Patients with AR visited doctors in the outpatient service in the Guo Yi Tang of Beijing University of Chinese Medicine. In this study, there were 24 subjects: 12 males and 12 females. They were divided into two groups: the AR group (patients with AR, 12 subjects: five males and seven females) and NP group (normal persons, 12 subjects: four males and eight females). With support/approval from the Ethics Committee of Beijing University of Chinese Medicine, this study was conducted while adhering to the principles of the Declaration of Helsinki. Patients in the AR group were positive for skin puncture test, including pollen, food, dust mites, paint, or molds as well as in specific IgE. Two weeks before study recruitment, these patients with AR received no topical or systemic corticosteroid therapy. We chose the study participants with no history of smoking or other immune system disorders, such as rheumatoid arthritis, systemic lupus erythematosus, and scleroderma.

The whole blood samples obtained from the two groups were stored in vacuum tubes with heparin, and PBMCs were isolated from these samples by lymphocyte separation solution (Tianjin Haoyang Biological Manufacture Co., Ltd.). Mononuclear cells were seeded in 12-well plates with the RPMI 1640 medium that contains 10% heat-inactivated fetal calf serum (FCS, GIBCO, Germany) and 2 mM of L-glutamine (R10 medium, Sigma, St. Louis, MO, United States). After incubation at 37°C for 2 h, the non-adherent cells were removed and the adherent cells were cultured within the medium containing 100 ng/ml of rhGM-CSF and 100 ng/ml of rhIL-4 (R&D Systems, Minneapolis, MN, United States). On the sixth day, 50 ng/ml of TNF-α (R&D Systems, Minneapolis, MN, United States) was added into the samples; the method had the same protocol in the study of Andreia et al. (2005). Immature DCs (imDCs) were collected on the fifth day, and the mature DCs (mDCs) were collected on the seventh day.

After being treated with 25 μg/mL of mitomycin at 37°C for 30 min, the DCs were placed at the concentrations of 2 × 10⁸ cells per well at a quantity of 200 μl and incubated with non-adherent PBMCs obtained from the same healthy people. These samples were later stimulated by lymphocytes in the concentration proportions of 1:10, 1:50, and 1:100. Thereafter, the samples were mixed and incubated with non-adherent PBMC from the same healthy persons at the same concentrations in triplicate. The cells were treated with 10% fetal bovine serum. Then, 10 μl of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) solution (5 mg/ml, medium dilution, Sigma–Aldrich Chemical Co., St. Louis, MO, United States) was added to each well, and the cells were incubated for 72 h in the incubator. Then, 150 μl of DMSO was added followed by the addition of enzymes after 4 h. The absorbance was detected using a spectrophotometer at 570 nm.

The CD14 (PerCP-Cy 5.5, BD Biosciences, United States), HLA-DR (APC, BD Biosciences, United States), and isotype mouse IgG2a-PE (PE, BioLegend, United States) were added in the samples of imDCs. Moreover, CD86-APC, CD80-PE, and isotype mouse IgG1–FITC (BioLegend, United States) were added in the samples of mDCs. The cells were then suspended with precooled PBS, counted under a microscope, and centrifuged at 1000 g for 5 min. Data were acquired using a FACSCalibur cytometer (BD Biosciences, United States) and the ratios of CD14⁺, HLA-DR⁺, CD80⁺, and CD86⁺ DCs were determined.

The RNA extraction, labeling, chip hybridization, and scanning were all finished following the use of Agilent Human lncRNA–mRNA profiling chip (4^∗180K, Design ID: 062918). All the chip results were detailed according to the processes of software operations. The screening criteria were to increase or decrease the fold change value ≥ 2.0 and p-value < 0.05.

The GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were used to study the differentially expressed (DE) mRNAs. The results of the target gene analysis of lncRNA and the mRNA expression results on the chip need to be correlated so that the upregulated and downregulated DE genes can be investigated. The corrected p-value < 0.05 by calculating the FDR and FDR < 0.05 was selected as the threshold.

The correlation of the lncRNA–mRNA expression in the imDCs and mDCs was calculated. Then, the relationship pairs of LncRNA and mRNA based on the abovementioned criteria (p-value < 0.05, FDR < 0.05) were screened. The co-expression network of lncRNAs and mRNAs was constructed, and the co-expression network of lncRNAs and mRNAs was then established.

Quantitative RT-qPCR was used to investigate the different expressions of imDCs’ and mDCs’ genes between the patients with AR and NP. The total RNA was extracted from imDCs and DCs according to the kit for cells. In all, 20 μg of total RNA was converted into cDNA by using oligo (dT) and reverse transcriptase (Thermo, United States) to analyze the qPCR results. The thermal cycler conditions were set as follows: amplificated at 95°C for 10 min, 95°C for 15 s, 60°C for 60 s, 40 cycles of denaturation (15 s, 94°C), 15 s at 95°C, and a combined process of annealing and extension (1 min, 60°C). Supplementary Table S1 shows the primers for these genes.

The SPSS 22.0 software was used for statistical analysis in this study. Results were expressed as mean ± standard deviation. The relationship of lncRNAs and mRNAs was determined by the Spearman’s correlation coefficient. p-values < 0.05 were considered significant values for this study.

In allergenic mixed lymphocyte reaction (MLR), the levels of T-cell proliferation were increased with the proportion of T cells. DCs in the patients with AR have a stronger stimulation ability than NP as shown in Figures 1A–C. When the ratio of DC cells to T cells was 1:50 and 1:100, the difference was significant (p < 0.001).

