Characterization of the human TARDBP gene promoter

Abstract

The expression of TDP-43, the main component of neuronal intracellular inclusions across a broad spectrum of ALS and FTD disorders, is developmentally regulated and studies in vivo have shown that TDP-43 overexpression can be toxic, even before observation of pathological aggregates. Starting from these observations, the regulation of its expression at transcriptional level might represent a further key element for the pathogenesis of neurodegenerative diseases. Therefore, we have characterized the human TARDBP promoter, in order to study the transcriptional mechanisms of expression. Mapping of cis-acting elements by luciferase assays in different cell outlined that the activity of the promoter seems to be higher in SH-SY5Y, Neuro2A, and HeLa than in HEK293. In addition, we tested effects of two SNPs found in the promoter region of ALS patients and observed no significant effect on transcription levels in all tested cell lines. Lastly, while TDP-43 overexpression did not affect significantly the activity of its promoter (suggesting that TDP-43 does not influence its own transcription), the presence of the 5'UTR sequence and of intron-1 splicing seem to impact positively on TDP-43 expression without affecting transcript stability. In conclusion, we have identified the region spanning nucleotides 451-230 upstream from the transcription start site as the minimal region with a significant transcription activity. These results lay an important foundation for exploring the regulation of the TARDBP gene transcription by exogenous and endogenous stimuli and the implication of transcriptional mechanisms in the pathogenesis of TDP-43 proteinopathies.

Figure 1

Sequence of the human TARDBP promoter and map of the TARDBP TSSs in adult and fetal tissues. (A) Proximal sequences (− 1316/ + 122) of the human TARDBP promoter are shown. Transcript reference: NM_007375.4. The promoter sequences are numbered (left side) relative to the transcription start site (TSS, + 1). (B) Graphical overview of the multiple TARDBP TSSs mapped in adult and fetal tissues presented in the DataBase of Transcriptional Start Sites (DBTSS). The top turquoise (5′-region) box correspond to the RefSeq NM_007375. Red boxes represent DBTSS clones of Adult tissues, while blue boxes represent DBTSS clones of Fetal tissues.

Figure 2

Genomic alignment of putative promoter regions from Homo sapiens, Pan troglodytes and Callithrix jacchus TARDBP genes. The alignment of the 1316nt—sequence upstream and 78 bp downstream of the transcription starting site of Human (Homo sapiens, ENSG00000120948, NM_007375.3) was carried out versus: Marmoset (Callithrix jacchus, ENSCJAG00000002381), Macaque (Macaca mulatta, ENSMMUG00000007456), Olive baboon (Papio anubis, ENSPANG00000017459), Gibbon (Nomascus leucogenys, ENSNLEG00000009512) and Bonobo (Pan paniscus, ENSPPAG00000039086) TARDBP transcripts, by using the MUSCLE alignment program (http://www.ebi.ac.uk/Tools/muscle/index.html).

Figure 3

Global and local alignment of putative promoter regions from Homo sapiens, Mus musculus, Rattus norvegicus and Drosophila melanogaster TARDBP ortholog genes. (A) The alignment of the 1316nt—sequence upstream of the transcription starting site of Human (RefSeq: NM_007375.3), Mouse (ENSMUST00000084125) and Rat (ENSRNOT00000049822) TARDBP transcripts was generated with MUSCLE alignment software (http://www.ebi.ac.uk/Tools/muscle/index.html). (B) Pairwise Sequence Alignment was used to identify regions of identity/similarity between the promoter sequences of the Human/Mouse and Human/Rat TARDBP orthologs. The local similarities between the sequences were identified by using the EMBOSS Matcher Local alignment tool (https://www.ebi.ac.uk/Tools/psa/).

Figure 4

Deletion analysis of the human TARDBP promoter. Firefly luciferase activity of TARDBP promoter deletants in different cell lines (HEK293, HeLa, SH-SY5Y and Neuro2A) was measured in the cell lysate and values were normalized against Renilla. Activity of deletants is expressed as-fold against the cells transfected with the 1316 construct (= 1). Bars indicate the mean value of three independent assays. Error bars indicate standard deviation: *p < 0.05, and **p < 0.01 (One way ANOVA with Tukey test).

Figure 5

Comparison of the human TARDBP promoter activity in different cell lines. Firefly luciferase activity of 1316, 927 and 451 constructs was tested (normalized against Renilla) in different cell lines (HEK293, HeLa, SH-SY5Y and Neuro2A). Activity is expressed as-fold against transfections in HEK293 cells (= 1). Data are mean (± SD) of three independent assays: *p < 0.05, and **p < 0.01 (One way ANOVA with Tukey test).

