A Mycobacterium tuberculosis secreted virulence factor Rv1435c/hsr1 disrupts host snRNP biogenesis

Abstract

Transcriptional adaptation drives the host responses to Mycobacterium tuberculosis (Mtb) infection. However, Mtb alters host RNA splicing to quench host antibacterial responses, the mechanism for which remains unknown. Here, we report a mechanism whereby a secreted Mtb protein interferes with the biogenesis of key spliceosomal components. A high-throughput yeast-2-hybrid screen identified several Mtb-secreted proteins interacting with the host RNA splicing factors (SFs). Through custom-designed in-cell assays, we show that one of those proteins, Rv1435c/hsr1 (host splicing regulator 1), targets specific exon-skipping events. The Mtb Rv14345c/hsr1 facilitates direct interaction between Mtb phagosomes and U5 snRNA and SNRPF, key components of the snRNPs. Genetic deletion of Rv1435c/hsr1 reverses the specific exon-skipping events caused by WT Mtb infection. The Δhsr1 strain shows compromised growth during ex vivo infection in macrophages and in vivo infection in mice. Tissue sections from the WT Mtb or Δhsr1-infected mice showed significant hsr1-dependent SNRPF staining, a phenomenon also noted in the human intestinal tuberculosis (ITB) biopsies. Thus, hsr1 is a virulence factor that disrupts host snRNP biogenesis for pathogenesis. The splicing regulators from the host and pathogen are novel targets for antituberculosis therapy.

Keywords: RNA splicing; SNRPF; U5 snRNA; spliceosome; tuberculosis.

Fig. 1.

Dysregulation of RNA splicing and interactions between mycobacterial proteins and the host spliceosome in Mtb-infected macrophages. (A) A schematic model depicting the ordered pathway of mRNA transcription by RNAP II and splicing catalyzed by spliceosome complex. The U1snRNP binds the 5’ splice site and U2 snRNP binds the branch point adenine to form complex A. U4/U5.U6 snRNP binds to form complex B and causes conformational changes to form activated complex B. This is followed by the first transesterification reaction, releasing U1 and U4 snRNP, and the second transesterification reaction forming complex C, releasing the lariat intron and U2, U5, and U6 snRNP. The U snRNP are the major players in splicing catalysis. (B) The schematic of the spliceosome complex pulldown from the Mtb infected THP-1 macrophages 48 h postinfection for Mass Spectrometry. (C) Heatmap shows massive dysregulation in the level and cellular distribution of human proteins involved in RNA splicing and identified in MS of the spliceosome complex pulldown from the Mtb infected THP-1 macrophages 48 h postinfection. The three boxes depict proteins that show depletion (green), depletion and mislocalization (blue), and mislocalization (red) patterns in Mtb-infected cells compared to the uninfected cells. (D) Mtb proteins -host splicing factor (SF) interactome- the positive interactions for the Mtb proteins Rv1435c, Rv1566c, and Rv1643c from the Y2H screen were used to generate the network shown here. The diamond and circle represent the host and Mtb proteins, respectively, and the size corresponds to the number of interactions observed. The proteins colored yellow in the network are described further in the subsequent sections. (E) The Venn diagrams show proteins belonging to the major spliceosome components and their corresponding overlap with the interacting Mtb proteins.

Fig. 2.

Ex vivo infection and reporter assays revealing the role of Mtb Rv1435c protein in modulating host RNA splicing. (A) Designing of the custom in vivo fluorescent splicing reporter. Schematic for the constitutive splicing reporter pRint with a short adenovirus intron intervening the DsRed gene with a strong 5’ and 3’ splice sites. In the pGint splicing constructs, the alternative exon of ACADVL, DDX11, DDX42, ACSL1, and NFYA, along with the flanking 100 bp intronic sequences, were inserted within the eGFP reading frame. The exon disrupts the reading frame, however, skipping of the test exon reconstitutes EGFP fluorescence. (B) In cell reporter assay: the pRint expressing stable HEK293T cell lines were cotransfected with a pGint construct and pDest-FLAG construct encoding proteins Rv1252c (NIC), Rv1566c, Rv1435c, Rv1643, or empty vector control, followed by Flow cytometry. In the DsRed⁺ gate, the %FITC⁺ population was calculated to measure the extent of the test exon-skipping event. (C) %FITC⁺ population for each of the indicated reporter constructs in the presence of indicated mycobacterial protein, noninteracting control, or vector controls. Each dot represents an independent biological replicate (N = 3), indicating P values from one-way ANOVA. (D) Validation of physical interactions between indicated pairs of Rv1435c proteins (FLAG-tagged) and host proteins—(HA-tagged) in HEK293T cells. Data are representative of two or more biologically independent transfection and pulldown experiments. (E) The schematic representation of infecting THP-1 macrophages with Msmeg strains that express Rv1435c under IVN inducible promotor followed by QPCR for assessing target gene splicing event. (F) Isoform-specific qPCR to assess the impact of Rv1435c expressing Msmeg infection in THP-1 macrophages on exon skipping event in DDX11, N = 3, indicating P values from one-way ANOVA. (G) Q-PCR for the target transcripts as mentioned after infection with H37Rv, H37RvΔhsr1, and H37RvΔhsr1::WT strain. Data from three biological replicates (N = 3), indicating P-values from one-way ANOVA.

Fig. 3.