We evaluated the percentages of CD14, HLA-DR, CD80, and CD86 in the DCs. The immunophenotypic characteristics of DCs in patients with AR were compared with that of NP, as shown in Figure 2. The CD14 concentration of imDCs was lower in patients with AR than that in the NP group. The mean percentage of CD14⁺ DCs in the patients with AR was 0.84 ± 0.25% (median: 0.78%), and it was significantly lower than the value in the NP group (p = 0.0008), where the mean percentage of these cells was 0.012 ± 0.013% (median: 0.017%) (Figures 2A,B).

The mean proportion of HLA-DR⁺ mDCs in the patients with AR was 44.25 ± 8.64% (median: 44.8%) and was higher (p = 0.0006) than the NP group, where the mean percentage of these cells was 95.3 ± 3.84% (median: 99.0%) (Figures 2C,D).

The mean proportion of CD80⁺ mDCs in the patients with AR was 72.15 ± 7.64% (median: 67.78%) and was higher (p = 0.0003) than the NP group, where the mean percentage of these cells was 95.3 ± 3.84% (median: 96.92%) (Figures 2E,F).

The mean proportion of CD86⁺ mDCs in the patients with AR was 62.15 ± 7.64% (median: 64.69%) and was higher (p = 0.006) than the control group, where the mean percentage of these cells was 92.3 ± 5.84% (median: 95.12%) (Figures 2G,H).

In total, 308 DE mRNAs, including 175 upregulated mRNAs and 133 downregulated mRNAs, were found in the imDCs of patients with AR and NP. A clustergram (Figure 3A) and volcano plots (Figure 3C) are used to depict DE mRNAs. Table 1 shows 67 mRNAs with the largest fold changes. The list contains several genes, including HLA-C, MARCO, KIR2DS3, ITGAV, CD36, and IFNB1. Additionally, 168 DE mRNAs, including 77 upregulated mRNAs and 91 downregulated mRNAs, were found in the mDCs of patients with AR. Figures 4A,C show the clustergram and volcano plots of the DE mRNAs, respectively. Table 2 shows the 10 mRNAs with the greatest variation, and several genes, such as HLA-B, F11R, HLA-DQB1b, HLA-DQB1, and PTAFR, are also shown.

In total, 962 DE lncRNAs, including 434 upregulated and 528 downregulated lncRNAs, were found in patients with AR and NP. The clustergram (Figure 3B) and volcano plots (Figure 3D) show DE lncRNAs. In total, 601 lncRNAs, including 200 upregulated and 401 downregulated lncRNAs, were found in the DE mDCs of the patients with AR. The clustergram in Figure 4B and the volcano plots in Figure 4D show the DE lncRNAs. Tables 1, 3 show the top 10 lncRNAs of imDCs and mDCs with the largest fold changes. The pathways, including interferon-gamma-mediated signaling pathway, membrane repolarization, and peptide antigen binding, that contribute to the phagocytosis function in imDCs and antigen-presenting function of mDCs were also identified.

Figure 5A shows the interactions of proteins that were coded by DE mRNAs in imDCs. Additionally, Figure 5B shows the co-expression network between DE lncRNAs and mRNAs. TRIM69 has the maximum target mRNAs including 58 DE mRNAs, and SIDT2 has the maximum co-expressed lncRNAs. Table 2 shows the top 10 co-expression pairs in imDCs. lncRNAs exert their biological function as ceRNAs (Wilfried et al., 2016).

We identified 268 target genes after analyzing the possible DE lncRNAs target genes in imDCs. Figure 5C shows the target genes with a combined score of more than 0.9. HLA-C was the target gene of AC108142.1-005, and CD36 was the target gene of FR264384. Moreover, IFNB1 was the target gene of MIR3150B-210. The Venn diagram analysis showed that 95 mRNAs were coincided between the 166 DE mRNAs and the 95 DE lncRNA target genes (Figure 5D). The 95 DE lncRNAs were all included in the 166 DE mRNAs.

Figure 6A shows the interaction proteins that were coded by DE mRNAs in mDCs. In this network, HLA-B, HLA-DQB1, HLA-DQB2, PTAFR, and F11R are important genes that interact with many other DE mRNAs. Furthermore, Figure 6B shows the co-expression network of DE lncRNAs and mRNAs. TRIM77P has the maximum target numbers including 39 DE mRNAs, in which FAM153A and ZNF396 have the maximum co-expressed lncRNAs. Table 4 shows the top 10 co-expression pairs of mDCs.

Additionally, we also investigated the possible presence of DE lncRNA target genes in mDCs. In our study, 99 target genes were identified in these DE lncRNAs. Figure 6C shows the target genes with the combined score of more than 0.9. In the figure, HLA-B has three target genes of lincRNAs, namely, DHRS3, FCHO2, and linc-PRR5-1. Figure 6D shows that 42 mRNAs coincided between the 95 DE mRNAs and the 42 DE lncRNA target genes in the Venn diagram analysis. Moreover, the 42 DE lncRNAs were included in the 95 DE mRNAs. It also includes some known inflammatory-related molecules.

RT-qPCR was used to evaluate DE mRNAs and lncRNAs to verify our RNA chip results. We randomly detected three lncRNAs and 10 mRNAs. MARCO, KIR2DS3, F11R, HLA-B, HLA-C, NON-HSAG046717, and NON-HSAT089067 were upregulated, whereas ITGAV, CD36, IFNB1, PTAFR, HLA-DQB1, NON-HSAT 059748, NON-HSAT024276, and NON-HSAT098958 were downregulated. The results of RT-qPCR were consistent with those of RNA chip results, hence confirming that our chip data were reliable (Figure 7).