Up-Regulated Expression of Pro-Apoptotic Long Noncoding RNA lincRNA-p21 with Enhanced Cell Apoptosis in Lupus Nephritis

Accelerated cell apoptosis with dysregulated long noncoding RNAs is the crucial pathogenesis in lupus nephritis (LN). Pro-apoptotic lincRNA-p21 was studied in LN patients, cell lines with lentivirus-mediated overexpression and CRISPR interference (CRISPRi)-conducted repression, and a mouse model. Clinical samples were from patients and age/sex-matched controls. Expression of lincRNA-p21 and endogenous RNA target miR-181a, were examined in mononuclear and urine cells. Guide RNA sequences targeting lincRNA-p21 were cloned into CRISPRi with dCas9/ Krüppel-associated box (KRAB) domain. LincRNA-p21-silened transfectants were investigated for apoptosis and miR-181a expression. LincRNA-p21-overexpressed cells were evaluated for apoptosis and p53-related down-stream molecules. Balb/C mice were injected with pristane to induce LN and examined for apoptosis and lincRNA-p21. Higher lincRNA-p21 levels were found in LN mononuclear and urine cells, positively correlated with activity. There were lower miR-181a levels in LN mononuclear cells, negatively correlated with activity. Doxorubicin-induced apoptotic cells had up-regulated lincRNA-p21 levels. CRISPRi with dCas9/KARA domain showed efficient repression ability on transcription initiation/elongation. CRISPRi-conducted lincRNA-p21-silenced transfectants displayed reduced apoptosis with up-regulated miR-181a levels, whereas lentivirus-mediated lincRNA-p21-overexpressed cells revealed enhanced apoptosis with up-regulated downstream PUMA/Bax expression. LN mice had glomerular apoptosis with progressive increased lincRNA-p21 levels. Our results demonstrate up-regulated lincRNA-p21 expression in LN, implicating a potential diagnostic marker and therapeutic target.


Introduction
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by a loss of immune tolerance with the formation of immune complexes (IC) containing nuclear autoantigens, resulting in inflammation at various organs and tissues, of which damage to the kidney as a consequence of lupus nephritis (LN) is the most common cause of morbidity [1,2]. The current understanding of the crucial pathogenesis involves an imbalance between production of apoptotic cells and disposal of apoptotic materials [3]. Furthermore, apoptotic cell death with inefficient clearance results in the accumulation of self-double strand DNA (dsDNA), followed by a break of tolerance to induce production

Up-Regulated Expression of LincRNA-p21 in LN Patient
Firstly, we examined MMCs from SLE patients and healthy controls (HCs) for the expression of lincRNA-p21 and H19. Significantly higher lincRNA-p21 rather than H19 levels were found in SLE patients in comparison with HCs ( Figure 1A, p = 0.002). LN patients or those with class IV histopathology, had higher levels of lincRNA-p21 than those without renal involvement ( Figure 1B, LN versus Nil, p = 0.013, LN-IV versus Nil, p = 0.016). There were no differences in H19 levels between SLE patients without renal involvement and those with LN, either class IV histopathology or others ( Figure 1C). Moreover, there was a significant positive correlation between lincRNA-p21 levels and SLEDAI-2K scores ( Figure 1D, r = 0.423, p = 0.013) or daily proteinuria amounts ( Figure 1E, r = 0.395, p = 0.021). No correlation was identified between H19 levels and SLEDAI-2k scores or proteinuria amounts ( Figure 1F,G).

