Amyotrophic lateral sclerosis (ALS) is a fatal progressive neurodegenerative disease characterized by degeneration of motor neurons (MNs) located in the brain stem motor nuclei, spinal anterior horn neurons, and corticospinal tract, resulting in paralysis and death within 3–5 years from the onset of symptoms. In different population, the proportion of familial ALS (fALS) ranges from 1 to 19%, while the remains are sporadic ALS (SALS) with unknown cause (Rowland and Shneider, 2001). To date, hundreds of mutations in more than 20 genes have been implicated in the pathology of ALS, such as mutations in superoxide dismutase (SOD1), TAR DNA-binding protein 43 (TARDBP), fusion (FUS), and a hexanucleotide expansion on chromosome 9 in open reading frame 72 (C9ORF72). However, these causative genes explain only two-thirds of fALS patients and about 11% of SALS cases in Caucasian populations (Renton et al., 2014).

Multiple mechanisms, including glutamate-mediated excite-toxicity effects, abnormal astrocyte and microglial activation, deficiency of neurotrophic factor secretion, protein misfolding, and aggregations, mitochondrial dysfunction, rupture in the axonal passage, destruction in calcium metabolism, changes in skeletal proteins account for the selective vulnerability of MNs, were implicated in the pathogenesis of ALS. However, the exact mechanism underlying MNs death in ALS is still unknown. Accumulating evidence indicates that dysregulation of RNA processing pathway plays an important role in ALS (Droppelmann et al., 2014).

RNA processing is a tightly regulated, highly complex pathway, which includes RNA transcription, pre-mRNA splicing, editing, transportation, translation, and degradation of RNA (Bolognani and Perrone-Bizzozero, 2008). Several identified causative genes of ALS are related to RNA metabolism, such as angiogenin (ANG) (Gao and Xu, 2008), FUS (Kwiatkowski et al., 2009), TARDBP gene (Strong et al., 2007). In addition, several ALS-associated proteins, such as ELP3, SETX, were reported to be involved in RNA processing (Strong, 2010). Furthermore, the recent discovery of miRNAs, which are small endogenous non-coding RNAs of 19–25 nucleotides functioned as inhibitor mRNA translation of target protein-coding genes or induce its degradation, has provided new insight into the regulation of gene expression. Functional studies indicate that miRNA plays a significant role in a broad range of cellular and developmental processes, such as stem cell maintenance, differentiation, development and energy metabolism (Bartel, 2009; Friedman et al., 2009).

Aberrant miRNAs expression has been implicated in several neurodegenerative diseases, including ALS (Borovecki et al., 2005; Hebert and De Strooper, 2009; Miñones-Moyano et al., 2011; Cloutier et al., 2015). Different miRNA expression profiles in ALS have been reported in several studies (Cloutier et al., 2015). Up-regulation of miR-206 can slow down the progression of ALS and promote regeneration of neuromuscular synapses, but this up-regulation was also identified in other diseases, such as schizophrenia (Hansen et al., 2007) and cerebral ischemia (Jeyaseelan et al., 2008), suggesting that up-regulation of miR-206 is not specific for ALS. In addition, the limitations of previous studies on miRNAs in ALS include small sample size (De Felice et al., 2012; Campos-Melo et al., 2013; Ishtiaq et al., 2014), the use of monogenic animal models (Butovsky et al., 2012; Freischmidt et al., 2013; Marcuzzo et al., 2014; Nolan et al., 2014), or human biopsy tissue from skeletal muscle (Russell et al., 2012), spinal cord (Campos-Melo et al., 2013; Ishtiaq et al., 2014) or frontal cortex (Shioya et al., 2010) which are ethically unacceptable in life. Thus, a better understanding of the involvement of miRNAs in ALS is required. On the other hand, early diagnosis and management in specialized ALS clinics providing multidisciplinary patient care has been shown to positively impact quality of life and prolong survival of ALS patients. However, delay in diagnosis for ALS has been widely reported due to its significant overlap of clinical manifestations with some clinically alike conditions at the early stage of ALS (Andersen et al., 2012). Therefore, exploring the robust biomarkers is essential for early diagnosis of ALS. Unlike other biomarkers, miRNAs can be easily secreted into the extracellular space from neurons and other CNS cells. Also, miRNAs are isolated and quantified from a variety of biofluids, allowing miRNAs as ideal potential biomarkers for ALS. However, in the studies mentioned above, the samples were mostly obtained at the relatively end stage of the disease (De Felice et al., 2012; Ishtiaq et al., 2014).

Considering the limited stability of specific brain-enriched miRNAs due to the long interval for tissue samples to be obtained, and the difficulties of obtaining samples from brain, spinal cord or muscle (Sethi and Lukiw, 2009), it seems to be of great advantage to test for biomarkers presented in peripheral blood for diseases. Further evidence suggest that peripheral blood gene and miRNAs expression could be used as indicators in neurological diseases, such as Parkinson's disease (PD) (Serafin et al., 2015), Huntington's disease (Borovecki et al., 2005), and Alzheimer's disease (Cheng et al., 2014). Moreover, in order to avoid the effect of anticoagulation on the expression level of miRNAs (Hastings et al., 2012), we investigated potentially differential expression of miRNAs in leukocytes in SALS patients.

