LncRNA‐HOTAIR promotes endothelial cell pyroptosis by regulating the miR‐22/NLRP3 axis in hyperuricaemia

Abstract Long non‐coding RNA (lncRNA) plays an important role in the renal inflammatory response caused by hyperuricaemia. However, the underlying molecular mechanisms through which lncRNA is involved in endothelial injury induced by hyperuricaemia remain unclear. In this study, we investigated the regulatory role of lncRNA‐HOTAIR in high concentration of uric acid (HUA)–induced renal injury. We established hyperuricaemia mouse model and an in vitro uric acid (UA)–induced human umbilical vein endothelial cell (HUVEC) injury model. In HUA‐treated HUVECs and hyperuricaemia mice, we observed increased HOTAIR and decreased miR‐22 expression. The expression of pyroptosis‐associated protein (NLRP3, Caspase‐1, GSDMD‐N, GSDMD‐FL) was increased. The release of LDH, IL‐1β and IL‐18 in cell supernatants and the sera of model mice was also increased. The proliferation of HUVECs stimulated by HUA was significantly inhibited, and the number of TUNEL‐positive cells in hyperuricaemia mouse kidney was increased. Bioinformatics analysis and luciferase reporter and RIP assays confirmed that HOTAIR promoted NLRP3 inflammasome activation by competitively binding miR‐22. In gain‐ or loss‐of‐function experiments, we found that HOTAIR and NLRP3 overexpression or miR‐22 knock down activated the NLRP3 inflammasome and promoted pyroptosis in HUA‐treated HUVECs, while NLRP3 and HOTAIR knockdown or a miR‐22 mimic exerted the opposite effects. Furthermore, in vivo experiments validated that HOTAIR knockdown alleviated renal inflammation in hyperuricaemia mice. In conclusion, we demonstrated that in hyperuricaemia, lncRNA‐HOTAIR promotes endothelial cell pyroptosis by competitively binding miR‐22 to regulate NLRP3 expression.


| INTRODUC TI ON
Hyperuricaemia is a disease caused by purine metabolism disorders.
It has become a risk factor for cardiovascular disease and chronic kidney disease, in addition to hypertension, diabetes, obesity, hyperlipidaemia, etc. [1][2][3] Current studies have confirmed that high UA levels can cause microvascular disorders in the kidney, vascular endothelial dysfunction and the release of inflammatory cytokines, [4][5][6] subsequently causing kidney damage. Renal biopsies from patients with hyperuricaemia nephropathy showed interstitial infiltration of inflammatory cells and elevated serum levels of inflammatory markers, indicating that hyperuricaemia can induce an inflammatory response and cause kidney injury. 7,8 Inflammatory damage in the kidney is inseparable from the activation of inflammasomes. In particular, the nucleotide-binding domain and leucine-rich repeat (NLR) protein 3 (NLRP3) inflammasome has been shown to cause acute and chronic kidney diseases by regulating canonical and non-canonical mechanisms of inflammation. [9][10][11] Therefore, the involvement of NLRP3 in kidney damage caused by high serum UA stimulation cannot be ignored in clinical practice. The NLRP3 inflammasome is composed of NLRP3, apoptosis-associated speck-like protein containing a CARD (ASC) and active caspase-1, which regulates interleukin-1β (IL-1β) maturation. 12,13 Its activation promotes caspase-1-mediated cytokine maturation and simultaneously enhances the release of the N-terminal domain of gasdermin D from the self-inhibiting C-terminal domain and subsequent transfer to the plasma membrane, forming large nanopores in the plasma membrane, thereby causing pyroptosis. [14][15][16][17] This process is accompanied by the release of a large number of pro-inflammatory cytokines, further aggravating tissue inflammatory injury. Studies have shown that pyroptosis is associated with the occurrence and development of kidney diseases. [18][19][20][21] The major kidney damage caused by hyperuricaemia is endothelial cell dysfunction. However, the molecular mechanism by which HUA triggers the activation of inflammasomes and thereby promotes pyroptosis has not been elucidated.
Long non-coding RNA (lncRNA) is a class of non-coding RNA molecules over 200 nt in length. lncRNA plays important roles in cell differentiation, proliferation, apoptosis and pyroptosis. [22][23][24] Its mechanism relies on binding with microRNAs (miRNAs) to prevent the latter from binding to target genes, thereby regulating gene expression. 25,26 Similarly, lncRNAs play important roles in the occurrence and development of kidney diseases. For example, Gm4419 increases the inflammation-mediated inflammatory response in diabetic nephropathy through the nuclear factor-kappa B (NF-κB)/ NLRP3 pathway. 27 Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) regulates high glucose-induced inflammatory responses in endothelial cells. 28 Our previous study also confirmed that HOX transcript antisense RNA (HOTAIR) promotes the proliferation and migration of clear cell renal cell carcinoma (ccRCC) by increasing hypoxia-inducible factor (HIF)-1α expression via competing with miR-217. 29 lncRNAs can also promote the angiogenesis function of endothelial cells. 30 These results suggest that HOTAIR might be involved in vascular endothelial cell injury induced by HUA. In acute renal injury caused by sepsis, HOTAIR promotes the apoptosis of HK-2 cells through the miR-22/HMGB1 pathway, thereby causing kidney injury. 31 However, whether the interaction between HOTAIR and miR-22 participates in vascular endothelial injury caused by HUA levels has not been reported.
Therefore, in this study, by culturing vascular endothelial cells and constructing a hyperuricaemia animal model, we attempted to confirm the molecular mechanism underlying kidney injury; that is, in an HUA environment, HOTAIR induces the activation of the NLRP3 inflammasome and pyroptosis by binding to miR-22, ultimately promoting endothelial cell injury and causing kidney injury.