Up-Regulated Expression of LincRNA-p21 in LN Patient
Firstly, we examined MMCs from SLE patients and healthy controls (HCs) for the expression of lincRNA-p21 and H19. Significantly higher lincRNA-p21 rather than H19 levels were found in SLE patients in comparison with HCs ( Figure 1A, p = 0.002). LN patients or those with class IV histopathology, had higher levels of lincRNA-p21 than those without renal involvement ( Figure 1B, LN versus Nil, p = 0.013, LN-IV versus Nil, p = 0.016). There were no differences in H19 levels between SLE patients without renal involvement and those with LN, either class IV histopathology or others ( Figure 1C). Moreover, there was a significant positive correlation between lincRNA-p21 levels and SLEDAI-2K scores ( Figure 1D, r = 0.423, p = 0.013) or daily proteinuria amounts ( Figure  1E, r = 0.395, p = 0.021). No correlation was identified between H19 levels and SLEDAI-2k scores or proteinuria amounts ( Figure 1F,G). In Figure 2A, there were higher lincRNA-p21 levels in urine cells from LN patients than HCs (LN versus HC, 150.4 ± 78.0 versus 100.0 ± 27.1 %), and patients with class IV histopathology had significantly higher lincRNA-p21 levels than HCs (p = 0.028). No differences were found in H19 levels between HCs and LN patients or those with class IV histopathology ( Figure 2B). Next, we analyzed the expression of miR-181a in MNCs from SLE patients. There were decreased miR-181a levels in SLE patients ( Figure 2C, SLE versus HC, 56.8 ± 16.0 versus 100.0 ± 29.6 %), and LN patients had significantly lower levels than SLE without renal involvement ( Figure 2C, p = 0.011). A significant negative correlation existed between miR-181a levels and SLEDAI-2K scores ( Figure 2D, r = −0.383, p = 0.026). LincRNA-p21 levels in SLE patients without renal involvement and LN patients, including IV histopathology and others. (C) H19 levels in SLE without renal involvement and LN patients including IV histopathology and others. (D,E) A significant positive correlation between lincRNA-p21 levels and SLEDAI-2K scores or daily proteinuria amounts in SLE patients. (F,G) No correlation between H19 levels and SLEDAI-2K scores or daily proteinuria amounts in SLE patients. Horizontal lines in (A-D) are mean values from HCs and patients. n = 30 for HCs, n = 34 for SLE patients, n = 17 for LN patients, n = 17 for SLE without renal involvement, n = 8 for LN patients with IV histopathology, and n = 9 for other LN patients. * p < 0.05, ** p < 0.01.
In Figure 2A, there were higher lincRNA-p21 levels in urine cells from LN patients than HCs (LN versus HC, 150.4 ± 78.0 versus 100.0 ± 27.1 %), and patients with class IV histopathology had significantly higher lincRNA-p21 levels than HCs (p = 0.028). No differences were found in H19 levels between HCs and LN patients or those with class IV histopathology ( Figure 2B). Next, we analyzed the expression of miR-181a in MNCs from SLE patients. There were decreased miR-181a levels in SLE patients ( Figure 2C, SLE versus HC, 56.8 ± 16.0 versus 100.0 ± 29.6 %), and LN patients had significantly lower levels than SLE without renal involvement ( Figure 2C, p = 0.011). A significant negative correlation existed between miR-181a levels and SLEDAI-2K scores ( Figure 2D, r = −0.383, p = 0.026). modulator of T cell receptor (TCR) signaling, in primed T-lymphocytes can up-regulate the expression of IL-2 [19]. Lower expression of TCR-ζ chain has been identified in T cells from SLE patients with poor IL-2 production and refilling this molecule can normalize IL2 levels in vitro [20]. Further analyses of CD4+T cells from LN patients revealed lower levels of miR-181a, IL-2 and TCR-ζ chain as compared with those from HCs ( Figure 2F). Collectively, these ex vivo findings from clinical samples in Figures 1 and 2 demonstrated upregulated expression of lincRNA-p21 in LN patients.

