The implication of circulating long non-coding RNA MALAT1 in diagnosis, disease surveillance and prognosis of acute ischemic stroke

Background We aimed to investigate predictive value of long non-coding RNA metastasis-associated lung adenocarcinoma transcript 1 (lnc-MALAT1) for acute ischemic stroke (AIS) risk, and the association of lnc-MALAT1 expression with disease severity, inflammation as well as recurrence free survival (RFS) in AIS patients. Methods 120 AIS patients and 120 controls were recruited. Venous blood samples from AIS patients (within 24 hours after symptoms onset) and controls (at entry to study) were collected to detect plasma lnc-MALAT1 expression by real-time quantitative polymerase chain reaction. For AIS patients, AIS severity was assessed by NIHSS score; plasma concentrations of inflammation factors (including C-reactive protein (CRP), tumor necrosis factor α (TNF-α), interleukin (IL)-6, IL-8, IL-10, IL-17 and IL-22) were measure; and RFS was calculated. Results Lnc-MALAT1 expression was decreased in AIS patients compared to controls, and it had a good predictive value for AIS risk (AUC=0.791, 95% CI: 0.735-0.846). For disease condition, lnc-MALAT1 expression negatively correlated with NIHSS score and pro-inflammatory factors expressions (including CRP, TNF-α, IL-6, IL-8 and IL-22), while it positively correlated with anti-inflammatory factor IL-10 expression. Besides, lnc-MALAT1 expression was elevated in AIS complicated with diabetes but numerically reduced in AIS complicated with hepertension. For prognosis, lnc-MALAT1 high expression numerically correlated with longer RFS, but without statistical significance. Conclusion lnc-MALAT1 is downregulated and has a good predictive value of AIS risk, and its high expression correlates with decreased NIHSS score, reduced inflammation, as well as numerically better RFS in AIS patients.

3 Acute ischemic stroke (AIS), a common disorder caused by embolic or thromboembolic occlusion in an artery supplying the brain, results in irreversible infarction of brain tissue and functional impairment, which brings economic and social burden for families. [1][2][3] Current treatments of AIS mainly focus on recovering cerebral circulation and lead to a great likelihood of independent living following AIS, while several restrictions exist in current therapy (such as narrow treatment time window and low recanalization rate), and the prognosis of AIS is still far from satisfactory. [4][5][6] Increasing studies disclose that early identification of AIS risk and prediction of prognosis may help improve treatment outcomes of AIS patients. [7,8] Thus, searching for convincing biomarkers that monitoring AIS risk and predicting prognosis is urgently needed.
Long non-coding RNAs (lncRNAs) are transcripts with more than 200 nucleotides and rarely encode proteins, which play essential roles in regulation of protein-coding genes and signaling pathways related to development of diseases. [8] LncRNA metastasis-associated lung adenocarcinoma transcript 1 (lnc-MALAT1), located on human chromosome 11q13.1, is a lncRNA abundantly expressed and evolutionarily conserved throughout a variety of mammalian species. [7] Accumulating data from basic and clinical studies have disclosed that lnc-MALAT1 could protect ischemia-induced brain microvascular endothelial cells (BMECs), implying that lnc-MALAT1 might regulate cerebrovascular pathologies in stroke. [9,10] Meanwhile, lnc-MALAT1 has been reported to be downregulated and primarily plays an anti-inflammatory role in neurological diseases patients as well as cardiovascular and cerebrovascular diseases patients. [11,12] Considering the participation of lnc-MALAT1 in cerebrovascular pathologies of stroke and its anti-inflammation effect on cardiovascular and cerebrovascular diseases, we hypothesized that lnc-MALAT1 might serve as a biomarker for disease risk and progression in AIS through affecting inflammation level, whereas related evidences were seldomly reported. Hence, we conducted this study to hypertension, diabetes mellitus, hyperlipidemia, hypeluricemia and chronic kidney disease (CKD). Severity of AIS was assessed within the day of admission by use of the National Institutes of Health Stroke Scale (NIHSS) score. The NIHSS score aimed at assessing neurological impairment ranging from 0 to 42, a higher score indicated a more serious nerve damage, and the classification of severity was as follows: 0-1 point, normal or near normal; 2-4 points, mild stroke; 5-15 points, moderate stroke; 16-20 points, moderatesevere stroke; 21-42 points, severe stroke. Besides, controls' basic characteristics including age, gender, BMI, smoke, hypertension, diabetes mellitus, hyperlipidemia, hypeluricemia and CKD were also documented on the enrollment.

Blood samples collection and determination
Venous blood samples were collected from AIS patients (within 24 hours after symptoms onset) and controls using ethylene diamine tetraacetic acid (EDTA) tubes, and subsequently centrifuged at 1600 g for 10 min (within 30 min) to acquire supernatant, then the supernatant was further centrifuged at 16000g for 10 min to obtain the plasma, which was scored at -80 ℃ for further analysis. Lnc-MALAT1 relative expression in the plasma of AIS patients and controls was determined by real-time quantitative polymerase chain reaction (RT-qPCR). AIS patients' CRP concentration in plasma was detected using Fully automatic POCT fluorescence immunoassay analyzer (GeteinBiotech, Nanjing, Jiangsu, China), and the plasma levels of inflammatory cytokines including tumor necrosis factor α (TNF-α), interleukin-6 (IL-6), IL-8, IL-10, IL-17 and IL-22 were measured by human enzyme-linked immunosorbent assay (ELISA) kits (Thermo Fisher Scientific, Waltham, Massachusetts, USA) according to the manufacturer's recommendations.

