Mammalian TatD DNase domain containing 1 (TATDN1) is a proteostasis-responsive gene with roles in ventricular structure and neuromuscular function

Abstract

The characterization of highly conserved but poorly understood genes often reveals unexpected biological roles, advancing our understanding of disease mechanisms. One such gene is Mammalian TatD DNase domain containing 1 (Tatdn1), the mammalian homolog of bacterial Twin-arginine translocation D (TatD), a protein proposed to have roles either in DNA degradation or protein quality control in unicellular organisms. Despite its association with different pathologies, including several cancer types and cardiovascular diseases, the role of TATDN1 in mammals remains unexplored. Here, we demonstrate that Tatdn1 encodes a cytoplasmic protein that does not participate in DNA degradation but is upregulated in cells under proteostasis stress. Tatdn1-deficient mice exhibit dysregulated expression of genes involved in membrane and extracellular protein biology, along with mild dilated cardiomyopathy and impaired motor coordination. These findings identify TATDN1 as a key player in cytosolic processes linked to protein homeostasis, with significant physiological implications for cardiac and neurological function.

Keywords: TatD; Tatdn1; cardiomyopathy; motor control; neurobehavior; ventricle dilation.

Fig. 1

Tatdn1 gene is ubiquitously expressed and FLAG‐TATDN1 has a cytoplasmic localization. (A) RT‐qPCR of Tatdn1 in 3‐ to 5‐day‐old neonatal (blue) and 4–5‐month‐old adult (orange) mouse tissues: Br, brain; Cb, cerebellum; Ht, heart; K, kidney; Lu, lung; Li, liver; Sp, spleen. Each individual value from 3 to 7 replicates is shown plus mean ± SD. (B) Representative TATDN1 western blot from total protein extracts obtained from mice in (A). NB, Naphthol blue stained membrane for loading control. Densitometric quantifications of TATDN1 band for each tissue referred to neonatal heart signal are depicted in the graph below. Each value from triplicates is shown plus mean ± SD (C) Representative western blot and confocal immunofluorescence images of overexpressed FLAG‐TATDN1 in HEK293 cells. Green: FLAG‐TATDN1, blue: Hoechst staining (nuclei) and merged images are shown. (D) Western blot of protein extracts and subcellular fractions of HEK293 cells transfected with empty vector (ø) or FLAG‐TATDN1 vector. Cyto, cytoplasm; Mito, mitochondria; Nuc, nucleus. Markers: CytC, cytochrome c for mitochondria; GAPDH for cytoplasm; Lamin A + C for nuclei. NB, Naphthol blue membrane staining. For A and B, two‐way ANOVA followed by Sidak's test was run. Inter. = Interaction. (E) Bar plots: ± SEM showing TATDN1 levels as raw intensity obtained from mass spectrometry analysis of six subcellular fractions in HeLa cells (n = 4), mouse liver (n = 4) and mouse muscle (n = 3) from Martinez‐Val et al. [24].

Fig. 2

TATDN1 cytoplasmic localization is not altered during cell death and its expression is not required for caspase‐independent and caspase‐dependent DNA degradation. Representative western blot showing endogenous TATDN1 expression in total (T), cytosolic (C), nuclear (N) and mitochondrial (M) subcellular fractions of (A) primary neonatal rat ventricular cardiomyocytes cultured in normal conditions or submitted to 12 h ischemia; or (B) primary neonatal dermal fibroblasts treated with 1 μmol·L⁻¹ staurosporine (STS) for 14 h. N = 3 independent experiments. Subcellular fraction purity was assessed by blotting fragments of the membranes with antibodies against lactate dehydrogenase (LDH; cytoplasm), Lamin A + C (Lamin, nucleus), cytochrome oxidase‐IV (COXIV, mitochondrial membrane), cytochrome c (CytC, mitochondria but released during cell death). NB, Naphthol blue total protein staining of the membrane. (C) TATDN1 western blot for the assessment of the efficiency of lentiviral‐driven Tatdn1 gene silencing in primary rat neonatal cardiomyocytes (upper panel) and DNA agarose gel electrophoresis images of total DNA extracts from cardiomyocytes that were cultured in control conditions or in experimental ischemia for 8, 16, and 24 h. (D) TATDN1 western blot for the assessment of the efficiency of lentiviral‐driven Tatdn1 gene silencing in Rat2 cells (upper panel) and DNA agarose gel electrophoresis images of total DNA extracts from control fibroblasts and cultures treated with 1 μmol·L⁻¹ staurosporine (STS) and STS + 5 μmol·L⁻¹ Q‐VD‐OPh (QVD; pan‐caspase inhibitor) for 8 h. NT, not transduced; Scr, scrambled‐transduced; shRNA, Tatdn1‐specific silencing. HMWF, DNA high molecular weight fragmentation; LMWF, DNA low molecular weight fragmentation. Representative images are shown from three independent experiments. Marker: 50 kb DNA molecular weight marker (for CHEF) and 1 kb DNA marker (for conventional electrophoresis). NB, Naphthol blue staining.