Transcription Repression Ability in pAll-dCas9-KRAB.pPuro with dCas9/KRAB Domain
pAll-dCas9-KRAB.pPuro vector is shown in Figure 3A. Green fluorescent protein (GFP) silencing effects in gRNA sequences are demonstrated in Figure 3B with more than half no less than 70% efficacy ( Figure 3C). No differences were found in sequences targeting distance from TSS ( Figure 3D) or guanine-cytosine contents ( Figure 3E). CMV promoter silencing effects in gRNA sequences are shown in Figure 2F with more than half no less than 75% efficacy. These findings indicated that pAll-dCas9-KRAB.pPuro with dCas9/KRAB domain has efficient repression ability on transcription initiation and elongation through targeting promoter and coding regions, respectively. MNC subpopulations from LN patients and HCs were examined for the expression of lncRNAs. CD4+T cells from LN patients had higher levels of lincRNA-p21 and H19 in comparison with HCs ( Figure 2E). Interestingly, overexpressing miR-181a, an intrinsic modulator of T cell receptor (TCR) signaling, in primed T-lymphocytes can up-regulate the expression of IL-2 [19]. Lower expression of TCR-ζ chain has been identified in T cells from SLE patients with poor IL-2 production and refilling this molecule can normalize IL2 levels in vitro [20]. Further analyses of CD4+T cells from LN patients revealed lower levels of miR-181a, IL-2 and TCR-ζ chain as compared with those from HCs ( Figure 2F). Collectively, these ex vivo findings from clinical samples in Figures 1 and 2 demonstrated up-regulated expression of lincRNA-p21 in LN patients.

Transcription
Repression Ability in pAll-dCas9-KRAB.pPuro with dCas9/KRAB Domain pAll-dCas9-KRAB.pPuro vector is shown in Figure 3A. Green fluorescent protein (GFP) silencing effects in gRNA sequences are demonstrated in Figure 3B with more than half no less than 70% efficacy ( Figure 3C). No differences were found in sequences targeting distance from TSS ( Figure 3D) or guanine-cytosine contents ( Figure 3E). CMV promoter silencing effects in gRNA sequences are shown in Figure 2F with more than half no less than 75% efficacy. These findings indicated that pAll-dCas9-KRAB.pPuro with dCas9/KRAB domain has efficient repression ability on transcription initiation and elongation through targeting promoter and coding regions, respectively.

Up-Regulated Expression of LincRNA-p21 in Apoptotic Human T-Lymphocyte and Kidney Cell Lines
Under Dox-induced DNA damage to trigger p53-dependent cell apoptosis [11], we investigated lincRNA-p21 expression in a T-lymphocyte and two kidney cell lines. All of the results in Figure 4 and Figure 5 were representative of at least 2 independent experiments with similar findings.
By culturing Jurkat cells in the presence of Dox, there were dose-dependent up-regulated lincRNA-p21 levels and apoptotic cell ratios, and reciprocal down-regulation of miR-181a expression with reduced TCR-ζ chain and IL-2 levels as well as an increase in expression of caspase 3 and p21 ( Figures 4A-F). Moreover, in two CRISPRi-lincRNA-p21 transfectants, the one (71i) with a higher silenced efficacy (70% knockdown in comparison with CRISPRi-GFP transfectants) had decreased Dox-induced apoptotic cell ratios and increased miR-181a levels ( Figure 4G).
Notably, TNF-α has been identified to possess the strongest correlation with SLE activity among different tested plasma cytokines [21] and regulate the expression of abundant lncRNAs [22]. LncRNA-p21 levels were up-regulated in a dose-dependent manner by adding TNF-α into Jurkat cells culture ( Figure 4H). In the presence of Dox, despite a simultaneous up-regulated expression of lincRNA-p21 and caspase 3, the addition of a caspase 3 inhibitor (Z-DEVD-FMK) could reduce lincRNA-p21 levels ( Figure 4I), suggesting that caspase 3 activation can provide a feedback to enhance lincRNA-p21 expression in the apoptotic cell process.

