SGECs (cat. #T9289, Applied Biological Materials Inc.) were thawed and cultured in complete media (DMEM:F12 with 10% FBS and PSN) and maintained at 37°C in a humidified atmosphere with 5% CO₂. The cells were passaged at approximately 80% confluency using 0.05% trypsin-EDTA. Jurkat cells (cat. #TIB-152) were maintained in RPMI-1640 with 10% FBS.

Transwell culture systems with cell inserts (pore size: 0.8 µm) were used to co-culture SGECs and Jurkat cells. Jurkat cells (1*10⁶ cells) were seeded on the upper chamber of the cell insert transwell, and SGECs (1*10⁶ cells) were seeded in the lower chamber of the transwell. This configuration physically separates the SGECs and Jurkat cells while allowing for the exchange of soluble factors, thereby simulating a paracrine interaction without direct cell-to-cell contact. The SGECs, which were seeded in the bottom chamber, were physically separated from the Jurkat cells, which were cultured in the upper chamber, and both were collected separately. Total RNA was then isolated independently from each distinct cell population (i.e., from the SGECs and the Jurkat cells) using standard RNA isolation protocols. RT-PCR was performed on purified cell-type-specific RNA samples and not on a mixed-cell RNA extract. This approach inherently ensured that the quantified mRNA signals were unequivocally derived from either SGECs or Jurkat cells, allowing us to accurately distinguish cell-type-specific transcriptional responses to the co-culture conditions. The co-cultures were kept as the control group and LPS (10 ng/mL)-treated group for SGECs with each risk alleles of rs6920220 (G/A) and rs2230926 (T/C/G) in replicates. LPS was used as an inflammatory inducer, which activates various immune pathways primarily through Toll-like receptor 4 (TLR4), expressed on many immune cells (Liu et al., 2023).

To incorporate the desired SNPs into the genome of SGECs, homology-directed repair (HDR) using the CRISPR-Cas9 genome editing method has been used in this work. First, the donor Ultramer quality Alt-R HDR oligos were designed using the IDT online Alt-R HDR design tool.

The protocol involves co-transfection of gRNA (guide RNA) and HDR donor oligos (CRISPR-Cas9 reagents are stored at −20°C and always kept on ice while using). (HDR oligos and gRNA sequences are shown in Table1.)

The oligos were resuspended in nuclease-free IDTE, and the guide RNA complex was prepared by combining crRNA and tracrRNA to a final concentration of 50 μM. The mixture was heated at 95°C for 5 min and cooled to room temperature (25°C). Next, the gRNA and Cas9 nuclease were combined to form a ribonucleoprotein (RNP) complex and incubated at room temperature for 15 min. Lipofectamine CRISPRMAX (cat. #CMAX00001) reagent was used to transfect cells in a 96-well plate. The RNP complex was mixed with HDR donor oligo and Cas9 Plus reagent (tube 1). CRISPRMAX reagent was diluted in media (tube 2). The solution in tube 1 was mixed into tube 2 and incubated for 15 min. The final solution was added to the cells and kept in an incubator for 72 h.

Genomic DNAs were isolated from SGEC lines using a Qiagen DNeasy kit (cat. #69506). To detect and quantify the desired SNP modifications by CRISPR-Cas9, primer- and probe-based real-time quantitative PCR was performed. PrimeTime Gene Expression Master Mix (cat. #1055770) from IDT was used. It is a 2X master mix solution containing a hot-start antibody, Taq polymerase, and other components needed for probe-based qPCR in two-step RT-PCR experiments with the reference dye included separately. In this 5′ nuclease-based assay, the primers and probe hybridize in a sequence-dependent manner to the complementary DNA (cDNA) strand. Because the probe is intact, the fluorophore and quencher are in proximity, and the fluorescence emitted from the fluorophore is absorbed by the quencher. The exonuclease activity of the polymerase cleaves the hybridized probe, and the fluorophore is separated from the quencher and fluoresces. These steps are repeated for each PCR cycle, which allow the detection of specific products. For the rs2230926 (T/C/G) SNP: for the T probe, Affinity Plus^® Mini Probe 5′6-FAM™/3′IB®FQ detected by the FAM/SYBR channel; for the C probe, Affinity Plus^® Mini Probe 5′SUN/3′IB®FQ detected by the VIC channel; and for the G probe, Affinity Plus^® Mini Probe 5′YAK^®/3′IB®FQ detected by the ROX channel. For rs6920220 (G/A) SNP: for the WT probe, Affinity Plus^® Mini Probe 5′6-FAM™/3′IB®FQ detected by the FAM/SYBR channel and for the mutant (Mut) probe, Affinity Plus^® Mini Probe 5′YAK^®/3′IB®FQ detected by the ROX channel. The probe sequences are provided in Table 1.