This study was approved by the Ethics Committee of West China Hospital of Sichuan University, Chengdu, P.R. China, and written informed consent was obtained from all subjects before blood collection.

In this study, we were the first to report a comprehensive array-based miRNA expression profile in the peripheral blood of SALS patients from Han Chinese. We observed four down-regulated miRNAs in SALS contrasted with that in healthy controls and found that hsa-miRNA-183 might be a specific biomarker for SALS.

According to the results of our previous study (Yang et al., 2011), the mean survival time of SALS patients was 32.44 ± 19.63 months and the median survival time was 29.67 months. However, the survival time of ALS patients is much shorter in our cohort than that in Caucasian populations (Martínez et al., 2011). Besides the disparity of the medical condition, delay in diagnosis is an important factor affecting the mean survival time. Previous retrospective studies found a delay from symptom onset to diagnosis ranges from 8 to 15.6 months (Cellura et al., 2012). Another study reported that more than half of ALS patients received an alternative diagnosis and each patient saw an average of three different physicians before ALS diagnosis was confirmed (Paganoni et al., 2014). Moreover, time to diagnosis are longer for patients with limb-onset, older age, or sporadic form (Paganoni et al., 2014). In China, fewer neurologists were available for the same population-based community comparing with that in the developed countries. Therefore, discovery of an accurate and sensitive biomarker is of utmost importance in current ALS research. Biomarkers in blood are the primary consideration due to its advantages of non-invasive, easy handling and multiple testing at a low cost. Further, the steady miRNAs is a more promising potential biomarker than cytokines and growth factors in plasma and serum.

MiRNAs, as a key regulator, was first reported by Williams et al. That study showed that miR-206 slowed ALS progression and promoted regeneration of neuromuscular synapses in mice model (Williams et al., 2009). Ever since, numerous studies about miRNAs in ALS have been conducted, the samples were obtained from cultured cells, animal model and ALS patients, and the tissues included blood, cerebral spinal fluid (CSF), muscle, spinal cord, and brain (Cloutier et al., 2015). However, as a biomarker, miR-206 was reproduced by only one study, which reported that miR-206 was up regulated in serum from SOD1 ALS mice and ALS patients (Toivonen et al., 2014). Except for this study, few miRNAs were reproduced in the following studies whether microarray or qRT-PCR based. Possible explanations for these different results included detection methods, sample size, resource of tissue, disease course, familial or sporadic form, and ethnic origin. Some studies only used microarray method without verifying the facticity; some studies only applied qRT-PCR method without discovering the novel miRNAs; while other studies recruited patients or controls less than 10, which easily resulted in false positive or false negative findings. Furthermore, miRNAs expressions may varied at different stages as the disease progresses, the different expression of miRNAs in fALS patients could not be reproduced in SALS patients probably due to the specific defective genes in fALS, and a more heterogeneous of miRNAs in SALS compared with fALS were identified (Freischmidt et al., 2015). Thus, the expression of miRNAs might be regulated by some upstream genes, the epigenetic modifications, which were by mediated in different genetic background. In the present study, using the methods of combining microarray and qRT-PCR, we first detected miRNAs in Chinese SALS population. These miRNAs, including hsa-miR-183, hsa-miR-193b, hsa-miR-451, and hsa-miR-3935, were down regulated in SALS patients relative to healthy controls, with AUC values of 0.69, 0.62, 0.81, and 0.57, respectively. Interestingly, in the subgroup analysis, no significant differences in the expression levels of these four miRNAs were found, according to clinical presentations of ALS, such as site of onset (spinal or bulbar), age of onset (early or late), disease duration (within one year and within 1–2 years), and gender. Our finding may indicate that these miRNAs could provide a new promising opportunity for detection of SALS. Combining the expression of these four miRNAs in PD, the results suggested that hsa-miRNA-183 might be a specific biomarker for SALS, whereas hsa-miR-451 and hsa-miR-3935 might be common biomarkers linked to neurodegenerative diseases, such as ALS and PD. However, these need to be confirmed in further related studies with other neurodegenerative diseases and a larger number of participants.