| Patients and controls
Seventeen hyperuricaemia patients at the Chinese PLA General Hospital were screened by the following criteria. The inclusion criteria for 17 hyperuricaemia patients (male 12 and female 5) were as follows: subjects are between the age of 21 and 85, with an average age of 43.5 ± 3.7 years; subjects meet the diagnostic criteria of hyperuricaemia (serum UA concentration >440 μmol/L for males and >367 μmol/L for females); first onset; informed consent. The exclusion criteria were as follows: patients with stage 3 hypertension, diabetic nephropathy and anti-UA therapy in the previous 3 months.
Among those included, 12 were male, and five were female, with an average age of 43.5 years. Additionally, we selected 17 age-matched healthy volunteers as controls. Fasting blood samples were collected from all subjects and centrifuged at 3000 g for 10 min at 4°C. Serum samples were then collected for detecting the expression levels of HOTAIR, miR-22, IL-1β and IL-18 using qPCR analysis and ELISA.
This study was approved by the Ethics Committee of PLA General Hospital, and informed consent was obtained from all participants.
In addition, lentivirus vectors psiH1-HOTAIR and psiH1-NLRP3 were used to knock down the above 2 genes. A miR-22 mimic and inhibitor (AMO-22) were designed and manufactured by Sangon Biotech.
HUVECs were infected with the lentiviral vectors when they reached 70%-80% confluence.
For the reporter assays, HUVECs were harvested and lysed and then the assays were conducted according to manufacturer's instructions (Promega, cat. no. E1910). Firefly and Renilla luciferase activities were measured using a dual-luciferase reporter assay system. Firefly luciferase activity (F value) and Renilla luciferase activity (R value) were detected using a fluorescence/chemiluminescence instrument.
The ratio of Renilla luciferase activity to firefly luciferase activity represented the relative expression level of reporter genes. Similarly, the region containing the target sequence in human lncRNA-HOTAIR was cloned into the vector psiCHECKTM-2, and similar experiments were carried out.

| Animals and treatment
Eight-to twelve-week-old C57/B6 male mice were provided by the Experimental Animal Center of the Academy of Military Medical Sciences (a total of 24 mice: six mice in HUA model group, six mice in HUA + Vector group and six mice in HUA + shHOTAIR group, six for control group); the animal were allowed to adapt for 1 week before the experiments. Hyperuricaemia was induced by gavage of potassium oxonate (PO; Sigma-Aldrich, cat. no. U156124) (250 mg/kg·day) and UA (Sigma-Aldrich, cat. no. U2625) solution (250 mg/kg·day) once a day for 10 consecutive days. A UA assay kit (abcam, cat. no. Ab65344) was used to detect serum UA levels in mice and determine whether modelling was successful.
The cells were washed twice with phosphate-buffered saline (PBS, Beyotime, cat. no. C0221A), and then, 10 µl of CCK-8 solution (Dojindo, cat. no. CK04) was added to each well. The cells were incubated at 37°C for 2 h, and then, the absorbance (450 nm) for each well was determined using an ELISA reader (Bio-Rad, iMARK).