Total RNA was extracted using a Quick-RNA Miniprep Kit (Zymo Research kit cat. #R1055) as per the manufacturer’s instructions. The extracted total RNA was quantified using the NanoDrop quantification system (Thermo Fisher Scientific). A measure of 500 ng of total RNA was converted to cDNA using an ABclonal ABScript Neo RT Master Mix synthesis kit (cat. #RK20433) according to the manufacturer’s instructions. Each real-time PCR used 1/20th of the cDNA, which was performed on the CFX96 Real-Time PCR system (Bio-Rad) using 2X Universal SYBR Green Fast qPCR mix (cat. #RK21203). The gene expression level was normalized with the housekeeping gene GAPDH as an endogenous reference and plotted as normalized expression using CFX96 Maestro software (Bio-Rad). Primers used for the real-time PCR analysis were as tabulated. The RT-PCR experiments were performed in replicates.

The data were analyzed using the t-test, and figures were constructed using GraphPad Prism. For statistical significance, p-values were determined and represented in graphs as * 0.01 < p < 0.05, ** 0.001 < p ≤ 0.01, *** 0.0001 < p ≤ 0.001, and **** p ≤ 0.0001. Each experiment was performed in replicates (n = 2) with technical triplicates (n = 3).

The validation of SNPs associated with TNFAIP3 as a risk factor through CRISPR-Cas9 technology offers a promising avenue for understanding the genetic underpinnings of a disease. Research has identified several SNPs within the TNFAIP3 gene that are associated with autoimmune diseases. For instance, rs6920220 (G/A) has been linked to various autoimmune diseases, highlighting the importance of TNFAIP3 in immune regulation (Wu et al., 2020; Ciccacci et al., 2019) (Figure 1A). Furthermore, other SNPs, such as the nonsynonymous SNP rs2230926 (T/C/G) (Figure 1B), have been shown to influence TNFAIP3 expression and its inhibitory activity on NF-κB signaling (Odqvist et al., 2019; Ciccacci et al., 2019; Han et al., 2016). This SNP is significantly associated with increased susceptibility to autoimmune conditions, including systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) (Ciccacci et al., 2019; Das et al., 2018; Han et al., 2016).

To investigate the functional impact of specific genetic variants associated with SD, we utilized CRISPR-Cas9 HDR to introduce risk alleles into the TNFAIP3 gene in human SGECs. We focused on the SNPs rs6920220 (G/A) and rs2230926 (T/C/G), which have been implicated in SD susceptibility.

For designing the single-nucleotide site-specific HDR donor templates and associated Cas9 guide RNAs, the IDT Alt-R HDR design tool was used. It is an easy-to-use, efficient online-based tool to design the oligonucleotides as donor templates and guide RNAs.

HDR donor templates were designed to introduce the desired SNP alleles (Table 1). For rs6920220 (G/A), the donor template (+) strand incorporated the A allele (ATC​TGC​TTC​CAT​CTG​TTA​GCA​GGT​AAC​TTC​TCC​ACT​AAA​AAG​ATA​TGG​TTC​TGT​AGA​ACA​ATG​GCA​TAT​GCA​GAC​AGT​GAT​C), whereas the (−) strand contained the T allele (GAT​CAC​TGT​CTG​CAT​ATG​CCA​TTG​TTC​TAC​AGA​ACC​ATA​TCT​TTT​TAG​TGG​AGA​AGT​TAC​CTG​CTA​ACA​GAT​GGA​AGC​AGA​T). The sgRNA sequence (ACT​TCT​CCA​CTA​AAA​GGA​TA) targeting the region surrounding rs6920220 (G/A) was selected for its predicted high on-target activity and minimal off-target potential.