Up-Regulated Expression of LincRNA-p21 in Apoptotic Human T-Lymphocyte and Kidney Cell Lines
Under Dox-induced DNA damage to trigger p53-dependent cell apoptosis [11], we investigated lincRNA-p21 expression in a T-lymphocyte and two kidney cell lines. All of the results in Figures 4 and 5 were representative of at least 2 independent experiments with similar findings.
By culturing Jurkat cells in the presence of Dox, there were dose-dependent upregulated lincRNA-p21 levels and apoptotic cell ratios, and reciprocal down-regulation of miR-181a expression with reduced TCR-ζ chain and IL-2 levels as well as an increase in expression of caspase 3 and p21 ( Figure 4A-F). Moreover, in two CRISPRi-lincRNA-p21 transfectants, the one (71i) with a higher silenced efficacy (70% knockdown in comparison with CRISPRi-GFP transfectants) had decreased Dox-induced apoptotic cell ratios and increased miR-181a levels ( Figure 4G). Furthermore, we induced apoptosis in HEK 293T kidney cells with Dox treatment. There were dose-dependent up-regulated lincRNA-p21 levels, apoptotic cell ratios and caspase 3 expression ( Figure 5A-C). CRISPRi-lincRNA-p21 transduced HEK 293T transfectants (71i) with a better silencing effect (86% knockdown as compared with control cells) demonstrated reduced Dox-induced apoptotic cell ratios and enhanced miR-181a expression levels ( Figure 5D).
Another renal tubular HK-2 cell line was treated with Dox to induce apoptosis, resulting in dose-dependent increases in lincRNA-p21 levels, apoptotic cell ratios, and expression of caspase 3 and p21 ( Figure 5E-G). Since lincRNA-p21 can provide a feedback to enhance the p53 transcriptional activity [10], we examined the expression of p53 downstream molecules in Dox-treated HK-2 cells. There was up-regulated expression of p53, lincRNA-p21, PUMA, and Bax ( Figure 5H). Moreover, lincRNA-p21-overexpressed HK-2 cells had increased expression of p53 downstream molecules PUMA and Bax ( Figure 5I) with enhanced Dox-induced apoptotic cell ratios ( Figure 5J). Instead of using transiently transfected cells, further studies can apply blasticidin selection process to create stable lincRNA-p21 tranfectants and examine whether there are higher expression levels of these downstream molecules.
Altogether, these in vitro results indicated that up-regulated expression of lincRNA-p21 could enhance apoptosis in human T-lymphocyte and kidney cell lines. Notably, TNF-α has been identified to possess the strongest correlation with SLE activity among different tested plasma cytokines [21] and regulate the expression of abundant lncRNAs [22]. LncRNA-p21 levels were up-regulated in a dose-dependent manner by adding TNF-α into Jurkat cells culture ( Figure 4H). In the presence of Dox, despite a simultaneous up-regulated expression of lincRNA-p21 and caspase 3, the addition of a caspase 3 inhibitor (Z-DEVD-FMK) could reduce lincRNA-p21 levels ( Figure 4I), suggesting that caspase 3 activation can provide a feedback to enhance lincRNA-p21 expression in the apoptotic cell process.
Another renal tubular HK-2 cell line was treated with Dox to induce apoptosis, resulting in dose-dependent increases in lincRNA-p21 levels, apoptotic cell ratios, and expression of caspase 3 and p21 ( Figure 5E-G). Since lincRNA-p21 can provide a feedback to enhance the p53 transcriptional activity [10], we examined the expression of p53 downstream molecules in Dox-treated HK-2 cells. There was up-regulated expression of p53, lincRNA-p21, PUMA, and Bax ( Figure 5H). Moreover, lincRNA-p21-overexpressed HK-2 cells had increased expression of p53 downstream molecules PUMA and Bax ( Figure 5I) with enhanced Dox-induced apoptotic cell ratios ( Figure 5J). Instead of using transiently transfected cells, further studies can apply blasticidin selection process to create stable lincRNA-p21 tranfectants and examine whether there are higher expression levels of these downstream molecules.
Altogether, these in vitro results indicated that up-regulated expression of lincRNA-p21 could enhance apoptosis in human T-lymphocyte and kidney cell lines.  Figure 5 were representative of at least two independent experiments with similar findings.

Up-Regulated Expression of LincRNA-p21 in a LN Mouse Model
Pristane-injected Balb/c mice were periodically examined for dsDNA antibody and urinary protein. Significantly elevated proteinuria amounts and higher anti-double strand DNA (dsDNA) titers were noted at 5 and 6 months after induction, respectively ( Figure  6A). Their kidneys were removed upon sacrifice for histopathological and in situ apoptosis analyses. In Figure 6B, kidneys from LN mice had GN with glomerular hyper-cellularity/mesangial expansion and the presence of in situ apoptotic cells at 6 months after pristane injection, an induced model with the presence of Fas-independent transferase dUTP nick end labeling (TUNEL)-positive tissue cells [23]. Significantly up-regulated lin-cRNA-p21 and down-regulated miR-181a levels were shown in kidney cells from LN mice at 6 months after induction ( Figure 6C,D). Furthermore, there were increasingly up-regulated lincRNA-p21 levels in CD4 + T cells from LN mice after pristane induction with significant higher levels at 6 months, as well as increased expression levels of caspase 3 and p21 ( Figure 6E). Taken together, these ex vivo data from LN mice indicated progressively up-regulated lincRNA-p21 expression in kidney and CD4 + T cells.  Figure 5 were representative of at least two independent experiments with similar findings.