RT-qPCR
Using QIAamp RNA Blood Mini Kit (Qiagen, Duesseldorf, Nordrhein-Westfalen, German), total RNA was extracted from plasm samples. Then, reverse transcription to cDNA was was applied to the graphs plotting. Continuous variables were expressed as mean ± standard deviation (SD) or median and interquartile range (IQR), and the categorical variables were expressed as number (percentage). Differences between groups were determined by Student's t test, Wilcoxon rank sum test or Chi-square test. Correlations between variables were analyzed by Spearman rank test. Diagnostic value of variable was assessed by receiver operating characteristic (ROC) curve analysis and the derived area under the curve (AUC) as well as 95% confidence interval (CI). RFS profiles were illuminated by plotting the Kaplan-Meier (K-M) curve, and the difference of RFS between groups was determined by the log-rank test. Factors affecting RFS were analyzed by univariate and multivariate Cox's proportional hazards regression models. All tests were 2-

Characteristics in AIS patients and controls
120 AIS patients and 120 controls were enrolled in this study ( Table 1) Table 1.

Correlation of lnc-MALAT1 expression with NIHSS score in AIS patients
Spearman rank test was used to determine the correlation of lnc-MALAT1 expression with NIHSS score, which showed that lnc-MALAT1 expression was negatively correlated with NIHSS score in AIS patients (P < 0.001, r = -0.437) (Figure 2).  Figure 4E). These data indicated that lnc-MALAT1 expression was positively correlated with diabetes occurrence in AIS patients.

Comparison of RFS between lnc-MALAT1 high expression patients and lnc-MALAT1 low expression patients
Patients were divided into lnc-MALAT1 high expression group and lnc-MALAT1 low expression group according to the median value of lnc-MALAT1 expression. RFS in AIS patients with lnc-MALAT1 high expression was numerically longer compared to AIS patients with lnc-MALAT1 low expression, but without statistical significance (P = 0.053) ( Figure 5).

Analysis of factors affecting RFS in AIS patients
Univariate Cox's regression analysis displayed that lnc-MALAT1 high expression was numerically associated with better RFS in AIS patients (although no statistical significance) (P = 0.063), while CRP (≥29.0 vs. <29.0 mg/L) was associated with worse RFS (P = 0.006) ( Table 3). Moreover, multivariate Cox's regression analysis showed that lnc-MALAT1 high expression was not an independent predictive factor for RFS, whereas CRP (≥29.0 vs. <29.0 mg/L) was an independent factor predicting reduced RFS (P = 0.024).

