Proteolytic cleavage and inactivation of the TRMT1 tRNA modification enzyme by SARS-CoV-2 main protease

Abstract

Nonstructural protein 5 (Nsp5) is the main protease of SARS-CoV-2 that cleaves viral polyproteins into individual polypeptides necessary for viral replication. Here, we show that Nsp5 binds and cleaves human tRNA methyltransferase 1 (TRMT1), a host enzyme required for a prevalent post-transcriptional modification in tRNAs. Human cells infected with SARS-CoV-2 exhibit a decrease in TRMT1 protein levels and TRMT1-catalyzed tRNA modifications, consistent with TRMT1 cleavage and inactivation by Nsp5. Nsp5 cleaves TRMT1 at a specific position that matches the consensus sequence of SARS-CoV-2 polyprotein cleavage sites, and a single mutation within the sequence inhibits Nsp5-dependent proteolysis of TRMT1. The TRMT1 cleavage fragments exhibit altered RNA binding activity and are unable to rescue tRNA modification in TRMT1-deficient human cells. Compared to wildtype human cells, TRMT1-deficient human cells infected with SARS-CoV-2 exhibit reduced levels of intracellular viral RNA. These findings provide evidence that Nsp5-dependent cleavage of TRMT1 and perturbation of tRNA modification patterns contribute to the cellular pathogenesis of SARS-CoV-2 infection.

Keywords: Mpro; Nsp5; SARS-CoV-2; TRMT1; tRNA.

Figure 1.

Human cells infected with SARS-CoV-2 exhibit a reduction in TRMT1 levels and perturbations in tRNA modification patterns. (A) Immunoblot analysis of lysates prepared from MRC-5-ACE2 human cells that were mock-infected or infected with SARS-CoV-2 at MOI of 5 for 24 or 48 hours. The immunoblot was probed with anti-TRMT1, actin, or SARS-CoV-2 nucleocapsid (N) antibodies. Circle represents endogenous full-length TRMT1. Asterisk (*) denotes a non-specific band. Size markers are noted in kiloDalton. (B) Quantification of TRMT1 signal intensity normalized to actin in the mock or SARS-CoV-2-infected cell lines. TRMT1 protein levels are expressed relative to mock-infected samples for each time point. (C) m2,2G levels in small RNAs isolated from MRC5 cells that were either mock-infected or infected with SARS-CoV-2 at MOI of 5 for 24 or 48 hours. m2,2G levels were normalized to A, C, G, and U. Samples were measured in biological replicates. Statistical significance for (B) and (C) was determined by two-way ANOVA with multiple comparisons test. ***p<0.001; ****p<0.0001; ns, non-significant. (D) Levels of the indicated RNA modifications in small RNAs isolated from MRC5 cells that were either mock-infected or infected with SARS-CoV-2 for 24 or 48 hours. RNA modification levels were normalized to A, C, G, and U. Y-axis represents the log2 fold change in the levels of the indicated tRNA modification between SARS-CoV-2 infected versus mock-infected MRC5 cells.

Figure 2.

SARS-CoV-2 Nsp5 binds TRMT1 in human cells. (A) Schematic of human TRMT1 primary structure with predicted Nsp5 cleavage site. Mitochondrial targeting signal (MTS), methyltransferase (MT) domain and zinc finger motif are denoted. (B) Consensus sequence logo of cleavage sites in SARS-CoV-2 polyproteins. (C) Alpha-fold predicted structure of human TRMT1 with putative Nsp5 cleavage site denoted in magenta and arrowhead. (D) Immunoblot of input and streptactin purifications from human cells expressing empty vector, wildtype (WT) Nsp5, or Nsp5-C145A fused to the Strep tag without or with co-expression with TRMT1-FLAG. The immunoblot was probed with anti-Strep, FLAG and actin antibodies. Square represents TRMT1-FLAG, circle represents endogenous TRMT1. Size markers are noted to the left in kiloDalton. The experiment in (D) was repeated in Supplemental Figure 2.

Figure 3.

Nsp5 expression induces cleavage of TRMT1 in human cells. (A) Immunoblot of lysates prepared from human 293T cells expressing GFP, Nsp5 or Nsp5-C145A. The immunoblot was probed with anti-TRMT1, Strep, or actin antibodies. Hours post represents the time post-transfection. Circle represents endogenous TRMT1. Arrow represents the N-terminal (N)-TRMT1 cleavage fragment. * denotes a non-specific band. Size markers to the left in kiloDalton. (B, C) Quantification of endogenous TRMT1 or N-terminal (N)-TRMT1 cleavage product in transfected cells. TRMT1 and N-TRMT1 signal was normalized to actin. (D) Immunoblot of lysates prepared from wildtype or TRMT1-knockout (KO) human cell lines expressing Nsp5. Experiments in (A) and (D) were repeated three times (see Source data).

Figure 4.