Up-Regulated Expression of LincRNA-p21 in a LN Mouse Model
Pristane-injected Balb/c mice were periodically examined for dsDNA antibody and urinary protein. Significantly elevated proteinuria amounts and higher anti-double strand DNA (dsDNA) titers were noted at 5 and 6 months after induction, respectively ( Figure 6A). Their kidneys were removed upon sacrifice for histopathological and in situ apoptosis analyses. In Figure 6B, kidneys from LN mice had GN with glomerular hyper-cellularity/ mesangial expansion and the presence of in situ apoptotic cells at 6 months after pristane injection, an induced model with the presence of Fas-independent transferase dUTP nick end labeling (TUNEL)-positive tissue cells [23]. Significantly up-regulated lincRNA-p21 and down-regulated miR-181a levels were shown in kidney cells from LN mice at 6 months after induction ( Figure 6C,D). Furthermore, there were increasingly up-regulated lincRNA-p21 levels in CD4 + T cells from LN mice after pristane induction with significant higher levels at 6 months, as well as increased expression levels of caspase 3 and p21 ( Figure 6E). Taken together, these ex vivo data from LN mice indicated progressively up-regulated lincRNA-p21 expression in kidney and CD4 + T cells. Values are the mean ± SEM of 3 mice per group for measurement of lincRNA-p21 levels, and 3 mice per group for caspase 3/p21 immunoblot assay. All of the results in Figure 6 were representative of two independent experiments with similar findings.

Discussion
LN has multiple pathogenic pathways including aberrant apoptosis, autoantibody production and IC deposition with complement activation [1,2]. Apoptotic cell death with roles in tissue damage and immune dysregulation, is involved in the generation of autoantigens and the externalization of modified nuclear antigens [3]. In the development of LN, there are accelerated cell apoptosis in circulating lymphocytes [24], kidney cells (renal tubular and glomerular parenchymal cells) [25], and phagocytes for clearance of apoptotic cells [26]. LncRNAs are emerging as key players in controlling the cellular apoptotic process [6], and these molecules participate in the LN pathogenesis with aberrant expression levels [8]. In this study, up-regulated expression of pro-apoptotic lincRNA-p21, rather than anti-apoptotic H19, was identified in MNCs, especially in CD4+ T cells, from LN patients as well as a human T-lymphocyte line receiving Dox treatment to induce apoptosis. Moreover, higher lincRNA-p21 levels were detected in urine cells from LN patients and human kidney cell lines under the DNA damage response. Notably, transcriptomewide studies have demonstrated that the expression of different lncRNAs is specific for cell types to exert their distinct regulatory functions [22]. Pro-inflammatory cytokines, TNF-α in particular, have been demonstrated to regulate the expression of lncRNAs, some of which have up-regulated levels in a NF-κB dependent manner [27], as demonstrated in this study with dose-dependent increases in lincRNA-p21 expression levels in Jurkat cells Values are the mean ± SEM of 3 mice per group for measurement of lincRNA-p21 levels, and 3 mice per group for caspase 3/p21 immunoblot assay. All of the results in Figure 6 were representative of two independent experiments with similar findings.