Discussion
Our results indicated that: (1) lnc-MALAT1 expression was decreased in AIS patients compared to controls, and it had a good predictive value of AIS risk; (2) lnc-MALAT1 high expression correlated with decreased NIHSS score as well as reduced inflammatory factors levels in AIS patients; (3) lnc-MALAT1 high expression numerically correlated with prolonged RFS in AIS patients, but without statistical significance.
LncRNAs are involved in a variety of biological processes (such as genes transcription, organizing RNA-protein complex, and management in protein activity). [14][15][16][17] As one of the frequently investigated lncRNAs, lnc-MALAT1 has various molecular functions (including alternative splicing, transcriptional regulation and post-transcriptional regulation), and also participates in multiple physiological functions (including neural development, skeletal myogenesis and vascular growth). [7] As to its role in neurological diseases or cerebrovascular diseases, previous studies demonstrate that lnc-MALAT1 has protective effects in these diseases via repressing proapoptotic or pro-inflammatory factors. [10,11,18,19] For instance, a study displays that lnc-MALAT1 protects human brain vascular endothelial cells from OGD-induced apoptosis through activating phosphatidylinositol 3-kinase (PI3K). [19] In addition, lnc-MALAT1 expression is decreased in the spinal cords of mice with experimental autoimmune encephalomyelitis, and its knockdown raises levels of inflammatory cytokines (including IL-1 and IL-6). [11] Also, silencing of lnc-MALAT1 raises the levels of proapoptotic factor Bim and pro-inflammatory cytokines (including MCP-1 as well as E-selectin) in BMECs following oxygen-glucose deprivation (OGD), which is an in vitro mimic of ischemic stroke conditions. [10] Another study discloses that lnc-MALAT1 knockdown aggravates OGD-induced overexpression of pro-inflammatory cytokines including MCP-1 and IL-6 in mouse cerebral microvascular endothelial cells. [18] These studies reveal that lnc-MALAT1 has protective effects in neurological diseases or cerebrovascular diseases through decreasing inflammation level or inhibiting cell apoptosis.
Previous clinical trials mainly focus on the exploration of lnc-MALAT1 in cancer patients, which display that lnc-MALAT1 plays a tumor-promotive role in these cancer patients, while the information about the role of lnc-MALAT1 in neurological diseases patients or cerebrovascular diseases patients is still limited. [20][21][22] Just a few studies display that lnc-MALAT1 expression is reduced in central nervous system tissues from multiple sclerosis patients and carotid plaques from atherosclerosis patients. [11,12] As for the correlation of lnc-MALAT1 with disease severity and inflammation in neurological diseases or cerebrovascular diseases patients, only a study shows that elevated lnc-MALAT1 expression is correlated with less advanced lesions in atherosclerosis patients. [12] Considering that lnc-MALAT1 might participate in the cerebrovascular pathologies of stroke according to previous studies, meanwhile, it presented anti-inflammatory effect in cardiovascular and cerebrovascular diseases, we hypothesized that lnc-MALAT1 expression might be related to the disease risk, progression or inflammation level of AIS, while related explorations in AIS are seldomly reported. [9][10][11][12]23] In our study, we assessed the predictive value of lnc-MALAT1 for AIS risk, and we discovered that lnc-MALAT1 expression was lower in AIS patients compared to controls, and lnc-MALAT1 expression presented good diagnostic value for AIS, which might due to its protective effect in these diseases via repressing cell apoptosis of BMECs. Moreover, we investigated the correlation of lnc-MALAT1 expression with disease severity as well as inflammation in AIS patients, and we observed that lnc-MALAT1 high expression was associated with decreased NIHSS score, reduced levels of inflammatory factors (including CRP, TNF-α, IL-6, IL-8 and IL-22), meanwhile it reduced proportion of diabetes but numerically increased proportion of hypertension (without significant difference) in AIS patients. The possible reasons of these results might be as follows: (1) lnc-MALAT1 had protective influence on brain microvascular endothelial cells through inhibiting cell apoptosis, thereby contributed to maintaining a healthy brain endothelium that was essential for normal cerebrovascular physiology, therefore the severity of AIS was attenuated and decreased NIHSS score was observed [7,18]; (2) lnc-MALAT1 might decrease the expression of transcriptional factor nuclear factor-kappa B (NF-κB) that drove the transcription of a series of inflammatory factors, thus led to reduced inflammatory factors levels and alleviative inflammation in AIS patients [24]; (3) lnc-MALAT1 protected the brain microvascular endothelial cells against apoptosis, thereby facilitated microvascular permeability and helped restored cerebral vasoreactivity, thus it might attenuate hypertension in AIS patients [23] (4) lnc-MALAT1 was found to interact with transcription factor Foxo1 and SIRT1 transcription to induce poor glycemic control and insulin resistance, which might promote the occurrence of diabetes. [25,26] As to the prognostic value of lnc-MALAT1 in human diseases, it has been identified as a valuable biomarker for disease prognosis in several cancers, while limited clinical studies have been found in AIS, only a study displays that lnc-MALAT1 high expression is associated with prolonged main adverse cardiovascular and cerebrovascular events (MACCE)-free survival in atherosclerosis patients. [12] To our knowledge, there was still no evidence about the influence of lnc-MALAT1 on the prognosis in AIS patients. In our study, we observed that lnc-MALAT1 expression was numerically positively correlated with RFS in AIS patients, but without significant difference. These results might be due to: (1) lnc-MALAT1 reduced the production of inflammation cytokines and further decreased inflammation and attenuated disease progression, thus led to better RFS in AIS patients [10,18]; (2) lnc-MALAT1 might facilitate sensitivity to treatment in AIS, thereby increased treatment efficacy and resulted in longer RFS in AIS patients.
Some limitations existed in our study: (1) sample size of 120 AIS patients was relatively small, thus the statistical power might be low; (2) the median follow-up duration (25.0 months (range: 1.0-42.0 months)) was relatively short, thus the correlation of lnc-MALAT1 expression with long-term prognosis was not investigated; (3) detailed mechanism of lnc-MALAT1 in AIS was still unclear, and further study was still needed.

Conclusions
In conclusion, lnc-MALAT1 is downregulated and has a good predictive value of AIS risk, and its high expression correlates with decreased NIHSS score, reduced inflammation, as well as numerically better RFS in AIS patients.

Ethics approval and written consent to participate
The Institutional Review Board of HanDan Central Hospital approved the study protocol, and all participants or their guardians provided written informed consents.

Authors contributions
HR designed and approved the entire study; HR prepared the draft; FW, BL and ZS collected and analyzed the data; DQ revised the manuscript; All authors have read and approved the final manuscript.

Funding
None.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Consent for publication
Not applicable.    Correlation of lnc-MALAT1 expression with NIHSS score. Lnc-MALAT1 high expression was correlated with decreased NIHSS score. Correlation of lnc-MALAT1 expression with NIHSS score was determined by Spearman rank test. Lnc-MALAT1, long non-coding RNA metastasis-associated lung adenocarcinoma transcript 1; NIHSS score, National Institutes of Health Stroke Scale score. P < 0.05 was considered significant.