C57BL/6 mice were from by the Model Animal Research Center of Nanjing University. LncOlfr29 knockout (KO) mice on a C57BL/6J background were generated using CRISPR-Cas9 system by the Model Animal Research Center of Nanjing University (Nanjing, Jiangsu, China). NLRP3 knockout (-/-) mice were from Prof. Meng, Pasteur Institute, Shanghai. B6.SJL-CD45a(Ly5a) (CD45.1) mice were from the Model Animal Research Center of Beijing. All mice were maintained in specific pathogen-free (SPF) conditions in the Animal Center of Nankai University. Murine experiments were performed according to Nankai University Guide for the Care and Use of Laboratory Animals. Human macrophage cell line THP-1 was from the American Type Culture Collection.

Salmonella Typhimurium (S. T) infection was performed according to the previously reported method (17). Briefly, mice were first treated by gavage with 7.5 mg of streptomycin, and then infected with S. T (1 × 10³CFUs/mouse for chronic infection, 1 × 10⁷ CFUs/mouse for acute infection) at 20 hours after streptomycin treatment. Weight changes in mice was calculated as: % weight change = (weight at day X-day 0/weight at day 0) × 100. The colon and lung tissues were embedded in OCT compound for immuno-staining, or in paraffin for hematoxylin/eosin (H&E) staining. Spleen, Lung and liver were homogenized for 2 mins in PBS. CFUs were quantified after plating lysates onto LB agar for 24 hrs.

DSS-mediated colitis was carried out according to previously reported method (18). Briefly, mice (15 mice per group) were received 2.5% (w/v) DSS in their drinking water for 7 days, and then switched to regular drinking water. Body weights were detected and calculated as: % weight change = (weight at day X-day 0/weight at day 0) × 100. Mice were also investigated for rectal bleeding, diarrhea, and signs of morbidity such as hunched posture and failure to groom. Disease activity index (DAI) and histological evaluation were assessed according to methods (19, 20). DAI was the average of combined scores of stool consistency, bleeding and weight loss: Weight loss: 0, none; 1, 1–5%; 2, 5–10%; 3, 10–15%; 4, >15%; Bloods: 0, normal; 2, slight bleeding; 4, gross bleeding; Diarrhea: 0, normal; 2, loose stools; 4, watery diarrhea.

Histology scores (E + I) were assessed as followings: Infiltration (I), 0 = no infiltrate; 1 = infiltrate around the crypt basis; 2 = infiltrate reaching the lamina muscularis mucosae; 3 = extensive infiltration reaching the lamina muscularis mucosae and thickening of the mucosa with abundant edema; 4 = infiltration of the lamina submucosa; Epithelium (E): 0 = normal morphology; 1 = goblet cell loss; 2 = goblet cell loss in large areas; 3 =  crypt loss; 4 =  crypt loss in large areas.

For macrophage transplanted experiments. Bone marrow cells (BMCs) were collected from WT or LncOlfr29-/- mice, and then the monocytes/macrophages were isolated from BMCs through sorting after staining using CD11b antibody. Isolated monocytes/macrophages were then injected into different recipient mice, which were irradiated (8 Gy, a single dose) using a Shepherd Mark I Cesium Irradiator (J.L. Shepherd and Associates). After 3 weeks, DSS model was performed in the recipient mice.

Murine macrophages were generated from the peritoneal cavity of mice, which were intraperitoneally injected with 4 mL of 3% thioglycollate medium. The cells from the peritoneal cavity were seeded in RPMI containing 10% FBS. After removing non-adherent cells, adherent cells were collected.

For bone marrow derived macrophages (BMDMs), bone marrow cells were first obtained from bone marrow of the tibia and femur, and cultured for 6 days in RPMI with 10% FBS, 20 ng/ml mouse M-CSF and 1% penicillin/streptomycin.

For human peripheral blood cells derived macrophages (PBDMs), after collecting human peripheral blood, the cells were isolated using Percoll. Then isolated cells were cultured in RPMI containing 10% FBS. After removing non-adherent cells, adherent cells were cultured in the medium with GM-CSF and IL-4 for 5 days.