Discussion
LN has multiple pathogenic pathways including aberrant apoptosis, autoantibody production and IC deposition with complement activation [1,2]. Apoptotic cell death with roles in tissue damage and immune dysregulation, is involved in the generation of autoantigens and the externalization of modified nuclear antigens [3]. In the development of LN, there are accelerated cell apoptosis in circulating lymphocytes [24], kidney cells (renal tubular and glomerular parenchymal cells) [25], and phagocytes for clearance of apoptotic cells [26]. LncRNAs are emerging as key players in controlling the cellular apoptotic process [6], and these molecules participate in the LN pathogenesis with aberrant expression levels [8]. In this study, up-regulated expression of pro-apoptotic lincRNA-p21, rather than anti-apoptotic H19, was identified in MNCs, especially in CD4+ T cells, from LN patients as well as a human T-lymphocyte line receiving Dox treatment to induce apoptosis. Moreover, higher lincRNA-p21 levels were detected in urine cells from LN patients and human kidney cell lines under the DNA damage response. Notably, transcriptome-wide studies have demonstrated that the expression of different lncRNAs is specific for cell types to exert their distinct regulatory functions [22]. Pro-inflammatory cytokines, TNF-α in particular, have been demonstrated to regulate the expression of lncRNAs, some of which have up-regulated levels in a NF-κB dependent manner [27], as demonstrated in this study with dose-dependent increases in lincRNA-p21 expression levels in Jurkat cells upon in vitro TNF-α stimulation ( Figure 4H). Indeed, in SLE patients, elevated cytokines levels can influence the expression of lncRNAs at different tissues and organs, leading to heterogeneous clinical involvement [16,21,22]. Furthermore, we observed increases in apoptotic cell ratios and expression levels of caspase 3 and p21 in Dox-treated T-lymphocyte and kidney cell lines. Up-regulated expression of lincRNA-p21 contributes to apoptotic cell death in circulating lymphocytes and renal cells, followed by production of autoantibodies, resulting in in situ IC accumulation and the formation of GN in SLE. Further experiments can use T lymphocytes and renal tubular cells from LN patients to improve the clinical relevance of this study.
Owing to complex disease presentations and inherent limitations in clinical research, there are difficulties in performing direct studies in lupus patients. SLE mouse models have been developed to dissect pathogenic mechanisms and identify therapeutic targets [28]. In addition to spontaneous lupus models like NZB/W F1 and MRL/lpr mice, induced mouse models, particularly the pristane-induced mice with renal IC deposition causing GN, are useful tools to investigate the molecular pathogenesis with dysregulated signaling pathways and to screen therapeutic modalities in LN [28,29]. In this study with Balb/c female mice, LN developed with increased anti-dsDNA levels, elevated proteinuria amounts, and the formation of GN after pristane induction. The pristane-induced model is driven by a strong type I IFN response [29], a well-known inducer of lncRNA expression in immune responses, which is much weaker in spontaneous mouse models like NZB/W F1 and MRL/lpr mice. Therefore, such an induced model is suitable to analyze the role of lincRNA-p21 in LN-related immune processes. Moreover, there were in situ apoptosis with up-regulated expression of lincRNA-p21 in CD4+ lymphocytes and kidney cells as well as elevated caspase 3 and p21 levels. By using the lupus mouse model through a proof-of-concept approach, we demonstrated a progressive increase in renal expression of lincRNA-p21 during the development of LN. In vitro experiments by using CRISPRi-lincRNA-p21 transfected cell lines revealed lower apoptotic cell ratios in the presence of DNA damage response, implicating a therapeutic strategy to treat LN by knocking down the expression of lincRNA-p21 to reduce cell apoptosis. Nevertheless, further efforts are needed to elucidate the potential of lincRNA-p21 as a therapeutic candidate by silencing its renal expression to examine whether there is improvement of GN in the LN mouse model. RNA interference can attack mature cytosolic RNA for degradation, but not effective in targeting nuclear lncRNA [30]. Despite a complete gene knockout, CRISPR/Cas9 editing can introduce the DNA cleavage with a risk of error-prone repair, whereas CRISPR/dCas9based reversible gene repression allows specific transcriptional and epigenomic modulation at targeted loci with a less off-target effect, not only enabling the time-resolved investigation in gene functions but also predicting the outcome of pharmacological inhibition on gene products [17,30]. In this study with CRISPRi containing dCas9/KRAB repression domain for in vitro transfection, we demonstrate that silencing lincRNA-p21 inactivates p21 and caspase 3 to reduce the apoptotic cell death in human cell lines. Interestingly, dCas9 coupling a bipartite repressor KRAB-MeCP2 has recently been demonstrated to hold a higher transcription repression ability than KRAB alone [31], raising a possibility to create an all-in-one CRISPRi vector with this bipartite domain for more efficiently silencing the expression of lincRNA-p21 to enhance the inhibition efficacy on cell apoptosis.
Transcript RNAs with specific MREs can communicate with others via the miRNA messenger and may serve as ceRNAs to de-repress the activity of other RNAs with similar MREs by competing for the same miRNAs [32]. LncRNAs harboring the MREs can serve as ceRNAs with the function to sponge or sequestrate miRNAs, and growing evidence has demonstrated that, an interaction between lncRNAs and miRNAs can regulate miscellaneous cellular processes to affect human disease states [13,32]. MiR-181a levels in PBMNCs have been shown to be down-regulated in SLE patients with higher disease activity [15], and also demonstrated in LN patients with a negative SLEDAI-2K correlation from this study. Furthermore, in cell lines under the DNA damage response, there were increases in lincRNA-p21 levels with a reciprocal decrease in miR-181a expression, and CRISPRi-mediated repression of lincRNA-p21 could restore the down-regulated expression of miR-181a. In addition, decreased mir-181a levels were observed in kidney cells form LN mice in this study. By transducing primed T cells with retroviral vector carrying miR-181a, IL-2 production can be up-regulated through targeting multiple negative regulators to augment T cell activation [19]. Impaired IL-2 production has been shown in T cells from SLE patients, and IL-2-treated LN mice have decreased autoantibody levels and reduced GN severity [33]. Notably, low-dose IL-2 treatment has recently been demonstrated to have the therapeutic efficacy in LN patients [34]. Accordingly, these findings suggest that the disease progression in LN can be reduced by inhibiting the expression of lincRNA-p21 to raise miR-181a levels for IL-2 restoration.