| LDH release assay
To test cell cytotoxicity, we used a lactate dehydrogenase assay kit (abcam, cat. no. ab102526) to detect LDH activity in serum and cell culture supernatant. Samples were placed into a 96-well plate, and 50 µl of reaction mixture (LDH assay buffer, 48 µl; LDH substrate mix, 2 µl) was added. After incubating 45 min at 37°C, the absorbance (450 nm) for each well was determined using an ELISA reader (Bio-Rad, iMARK).

| Western blot
Total protein was extracted from mouse kidney tissue and HUVECs using RIPA lysis buffer (Beyotime, cat. no. P0013B) on ice; the obtained lysate was centrifuged at 12,000 × g at 4°C for cat. no. ab215203). Next, the membranes were incubated with horseradish peroxidase-conjugated anti-rabbit IgG secondary antibodies for 2 h at room temperature. Bands were detected using a ChemiDoc MP Imaging System (Bio-Rad), and ImageJ 1.51 software was used for the quantitative analysis of all protein bands.

| Terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) staining
Paraffin sections of mouse kidney tissue were dewaxed in water.
TUNEL staining was performed according to the kit instructions.
The sections were stained with 3, 3′-diaminobenzidine (DAB; Servicebio, cat. no. G2111), and the nuclei were stained with haematoxylin. TUNEL-positive cells, which were stained brown, were counted under 400× magnification. The pyroptosis rate was calculated as the percentage of TUNEL-positive cells relative to the total number of renal tubular cells. Nuclei stained with haematoxylin were blue, and pyroptosis nuclei stained with DAB were brownish-yellow.

| Immunohistochemical staining
Renal samples were fixed in 4% paraformaldehyde and embedded in paraffin. The samples were then cut into 5µm-thick sections and incubated with primary antibody, that is anti-CD68 (abcam, cat. no. ab955), at 4°C overnight, followed by incubation with the appropriate secondary antibody. Then, the sections were stained with diaminobenzidine, and images were captured using a fluorescence microscope (Olympus).

| ELISA
Twenty-four hours after transfection, the concentrations of IL-1β (abcam, cat. no. ab100562) and IL-18 (Novus, cat. no. KA0561) in HUVEC culture supernatants or HUA mouse sera were determined by ELISA. The detection was performed strictly according to the kit instructions.

Forward
Reverse

| Statistical analysis
Statistical analysis was performed using IBM SPSS Statistics 24.0 software (IBM Corporation). The quantitative data are expressed as the mean ± standard deviation (SD). All data were tested for a normal distribution using the Shapiro-Wilk test. Comparisons between two groups were performed with Student's t test, and multiple comparisons of more than two groups were performed with one-way analysis of variance (ANOVA). A p value less than 0.05 was considered to be statistically significant.

F I G U R E 1 Effect of HUA on inflammation and the expression of lncRNA-HOTAIR in patients, hyperuricaemic mice and HUVECs. (A)
Serum levels of HOTAIR in normal controls and hyperuricaemia patients, measured by qPCR. HOTAIR expression was significantly increased in hyperuricaemia group; * p < 0.05 compared to the normal controls; n = 17 in each group. (B) Serum levels of IL-1β and IL-18 in normal controls and hyperuricaemia patients, measured by ELISA. Serum IL-1β and IL-18 levels were significantly higher in hyperuricaemia group; * p < 0.05 compared to the normal controls; n = 17 in each group. (C) Serum levels of HOTAIR in control mice and hyperuricaemia mice, measured by qPCR. HOTAIR expression was significantly higher in the hyperuricaemia mice; * p < 0.05 compared to the control group; n = 6 in each group. (D) Serum levels of IL-1β and IL-18 in control mice and hyperuricaemia mice, measured by and ELISA. IL-1β and IL-18 levels were significantly higher in hyperuricaemia group; * p < 0.05 compared to the control group; n = 6 in each group. (E) Immunohistochemistry of CD68 in renal tissue from hyperuricaemia mice and control mice. Positive CD68 staining implies the infiltration of mononuclear macrophages in the renal interstitium and glomerular mesangial area in hyperuricaemia group; * p < 0.05 compared to the control group; n = 6 in each group (scale bar, 100 μm; magnification, 400×). (F) HOTAIR levels in HUA-treated HUVECs, measured by qPCR. HOTAIR expression was significantly higher in the HUA group; * p < 0.05 compared to the control group; n = 3 in each group. (G) IL-1β and IL-18 levels in HUA-treated HUVECs, measured by ELISA. IL-1β and IL-18 levels were significantly higher in HUA group; * p < 0.05 compared to the control group; n = 3 in each group