For rs2230926 (T/C/G), the donor template (+) strand incorporated the G allele (GGC​GTT​CAG​GAC​ACA​GAC​TTG​GTA​CTG​AGG​AAG​GCG​CTG​TGC​AGC​ACG​CTC​AAG​GAA​ACA​GAC​ACA​CGC​AAC​TTT​AAA​TTC​CGC​TGG​CA), whereas the (−) strand contained the C allele (TGC​CAG​CGG​AAT​TTA​AAG​TTG​CGT​GTG​TCT​GTT​TCC​TTG​AGC​GTG​CTG​CAC​AGC​GCC​TTC​CTC​AGT​ACC​AAG​TCT​GTG​TCC​TGA​ACG​CC).

The sgRNA sequence (GGC​GCT​GTT​CAG​CAC​GCT​CA) targeting the region surrounding rs2230926 (T/C/G) was selected for its predicted high on-target activity and minimal off-target potential.

SGECs were transfected with Cas9 RNP complexes and the respective HDR donor templates. Control SGEC lines were also generated by SGEC transfection with a non-targeting control sgRNA using an Alt-R CRISPR-Cas9 Control Kit, Human, 2 nmol (cat. #1072554).

To confirm the successful introduction of the rs6920220 (G/A) and rs2230926 (T/C/G) SNPs in SGECs via CRISPR-Cas9 HDR, an allele-specific probe-based PCR was performed. The data demonstrate differential amplification across control and edited cell lines, indicative of successful gene editing. For desired SNP detection, we designed the IDT’s Affinity Plus SNP genotyping assay. This strategy is similar to the 5′ nuclease-based TaqMan assay. This assay employs two primers to amplify the sequence of interest containing the desired SNP, and two or three probes are designed to detect the SNP of interest. This strategy works on the competitive binding assay in which the probes compete to bind to the same sequence. The probe that perfectly complements the target sequence will outcompete other probes that mismatch the sequence. The probes are labeled with different fluorophores to detect the specific probe binding and SNP genotype detection (Table 1).

For rs6920220 (G/A), we used the abovementioned approach with specific probes for the A and G alleles. WT probe: 5′-TAAA + A + G + GA + TA + T + GGT-3’; Mut probe: 5′-TCT + C + CA + CTAAA + A + A + GA-3’; forward primer: 5′-TAC​GGC​AGC​GTA​ACA​TAG​TA-3′; and reverse primer: 5′-GTC​TGC​ATA​TGC​CAT​TGT​TCT​A-3′ to amplify the sequence of interest with an amplicon size of 125 bp. Similarly, for rs2230926 (T/C/G), probes were designed for the T probe: 5′-TG + CT + G + A + AC + AGCG-3’; C probe: 5′-TGCT + G + G + A + CAGC-3’; and G probe: 5′-TGCT + G + C + A + CA + GC-3′ with the forward primer: 5′-CAT​GCC​ACT​TCT​CAG​TAC​ATG-3′ and reverse primer: 5′-CGT​GTG​TCT​GTT​TCC​TTG​AG-3′ to amplify the sequence of interest with an amplicon size of 92 bp. The “+” symbol preceding a nucleotide in the probe sequence denotes synthetic modification of that nucleotide for better stability and hybridization efficiency. Probe-based RT-PCR analysis using genomic DNA from edited and control SGECs revealed that for the rs6920220 (G/A) variant, the Mut probe targeting the A allele showed a 2,799-fold increase in relative fluorescence compared to that in the control. In contrast, the WT probe targeting the G allele exhibited no significant difference in fluorescence relative to that in the control (Figure 2A).

Data for rs2230926 (T/C/G) demonstrated significant upregulation of the G probe (80-fold) and C probe (948-fold), whereas there is no significant upregulation of the T probe (Figure 2B). These data elaborated the presence of the targeted SNP alleles at the desired locations (Figure 2).