LN Patients and Age/Sex-Matched Control Subjects
Thirty-four patients fulfilling the American College of Rheumatology revised Criteria for SLE [35], 30 females and 4 males aged from 28 to 65 years (44.4 ± 1.6), and age/sexmatched healthy controls (HCs) were enrolled into this study. Their venous blood samples were collected for further examination. Medical records were reviewed for demographic, clinical and laboratory data, and the disease activity at the time of sample collection were assessed by SLEDAI-2K. Seventeen patients in this study had LN, 15 females and 2 males aged from 28 to 60 years (44.0 ± 2.3), including 8 with class IV, 5 with class III, 3 with class V, and one with class II histopathological findings. All SLE patients were under corticosteroids therapy. In class IV patients, 4 received cyclophosphamide, 2 used azathioprine, and 2 under mycophenolate mofetil treatment. In class III patients, 3 received azathioprine and 2 under mycophenolate mofetil therapy. Three class V patients received azathioprine therapy, and immunosuppressants were not prescribed in one class II case. In 17 SLE patients without renal involvement, 15 received azathioprine therapy, and 2 not under immunosuppressants treatment. Notably, LN patients had significantly higher SLEDAI-2K scores than SLE without renal involvement (8.4 ± 1.2 versus 3.5 ± 0.7, p < 0.001). Both renal involvement and kidney sparing groups has no differences in their age/sex distribution. Fresh urine specimens were collected from all LN patients and age/sex-matched HCs. This study was approved by the Institutional Review Board of National Cheng Kung University Hospital (approval number A-ER-108-455) with the informed consent from each participant.

Pristane-Induced LN Mouse Model
Eight-week old female BALB/c mice were purchased from the Laboratory Animal Center of our medical college, and housed under specific pathogen-free conditions. Animal experiments were approved by the Institutional Animal Care and Use Committee of our university. Mice were intraperitoneally (i.p.) singly injected with 0.5 mL pristane (Sigma-Aldrich, St. Louis, MO, USA) to induce LN, whereas the control group was i.p. singly injected with 0.5 mL of phosphate-buffered saline (PBS) [29,36]. Except 5 mice per pristaneinjected or control group for measuring renal lincRNA-p21 and mir-181a levels, there were 3 mice per group in all animal experiments. Blood samples were collected monthly for examining anti-dsDNA levels until 6 months after induction, while urine specimens were harvested for measuring proteinuria concentrations at 0, 3, and 5 months. Mice were sacrificed at 0, 3, 5, and 6 months for removing the kidneys to measure lincRNA-p21/mir-181a levels and analyze histopatological findings, and at 0.5, 3, 5, 6, and 7 months for obtaining the spleens to measure lincRNA-p21 levels and perform immunoblotting assay.