Using BLOCK-iT™ RNAi Designer (Invitrogen), short hairpin RNA (shRNA) target sequences were selected. Through pGreenPuro™ cloning and expression lentivector kit, shRNA or lncOlfr29 constructs were generated according to the manual. The control luciferase control RNA is from the kit. For packaging lentivirus particles, shRNA, lncOlfr29 or control lentivector together with packaging plasmids pMD2.G and psPAX2 were co-transfected into 293T cells. Macrophages were infected by centrifugation with the lentivirus in the presence of 8 μg/ml polybrene.

For inflammasome activation, macrophages were primed with LPS (2 μg/mL) for 4 h, and then stimulated with Dotap to activate caspase 11(mice)/caspase 4(human) or nigericin (5 μM) to activate NLRP3 inflammasome, DOTAP-transfected flagellin (5μg/mL) to activate NLRC4 inflammasome.

Macrophages transfected with ZFP36 siRNA (siZFP36) and exogenous lncOlfr29 (oeOlfr29) were treated with actinomycin D (0.5 mg/mL) for indicated time periods to initiate the time-course experiment. NLRP3 mRNA levels were detected by qRT-PCR.

RNA fluorescence in situ hybridization (RNA-FISH) was carried out according to reported protocol (21). Briefly, cells were first slicked on 0.01% poly-lysine-treated slides. For cell membrane perforation, cells were processed with CSK+0.4% Triton X-100 buffer for 30 s. Pre-warmed 5% goat serum was used for blocking for 30 mins at 37°C. The slides incubated with primary antibody at 37°C for 1 hour. The slides were incubated with secondary antibody at 37°C for 30 mins after washing. and then dehydrated through ethanol series (85, 95, and 100% ethanol), and hybridized using the indicated probes overnight at 37°C in a humid chamber. After washing, DAPI dye was added.

RNA immunoprecipitation (RIP) was carried out according to reported protocol (22). Briefly, cells were first added into ice cold IP lysis buffer containing 0.5% ribonuclease inhibitor, and incubated on ice for 5 mins with mixing. The lysates were centrifuged at 13,000 g for 10 mins to pellet the cell debris at 4°C. For preclearing, protein G agarose was added and incubated for 1 hour at 4°C with rotation. After that, immune-precipitating antibodies were added and incubated overnight at 4°C with rotation. Then protein G agarose was pelleted and washed with IP lysis buffer, containing 0.5% ribonuclease inhibitor. Finally, RNA was extracted from the complexes of protein/RNA and quantified by qPCR.

The cells were first incubated with hypotonic buffer, containing 25 mM Tris-HCl, pH 7.4, 1 mM MgCl2, 5 mM KCl on ice for 5 min, and then an equal volume of hypotonic buffer containing 1% NP-40 was added. The supernatants were collected as the cytosolic fraction after centrifugation at 5000 × g for 5 mins. The pellet was re-suspended in buffer, containing 20 mM HEPES, pH 7.9, 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT and1 mM PMSF, and then incubated at 4°C for 30 mins. Nuclear fraction was collected by centrifugation at 12,000 g for 10 mins.

Total RNA extraction, qRT-PCR, Western blotting, cell isolation, flow cytometry and immunoprecipitation (IP) was performed according to our previously method (22, 23).

ONE-way ANOVA Bonferroni’s Multiple Comparison text and two side student’s t-test were used. The survival curves were compared using the generalized Wilcoxon’s test. All of these were performed by GraphPad Prism 5 software. A 95% confidence interval was considered as significance and defined as p <0.05.

To discover a new factor that can regulate the function of macrophages, we reanalyzed the previous RNA sequencing data (17, 24). Data showed that lncRNA Olfr29-ps1 (lncOlfr29) was significantly highly expressed in myeloid derived suppressor cells (MDSCs) ( Figure 1A ). Compared with other immune cells, the expression of lncOlfr29 in mouse macrophages was significantly higher ( Figure 1B ). This lncOlfr29 could be detected in both the cytoplasm and the nucleus, but it was more common in the cytoplasm of macrophages ( Figure 1C ). Bioinformatics analyses revealed that lncOlfr29 was located in the chromosome 4 in mice and chromosome 16 in human ( Supplementary Figure 1 ). There had widely epigenetic modification on the promoter region of lncOlfr29 in the macrophages not only in human but also in mice, which could promote the expression of genes ( Supplementary Figure 1 ), supporting the high expression of this lncRNA in macrophages. We also sought out the factor(s) that might affect the expression of lncOlfr29. Since gut microbiota plays a critical role in human healthy and diseases (25), we observed the effects of several metabolites from the gut microbiota on the expression of lncOlfr29 in the macrophages. Data showed that LPS could significantly promote the expression of lncOlfr29 ( Figure 1D ). Furthermore, LPS mediated lncOlfr29 was time- and dose-dependent ( Figures 1E–H ). The expression level of lncOlfr29 was the highest 5 hours after LPS stimulation, and the expression level of lncOlfr29 increased with the increase of LPS concentration ( Figures 1E–H ). Thus, we identify a new lncRNA lncOlfr29 in macrophages, which can be promoted by LPS.