To investigate the functional impact of specific genetic variants associated with SD, we utilized CRISPR-Cas9 gene editing to generate SGEC lines carrying the risk alleles. We focused on variants within or near the genes implicated in SD pathogenesis, specifically examining the impact of rs6920220 (G/A) and rs2230926 (T/C/G). Following editing, we assessed the expression of key genes involved in inflammation and immune response in both SGECs and co-cultures of SGECs with Jurkat cells in a transwell setup, a model for immune cell interaction with the salivary gland epithelium. The cells were treated with LPS (10 ng/mL) in complete media to mimic and induce an inflammatory response. In SGECs, we observed a significant downregulation of TNFAIP3 mRNA in both the rs6920220 (G/A) and rs2230926 (T/C/G) cells edited using CRISPR-Cas9 technology compared to the control SGECs (Figure 3Ai). This decrease was most pronounced in the rs6920220 (G/A) cells. NF-κB mRNA levels were markedly increased in rs6920220 (G/A) cells, indicative of increased NF-κB activity, a key driver of inflammation in SD. NF-κB mRNA expression in rs2230926 (T/C/G)-edited cells was not increased and instead showed a slight downward trend relative to that in the controls (Figure 3Ai). TGF-β expression, however, did not show significant changes across the groups.

Analysis of pro-inflammatory cytokines revealed a significant upregulation of IL-6 and IL-8 mRNA in both rs6920220 (G/A)- and rs2230926 (T/C/G)-edited SGECs (Figure 3Aii). IL-6 and IL-8 are known to be involved in pro-inflammatory and chronic inflammation signature as in the case of SD. IL-1β expression was also elevated, though not significantly. These data suggest that the presence of the risk alleles in SGECs leads to increased expression of pro-inflammatory cytokines, potentially contributing to the inflammatory milieu observed in SD. We also examined the expression of genes involved in immune responses, including PKR, IRF1, IRF5, and CXCL10 (Figure 3Aiii). We observed a trend toward increased expression of IRF5 in rs2230926 (T/C/G)-edited cells, but this increase did not reach statistical significance (p = 0.754) (Figure 3Aiii). PKR is the serine/threonine kinase that plays an important role in innate immune response. IRF1 and IRF5 are the members of the interferon regulatory factor (IRF) family of transcription factors involved in immune response regulation. CXCL10 is the chemokine responsible for monocyte, natural killer cell, and T-cell migration. Further analysis focused on OAS1 and interleukins IL-4, IL-10, IL-12, and IL-13 (Figure 3Aiv). OAS1 is involved in innate cellular immune response. IL-4, IL-10, IL-12, and IL-13 are chemokines involved in pleiotropic effects such as immune response modulation, inflammation and related pathway, and B-cell and T-cell activity regulation. Similar to the previous set of genes, we observed some fluctuations, but none reached statistical significance. To mimic the interaction between SGECs and immune cells in the context of SD, we co-cultured the CRISPR-Cas9-edited SGECs with Jurkat cells. In the co-culture setting, we again observed a trend toward decreased TNFAIP3 expression in the edited SGECs, although it did not reach statistical significance (Figure 3Bi). NF-κB expression showed a similar pattern of increase as seen in the SGEC monocultures without statistical significance. TGF-β expression remained unchanged. Co-culturing with Jurkat cells significantly impacted the expression of inflammatory cytokines. IL-1β, IL-6, and IL-8 were all markedly upregulated in the co-cultures with SGECs carrying the risk alleles, particularly rs6920220 (G/A) (Figure 3Bii). This suggests a synergistic effect of the genetic variants and immune cell interaction on inflammatory cytokine production. We also examined the expression of PKR, IRF1, IRF5, and CXCL10 in the co-culture setting (Figure 3Biii). Similar to the monocultures, IRF5 and CXCL10 showed a significant upregulation in rs2230926 (T/C/G) cells co-cultured with Jurkat cells. In the co-culture, we observed a significant increase in IL-12 mRNA levels in the presence of the rs6920220 (G/A)-edited SGECs (Figure 3Biv).

Other interleukins (IL-4, IL-10, and IL-13) showed some changes but were not statistically significant. These data demonstrate that the CRISPR-Cas9-mediated introduction of SD-associated risk alleles in SGECs impacts the expression of key inflammatory regulators. The presence of the A allele at rs6920220 (G/A) in particular and G and C alleles at rs2230926 (T/C/G) leads to decreased TNFAIP3 expression and increased NF-κB activity, resulting in elevated pro-inflammatory cytokine production. Co-culturing with Jurkat cells further augmented this inflammatory response, highlighting the interplay between genetic predisposition and immune cell activation in the context of SD. These findings provide functional evidence supporting the role of these genetic variants in the pathogenesis of SD and suggest potential therapeutic targets for this debilitating autoimmune disease. All the RT-PCR primer sequences are listed in Table 2.