IM, et al. T lymphocytes potentiate endogenous neuroprotective inflammation in a mouse model of ALS. Proc Natl Acad Sci U S A. 2008;105(46):17913–17918. doi: 10.1073/pnas.0804610105. [PubMed] [Cross Ref]
12. Gordon PH, et al. Efficacy of minocycline in patients with amyotrophic lateral sclerosis: a phase III randomised trial. Lancet Neurol. 2007;6(12):1045–1053. doi: 10.1016/S1474-4422(07)70270-3. [PubMed] [Cross Ref]
13. Stommel EW, et al. Efficacy of thalidomide for the treatment of amyotrophic lateral sclerosis: a phase II open label clinical trial. Amyotroph Lateral Scler. 2009. 10 (5-6): 393 404 10.3109/17482960802709416 [PMC free article] [PubMed] [Cross Ref]
14. Cudkowicz ME, et al. Trial of celecoxib in amyotrophic lateral sclerosis. Ann Neurol. 2006;60(1):22–31. doi: 10.1002/ana.20903. [PubMed] [Cross Ref]
15. Brown RH, Jr, Hauser SL, Harrington H, Weiner HL. Failure of immunosuppression with a ten- to 14-day course of high-dose intravenous cyclophosphamide to alter the progression of amyotrophic lateral sclerosis. Arch Neurol. 1986;43(4):383–384. doi: 10.1001/archneur.1986.00520040063021. [PubMed] [Cross Ref]
16. Smith SA, Miller RG, Murphy JR, Ringel SP. Treatment of ALS with high dose pulse cyclophosphamide. J Neurol Sci. 1994;124(suppl):84–87. doi: 10.1016/0022-510X(94)90188-0. [PubMed] [Cross Ref]
17. King IL, Dickendesher TL, Segal BM. Circulating Ly-6C+ myeloid precursors migrate to the CNS and play a pathogenic role during autoimmune demyelinating disease. Blood. 2009;113(14):3190–3197. doi: 10.1182/blood-2008-07-168575. [PubMed] [Cross Ref]
18. Dimitrijevic OB, Stamatovic SM, Keep RF, Andjelkovic AV. Absence of the chemokine receptor CCR2 protects against cerebral ischemia/reperfusion injury in mice. Stroke. 2007;38(4):1345–1353. doi: 10.1161/01.STR.0000259709.16654.8f. [PubMed] [Cross Ref]
19. Nahrendorf M, et al. The healing myocardium sequentially mobilizes two monocyte subsets with divergent and complementary functions. J Exp Med. 2007;204(12):3037–3047. doi: 10.1084/jem.20070885. [PMC free article] [PubMed] [Cross Ref]
20. Mildner A, et al. Microglia in the adult brain arise from Ly-6C(hi)CCR2(+) monocytes only under defined host conditions. Nat Neurosci. 2007;10(12):1544–1553. doi: 10.1038/nn2015. [PubMed] [Cross Ref]
21. Kim YG, et al. The Nod2 Sensor Promotes Intestinal Pathogen Eradication via the Chemokine CCL2-Dependent Recruitment of Inflammatory Monocytes. Immunity. 2011;34(5):769–780. doi: 10.1016/j.immuni.2011.04.013. [PMC free article] [PubMed] [Cross Ref]
22. Nagata T, et al. Elevation of MCP-1 and MCP-1/VEGF ratio in cerebrospinal fluid of amyotrophic lateral sclerosis patients. Neurol Res. 2007;29(8):772–776. doi: 10.1179/016164107X229795. [PubMed] [Cross Ref]
23. Ransohoff RM, Cardona AE. The myeloid cells of the central nervous system parenchyma. Nature. 2010;468(7321):253–262. doi: 10.1038/nature09615. [PubMed] [Cross Ref]
24. Mildner A, et al. CCR2+Ly-6Chi monocytes are crucial for the effector phase of autoimmunity in the central nervous system. Brain. 2009;132(pt 9):2487–2500. doi: 10.1093/brain/awp144. [PubMed] [Cross Ref]