Purification of Human and Mouse Cells
Human MNCs were isolated from blood samples by Ficoll-Paque PLUS (GE Healthcare, Chicago, IL, USA) and incubated with CD14 microbeads. CD14+ cells were eluted from the positive selection column of Magnetic Cell Sorter (Miltenyi Biotec, Germany). CD14-cells were incubated with CD4 microbeads, and CD4+ cells were eluted from the column. Mouse spleens were homogenized by using syringe plunger and mesh strainer. Mouse MNCs were further incubated with PE-Cy5 anti-CD4 (BD Pharmingen, San Diego, CA, USA) or FITC anti-CD19 (BD Pharmingen), and sorted by Moflo XDP Cell Sorter (Beckman Coulter, Mountain View, CA) to obtain CD4+ or CD19+ cells. Purity of cell subpopulation was confirmed to be up to 95 % by flow cytometric analyses. Human urine cells were isolated from urine specimens by centrifugation and washing procedures to obtain cell pellets [37]. After removing capsules, mouse kidneys were minced into tiny pieces to obtain cortex tissues, followed by incubation with digestion buffer with collagenase (Sigma-Aldrich), and centrifuged to collect cell pellets [38].

Quantitative Real Time Polymerase Chain Reaction (qRT-PCR)
Total RNAs from human or mouse cells were extracted by TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and complementary DNAs were obtained by using reverse transcriptase (Applied Biosystems, Foster City, CA, USA). qRT-PCR was performed to quantify the target RNAs levels by using the SYBR qPCR Mix Kit (TOOLS) [39]. The condition of PCR was: 95 • C for 5 min, 95 • C for 15 s, primer-melting temperature (Tm) for 1 min with 40 cycles, and elongation at 72 • C for 20 s. Primer sequences were as follows.
The relative abundance of a measured gene expression was normalized by GAPDH gene from each sample. The average levels of human HCs or PBS-injected control mice, and expression levels of cell lines without stimulation, CRISPRi-GFP-silenced transfectants, and LV-SFFV-Blast-overexpressed cells were determined as 100%.
For analyzing the expression levels of human and mouse miR-181a, total RNAs were reverse transcription (RT) by using the reverse transcriptase kit (Applied Biosystems) with 10 ng purified RNA, dNTP, MultliScribe reverse transcriptase, RT buffer, RNase inhibitor, random primers, and gene-specific stem-loop RT primer with a TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems) in Smart Cycler (Cepheid, Sunnyvale, CA, USA) [39]. The reagents were incubated with 16 • C for 30 min, 42 • C for 30 min and 85 • C for 5 min. The condition of PCR was: 95 • C for 10 min, 95 • C for 15 sec and 60 • C for 1 min with 40 cycles. Quantitative expression levels of miR-181a were analyzed with RNU6B small RNA (Applied Biosystems) as an endogenous control. The average levels of human HCs or PBS-injected control mice and expression levels of cell lines without stimulation and CRISPRi-GFP-silenced transfectants were determined as 100%.

Conclusions
A better understanding of the LN pathogenesis can help to develop disease biomarkers in diagnosis and prognosis, and to identify novel therapeutics other than conventional immunosuppressive agents with significant failures and adverse effects. In this study, by using clinical samples, human cell lines, and a mouse model, we demonstrate up-regulated expression of lincRNA-p21 in LN, implicating this pro-apoptotic lncRNA as a potential diagnostic biomarker and therapeutic target.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data of this study can be provided to researchers from the corresponding author upon reasonable request.

Acknowledgments:
The authors are indebted to Didier Trono (École Polytechnique Fédérale de Lausanne, Switzerland) for providing psPAX2 and pMD2.G plasmids.

Conflicts of Interest:
The authors declare that they have no conflicts of interest.