Optimization of gluconeogenesis analysis

We previously showed that the production of hexose 6-phosphates (H6P) from proline by PCF trypanosomes is strongly impaired by the absence of FBPase (Tb927.9.8720) in the Δfbpase mutant cell line and rescued in the Δfbpase/FBPase.i cell line (.i standing for “tetracycline-induced”), which re-expresses an ectopic copy of the FBPase gene in the Δfbpase background using the tetracycline-inducible pLew100 vector [23]. These data, previously described in [23] and included in Fig 2 (columns 4–6, labeling with [U-¹³C]-proline), demonstrated that FBPase is involved in gluconeogenesis. Briefly, this was determined by ionic-exchange chromatography coupled with tandem mass spectrometry (IC–MS/MS) analyses of the incorporation levels of ¹³C atoms from uniformly [¹³C]-enriched proline ([U-¹³C]-proline) into key gluconeogenic intermediates. Incorporation of [¹³C]-enriched atoms from proline into triose phosphates (PEP, 2/3PG, 1,3BPG, and Gly3P, see Fig 1 for abbreviations) and F1,6BP are equivalent in the parental, ∆fbpase, and ∆fbpase/FBPase.i cell lines. However, incorporation of [¹³C]-enriched atoms in F6P (the product of the reaction catalyzed by FBPase), as well as in metabolites produced from F6P, i.e., G6P, mannose 6-phosphate (M6P) and 6-phosphogluconate (6PG), was strongly reduced in the ∆fbpase mutant compared to the parental and rescue cell line, but was not abolished. This suggests the existence of an alternative step to FBPase producing F6P from F1,6BP. However, the residual levels of [¹³C]-incorporation into H6P in the ∆fbpase cell line were too low to investigate these alternatives. To increase the efficiency of [¹³C]-incorporation through gluconeogenesis, [U-¹³C]-proline was replaced by [U-¹³C]-glycerol in the same experimental set-up (Fig 2). Indeed, for the same incubation time in the presence of the ¹³C-carbon source (2h), [U-¹³C]-glycerol is directly incorporated into gluconeogenic intermediates (see Fig 1), whereas incorporation of [U-¹³C]-proline into gluconeogenic intermediates may be significantly delayed by the 9-step metabolic pathway, including 7 mitochondrial steps, leading to PEP from proline [14]. As expected, 91.0 ± 2.2% of F1,6BP in the parental cells was fully [¹³C]-enriched (m6) from [U-¹³C]-glycerol, compared to 48.5 ± 5.9% from [U-¹³C]-proline (Fig 2; F1,6BP; columns 1 versus 4) and incorporation of [U-¹³C]-glycerol in F6P is significantly reduced in the Δfbpase cell line (Fig 2; F6P; columns 1 versus 2). More importantly, the amounts of fully [¹³C]-enriched H6P (m6) extracted from the Δfbpase cell line were 40-times higher with [U-¹³C]-glycerol than with [U-¹³C]-proline (78.3 ± 4.5% versus 2.0 ± 4.0%) (Fig 2). This clearly confirmed that gluconeogenesis occurs in the absence of FBPase and that glycerol is much more efficient to feed gluconeogenesis than proline.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Comparative IC–MS/MS analysis of intracellular metabolites after isotopic labeling with [U-¹³C]-proline or [U-¹³C]-glycerol.

The EATRO1125.T7T parental (WT), Δfbpase and tetracycline-induced Δfbpase/FBPase (Δfbpase/FBPase.i) cell lines were incubated for 2 h in PBS containing 2 mM [U-¹³C]-glycerol (columns 1–3) or 2 mM [U-¹³C]-proline (columns 4-6) prior to metabolite extraction. The figure shows enrichment of key gluconeogenic intermediates at 0 to 6 carbon positions (m0 to m6, with a color code on the side) with [¹³C]-incorporation expressed as percentage of all corresponding molecules (MID; Mass Isotopomer Distribution). The color of the arrows corresponds to metabolism of glycerol (blue), proline (orange) or the two ¹³C-enriched carbon sources (black). The solid and dotted arrows correspond to a single or at least two enzymatic steps. For abbreviations, see Fig 1. 2/3PG means that 2-phosphoglycerate (2PG) and 3-phosphoglycerate (3PG) are undistinguished by IC–MS/MS. The inset shows a western blot analysis of the three cell lines with the anti-FBPase and anti-PPDK (control) immune sera. Data obtained with [U-¹³C]-proline, previously presented in Wargnies and colleagues [23], have been included here for comparison with the labeling with [U-¹³C]-glycerol. Each value corresponds to the mean of three independent experiments (biological replicates) and a t test was performed by assuming equal variances to determine the p-value (^*p < 0.1; ^**p < 0.01; ^***p < 0.001; ^****p < 0.0001; ns, not significant). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g002

