Tristetraprolin/ZFP36 Regulates the Turnover of Autoimmune-Associated HLA-DQ mRNAs

Figure 1. RNA binding proteins interaction. (A) REMSA experiments performed using 3DQA101 (Lane 1) and 3DQA105 (Lane 9) riboprobes. Lanes 2 and 8 show the digestion of riboprobes with T1 RNase. The binding of DQA101 with M14 extract is in Lane 3 and with B-LCL#5 extract in Lane 6. The binding of 3DQA105 with M14 extract is in Lane 9 and with B-LCL#5 extract in Lane 12. The interaction of 5 and 1 µg of rTTP with 3DQA101 is shown in Lanes 4 and 5 and with 3DQA105 in Lanes 10 and 11. (B) Western blot analysis of biotin pull-down assay carried by using 3DQA101, 3DQA105, 3DQB102, and 3DQB105 riboprobes. The antibodies used for the immunoblot were anti-NF90, anti-TTP, and anti-EBP1. Molecular weights are as indicated.

Figure 2. Sequences and structures comparison between 3′UTR of DQA1* and DQB1* genes. (A) Comparison of the 3′UTR sequences for the DQA1*01 and DQA1*05 alleles indicates 94.1% sequence identity. There is a slightly higher proportion of GC content in the DQA1*01 allele (0.46) as compared to DQA1*05 (0.44). (B) The DQB1*05 and DQB1*02 alleles show a lower sequence identity (92.3%), but identical GC proportions (0.55). (C,D) RNA secondary structure prediction reveals the sequence differences between alleles impacts the minimum-free energy folding (RNAfold). Common structural motifs from FOLDALIGN are mapped onto the structures as well as canonical ARE (AUUUA/AUUUUA) and half-ARE motifs (e.g., UAUU).

Figure 3. AU-rich motif analysis of the DQA1* and DBQ1* genes reveals key differences. (A) All possible di-, tri-, tetra-, penta- and hexa-mers are assessed for their observed frequencies against their expected frequencies. AU-rich motifs are colored, non-AU rich in grey. All AU-rich kmer motifs have higher enrichments in the CD associated DQA1*05 and DQB1*02 alleles as compared to the non-CD associated DQA1*01 and DQB1*05 alleles. (B) Mapping the AU-rich motifs to their position within the 3′UTR sequences reveals a clear enrichment in AU-motifs throughout the CD associated alleles DQA1*05 and DQB1*02 (Blue) as compared to the non-CD associated DQA1*01 and DQB1*05 alleles (Red). Nested motifs are indicated on the y-axis; for example, two overlapping motifs will show a score of 2. A structure accessibility score indicates the likelihood the RNA single stranded, an absence of base-pairing, across its length. A score of 1.0 indicates the structure is single stranded.

Figure 4. TTP knockdown in M14. (A) Western blot performed with anti-TTP for the assessment of protein depletion 48 h after silencing with siCtlr or siTTP (B) Fold variation of DRA, DRB1, and DQA1 mRNAs. HLA-a,b,c are the class I mRNAs, CIITA is the HLA class II transcriptional transactivator.

Figure 5. TTP knockdown in B-LCL#5. (A). Western blot performed with anti-TTP for the assessment of protein depletion 48 h after silencing with siCtlr or siTTP nucleofection. (B) Cytofluorimetric analysis of HLA-DQ surface expression, reported as fold change of MFI (Mean Fluorescence Intensity). (C) Fold variation of DRA, DRB1, DQA1, DQB1, and HLA-A, -B, and -C (class I) mRNAs. (D) Fold variation of DQA1*01, DQA1*05, DQB1*02, and DQB1*05 mRNAs.