Next, we studied the role of lncOlfr29 in macrophages. To this end, we first generated lncOlfr29 silencing (lncOlfr29 shRNA) and lncOlfr29 overexpressing (exogenous lncOlfr29) lentiviruses, both of which demonstrated significant effects on lncOlfr29 expression levels ( Supplementary Figure 2 ) Then macrophages were transfected with these lentivirus, and stimulated with multiple ligands, which could activate the inflammasomes, including LPS, LPS plus Dotap (activator of caspase-11) (26), LPS with flagellin (activator of NLRC4) (27), LPS with nigericin (activator of NLRP3) (28, 29). Data showed that NLRP3 ligand LPS plus nigericin significantly reduced the production of mature IL-1β (mIL-1β) but not IL-1β mRNA in lncOlfr29 silenced macrophages, while LPS plus Dotap or LPS plus flagellin had no significant effects ( Figures 2A, C ). In lncOlfr29 overexpressed macrophages, mIL-1β was significantly increased in response to LPS plus nigericin, but not to LPS plus Dotap or LPS plus flagellin ( Figures 2B, D ). Since IL-1β maturation often company with pyroptosis (8, 9), which is a lytic cell death induced by pathogen infection or endogenous challenge (30), we also observed the effects of silencing or overexpressing lncOlfr29 on the pyroptosis. Silencing lncOlfr29 could significantly alleviate LPS plus nigericin mediated pyroptosis on macrophages, but LPS plus Dotap or LPS plus flagellin mediated pyroptosis was not affected ( Figure 2E ). The pyroptosis of lncOlfr29 overexpressed macrophages was also more severe upon exposure to LPS plus nigericin than to LPS plus Dotap or LPS plus flagellin ( Figure 2E ). Immunofluorescence analysis of cleaved caspase-1 (CC-1) for detecting pyroptosis (31, 32) also showed that silencing lncOlfr29 could alleviate pyroptosis, while overexpression of lncOlfr29 aggravated pyroptosis ( Figure 2F ). Thus, lncOlfr29 can promote the pyroptosis of macrophages and the maturation of IL-1β upon exposure to NLRP3 ligand.

To further determine the functions of lncOlfr29, we generated lncOlfr29 KO mice (lncOlfr29 -/-). Since lncOlfr29 transfected or silenced macrophages exhibited difference in mature IL-1β production and pyroptosis only after exposed to NLRP3 ligand LPS with nigericin, this suggests that lncOlfr29 mediated mature IL-1β and pyroptosis might be through NLRP3. Thus, we also observed whether there existed similar roles between lncOlfr29 and NLRP3. Interestingly, while lncOlfr29 KO and NLRP3 KO macrophages were exposed to LPS, LPS plus DOTAP, LPS plus nigericin and LPS plus flagellin, they exhibited similar results. Both lncOlfr29 -/- mice and NLRP3 -/- macrophages exhibited reduced mature IL-1β but not IL-1β mRNA upon exposure to NLRP3 ligands LPS plus nigericin as compared to the controls ( Figures 3A–C ). LncOlfr29 -/- and NLRP3 -/- macrophages also exhibited resistance to LPS plus nigericin mediated pyroptosis ( Figure 3D ). Consistent with LPS plus nigericin mediated pyroptosis, the number of CC1+ macrophages was significantly reduced in both lncOlfr29 -/- and NLRP3 -/- macrophages as compared to controls ( Figure 3E ). Whereas, there was no difference between lncOlfr29 -/- and NLRP3 -/- macrophages ( Figures 3A–E ). Finally, the release of lactate dehydrogenase (LDH) also decreased significantly in both lncOlfr29 -/- and NLRP3 -/- macrophages as compared to control macrophages ( Figure 3F ). Again, there are no difference between lncOlfr29 -/- and NLRP3 -/- macrophages ( Figure 3F ). Thus, similar to NLRP3, lncOlfr29 can promote pyroptosis of macrophages and production of mature IL-1β.