Sequence-dependent cleavage of TRMT1 by SARS-CoV-2 Nsp5. (A) Immunoblot of lysates from human cells expressing empty vector, wildtype (WT) Nsp5-Strep, or Nsp5-C145A-Strep without or with co-expression with TRMT1-FLAG. The immunoblot was probed with anti-Strep, FLAG and actin antibodies. Square represents TRMT1-FLAG, * denotes a non-specific band, arrow represents N-terminal TRMT1 cleavage product and arrowhead indicates the C-terminal TRMT1 cleavage product. (B) Schematic of human TRMT1 with predicted Nsp5 cleavage site and Q530N mutation. (C) Immunoblot of lysates from human cells expressing empty vector, wildtype (WT) Nsp5-Strep, or Nsp5-C145A-Strep without or with co-expression with TRMT1-FLAG or TRMT1-FLAG Q530N. Experiments in (A) and (C) were repeated three times with comparable results (see Source data).

Figure 5.

N- and C-terminal TRMT1 cleavage fragments exhibit alterations in RNA binding and tRNA modification activity. (A) Schematic of wildtype TRMT1 and predicted TRMT1 fragments resulting from Nsp5 cleavage at Q530N. (B) Immunoblot analysis of anti-FLAG purifications from human cells expressing vector control, full-length TRMT1, or TRMT1 cleavage fragments fused to the FLAG tag. The immunoblot was probed with anti-FLAG and anti-actin antibodies. (C) Nucleic acid stain of RNAs extracted from the indicated input or purified samples after denaturing PAGE. The migration pattern of 5.8S rRNA (~150 nt), 5S rRNA (~120 nt) and tRNAs (~70–80 nt) are denoted. (D) Immunoblot of TRMT1 expression in either control-WT or TRMT1-KO human 293T cell lines. (E, F) Representative gel of primer extension assays to monitor the presence of m2,2G in tRNA-Met-CAU or mt-tRNA-Ile-UAU from the cell lines transfected with the indicated TRMT1 constructs. D, dihydrouridine; m1G, 1-methylguanosine; >, labeled oligonucleotide used for primer extension. Protein-RNA purification was repeated with comparable results (see Source data for repeat).

Figure 6.

The expression of TRMT1 affects the levels of SARS-CoV-2 RNA replication in human cells. (A) Immunoblot of lysates prepared from 293T control-wildtype (WT) or TRMT1-KO cell lines that were mock-infected (MOI of 0) or infected with SARS-CoV-2 at MOI of 0.2 or 0.4 for 24 hours. The immunoblot was probed with antibodies against TRMT1 or actin. Circle represents endogenous full-length TRMT1. Asterisk denotes a non-specific band. Size markers to the right in kiloDalton. (B) Normalized TRMT1 signal intensity relative to mock-infected cells (MOI of 0). (C) Immunoblot of lysates prepared from 293T control-wildtype (WT) or TRMT1-KO cell lines that were mock-infected (MOI of 0) or infected with SARS-CoV-2 at MOI of 0.1 or 5.0 for 24 hours. The immunoblot was probed with antibodies as in (A). (D) Normalized TRMT1 signal intensity relative to mock-infected cells (MOI of 0). (E, F) SARS-CoV-2 RNA copy number in control-WT or TRMT1-KO human 293T cell lines after infection at the indicated MOI for 24 hours. Viral copy number was measured by QRT-PCR and normalized to GAPDH. Samples were measured in triplicate. *p < 0.05; **p<0.01; ***p<0.001; ****p < 0.0001; ns, non-significant.

Figure 7.

TRMT1 is required for efficient SARS-CoV-2 replication in human cells. (A) Immunoblot of lysates prepared from the indicated 293T TRMT1-KO cell lines that were mock-infected (MOI of 0) or infected with SARS-CoV-2 for 24 hours. The immunoblot was probed with antibodies against TRMT1 or actin. Square represents full-length TRMT1-FLAG. Asterisk denotes a non-specific band. Size markers to the left in kiloDalton. (B) Normalized TRMT1-WT signal intensity relative to mock-infected cells (MOI of 0). (C) Normalized TRMT1-Q530N signal intensity relative to mock-infected cells (MOI of 0). (D) SARS-CoV-2 RNA copy number in control-WT or TRMT1-KO human 293T cell lines after infection at the indicated MOI. Viral copy number was measured by QRT-PCR and normalized to GAPDH. *p < 0.05; **p<0.01; ****p < 0.0001; ns, non-significant.

Figure 8.

Viral infectivity measurement of supernatants collected from the indicated cell lines infected with SARS-CoV-2 for 24 hours. (A) Viral titer of supernatants collected from the indicated cell lines infected with SARS-CoV-2. Infectious titer was determined by TCID₅₀ endpoint dilution assay in VeroE6 cells and expressed in focus forming units per mL of supernatant (FFU/mL). (B) Infectivity of SARS-CoV-2 particles generated from cell lines in (A). Infectivity of viral particles was calculated with the formula [(FFU/mL)/(viral genomic RNA copies/mL)], and expressed in FFU per 100 genomic copies.