25. Saederup N, et al. Selective chemokine receptor usage by central nervous system myeloid cells in CCR2-red fluorescent protein knock-in mice. PLoS One. 2010;5(10):e13693. doi: 10.1371/journal.pone.0013693. [PMC free article] [PubMed] [Cross Ref]
26. Deaglio S, et al. Adenosine generation catalyzed by CD39 and CD73 expressed on regulatory T cells mediates immune suppression. J Exp Med. 2007;204(6):1257–1265. doi: 10.1084/jem.20062512. [PMC free article] [PubMed] [Cross Ref]
27. Braun N, et al. Assignment of ecto-nucleoside triphosphate diphosphohydrolase-1/cd39 expression to microglia and vasculature of the brain. Eur J Neurosci. 2000;12(12):4357–4366. [PubMed]
28. Guttman M, et al. lincRNAs act in the circuitry controlling pluripotency and differentiation. Nature. 2011;477(7364):295–300. doi: 10.1038/nature10398. [PMC free article] [PubMed] [Cross Ref]
29. Malkov VA, et al. Multiplexed measurements of gene signatures in different analytes using the Nanostring nCounter Assay System. BMC Res Notes. 2009;2:80. doi: 10.1186/1756-0500-2-80. [PMC free article] [PubMed] [Cross Ref]
30. Kulkarni MM. Digital multiplexed gene expression analysis using the NanoString nCounter system. Curr Protoc Mol Biol. 2011. Chapter 25: Unit25B .10. [PubMed]
31. Geiss GK, et al. Direct multiplexed measurement of gene expression with color-coded probe pairs. Nat Biotechnol. 2008;26(3):317–325. doi: 10.1038/nbt1385. [PubMed] [Cross Ref]
32. Park JM, et al. Signaling pathways and genes that inhibit pathogen-induced macrophage apoptosis--CREB and NF-kappaB as key regulators. Immunity. 2005;23(3):319–329. doi: 10.1016/j.immuni.2005.08.010. [PubMed] [Cross Ref]
33. Hirata Y, et al. Sp1 is an essential transcription factor for LPS-induced tissue factor expression in THP-1 monocytic cells, and nobiletin represses the expression through inhibition of NF-kappaB, AP-1, and Sp1 activation. Biochem Pharmacol. 2008;75(7):1504–1514. doi: 10.1016/j.bcp.2007.12.019. [PubMed] [Cross Ref]
34. Hsieh TH, et al. Phthalates induce proliferation and invasiveness of estrogen receptor-negative breast cancer through the AhR/HDAC6/c-Myc signaling pathway. FASEB J. 2012;26(2):778–787. doi: 10.1096/fj.11-191742. [PubMed] [Cross Ref]
35. Kim DW, Gazourian L, Quadri SA, Romieu-Mourez R, Sherr DH, Sonenshein GE. The RelA NF-kappaB subunit and the aryl hydrocarbon receptor (AhR) cooperate to transactivate the c-myc promoter in mammary cells. Oncogene. 2000;19(48):5498–5506. doi: 10.1038/sj.onc.1203945. [PubMed] [Cross Ref]
36. Chen CZ, Li L, Lodish HF, Bartel DP. MicroRNAs modulate hematopoietic lineage differentiation. Science. 2004;303(5654):83–86. doi: 10.1126/science.1091903. [PubMed] [Cross Ref]
37. Thum T, et al. MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts. Nature. 2008;456(7224):980–984. doi: 10.1038/nature07511. [PubMed] [Cross Ref]
38. Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature. 2007;449(7163):682–688. doi: 10.1038/nature06174. [PubMed] [Cross Ref]
39. Hebert SS, et al. Loss of microRNA cluster miR-29a/b-1 in sporadic Alzheimer’s disease correlates with increased BACE1/beta-secretase expression. Proc Natl Acad Sci U S A. 2008;105(17):6415–6420. doi: 10.1073/pnas.0710263105. [PubMed] [Cross Ref]