| High UA levels promote an increase in HOTAIR expression and induce inflammation
Serum from patients with hyperuricaemia and healthy individuals was collected for qPCR analysis. Compared with that in serum from healthy individuals, the expression of HOTAIR in serum from hyperuricaemia patients was significantly higher ( Figure 1A). Serum levels of inflammatory cytokines were determined by ELISA. The results showed that compared to those in the healthy control group, the serum inflammatory cytokine levels of IL-1β and IL-18 in hyperuricaemia patients were significantly higher ( Figure 1B). Similarly, in the hyperuricaemia mouse model, the expression of HOTAIR, as determined by qPCR, in the kidney tissue of model mice was significantly higher than that in tissue from the healthy control group mice; statistical analysis indicated that the difference between the groups was significant ( Figure 1E).
Subsequently, HUVECs were cultured and stimulated with 600 μmol/L UA for 24 h. qPCR showed that the expression of HOTAIR in cells stimulated with HUA group was significantly higher than that in cells of the control group ( Figure 1F); furthermore, the levels of the inflammatory cytokines IL-1β and IL-18 in the culture supernatant was also significantly higher in the HUA group ( Figure 1G).

| HUA levels induce NLRP3 inflammasome activation and pyroptosis in endothelial cells
To determine whether high UA levels activate the NLRP3 inflam- LDH release in the experimental group was also significantly higher ( Figure 2C).
Furthermore, we examined the status of inflammasome activation and pyroptosis in kidney tissue from hyperuricaemia mice.
The results showed that the protein expression levels of NLRP3, caspase-1, GSDMD-FL and GSDMD-N in kidney tissue from hyperuricaemia mice were significantly higher than those in kidney tissue from mice in the saline control group ( Figure 2D). Tissue immunofluorescence results indicated that the expression of NLRP3, caspase-1 and GSDMD-FL was distributed throughout the glomeruli of the hyperuricaemia group; the glomerular fluorescence intensity in the hyperuricaemia group was significantly higher than that in the control group (Figure 2E-G). Similarly, compared with that in the control group, serum LDH in hyperuricaemia mice was significantly higher ( Figure 2H), and the number of kidney TUNEL-positive cells was significantly higher ( Figure 2I). Both in vitro and in vivo experimental results indicated that HUA induced NLRP3 inflammasome activation and pyroptosis in the cells.

| miR-22 binds to HOTAIR and NLRP3
To explore the regulatory relationship between HOTAIR and NLRP3, and NLRP3 and miR-22 have potential binding sites ( Figure 3A).
Subsequently, a luciferase reporter assay confirmed that the miR-22 mimic significantly inhibited luciferase activity in wild-type psiCHECK-WT-NLRP3 but had no significant inhibitory effect on the luciferase activity of psiCHECK-MT-NLRP3 ( Figure 3G).
miRNAs form an RNA-induced silencing complex (RISC) with RNA via Ago2 and then induce gene silencing. 32 Based on the above results, we speculated that HOTAIR might bind with miR-22 and Ago2 to form a RISC, which was further verified by the RIP experiment. The results showed that the amounts of HOTAIR and miR-22 that immunoprecipitated with Ago2 were significantly higher than those that precipitated with the IgG control ( Figure 3H).
Finally, we applied a biotin-labelled pulldown assay to further explore whether miR-22 and HOTAIR could direct binding. Consistent with luciferase results, miR-22 was precipitated by wild-type HOTAIR but not HOTAIR mutant ( Figure 3I).