Previous studies showed that NLRP3 played an important role to Salmonella infection (33–35). Since lncOlfr29 -/- and NLRP3 -/- macrophages had similar functions, we next used a mouse model of S. T infection to test the effect of lncOlfr29. Similar to NLRP3 -/- mice, lncOlfr29 -/- mice had also reduced resistant to Salmonella infection. Both Salmonella-infected lncOlfr29 -/- and NLRP3 -/- mice had less weight loss and lower mortality rates as compared to wild type (WT) mice ( Figures 4A, B ). However, there were not different in body weight loss and mortality between lncOlfr29 -/- and NLRP3 -/- mice. Since Salmonella could invade organs other than the intestinal tract, such as the liver, spleen and lung tissues, we also detected the number of Salmonella in these organ tissues. The number of S. T bacteria in lncOlfr29 -/- and NLRP3 -/- mice was significantly lower than those in WT mice ( Figure 4C ). Again, S. T numbers were no differences in these organs between lncOlfr29-/- and NLRP3 -/- mice ( Figure 4C ). The colon lengths of lncOlfr29 -/- and NLRP3 -/- mice were also longer than those of WT mice ( Figure 4D ), indicating that lncOlfr29 and NLRP3 KO can alleviate intestinal inflammation. The inflammatory cytokine IL-1β in lncOlfr29 -/- was also significantly lower than that in WT mice ( Figure 4E ). We also analyzed the inflammatory cells in the lamina propria of the colon of mice. The results indicated that the proportion of neutrophils in lncOlfr29 -/- and NLRP3 -/- mice was less than that in WT mice ( Figure 4F ). H/E staining also showed that the inflammation in lung tissue was significantly reduced in lncOlfr29 -/- mice as compared to control, consistent with that in NLRP3 -/- mice ( Figure 4G ). The inflammation in the colon of lncOlfr29 -/- and NLRP3 -/- mice was also decreased as compared to WT mice ( Figure 4H ). Similar results were also found in S. T mediated acute infection model ( Supplementary Figure 3 ). Therefore, similar to NLRP3, lncOlfr29 promotes resistance to Salmonella infection.

Macrophages are essential for the maintenance of intestinal homeostasis, and yet appear to be drivers of inflammation in the context of IBD (36, 37). Next, we further investigated the effects of lncOlfr29 on colitis using a mouse model of DSS-mediated colitis. DSS was administered for 7 days and normal water was given for 3 days to induce obvious symptoms of acute colitis in the mice. The results demonstrated that there had less weight loss and lower mortality rates in both lncOlfr29 -/- and NLRP3 -/- mice as compared to the WT mice ( Figures 5A, B ). Disease activated index (DAI) in lncOlfr29 -/- and NLRP3 -/- mice was also lower than WT mice ( Figure 5C ). The colon lengths of lncOlfr29 -/- and NLRP3 -/- mice were longer than those of WT mice, and the cecum was larger, indicating that the condition of intestinal inflammation has lessened ( Figure 5D ). The proportion of neutrophils in lncOlfr29 -/- and NLRP3 -/- mice also was lower than that those in WT mice ( Figure 5E ). The inflammatory courses of the colon in lncOlfr29 -/- and NLRP3 -/- mice were mush lighter than those in WT mice ( Figure 5F ). Notably, there were not different in DSS-mediated colitis between lncOlfr29 -/- and NLRP3 -/- mice. Thus, similar to NLRP3, lncOlfr29 promotes sensitivity to DSS-induced colitis.

