Research ArticleCELL BIOLOGY

Escherichia coli as a platform for the study of phosphoinositide biology

See allHide authors and affiliations

Science Advances  27 Mar 2019:
Vol. 5, no. 3, eaat4872
DOI: 10.1126/sciadv.aat4872
  • Fig. 1 Synthesis of PIP2 in bacteria.

    Metabolic pathway for phospholipid production in E. coli and construct design for the expression of the required enzymes. (A) E. coli phospholipid metabolic pathway and designed PIP2 synthesis pathway. CDP-DAG, cytidine diphosphate diacylglycerol. (B) Map of the construct for the expression of the enzymes required for PIP2 synthesis. RBS, ribosome binding site.

  • Fig. 2 E. coli growth during phosphoinositide production.

    Effect of the different constructs on E. coli growth and synthesis of phosphorylated phosphoinositides. (A) Effects of inositol on the growth rate of E. coli expressing the different constructs. The data presented are the maximum growth rate observed for each construct during the exponential growth phase in LB media at 37°C. Error bars show the SD of three replicates. (B) Production of PI4P and PIP2 for the optimized constructs reported in this study. Cells were grown for 3 hours in LB media with 2 mM inositol. Data shown are the average of four independent measures with error bars showing the SD of the sample.

  • Fig. 3 Phosphoinositides and phospholipids in E. coli.

    Characterization of the changes in phosphoinositides and overall phospholipids produced in E. coli. (A) Time course of phosphorylated phosphoinositide production. Cells expressing the enzymes PIS, PI4K, and PI4P5K were assayed for PIP2 and PI4P after the addition of 2 mM inositol. The amount of culture pelleted was adjusted to obtain a pellet equivalent to 10 ml of culture at an optical density at 600 nm (OD600) = 1 for lipid extraction. Trend curves are from a second-order polynomial. Data shown are the average of four independent measures with error bars showing the SD of the sample. (B) Phosphoinositide production under different inositol concentrations. Cells expressing the biosynthetic enzymes PIS, PI4K, and PI4P5K were supplemented with varying concentrations of inositol. At 3 hours, the cell concentration was adjusted to obtain a pellet equivalent to 10 ml of culture at OD600 = 1 for lipid extraction. Trend curves are from an exponential fit of the form y = a + be(−c x). Data shown are the average of four independent measures with error bars showing the SD of the sample. (C) Change in the abundance of major phospholipid categories in E. coli upon production of phosphoinositides. The percentage of each phospholipid group is illustrated. For this experiment, cells were grown in LB media supplemented with different concentrations of inositol. PA, phosphatidic acid; PG, phosphatidyl glycerol; PE, phosphatidyl ethanolamine.

  • Fig. 4 PI4P and PIP2 localization.

    Localization of the PI4P- and PIP2-binding domains. The subcellular distribution of PI4P as assayed by the GFP-tagged SidM-P4M domain and of PIP2 as assayed by the GFP-tagged PLCδ-PH domain relative to cytosolic mCherry. (A) Line scans were 10 pixels wide (566 nm) and are indicated by the yellow line on the overlay image. Data from the line scans were standardized to show the same amplitude on all channels. Note the aggregation of protein toward the pole in some cells, in a way that excludes the cytoplasmic red marker, suggesting the formation of inclusion bodies even at the low level of expression used. Scale bar, 2 μm. (B) Subcellular distribution was quantified with a bimodality index analysis (see Materials and Methods) for the localization of PI4P in 40 cells for each construct. A one-way analysis of variance (ANOVA) indicated that treatments were different with P < 0.0001, and a Tukey honest significant difference (HSD) test showed that this is due to the construct where the PI4P-synthesizing enzymes are used in the presence of inositol as indicated in the figure. Error bars show the SD of the sample. An individual representative image for each treatment is provided in fig. S4A. (C) Subcellular distribution was quantified with a bimodality index analysis (see Materials and Methods) for the localization of PIP2 in 40 cells for each construct. A one-way ANOVA indicated that treatments were different with P < 0.0001, and a Tukey HSD test showed that this is due to the construct where the PIP2 synthesizing enzymes and the wild-type reporter are used in the presence of inositol as indicated in the figure. Error bars show the SD of the sample. An individual representative image for each treatment is provided in fig. S4B.

