Fig 1.
Morphology of LDs in uninfected and C. trachomatis-infected HeLa cells.
(A) A fresh suspension of LDs isolated from HeLa cells was stained with BODIPY 493/503 for 15 min and imaged by fluorescence microscopy (left panel) or processed for TEM (right panel). LDs appear as intact spherical structures. Since no morphological differences between LDs from uninfected and infected cells were observed, a representative image of LDs isolated from uninfected cells is shown (B) Representative confocal images of HeLa cells, uninfected (UI) or infected with C. trachomatis LGV-L2 for 20 h (20 hpi) or 40 h (40 hpi). Cells were stained with BODIPY 493/503 to visualize LDs (green) and Hoechst was utilized to visualize host and bacterial DNA (blue). Bacterial replicative vacuoles (inclusions) are indicated with white arrows. (C) TEM images of uninfected (left panel) and C. trachomatis infected (20 hpi, right panel) HeLa cells. LDs (asterisk) and C. trachomatis (black arrows) are indicated. Note the close apposition of LDs with mitochondria and ER-like membranes (dashed square, left panel) and with the inclusion (right panel). In all cases (A-C) HeLa cells were stimulated with oleic acid 100 μM for 14 h prior to fixation and/or staining to enhance LD production.
Table 1.
List of proteins represented abundantly in LD extracts from uninfected and C. trachomatis-infected HeLa cells.
Fig 2.
Cellular component analysis of LD-associated proteins.
(A) All LD-associated proteins identified and quantified in this study were classified based on their gene ontologies (GO) by “cellular component” using UniProtKB. The quantitative contribution of each protein group (average across all samples) was calculated and expressed as percent of the total mass of LD-associated proteins. Only the quantitatively more abundant categories are represented (see S5 Table for a detailed list). Because a given protein can be classified in more than one GO category, there is some degree of overlap between the groups and thereby the sum of all categories exceeds one hundred percent. The quantitative contribution of proteins belonging to the “lipid droplet” category (GO: 0005811) as well as that corresponding to proteins “shared” and “not shared” between the “lipid droplet” and the other specified categories are indicated. (B, C) Venn diagrams show the number of overlapping proteins between “lipid droplets” and other selected cellular components.
Fig 3.
Immunoblot analysis of selected proteins identified in LD-extracts by proteomics.
LD protein extracts were obtained from HeLa cells uninfected (UI) or C. trachomatis-infected for 20 h (20 hpi) and 40 h (40 hpi). Protein content was assessed by Bradford assay, equal amounts per lane were separated by SDS-PAGE and transferred to nitrocellulose membranes. The relative abundance of a subset of proteins associated to LD-extracts was assessed by immunoblot densitometry analysis (using ImageJ software) and compared with quantitative changes obtained from mass spectrometry (MS)-based quantification (expressed as the ratio, infected/uninfected). Proteins were considered “increased” or “decreased” when the Inf/Uninf ratio was > 1.5 or < 0.67 (this ratio representing a fold change greater or equal to -1.5), respectively, as shown in the “Enrichment” columns. Host proteins assessed were: ACSL3 (long-chain-fatty-acid-ligase 3); ACSL4 (long-chain-fatty-acid-ligase 4); NSDHL (sterol-4-alpha-carboxylate 3-decarboxylating); PLIN2 (perilpin-2); Rab1 (ras-related protein Rab-1a); Rab11 (ras-related protein Rab-11b); CALR (calreticulin). C. trachomatis RpoB (RNA polymerase subunit beta) was used as a marker of infection. A representative image is shown. N.A., “not applicable”.
Fig 4.
LDs isolated from C. trachomatis-infected cells are enriched in proteins related to lipid metabolism.
(A) The expression levels of all proteins belonging to “lipid metabolic processes” (GO: 0006629) increased in LD fractions at 40 hpi are represented. Bars represent the mean (fmol μg-1) resulting from six mass spectrometry-based determinations (as detailed in supporting information S3 Table). Error bars indicate the standard deviation. Asterisks (***) indicate statistically significant differences (p < 0.001, two-tailed T-test). Fold changes (40 hpi/Uninfected, average across 6 replicates) are indicated. (B) Venn diagram summarizing the overlap between the protein groups “Lipid metabolic processes” (GO: 0006629), “Lipid droplet” (GO: 0005811) and “Lipid biosynthetic processes” (GO: 0008610) identified in this study (see S5 Table for a detailed list). (C) Proteins were grouped into the indicated GO categories and the total mass for each category was calculated. Bars represent the mean (fmol μg-1) resulting from six MS-based determinations. Error bars represent the standard deviation. Asterisks (***) indicate statistically significant differences (p < 0.001, two-tailed T-test). ACSL3 (long-chain-fatty-acid-ligase 3); PCAT1 (lysophosphatidylcholine acyltransferase 1); NSDHL (sterol-4-alpha-carboxylate 3-decarboxylating); PLIN2 (perilpin-2); DHB11 (estradiol 17-beta-dehydrogenase 11); ACSL4 (long-chain-fatty-acid-ligase 4); DHDDS (dehydrodolichyl diphosphate synthase); PCAT2 (lysophosphatidylcholine acyltransferase 2); LPCT4 (lysophospholipid acyltransferase LPCAT4).
