Cytotoxicity of 1-deoxysphingolipid unraveled by genome-wide genetic screens and lipidomics in Saccharomyces cerevisiae

Abstract

Hereditary sensory and autonomic neuropathy (HSAN) types IA and IC (IA/C) are caused by elevated levels of an atypical class of lipid named 1-deoxysphingolipid (DoxSL). How elevated levels of DoxSL perturb the physiology of the cell and how the perturbations lead to HSAN IA/C are largely unknown. In this study, we show that C₂₆-1-deoxydihydroceramide (C₂₆-DoxDHCer) is highly toxic to the cell, while C₁₆- and C₁₈-DoxDHCer are less toxic. Genome-wide genetic screens and lipidomics revealed the dynamics of DoxSL accumulation and DoxSL species responsible for the toxicity over the course of DoxSL accumulation. Moreover, we show that disruption of F-actin organization, alteration of mitochondrial shape, and accumulation of hydrophobic bodies by DoxSL are not sufficient to cause complete cellular failure. We found that cell death coincides with collapsed ER membrane, although we cannot rule out other possible causes of cell death. Thus, we have unraveled key principles of DoxSL cytotoxicity that may help to explain the clinical features of HSAN IA/C.

FIGURE 1:

Elevated levels of DoxSL are toxic to yeast. (A) Schematic effect of HSAN IA/C mutations on the substrate promiscuity of SPT. The blue circles highlight the hydroxyl group that is missing in DoxSL. (B–D) Effect of sphingoid bases on the growth of the indicated strains evaluated by a spot assay. Sa, sphinganine; DoxSa, 1-deoxysphinganine; DoxmetSa, 1-deoxymethylsphinganine; DoxSB, 1-deoxysphingoid base; FA-CoA, fatty acyl-CoA; DHCer, dihydroceramide; DoxDHCer, 1-deoxydihydroceramide; DoxmetDHCer, 1-deoxymethyldihydroceramide; DoxCer, 1-deoxyceramide.

FIGURE 2:

C₂₆-DoxDHCer is more toxic than C₁₆- or C₁₈-DoxDHCer. (A) Schematic representation of the fatty acyl-CoA specificity of mammalian Cer synthase. The red numbers indicate the major species of fatty acyl-CoA and Cer. (B–D) Effect of sphingoid bases on the growth of the indicated strains evaluated by a spot assay. (E) Levels of DoxDHCer species in the indicated strains without or with a nontoxic DoxSa treatment (2 μM of DoxSa, 20 million cells/ml, 1.5 h) determined by MS. The red dashed line indicates the level of C₂₆-DoxDHCer in the CerS3 strain following the treatment; n = 3.

FIGURE 3:

Genome-wide genetic screens to reveal the MoA of DoxSL. (A) Schematic workflow of genome-wide genetic screen of knockout and DAmP mutants. (B) Changes in colony size of the mutants following a DoxSa treatment at the indicated concentrations; n = 2. (C) Schematic workflow of genome-wide genetic screen of transposon-insertion mutants (SATAY). (D) Changes in sequencing read of the mutants following a DoxSa treatment at the indicated concentrations; n = 2. (E, F) Proportions of genes that were in the bottom (E) or top (F) 200 of the fitness ranks and that were revealed by the two genetic screens. There is no GO enrichment in each set of genes.

FIGURE 4:

The dynamics of DoxSL species responsible for the toxicity over the course of DoxSL accumulation. (A) Effect of DoxSa on the growth of the indicated strains evaluated by a spot assay. (B, C) Levels of C₂₆-DoxDHCer (B) and total DoxDHCer (C) in the indicated strains without or with a nontoxic DoxSa treatment (2 μM of DoxSa, 20 million cells/ml, 1.5 h) determined by MS. The red dashed lines indicate the levels of C₂₆-DoxDHCer and total DoxDHCer in the CerS3 strain following the treatment; n = 3.

FIGURE 5:

Cellular phenotypes of DoxSL toxicity. (A) Ability of WT and the CerS3 strain to recover from the indicated DoxSa treatments. (B–E) Cellular phenotypes of DoxSL toxicity. Organization of F-actin (B), shape of mitochondria (C), hydrophobic bodies (D), and the ER membrane (E) in WT and the CerS3 strain following standardized DoxSa treatments (indicated concentrations of DoxSa, 1 million cells/ml, 1.5 h). F-actin was stained with phalloidin-Atto488. Mitochondria were visualized with Mdh1p-mCherry. Hydrophobic bodies were stained with Nile Red. The ER membrane was visualized with Sec61p-GFP. Each image is a maximum intensity projection of a 2-μm slice. The scale bars are 2 μm. (F–I) Size of globular structures stained with phalloidin-Atto488 (F), size of globular structures highlighted by Mdh1p-mCherry (G), size of hydrophobic bodies (H), or circularity (a value of 1.0 indicates a perfect circle) of the ER membrane (I) in WT and the CerS3 strain following the standardized DoxSa treatments. Black bars indicate medians, boxes indicate quartiles, whiskers indicate extreme values, and red dots indicate means. Numbers of images used for the quantification: two to six.

FIGURE 6:

DoxSa accumulation leads to depletion of major membrane lipids. Levels of different categories of phospholipid (A–E) and SL (F–I) in WT and the CerS3 strain following the standardized DoxSa treatments (indicated concentrations of DoxSa, 1 million cells/ml, 1.5 h) determined by MS. PC, phosphatidylcholine; PI, phosphatidylinositol; PS, phosphatidylserine; PE, phosphatidylethanolamine; CL, cardiolipin; Cer, ceramide; IPC, inositol-phosphoceramide; MIPC, mannose-inositol-phosphoceramide; M(IP)₂C, mannose-(inositol-P)₂-ceramide; n = 3.

FIGURE 7:

The rate of DoxSa accumulation determines the level of DoxDHCer accumulation. Levels of C₂₆-DoxDHCer (A) and total DoxDHCer (B) in WT and the CerS3 strain following the standardized DoxSa treatments (indicated concentrations of DoxSa, 1 million cells/ml, 1.5 h) determined by MS; n = 3. (C, D) Levels of C₂₆-DoxDHCer (C) and total DoxDHCer (D) in the CerS3 strain monitored by a time-course lipid analysis by MS following a DoxSa treatment at 6 μM for up to 3 h; n = 2. (E) Effect of DoxSa on the growth of the indicated strains evaluated by a spot assay. Deletion of ceramidase (Δypc1 Δydc1) did not modulate the sensitivity of the CerS3 strain to DoxSa, indicating that the accumulation of DoxDHCer is independent of ceramidase.

FIGURE 8:

Model of the cytotoxicity of DoxSL. Elevated levels of both DoxSa and DoxDHCer disrupt the organization of F-actin, alter the shape of mitochondria, induce the formation of hydrophobic bodies, and cause the ER membrane to collapse. The first three perturbations are not sufficient to cause complete cellular failure. The last perturbation coincides with cell death. This model is consistent with the hypothesis that DoxSL inhibits a multisubunit essential protein affecting multiple pathways. Alternatively, DoxSL could inhibit multiple pathways, directly or indirectly, such as by affecting the physical properties of cellular membranes.