  • Fig. 5 Export of FGF2 and FGF4.

    Tests for the export of FGF2 and FGF4. (A) Results for FGF2 export with PIP2. A three-way ANOVA shows that each factor (presence of phosphoinositide synthesizing enzymes, amount of inositol, and FGF2 construct) is significant with P < 0.05 and that the interactions of inositol × synthesizing enzymes and FGF2 construct × synthesizing enzymes are also significant. Data shown are the average and SD of the sample (n = 4). Results of a Tukey HSD for comparison of the samples are indicated in the figure. The mCherry results were excluded from the ANOVA to avoid artificially raising the relevance of the construct for the results. (B) Results for FGF4 export with PIP2. A three-way ANOVA shows that each factor (presence of phosphoinositide synthesizing enzymes, amount of inositol, and FGF4 construct) is significant with P < 0.05 and that the interaction of inositol × synthesizing enzymes is also significant. Data shown are the average and SD of the sample (n = 4). Results of a Tukey HSD for comparison of the samples are indicated in the figure. The mCherry results were excluded from the ANOVA to avoid artificially raising the relevance of the construct for the results. (C) Export change in luminescence index of FGF2 constructs depending on the phosphoinositides present. The comparison was performed as an independent experiment comparing export with no inositol or 5 mM inositol in the media. Data shown are the average and SD of the sample (n = 3). Raw data are shown in fig. S6.

  • Table 1 Plasmids used in this study.

    All plasmids, as well as their complete sequences, are available through Addgene under the numbers indicated in the last column.

    Plasmid namePromoter usedProteins expressedAddgene
    Plasmid type 1: Origin of replication P15A, 10 to 12 copies per cell, chloramphenicol resistance
    p15aC-GFPproDmyc-GFP (control)107862
    p15aC-1DproDmyc-PIS107863
    p15aC-4D1DDual proDmyc-PIS and myc-PI4K107864
    p15aC-1D-5proDmyc-PIS-RBS-myc-PI4P5K (as operon)107865
    p15aC-4D1D-5Dual proDmyc-PIS-RBS-myc-PI4P5K (as operon) and myc-PI4K107866
    p15aC-3D1D-5Dual proDmyc-PIS-RBS-myc-PI4P5K (as operon)
    and PI3K
    121050
    Plasmid type 2: Origin of replication CloDF13, 20 to 40 copies per cell, ampicillin resistance
    pClodAcytCh-PIP2PHGFPDual proDmCherry and GFP-PLCδ-PH domain107867
    pClodAcytCh-PIP2PHGFP-nbDual proDmCherry and GFP-PLCδ-PH domain nonbinding mutant107868
    pClodAcytCh-GFP-P4MDual proDmCherry and GFP-(2×)SidM-P4M
    domain
    113346
    pClodANL-CherryproDmCherry-nanoLuc107870
    pClodANL-FGF2proDFGF2wt-nanoLuc107871
    pClodANL-FGF2YEproDFGF2phosphomimic-nanoLuc107872
    pClodANL-FGF2NBAAproDFGF2minimum export-nanoLuc107873
    pClodANL-FGF4proDFGF4wt-nanoLuc107874
    pClodANL-FGF4exporterproDFGF4exporter-nanoLuc107875
    Plasmid type 3: Origin of replication pUC, ~500 copies per cell, ampicillin resistance (used during optimization)
    pC51T7-proDmyc-PI4P5K-RBS-myc-PIS (as operon)107869

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/3/eaat4872/DC1

    Fig. S1. Abnormal morphology observed.

    Fig. S2. Detection of PI4P.

    Fig. S3. Calculation of the bimodality index.

    Fig. S4. Representative images for the bimodality tests.

    Fig. S5. Pilot experiment on FGF2 export.

    Fig. S6. FGF2 export depending on the phosphoinositides present.

    Table S1. Lipid species detected.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Abnormal morphology observed.
    • Fig. S2. Detection of PI4P.
    • Fig. S3. Calculation of the bimodality index.
    • Fig. S4. Representative images for the bimodality tests.
    • Fig. S5. Pilot experiment on FGF2 export.
    • Fig. S6. FGF2 export depending on the phosphoinositides present.
    • Table S1. Lipid species detected.

    Download PDF

    Files in this Data Supplement:

Navigate This Article