| HOTAIR acts as a molecular sponge to adsorb miR-22 and cause endothelial cell inflammation and pyroptosis
We constructed and obtained a HOTAIR overexpression lentivirus, which was used to infect HUVECs that were stimulated and cultured  Figure 4N).

| High UA promoted endothelial cell inflammation and pyroptosis via HOTAIR/miR-22
To

| HUA cause endothelial cell inflammation and pyroptosis by regulating the miR-22/NLRP3 axis through HOTAIR
To further investigate whether HOTAIR induces inflammation and pyroptosis through the NLRP3 inflammasome, we used shRNAexpressing lentiviruses to knock down NLRP3 expression in HOTAIRoverexpressing endothelial cells, followed by HUA stimulation and culture. The results showed that after NLRP3 knockdown, the overexpression of HOTAIR did not induce the upregulation of pyroptosisassociated proteins caspase-1, GSDMD-N and GSDMD-FL, and the expression levels of these proteins were significantly lower than those in the HUA-only treatment group ( Figure 6A). Additionally, there were no significant differences in the release of the inflammatory cytokines IL-1β and IL-18 and LDH into the culture supernatant, but lower than those in the HUA-only treatment group ( Figure 6B,C), and there were no abnormal changes in cell proliferation ( Figure 6D). Furthermore, we conducted an experiment to verify that whether upregulating NLRP3 expression could rescue the effect of silencing of HOTAIR on cell pyroptosis. We used shRNA-expressing lentiviruses to knock down HOTAIR expression in NLRP3-overexpressing endothelial cells, followed by UA stimulation and culture. The results showed that after HOTAIR knockdown, the overexpression of NLRP3 induced the upregulation of pyroptosis-associated proteins, including NLRP3, caspase-1, GSDMD-N and GSDMD-FL, and the expression levels of these proteins were significantly higher than those in the HOTAIR knockdown group ( Figure 6E).
Additionally. The secretion of the inflammatory cytokines IL-1β and IL-18 and LDH significantly increased ( Figure 6F,G) in NLRP3-overexpressing group, and cell proliferation also significantly inhibited in NLRP3overexpressing group ( Figure 6H).
Next, we used AMO-22 (miR-22 inhibitor) to knock down miR-22 in cells, which were then stimulated with HUA. The qPCR results showed that HOTAIR expression did not change significantly ( Figure 7A). however, the expression levels of those proteins were significantly lower in the mimic group than in the HUA-only group ( Figure 7E).
The release of the inflammatory cytokines IL-1β and IL-18 and LDH into the culture supernatant was also significantly lower in the mimic group than in the high UA-only group ( Figure 7F,G).
Interestingly, the above phenomena were not observed when the miR-22 mimic was added to the NLRP3 overexpression group. Expression levels of the pyroptosis-related proteins NLRP3, caspase-1, GSDMD-N and GSDMD-FL were detected by Western blot, and the levels of these proteins were significantly higher in the NLRP3 overexpression group than in the no-load vector group ( Figure 7H). The release of inflammatory cytokines IL-1 and IL-18 and LDH was also significantly higher in the NLRP3 overexpression group than in the no-load vector group ( Figure 7I,J).

| Inhibition of HOTAIR attenuates inflammatory kidney injury in hyperuricaemia mice
To explore novel strategies for the targeted treatment of UA nephropathy, based on the above studies, we injected shHOTAIR lentiviruses, via tail vein, into hyperuricaemia mice and subsequently assessed the inflammation status in kidney tissues The results showed that the protein expression levels of NLRP3 and caspase-1 in the renal tissues of the treatment group were significantly lower than those in the unrelated lentivirus (scramble) group ( Figure 8A).
The immunofluorescence intensity of NLRP3 and caspase-1 in kidney tissue from the treatment group was significantly lower than that in kidney tissue from the model group ( Figure 8B,C). Furthermore, in the treatment group, serum inflammatory cytokines IL-1β and IL-18 decreased, with a statistically significant difference compared with the control group ( Figure 8D). Further immunohistochemical analysis of kidney sections revealed that compared with the model-only group, the use of shHOTAIR lentivirus to down-regulate HOTAIR significantly reduced the infiltration of CD68+ macrophages around the blood vessels of the glomerular mesangial area and renal interstitium ( Figure 8E).