Direct interaction of U5 snRNA with Mtb phagosomes and ESX1-dependent Hsr1 secretion sequesters SNRPF, impacting host RNA splicing. (A) Schematic representation of U snRNA nucleo-cytoplasmic trajectory and assembly of Sm core ring on snRNA by SMN complex to form snRNPs. (B) Representative images for RNA FISH against U5 snRNA in THP-1 macrophages upon Mtb (mCherry-H37Rv) infection at 24- and 48-h postinfection. The U5 snRNA probe was labeled with Alexa 647 (far-red, pseudocolor green), and the nucleus was stained with DAPI (blue). (C) Plots show quantification of U5 snRNA FISH MFI in uninfected and Mtb-infected macrophages at 24 and 48 hpi. (D) Direct interaction between H37Rv (red) and U5snRNA at 24- and 48-h postinfection. U5snRNA puncta within 0.2 microns from the H37Rv surface are shown as green dots on the 3-D constructed Mtb contours. The 3D reconstruction of the nucleus, U5 snRNA foci, and bacteria (Materials and Methods) was done to show the U5 snRNA Foci at <0.2 microns distance from the bacterial phagosomes. The quantification of interacting U5 snRNA puncta is shown in the plot at the Right. Approximately 100 cells were analyzed for this analysis, from three biological experiments, two or more fields from each. The P-value is from Student’s t test. (Scale bar, 2 μm.) (E) Representative image for the U5snRNA FISH in THP-1 macrophages infected with mCherry expressing H37Rv, Δhsr1, and Δhsr1::WT strains. The plots at the Right show the ratio of interacting to noninteracting U5snRNA with H37Rv, Δhsr1, and Δhsr1::WT strains. Data are from three biological experiments, two fields each; the P values are from one-way ANOVA. (F) Image showing SNRPF colocalizing with bacterial phagosomes. The 3D reconstruction of the nucleus, SNRPF foci, and bacteria (Materials and Methods) was done to show the SNRPF Foci at <0.2 microns distance from the bacterial phagosomes. Scale bar, 2 μm and P values indicate Student’s t test. (G) Representative images of SNRPF staining in THP-1 macrophages infected with H37Rv or Δhsr1, and Δhsr1::WT strains. For the plot below, the number of SNRPF puncta within 0.2 microns from the H37Rv surface was counted as interacting and shown as the ratio of interacting to noninteracting SNRPF with H37Rv, Δhsr1, and Δhsr1::WT strains. Data are from three biological replicates with two or more fields per slide; the P-values are from one-way ANOVA. (H) Validation of physical interactions between Rv1435c (FLAG-tagged) and SNRPF (HA-tagged) in HEK293T cells using coimmunoprecipitation. Data are representative of two or more biologically independent transfection and pulldown experiments. (I) Representative images of THP-1 macrophages stained with DAPI (nuclei, blue), SNRPF (green), and HSR1 (magenta), using Rv1435c/hsr1-specific antisera, following infection with mCherry-expressing H37Rv, Δhsr1, and Δhsr1::WT strains. White colocalization spots within the red boxes indicate overlap between SNRPF and hsr1 signals. The Right panel shows the Manders’ colocalization coefficient for SNRPF and hsr1 across the indicated strains. Data represent three independent biological replicates with two or more fields per slide; P-values were calculated using one-way ANOVA. (J) Representative images of THP1-macrophages stained with DAPI (nucleus) and Rv1435c antisera (green) upon infection with mCherry expressing WT H37Rv, Δhsr1, Δhsr1::WT, ΔCE, and ΔCE::WT strains. Images are representative of two independent biological experiments. (Scale bar, 2 μm.) (K) The exon-skipping event in ACADVL and ACSL1 transcripts upon infection with WT H37Rv, H37RvΔCE, and H37RvΔCE::WT at 48hpi from three biologically independent experiments. The P values are from one-way ANOVA.

Fig. 4.

Host alternative splicing regulation as a virulence mechanism and SNRPF dysregulation in granulomas of human TB patients. (A) THP-1 macrophages were infected with WT H37Rv or Δhsr1 strains. At 48 h postinfection, bacterial load was examined using CFU plating. Data are representative of three independent biological replicates. (N = 3); P values are from one-way ANOVA. (B) The plot shows quantification (MFI/cell) for staining of SNRPF in WT THP1 macrophages and THP1::SNRPF infected with WT H37Rv at 48hpi. Each dot represents a field analyzed, and an average of a hundred cells are quantified. The plot on the right shows the bacterial survival after 48 h postinfection in WT THP1 and THP1::SNRPF. Each dot represents an independent biological experiment. Indicated P values represent Student’s t test. (C) Schematic for study design using mice model of infection. (D) BALB/C mice (6 to 8 wk old) were infected with 200 CFU of WT H37Rv, Δhsr1, and Δhsr1::WT strains in separate experiments. At 4-wk postinfection, bacterial load in the lungs of the animals was determined by CFU plating. Increase in bacterial burden compared to Day 1 bacterial load was calculated and shown here as fold change. N = 6, representing P-values from one-way ANOVA. (E) The 4-wk postinfection lung tissue sections of uninfected, H37Rv, Δhsr1, and Δhsr1::WT strains were stained with DAPI (nucleus) and anti-SNRPF antibody. The plot on the Right shows quantification for SNRPF signal intensity across the groups, N = 5, representing P-values from one-way ANOVA. (Scale bar, 20 μm.) (F) RNAseq data from the monocytes of pulmonary TB cases (N = 20) and healthy contacts (N = 19) were analyzed to assess the expression of differential alternatively spliced transcripts. Genes belonging to the indicated functional classes were selected for the cluster analysis. Cluster 1 in each case shows transcripts upregulated in PTB cases. (G) Biopsy sections from human intestinal tuberculosis granulomatous lesions were stained with anti-SNRPF antibody. The arrows indicate multinucleated giant cells with intense staining for cytosolic SNRPF. Black arrows show areas of intense SNRPF staining in the granulomas and in the giant cells.

Fig. 5.

A model to show Mtb-mediated alteration of host RNA splicing by blocking the snRNP biogenesis and posttranscriptional gene regulation in the macrophages.