40. O’Connell RM, Rao DS, Chaudhuri AA, Baltimore D. Physiological and pathological roles for micro­RNAs in the immune system. Nat Rev Immunol. 2010;10(2):111–122. doi: 10.1038/nri2708. [PubMed] [Cross Ref]
41. Polikepahad S, et al. Proinflammatory role for let-7 microRNAS in experimental asthma. J Biol Chem. 2010;285(39):30139–30149. doi: 10.1074/jbc.M110.145698. [PMC free article] [PubMed] [Cross Ref]
42. Lehmann SM, et al. Nat Neurosci. An unconventional role for miRNA: let-7 activates Toll-like receptor 7 and causes neurodegeneration [published online ahead of print May 20, 2012]. doi: 10.1038/nn.3113 . [PubMed]
43. Jennewein C, von Knethen A, Schmid T, Brune B. MicroRNA-27b contributes to lipopolysaccharide-mediated peroxisome proliferator-activated receptor gamma (PPARgamma) mRNA destabilization. J Biol Chem. 2010;285(16):11846–11853. doi: 10.1074/jbc.M109.066399. [PMC free article] [PubMed] [Cross Ref]
44. Lu LF, et al. Function of miR-146a in controlling Treg cell-mediated regulation of Th1 responses. Cell. 2010;142(6):914–929. doi: 10.1016/j.cell.2010.08.012. [PMC free article] [PubMed] [Cross Ref]
45. Bandres E, et al. microRNA-451 regulates macrophage migration inhibitory factor production and proliferation of gastrointestinal cancer cells. Clin Cancer Res. 2009;15(7):2281–2290. doi: 10.1158/1078-0432.CCR-08-1818. [PubMed] [Cross Ref]
46. Li T, Morgan MJ, Choksi S, Zhang Y, Kim YS, Liu ZG. MicroRNAs modulate the noncanonical transcription factor NF-kappaB pathway by regulating expression of the kinase IKKalpha during macrophage differentiation. Nat Immunol. 2010;11(9):799–805. doi: 10.1038/ni.1918. [PMC free article] [PubMed] [Cross Ref]
47. Schaefer JS, Montufar-Solis D, Vigneswaran N, Klein JR. Selective upregulation of microRNA expression in peripheral blood leukocytes in IL-10–/– mice precedes expression in the colon. . J Immunol. 2011;187(11):5834–5841. doi: 10.4049/jimmunol.1100922. [PMC free article] [PubMed] [Cross Ref]
48. O’Connell RM, et al. MicroRNA-155 promotes autoimmune inflammation by enhancing inflammatory T cell development. Immunity. 2010;33(4):607–619. doi: 10.1016/j.immuni.2010.09.009. [PMC free article] [PubMed] [Cross Ref]
49. Etzrodt M, et al. Regulation of monocyte functional heterogeneity by miR-146a and Relb. Cell Rep. 2012;1(4):317–324. [PMC free article] [PubMed]
50. O’Connell RM, Taganov KD, Boldin MP, Cheng G, Baltimore D. MicroRNA-155 is induced during the macrophage inflammatory response. Proc Natl Acad Sci U S A. 2007;104(5):1604–1609. doi: 10.1073/pnas.0610731104. [PubMed] [Cross Ref]
51. Ceppi M, et al. MicroRNA-155 modulates the interleukin-1 signaling pathway in activated human monocyte-derived dendritic cells. Proc Natl Acad Sci U S A. 2009;106(8):2735–2740. doi: 10.1073/pnas.0811073106. [PubMed] [Cross Ref]
52. Louafi F, Martinez-Nunez RT, Sanchez-Elsner T. MicroRNA-155 targets SMAD2 and modulates the response of macrophages to transforming growth factor-{beta}. J Biol Chem. 2010;285(53):41328–41336. doi: 10.1074/jbc.M110.146852. [PMC free article] [PubMed] [Cross Ref]
53. Murugaiyan G, Benyon V, Mittal A, Joller N, Weiner HL. Silencing microRNA-155 ameliorates experimental autoimmune encephalomyelitis. J Immunol. 2011;187(5):2213–2221. doi: 10.4049/jimmunol.1003952. [PMC free article] [PubMed] [Cross Ref]