Figure 6

Effects of ALS-related SNPs on TARDBP promoter activity. (A) Schematic representation of the human TARDBP promoter variants. The diagram depicts the wild type and the mutated promoter sequences used for the luciferase assay. The mutated nucleotides (and their positions) are indicated. (B) Luciferase activity of TARDBP promoter variants in different cell lines (SH-SY5T, HEK293, HeLa and Neuro 2A). Luciferase activity (normalized against Renilla) is expressed as-fold against transfections of the haplotype [4439:T; 4901:T] (= 1). Data are mean (± SD) of three independent assays/cell line. The mean values did not differ to a statistically significant extent (p ≥ 0.05, one way ANOVA with Tukey test).

Figure 7

Effects of TDP-43 overexpression on TARDBP promoter activity. (A) Immunoblot analysis of TDP-43 expression after 48 h-Tetracycline induction (Tet) of the HEK293-Flp-In-TDP-43 wild type stable cell line. Both endogenous (lane 1) and Flag-tagged (lane 2) proteins were visualized using anti-TDP-43 polyclonal antibody (ProteinTech). As expected, overexpression of Flag-tagged wild-type TDP-43 silenced the expression of the endogenous protein by triggering autoregulatory loop. Beta-Tubulin was used as loading control to normalize the levels of detected proteins. (B) Effects of TDP-43 overexpression on TARDBP promoter activity. The luciferase activity of three different constructs (1316, 927 and 451) transfected in the HEK293-Flp-In-TDP-43 wild type stable cell line was assayed without (No-Tet) or after (Tet) Tetracycline induction for 48 h. No statistically significant differences were found. Data are mean (± SD) of three independent assays. Significance values refer to comparisons against control transfections (No-Tet) by using one way ANOVA with Tukey test.

Figure 8

Effects of 5′UTR and intron 1 on the transcriptional activity of the human TARDBP promoter. (A) Diagrams of the 5′UTR TDP-43 constructs generated using the pGL-451 vector. In the 451 + Ex1-IVS1-Ex2 wt construct, the 1086 bp genomic region of the human TARDBP gene spanning the 5′UTR, including Exon 1 (102 bp), Intron 1 (972 bp) to the first 12 nt of Exon 2) was (Hind III-Nco I) cloned in between the 451nt-promoter and the firefly luciferase ORF (Luciferase). In the 451 + Ex1-IVS1-Ex2 mut construct, the 3′ splicing site of Intron 1 was deleted. In the 451 + Ex1Ex2 construct, the fragment encompassing Exon 1 and Exon 2 (without Intron 1) was cloned in the 451 vector. (B) RT-PCR of the 5′UTR TDP-43 mRNA species (upper panel). Amplification of GAPDH (lower panel) was used as the endogenous control in the quantitative analysis of RT-PCR. (C) Quantitation of the 5′UTR TDP-43 mRNA species by Real Time PCR. GAPDH was used to confirm normalization of total RNA levels. Co-transfected renilla orf was used to normalize the luciferase qPCR. The 451 construct was used as reference (= 1). (D) Effects of transcriptional inhibition on mRNA stability. After 36 h post-transfection of the luciferase (451 + Ex1-IVS1-Ex2 wt and 451 + Ex1-IVS1-Ex2 mut) constructs, SH-SY5Y cells were treated with actinomycin D 5 µg/ml. The relative levels of the indicated mRNAs (wt and mut) were assessed at the designated time points, following shutoff of transcription using qRT-PCR to determine mRNA half-lives. The average half-lives are reported with SD from two independent experiments. Significance values refer to comparisons against t = 0 h, wt: *p < 0.05 (one way ANOVA with Tukey test). The mean values for 451 + Ex1-IVS1-Ex2 mut construct were negligible and did not differ to a statistically significant extent (p ≥ 0.05).

Figure 9

Effects of 5′UTR and intron 1 on the luciferase activity of the TARDBP promoter constructs. The luciferase activity of 5′UTR TDP-43 constructs depicted in Fig. 8A was assayed by transient cotransfections (luciferase and renilla) in SH-SY5Y cells. After 48 h, cells were assayed for firefly luciferase expression (normalized against renilla). The empty expression vector was tested and produced background levels of luciferase activity (data not shown). Data are mean (± SD) of three independent assays. Significance values refer to comparisons against control transfections (451): *p < 0.05 (one way ANOVA with Tukey test).