LncOlfr29 is expressed not only in macrophages but also in other immune cells and intestinal epithelial cells. In order to eliminate the influence of lncOlfr29 in other immune cells and intestinal epithelial cells on colitis, the macrophage transplantation experiment was conducted. Bone marrow derived monocytes/macrophages were isolated from lncOlfr29 -/- and WT mice, and then injected via the tail vein into radiation-exposed mice. After receiving monocytes/macrophages from WT mice, there was no difference in DSS-induced colitis in lncOlfr29 -/- and WT mice ( Figures 5G–L ); There was also no difference in DSS-induced colitis in lncOlfr29 -/- and WT mice after receiving the macrophages from lncOlfr29 -/- mice ( Figures 5G–L ). Whereas there were significant differences in WT mice that received mononuclear/macrophages derived from WT or lncOlfr29 -/- mice, and lncOlfr29 -/- mice that received macrophages from WT or lncOlfr29 -/- mice ( Figures 5G–L ). Therefore, these results indicate that lncOlfr29 - mediated colitis is macrophage-dependent.

Since the function of lncOlfr29 was similar to that of NLRP3, we believed that the function of lncOlfr29 was achieved through NLRP3. We next looked for the mechanism(s) by which lncOlfr29 could promote activation of NLRP3. LncRNA can directly bind to proteins (13). Thus, we first analyzed the potential binding proteins of lncOlfr29. However, data did not show the binding of lncOlfr29 to NLRP3 ( Figure 6F ). Indeed, while the positive control lncRNA Neat1 (38) could bind to NLRP3, RIP did not show binding of NLRP3 with lncOlfr29 ( Figure 6A ). RNA immunofluorescence in situ hybridization (FISH) also did not exhibit the binding of lncOlfr29 with NLRP3 ( Figure 6B ). All of these suggest that lncOlfr29 has no direct binding to NLRP3. LncOlfr29 can also encodes polypeptides that have an impact on the cellular function. So we also analyzed whether the lncOlfr29 could encode polypeptide(s). The prediction results showed that lncOlfr29 did not encode any polypeptides ( Supplementary Figure 4 ).

Next, we studied the effect of lncOlfr29 on the expression of NLRP3. Interestingly, the expression of NLRP3 in lncOlfr29 KO macrophages was lower than that in WT macrophages regardless of whether they were exposed to LPS ( Figure 6C ). QPCR and immunofluorescence assays also showed a reduction of NLRP3 in lncOlfr29 KO macrophages compared to WT macrophages ( Figures 6D, E ). The expression of NLRP3 can be regulated by various factors such as ZFP36, which can bind to the 3 ‘-UTR of mRNA to degrade NLRP3 mRNA (16, 39–41). Thus, we investigated whether lncOlfr29 could bind to ZFP36 to eliminate the degradation of NLRP3 mRNA by ZFP36. To do this, we again analyzed the potential binding protein of lncOlfr29. The prediction results show that there were multiple binding sites between lncOlfr29 and ZFP36 ( Figures 6F, G ). RIP experiment also showed a binding between lncOlfr29 and ZFP36 ( Figure 6H ). RNA-FISH analysis confirmed that lncOlfr29 could bind to ZFP36 ( Figure 6I ). These results suggest that lncOlfr29 can bind to ZFP36 and remove the degradation of NLRP3mRNA 3’UTR by ZFP36. Indeed, silencing ZFP36 not only upregulated protein but also mRNA levels of NLRP3 ( Figures 6J–M ). Critically, actinomycin D chase assays exhibited increased NLRP3 mRNA via time-course qPCR under conditions of ZFP36 silencin or lncOlfr29 overexpression ( Figure 6N ). Taken together, lncOlfr29 can combine with ZFP36 to promote NLRP3 expression.

Finally, we investigated whether the biological functions of human lncOlfr29 (hulncOlfr29) in macrophages were similar to those in mice.

The expression of human lncOlfr29 in macrophages and the macrophage cell line THP1 could also be detected ( Figures 7A, B ). Furthermore, LPS also promoted expression of human lncOlfr29 in macrophages ( Figure 7C ). RNA-FISH also showed increased expression of human lncOlfr29 in human macrophages after exposure to LPS ( Figure 7D ). Importantly, silencing human lncOlfr29 could not only reduce production of mIL-1β but also pyroptosis of human macrophages upon exposure to NLRP3 ligand LPS plus nigericin; whereas overexpressed hulncOlfr29 increased the production of mIL-1β and pyroptosis of macrophages ( Figures 7E–H ). In terms of mechanism, ZFP36 also combined with hulncOlfr29 ( Figures 7I, J ). Silencing ZFP36 also promoted the expression of NLRP3 in human macrophages ( Figures 7K–M ). Therefore, human lncOlfr29 has similar functions to mouse lncOlfr29 in macrophages.