There is growing evidence suggesting that specific miRNAs are involved in the development and progression of ALS through the regulation of broad signaling pathway, including inflammation (Butovsky et al., 2012; Parisi et al., 2013), neurofilament (Campos-Melo et al., 2013; Ishtiaq et al., 2014), mitochondrial function (Russell et al., 2012), endoplasmic reticulum (ER) stress (Nolan et al., 2014) and others complex network of miRNAs (Shinde et al., 2013). The miR-29a-special antagomir, demonstrated increased lifespan in male SOD1^G93A mice by reducing ER stress (Nolan et al., 2014); Inhibition of miR-23a ameliorates skeletal muscle mitochondrial function in ALS by increasing the PGC-1α signaling, which was involved in mitochondrial biogenesis and function (Russell et al., 2012); the anti-miR-155 significantly extends survival and disease duration in the SOD1^G93A mouse model (Koval et al., 2013), and the miR-206 delays ALS progression and promotes regeneration of neuromuscular synapses in mice partly through fibroblast growth factor and histone deactylase 4 signaling pathways (Williams et al., 2009). In the present study, we observed significantly lower expression of hsa-miR-183, hsa-miR-193b, hsa-miR-451, and hsa-miR-3935 in SALS patients compared with controls. By target prediction and GO analysis, we found low levels of these miRNAs may ease post-transcriptional suppression of their up-regulated mRNA targets related to several neurodegenerative signaling pathway, including PI3K-Akt signaling pathway (hsa-miR-183 andhsa-miR-193b), mTOR signaling pathway (hsa-miR-193b and hsa-miR-3935), regulation of actin cytoskeleton (hsa-miR-193b and hsa-miR-3935), axon guidance (hsa-miR-193b and hsa-miR-3935), MAPK signaling pathway (hsa-miR-183 andhsa-miR-3935) and so on. The differential expression of miR-183 was found in several cancers, including prostate cancer (Yuan et al., 2015), gastric cancer (Cao et al., 2014), breast cancer (Lowery et al., 2010), which may be involved in cell proliferation, migration and invasion (Zhu et al., 2012; Lu et al., 2015). The miR-183-96-182 cluster regulates multiple biological programs that converge to support the maintenance and metastatic potential of medulloblastoma coupled to the PI3K/AKT/mTOR pathway (Weeraratne et al., 2012). Previous studies also found that survival motor neuron (SMN) gene regulates axonal local translation via miR-183/mTOR pathway, which contributed to spinal muscular atrophy pathology (Kye et al., 2014). Thus, dysregulation of miR-183 may be involved in progressive loss of motor neurons in ALS. Similarly, the miR-193b was also reported to modulate cell proliferation (Chen et al., 2010; Gao et al., 2011), migration, and invasion (Hu et al., 2012), and downregulation of miR-193b contributes to enhanced tumor progression and invasion through its target gene-neurofibromin 1(NF1) (Li et al., 2009; Lenarduzzi et al., 2013). The NF1 is required for appropriate granule neuron progenitor expansion and migration in cerebellar development through regulation of RAS/ERK signaling (Sanchez-Ortiz et al., 2014). Moreover, previous study found the miR-193b may take part in the development of AD and is a potential blood-based biomarker of mild cognitive impairment and AD (Liu et al., 2014). However, the accurate mechanisms of the miR-193b causing neurodegeneration remained unknown. Previous studies found that the miR-451, which was differentially expressed in several diseases, such as multiple myeloma (Du et al., 2015), epithelial ovarian cancer (Ling et al., 2015), and central nervous system malignancies (Drusco et al., 2015), could represses cell proliferation and invasion (Zeng et al., 2014). In SH-SY5Y neuron-like cells, downregulation of miR-451 had neurotrophic, neuroprotective, anti-Oxidant, and anti-apoptotic effects (Alural et al., 2014). Interestingly, the down-regulation of hsa-miR-451 was also found in leukocytes from Italian ALS patients (De Felice et al., 2012). In the present study, we observed down-regulation of hsa-miR-451 in our SALS and PD patients, suggesting that the miRNA might be a biomarker for neurodegenerative diseases, which was not specific to SALS. However, little is known about the function and role of miR-3935. In the current study, the down-regulation of miR-3935 was found not only in SALS but also in PD compared with healthy controls, indicating that miR-3935 miRNA might be related to neurodegeneration.

However, we must consider some limitations of the current study. First, only one sample per patient was test in microarray analysis study, which was one of confounding factors resulting in the relatively high variability of miRNAs expression, although the quality of each sample of subjects were strictly tested. Second, for the potential candidate biomarker study, longitudinal data providing is important. Regrettably, the blood sample for each subject was collected only once in the current study. The expression profiling of ALS during disease progression should be tested in the future studies. Thirdly, only PD patients besides ALS were enrolled in the validation stage analysis for comparation, therefore other neurodegenerative disorders such as AD and FTD and so on should be compared. And whether alteration of these miRNAs in the CNS should be tested in the future studies.

In summary, our pilot study showed that peripheral blood leukocytes miR-183, miR-193b, miR-451, and miR-3935 are down regulated in SALS, which raises the potential clinical utility of leukocytes miRNA profiling in SALS diagnosis. Although, such approach is technically challenging, we believe that the leukocytes miRNAs could be a promising diagnostic tool for SALS. Further investigations are required to figure out the underlying mechanisms of how the change of miRNAs level in leukocytes reflect the pathology of motor neuron, as well as their biological functions.

HS planned the study. YC, QW, XC, CL, BC, RO, and HS collected and analyzed clinical data, and made patient follow-ups. YC conducted the molecular studies. SH interpreted data. YC wrote the article, and HS edited the paper.