Fig. 3

TATDN1 is dispensable for embryo and postnatal development, cell proliferation, and cell death‐associated DNA degradation in mice. (A) Strategy to generate the conditional Tatdn1 allele. Three‐dimensional model of human TATDN1 (UniProt Q6P1N9) associating the corresponding coding exons to their location in the polypeptide chain, Tatdn1 exonic structure indicating the Southern blot probes, and targeted allele showing location of the fragment integrated from the targeting vector. After homologous recombination, the Neomycin resistance cassette is located downstream exon 3, flanked by loxP and FRT sites. The Neomycin cassette was removed by crossing Tatdn1 floxed homozygous mice with ROSA‐FLPe mice. Cre‐mediated excision resulted in one loxP site in the place of exon 3. (B) TATDN1 protein expression in adult Tatdn1 ^+/+ and Tatdn1 ^−/− mouse tissues by western blot. Br, brain; Cb, cerebellum; E, eye; Ht, heart; K, kidney; Lu, lung; Li, liver; Sp, spleen; Ut, uterus. *Denotes unspecific band. NB, Naphthol blue protein staining of the membrane (representative image from 3 independent experiments). (C) Body weight of P5 pups (26–30 per gender and genotype), and P90 mice (4–10 per gender and genotype). (D) Organ weight of P5 (3 males/genotype) and P90 males (10 Tatdn1 ^+/+ , 4Tatdn1 ^−/− ). Individual values are plotted with mean ± SD (C, D). (E) cardiomyocyte number/heart counted as described in the Methods section, from 11 Tatdn1 ^+/+ and 6 Tatdn1 ^−/− P5 pups. (F) Area of P5 cultured cardiomyocytes calculated as described in the Methods section. Dots correspond to individual cardiomyocyte areas counted in several microscopic fields of two independent neonatal cardiomyocyte cultures per genotype (>50 cells/genotype are plotted). (G) Number of cell cycles completed in 72 h of primary dermal fibroblasts from Tatdn1 ^+/+ and Tatdn1 ^−/− neonatal mice. An equal number of primary dermal fibroblasts were simultaneously seeded, and 72 h later, cells were detached and counted. N = 4 independent experiments. Median ± interquartile range is also indicated (E–G). (H) Analysis of DNA degradation of primary dermal fibroblasts from Tatdn1 ^+/+ and Tatdn1 ^−/− neonatal mice. Fibroblasts were cultured and treated or not with 1 μmol·L⁻¹ staurosporine (STS) for 8 h in the presence or absence of 5 μmol·L⁻¹ of the pan‐caspase inhibitor Q‐VD‐OPh (QVD). DNA was extracted and electrophoresed and stained as described in the Methods section. HMWF, DNA high molecular weight fragmentation; LMWF, DNA low molecular weight fragmentation. N = 3 independent times with similar results. Plots show individual experimental values, medians, and interquartile ranges. Two‐way ANOVA followed by Sidak's test (C, D) and Mann–Whitney U‐test (E–G) were run. Inter., Interaction. Statistical differences are indicated in each graph.