| DISCUSS ION
Studies have confirmed that HOTAIR plays a role in diabetic retinopathy and leads to diabetic retinal endothelial cell dysfunction. 33 Related studies have also shown that HOTAIR is highly expressed in patients with diabetic nephropathy and in the kidney tissues of animals. 34 However, the role and mechanism of HOTAIR in HUAinduced endothelial injury are not fully understood. In this study, we found that the expression level of HOTAIR in hyperuricaemia patients was significantly higher than that in the normal population.
The same results were also observed in a hyperuricaemia mouse model and in HUVECs stimulated with HUA. lncRNAs can function as endogenous competitive RNAs of miRNAs to regulate miRNA expression and biological gene functions by sequence complementarity. 35  showed that HOTAIR could also pull down miR-22 by using a biotinlabelled specific HOTAIR probe through a pulldown system, indicating a sequence-specific recognition between miR-22 and HOTAIR.
RIP experiments suggested that Ago2 immunoprecipitated with a higher amount of HOTAIR and miR-22 than did IgG. Therefore, the competitive binding between HOTAIR and miR-22 transcripts is based on forming a RISC with Ago2, and HOTAIR might promote HUA-induced inflammatory kidney injury by competitively binding miR-22.
Actually, lncRNAs can regulate gene expression at a variety of levels. 39 In addition to acting as molecular sponge to bind miRNAs and prevent miRNAs from binding to their target mRNAs, 36 49 We examined the inhibitory effect on endothelial cell inflammation and pyroptosis in HUVEC through down-regulation of HOTAIR.
Compared to UA-only group, we observed that the expression levels of pyroptosis-related proteins were decreased, inflammatory cytokines and LDH release was decreased, and cell proliferation was significantly increased. We observed the same result when miR-22 was overexpressed. What is more, suppression of miR-22 restrained the inhibitory effects of HOTAIR knockdown on the inflammatory response and pyroptosis. We also observed that after the inhibition of miR-22 by AMO-22 in a HUA environment, the release of pyroptosis-related proteins, inflammatory cytokines and LDH further increased compared to that in the HUA-alone group; notably, the expression of HOTAIR was not affected by AMO-22. The results further suggest that miR-22 is a downstream target of HOTAIR.
In addition, a miR-22 mimic significantly reduced the expression of NLRP3 and other pyroptosis-related proteins (caspase-1, GSDMD-N and GSDMD-FL) in a HUA environment; furthermore, the release of inflammatory cytokines and LDH was also significantly reduced.
However, the miR-22 mimic could not reverse the inflammatory response and pyroptosis caused by NLRP3 overexpression, indicating that NLRP3 is a downstream target of miR-22. Therefore, we infer that lncRNA-HOTAIR induces pyroptosis of endothelial cells in a HUA environment through the miR-22/NLRP3 axis and induces inflammatory injury.
Currently, the treatments for HUA are primarily drugs that inhibit UA production and promote UA excretion, while simultaneously improving quality of life. 50 bind Snail and, in turn, trigger H3K27me3/EZH2-mediated repression of Snail epithelial target genes. HOTAIR-sbid was proven to reduce cellular motility, invasiveness, anchorage-independent growth and not impair the TGF-mediated mesenchymal gene expression.
They provide evidence on a lncRNA-based strategy to treat diseases caused by abnormal expression of LncRNAs. We will apply this strategy in our research in the future.
In summary, this study elucidated the mechanism underlying the promotion of inflammatory renal injury by HOTAIR in a HUA environment. As a ceRNA of miR-22, HOTAIR promotes the expression of the NLRP3 inflammasome in endothelial cells and induces inflammation and pyroptosis. Therefore, HOTAIR is proposed as a suitable therapeutic target for UA nephropathy, providing a novel theoretical basis for developing new drugs and correcting endothelial dysfunction in a HUA environment.

ACK N OWLED G EM ENT
We would like to acknowledge the reviewers for their helpful comments on this paper.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.