Fig 5.
C. trachomatis inclusion membrane proteins Cap1 and CTL0882 co-localize with LDs.
Representative confocal images of HeLa cells transfected with constructs encoding Cap1 or CTL0882 C-terminally fused to EGFP or N-terminally fused to Cherry, stimulated with oleic acid 200 μM for 8 h prior to fixation. LD labeling was done by either immunostaining of the LD structural protein PLIN3 (red channel) or by neutral lipid staining with BODIPY 493/503 (green channel). Hoechst staining was used to visualize DNA (blue). Bar sizes are shown in the merge panels.
Fig 6.
LDs associate with Cap1, CTL0882 and IncG at the inclusion membrane.
Representative confocal images of HeLa cells infected with C. trachomatis (40 hpi), stimulated with oleic acid (100 μM, 14h prior to fixation). Immunofluorescence staining of Cap1, CTL0882 and IncG was carried out (red channel) and LDs were visualized using the neutral lipid-specific BODIPY 493/503 staining (green channel). Hoechst staining was used to visualize DNA (blue). Bar sizes are shown in the merge panels. Arrowheads highlight LDs found in close apposition with Cap1, CTL0882 and IncG in the inclusion membrane, as indicated.
Fig 7.
C. trachomatis-infected cells display increased amounts of LDs.
(A) Representative fluorescence microscopy images of HeLa cells uninfected or infected with C. trachomatis (20 hpi or 40 hpi) and stimulated or not with the indicated micromolar concentrations of oleic acid (OA) for 14h prior to fixation, as specified. For visualization of LDs, neutral lipid-specific BODIPY 493/503 staining was used (green). Chlamydial inclusions are highlighted (white arrowheads). Bar sizes (20 μm) are indicated. (B) Fluorescence microscopy-based quantification of LD content was carried out using MBF-ImageJ software (as described in Materials and Methods) by measuring the total area of BODIPY 493/503 positive structures expressed as arbitrary units (a. u.) per cell. At least 30 cells per condition were analyzed in three independent experiments. Mean and SEM are represented. Asterisks indicate statistically significant differences (p<0.001) as determined by ANOVA-Bonferroni post-test.
Fig 8.
Production of C. trachomatis infectious progeny is impaired in cells devoid of LDs.
(A) Representative fluorescence microscopy pictures of wild type MEF (wt), the diacylglycerol-acyltransferases 1 and 2 double knock-out (dgat1/2-/-), dgat1/2-/- complemented with empty vector [dgat1/2-/- (Control)] and dgat1/2-/- complemented with DGAT1 [dgat1/2-/- (DGAT1)], stimulated or not with oleic acid (OA) 400 μM for 14 h, as indicated. For visualization of LDs, neutral lipid-specific BODIPY 493/503 staining was used (green). Hoechst staining was used to visualize DNA (blue). Bar sizes (20 μm) are shown. Note that dgat1/2-/- cells are devoid of LDs even under stimulation with OA, whereas LD formation is restored in dgat1/2-/- (DGAT1). (B) Impact of impaired LD biogenesis on chlamydial replication was evaluated by measuring bacterial yields expressed as inclusion forming unit per microliter (IFU/μl) in IFU assays, as detailed in Materials and Methods. Compared to wt cells, chlamydial replication was reduced in dgat1/2-/- whereas increased chlamydial yields were observed in dgat1/2-/- (DGAT1) compared to dgat1/2-/- (Control). Bars represent the mean and standard deviation. Data is representative of at least 3 independent experiments. Asterisks indicate statistically significant differences (***p<0.001; **p<0.01) as calculated by unpaired, two-tailed T-test. Input IFUs (mean ± standard deviation) were very similar for all cell types: wt, 19104 ± 1207; dgat1/2-/-, 18758 ± 1008; dgat1/2-/- (Control), 18691 ± 1163; dgat1/2-/- (DGAT1) 18484 ± 1128.