Fig. 4

DNA degradation in vitro assays do not support DNase activity of endogenous TATDN1. (A) Assessment of TATDN1 enrichment in cytosolic extracts (cytosol) compared to total cell lysates (Total) of rat primary dermal fibroblasts. Potential contamination of cytosolic fractions with other subcellular compartments was assessed by blotting different fragments of the same membranes with antibodies against lactate dehydrogenase (LDH; cytoplasm), Lamin A + C (Lamin, nucleus), cytochrome oxidase‐IV (COXIV, mitochondrial membrane). NB, Naphthol blue membrane protein staining. (A) Representative sample from 3 independent experiments used in DNase assays is shown. (B) 1 μg of BglII‐linearized pcDNA3.1 plasmid (5.4 kb) was incubated with different amounts of cytosolic extracts from Tatdn1 ^+/+ and Tatdn1 ^−/− dermal fibroblast cultures (10, 20, 30 μg of protein) for 2 h at 37 °C and pH 8 in the conditions described in Methods. (C) Cytosolic extracts (30 μg protein) from Tatdn1 ^+/+ and Tatdn1 ^−/− dermal fibroblast cultures or 30 μg of BSA were incubated with 1 μg of BglII‐linearized pcDNA3.1 plasmid for 2 h at 37 °C and pH 8 or 6. At the end of the incubation time, the samples and aliquots of BglII‐digested and nondigested plasmid were electrophoresed in 1% agarose gels stained with SYBR Safe, and images were captured under UV illumination. A marker of DNA molecular weight was loaded at the left of each gel (6 and 3 kb bands are identified in the figure). Each assay was repeated three times with equal results. (D) TATDN1 expression in total protein extracts of HEK293 transfected with the wild‐type Rps9 or the Rps9 D95N mutant form inducing mistranslation (N = 5 clones each for RPS9 WT and D95N; mean ± SD), unpaired t‐test was used to determine statistically significant differences. (E) TATDN1 expression in total protein extracts of HEK293 cells treated for 24 h with tunicamycin at two different concentrations to induce UPR (N = 5 DMSO; N = 6/tunicamycin 1.5 and 3 μmol·L⁻¹; mean ± SD). One‐way ANOVA followed by Dunnett's test was performed to assess statistically significant differences. Analysis of the endoplasmic reticulum chaperone α‐BiP was used to evidence activation of the unfolded protein response.

Fig. 5

Analysis of the gene expression changes induced by Tatdn1 deletion in the heart ventricle and brain cortex of adult mice. An expression microarray was performed comparing Tatdn1 ^+/+ (WT) and Tatdn1 ^−/− (KO) mice as described in the Methods section (n = 6/genotype). (A) Data mining analysis was done with GSEA software (FDR < 0.05, NES > 1.5). Representative enrichment plots of angiogenesis, epithelial mesenchymal transition, inflammatory response, and interferon alpha response gene sets are shown. FDR, false discovery rate; NES, normalized enrichment score. (B) Gene sets overrepresented in the heart and brain cortex. GSEA was used to test for significant enrichment of defined gene signatures. FDR, false discovery rate; NES, normalized enriched score. Threshold FDR < 0.1 and NES ≥ 2.0. Hallmark, C2, and C5 gene sets were obtained from the Molecular Signature Database (v2.5). (C) Heat map showing the top 100 genes from the ranked list of GSEA differentially expressed (upregulated and downregulated in heart and brain cortex and ordered by fold change). The Wilcoxon test was used to determine statistical significance. (D) Venn diagram showing the common upregulated and downregulated genes (FC >1.5, FDR <0.05) in the heart and brain cortex. Transcriptome analysis was carried out with Transcriptome Analysis Console software (Applied Biosystems, Thermo Fisher Scientific). Gene names and human orthologues of common genes are specified in Data S1.