54. Junker A, et al. MicroRNA profiling of multiple sclerosis lesions identifies modulators of the regulatory protein CD47. Brain. 2009;132(pt 12):3342–3352. doi: 10.1093/brain/awp300. [PubMed] [Cross Ref]
55. Lu LF, et al. Foxp3-dependent microRNA155 confers competitive fitness to regulatory T cells by targeting SOCS1 protein. Immunity. 2009;30(1):80–91. doi: 10.1016/j.immuni.2008.11.010. [PMC free article] [PubMed] [Cross Ref]
56. Geissmann F, Jung S, Littman DR. Blood monocytes consist of two principal subsets with distinct migratory properties. Immunity. 2003;19(1):71–82. doi: 10.1016/S1074-7613(03)00174-2. [PubMed] [Cross Ref]
57. Asano K, et al. CD169-positive macrophages dominate antitumor immunity by crosspresenting dead cell-associated antigens. Immunity. 2011;34(1):85–95. doi: 10.1016/j.immuni.2010.12.011. [PubMed] [Cross Ref]
58. Swirski FK, et al. Identification of splenic reservoir monocytes and their deployment to inflammatory sites. Science. 2009;325(5940):612–616. doi: 10.1126/science.1175202. [PMC free article] [PubMed] [Cross Ref]
59. Ajmo CT, Jr, et al. The spleen contributes to stroke-induced neurodegeneration. J Neurosci Res. 2008;86(10):2227–2234. doi: 10.1002/jnr.21661. [PMC free article] [PubMed] [Cross Ref]
60. Sargsyan SA, Blackburn DJ, Barber SC, Monk PN, Shaw PJ. Mutant SOD1 G93A microglia have an inflammatory phenotype and elevated production of MCP-1. Neuroreport. 2009;20(16):1450–1455. doi: 10.1097/WNR.0b013e328331e8fa. [PMC free article] [PubMed] [Cross Ref]
61. Mantovani S, et al. Immune system alterations in sporadic amyotrophic lateral sclerosis patients suggest an ongoing neuroinflammatory process. J Neuroimmunol. 2009;210(1–2):73–79. doi: 10.1016/j.jneuroim.2009.02.012. [PubMed] [Cross Ref]
62. Bjorkqvist M, et al. A novel pathogenic pathway of immune activation detectable before clinical onset in Huntington’s disease. J Exp Med. 2008;205(8):1869–1877. doi: 10.1084/jem.20080178. [PMC free article] [PubMed] [Cross Ref]
63. Fendrick SE, Xue QS, Streit WJ. Formation of multinucleated giant cells and microglial degeneration in rats expressing a mutant Cu/Zn superoxide dismutase gene. J Neuroinflammation. 2007;4:9. doi: 10.1186/1742-2094-4-9. [PMC free article] [PubMed] [Cross Ref]
64. Ginhoux F, et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science. 2010;330(6005):841–845. doi: 10.1126/science.1194637. [PMC free article] [PubMed] [Cross Ref]
65. Ajami B, Bennett JL, Krieger C, McNagny KM, Rossi FM. Infiltrating monocytes trigger EAE progression, but do not contribute to the resident microglia pool. Nat Neurosci. 2011;14(9):1142–1149. doi: 10.1038/nn.2887. [PubMed] [Cross Ref]
66. Taganov KD, Boldin MP, Chang KJ, Baltimore D. NF-kappaB-dependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc Natl Acad Sci U S A. 2006;103(33):12481–12486. doi: 10.1073/pnas.0605298103. [PubMed] [Cross Ref]
67. Zhou R, Hu G, Liu J, Gong AY, Drescher KM, Chen XM. NF-kappaB p65-dependent transactivation of miRNA genes following Cryptosporidium parvum infection stimulates epithelial cell immune responses. PLoS Pathog. 2009;5(12):e1000681. doi: 10.1371/journal.ppat.1000681. [PMC free article] [PubMed] [Cross Ref]