Fig. 6

Adult TATDN1‐deficient mice display left ventricle dilatation and narrow cardiomyocytes. Six‐month‐old Tatdn1 ^+/+ (7 females, 6 males) and Tatdn1 ^−/− (6 females, 6 males), and 16‐month‐old Tatdn1 ^+/+ (4 females, 10 males) and Tatdn1 ^−/− (4 females, 6 males) mice were subjected to echocardiography. (A) Representative images of M‐mode echocardiographic images from six‐month‐old mice and (B) Main cardiac functional parameters directly measured or calculated from M‐mode data. LVEDD, left ventricular end‐diastolic diameter; LVESD, left ventricular end‐systolic diameter; LVEF, left ventricle ejection fraction; SWT, septum wall thickness. Individual values and mean ± SD are shown. (C) representative confocal microscopic images obtained from paraffin‐embedded myocardial sections from 6‐month‐old mice stained with FITC conjugated wheat germ agglutinin (green) and counterstained with Hoechst (blue, nuclei) (scale bar, 25 μm). (D) Short axis cardiomyocyte areas were measured (n = 54 cardiomyocytes/genotype, 3 mice per genotype, mean ± SD. Student's t‐test analysis). (E) Upper panel: Venn diagram showing differentially expressed genes (DEGs) shared across three datasets. The red circle represents DEGs in the heart of Tatdn1 ^−/− vs. Tatdn1 ^+/+ mice (FC >2 or FC <0.5, P < 0.05, n = 118 genes at this stringency), the blue circle represents DEGs from Jurgens et al. [31] (n = 200), and the green circle represents DEGs from Zheng et al. [10] (n = 1579). Numbers within each section indicate the count of unique or shared DEGs. DEGs shared are highlighted, with genes in red showing upregulation and genes in blue showing downregulation. No DEGs were common to all three datasets. The diagram was generated using Interactivenn. Lower panel: Heatmap showing expression patterns of the selected DEGs in Tatdn1 ^−/− vs. Tatdn1 ^+/+ mice hearts. The color scale represents normalized expression values, with red indicating higher expression and blue indicating lower expression.

Fig. 7

Behavioral characterization of TATDN1 deficient mice. Five‐month‐old adult Tatdn1 ^+/+ (8 females, 10 males) and Tatdn1 ^−/− (8 females, 7 males) mice, and sixteen‐month‐old Tatdn1 ^+/+ (4 females, 10 males) and Tatdn1 ^−/− (4 females, 6 males) mice were subjected to a comprehensive behavioral characterization. In the open field, (A) Locomotor activity and (B) Parallel index were monitored for 5 min. Locomotor activity, two‐way ANOVA, genotype effect P = 0.006. Parallel index, genotype effect P = 0.0126. (C) In the wire hanging test, the time until the mouse falls was measured in a 1‐min session. Mann–Whitney t‐test: sum of ranks A: 283, B: 213, Mann–Whitney U: 108, P = 0.347. (D) Muscular strength was also measured by using the automatized grip strength test. Muscular strength (measured in grams) was obtained from the forepaws (left) or from all four paws (right), n = 10/genotype. Student t‐test: t = 0.162, df = 17, P = 0.873 and t = 0.316, df = 11, P = 0.757, respectively. (E) Latency to fall was evaluated in the accelerating rotarod paradigm. Two‐way ANOVA, genotype effect: F _(1,31) = 7.194, P < 0.0116. (F) In the plus maze, the time spent in the open arms was monitored for 5 min in the two groups of mice. Student t‐test: t = 0.4234, df = 31; P = 0.6749. (G) In the novel object recognition test, recognition long‐term memory was evaluated 24 h after a training trial as the percentage of time exploring the new object vs. the old object. Two‐way ANOVA, genotype effect: F _(1,60) = 0.0001, P > 0.9999. (H) In the spontaneous alternation in a t‐maze, arm preference was evaluated 2 h after a training trial as the percentage of time exploring the new arm vs. the old arm. Two‐way ANOVA, genotype effect: F _(1,62) = 0.0001, P > 0.9999. (I) In the passive avoidance paradigm, the latency (seconds) to step‐through was evaluated in the training trial and in the testing trial 24 h after receiving an electric shock (2 s/1 mA). Two‐way ANOVA, genotype effect: F _(1,62) = 1.28, P = 0.2615. The same mice were used in all the behavioral tests (n = 18 Tatdn1 ^+/+ and 15 Tatdn1 ^−/− five‐month‐old mice; n = 14 Tatdn1 ^+/+ and 10 Tatdn1 ^−/− 16‐month‐old mice). For A–K, Mean ± SEM is depicted where appropriate. (J) Immunoblotting for TatD and tubulin as a loading control in the striatum, cortex, hippocampus, and cerebellum of 5‐month‐old Tatdn1 ^+/+ mice. (K) Densitometry quantification of TATDN1 protein levels.