68. Akira S, Takeda K, Kaisho T. Toll-like receptors: critical proteins linking innate and acquired immunity. Nature Immunol. 2001;2(8):675–680. doi: 10.1038/90609. [PubMed] [Cross Ref]
69. Ischiropoulos H, Beckman JS. Oxidative stress and nitration in neurodegeneration: cause, effect, or association? J Clin Invest. 2003;111(2):163–169. [PMC free article] [PubMed]
70. Bruijn LI, et al. Aggregation and motor neuron toxicity of an ALS-linked SOD1 mutant independent from wild-type SOD1. Science. 1998;281(5384):1851–1854. doi: 10.1126/science.281.5384.1851. [PubMed] [Cross Ref]
71. Urushitani M, Sik A, Sakurai T, Nukina N, Takahashi R, Julien JP. Chromogranin-mediated secretion of mutant superoxide dismutase proteins linked to amyotrophic lateral sclerosis. Nat Neurosci. 2006;9(1):108–118. doi: 10.1038/nn1603. [PubMed] [Cross Ref]
72. Hayashi S, Okabe-Kado J, Honma Y, Kawajiri K. Expression of Ah receptor (TCDD receptor) during human monocytic differentiation. Carcinogenesis. 1995;16(6):1403–1409. doi: 10.1093/carcin/16.6.1403. [PubMed] [Cross Ref]
73. Wu D, et al. Activation of aryl hydrocarbon receptor induces vascular inflammation and promotes atherosclerosis in apolipoprotein E–/– mice. . Arterioscler Thromb Vasc Biol. 2011;31(6):1260–1267. doi: 10.1161/ATVBAHA.110.220202. [PMC free article] [PubMed] [Cross Ref]
74. Swarup V, et al. Deregulation of TDP-43 in amyotrophic lateral sclerosis triggers nuclear factor kappaB-mediated pathogenic pathways. J Exp Med. 2011;208(12):2429–2447. doi: 10.1084/jem.20111313. [PMC free article] [PubMed] [Cross Ref]
75. Rai D, Kim SW, McKeller MR, Dahia PL, Aguiar RC. Targeting of SMAD5 links microRNA-155 to the TGF-beta pathway and lymphomagenesis. Proc Natl Acad Sci U S A. 2010;107(7):3111–3116. doi: 10.1073/pnas.0910667107. [PubMed] [Cross Ref]
76. Roubelakis MG, et al. Human microRNA target analysis and gene ontology clustering by GOmir, a novel stand-alone application. BMC Bioinformatics. 2009;10(suppl 6):S20. [PMC free article] [PubMed]
77. Offner H, et al. Splenic atrophy in experimental stroke is accompanied by increased regulatory T cells and circulating macrophages. J Immunol. 2006;176(11):6523–6531. [PubMed]
78. Drachman DB, et al. Trial of immunosuppression in amyotrophic lateral sclerosis using total lymphoid irradiation. Ann Neurol. 1994;35(2):142–150. doi: 10.1002/ana.410350205. [PubMed] [Cross Ref]
79. Renton AE, et al. A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron. 2011;72(2):257–268. doi: 10.1016/j.neuron.2011.09.010. [PMC free article] [PubMed] [Cross Ref]
80. DeJesus-Hernandez M, et al. Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS. Neuron. 2011;72(2):245–256. doi: 10.1016/j.neuron.2011.09.011. [PMC free article] [PubMed] [Cross Ref]
81. Gabriely G, et al. MicroRNA 21 promotes glioma invasion by targeting matrix metalloproteinase regulators. Mol Cell Biol. 2008;28(17):5369–5380. doi: 10.1128/MCB.00479-08. [PMC free article] [PubMed] [Cross Ref]

Articles from The Journal of Clinical Investigation are provided here courtesy of American Society for Clinical Investigation