PFK is an important contributor to gluconeogenesis

As ATP-dependent PFK from T. brucei (Tb927.3.3270), as well as human, have been recently shown to operate in the gluconeogenic direction in vitro [26], PFK expression was down-regulated by RNAi in the parental and Δfbpase backgrounds. Western blot analyses showed that the tetracycline-dependent conditional expression was leaky since PFK expression was 95% and 98% down-regulated in the absence of tetracycline in the ^RNAiPFK.ni and Δfbpase/^RNAiPFK.ni cell lines, respectively, and after 3 days of induction, PFK expression was at most 1% of that of the parental cell line (Fig 3A). Unlike to the parental, ^RNAiPFK.ni and Δfbpase/^RNAiPFK.ni cells, tetracycline-induced ^RNAiPFK.i and Δfbpase/^RNAiPFK.i cells failed to grow and eventually died in glucose-rich medium (Fig 3B).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. IC–MS/MS analysis of intracellular metabolites extracted from FBPase and/or PFK mutant cell lines.

Panel A shows a western blot analysis of the EATRO1125.T7T parental (WT), Δfbpase, as well as the tetracycline-induced (.i) and non-induced (.ni) ^RNAiPFK (clone A4) and Δfbpase/^RNAiPFK (clone A9) cell lines using the anti-PFK and anti-PPDK (loading control) immune sera. Extracts of 5 × 10⁶ cells have been loaded in all lanes, except lanes 1/10 and 1/50 which correspond to 5 × 10⁵ and 10⁵ cells, respectively. The band corresponding to PFK is highlighted by an arrow and below the figure are the estimated amounts of PFK remaining, relative to WT cells (lane 1 with 5 × 10⁶ cells) based on signal intensity measurements. Panel B shows growth curves of the PCF parental and mutant cell lines, tetracycline-induced or not, in glucose-rich (+Glc) and glucose-depleted (−Glc) conditions. Cells were maintained in the exponential growth phase (between 10⁶ and 10⁷ cells ml⁻¹), and cumulative cell numbers reflect normalization for dilution during cultivation. Similar growth curves were obtained for the ^RNAiPFK clone A8 and Δfbpase/^RNAiPFK clone F8 (S1 Fig). Panel C shows ¹H-NMR analyses of succinate and acetate excreted after metabolism of [U-¹³C]-glucose by the parental (WT) and mutant cell lines incubated in PBS containing 4 mM [U-¹³C]-glucose. A part of each spectrum ranging from 1.6 to 2.6 ppm is shown in the left part. The doublet resonances corresponding to [¹³C]-enriched molecules produced from [U-¹³C]-glucose and the single resonances corresponding to non-enriched molecules produced from an uncharacterized internal carbon source [27] are highlighted by red and blue arrows, respectively. The nature of the single resonance highlighted by an asterisk is unknown. Each spectrum corresponds to one representative experiment from a set of least three whose quantitative data are presented in graph on the right. Each value corresponds to the mean of three independent experiments (biological replicates) and a t test was performed by assuming equal variances to determine the p-value (^*****p < 0.00001). In panel D, the parasites were incubated for 2 h in PBS containing 2 mM [U-¹³C]-glycerol prior to metabolite extraction. The figure shows enrichment of key gluconeogenic intermediates at 0 to 6 carbon positions (m0 to m6, with a color code on the side) with [¹³C]-labeled atoms expressed as percentage of all corresponding molecules (MID; Mass Isotopomer Distribution). The dotted arrows correspond to at least two enzymatic steps. Each value corresponds to the mean of three independent experiments (biological replicates). For abbreviations, see Figs 1 and 2. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g003

To determine the extent of impaired PFK activity in the ^RNAiPFK and Δfbpase/^RNAiPFK cell lines, we monitored glycolysis by quantifying the amounts of excreted [¹³C]-enriched end products from [U-¹³C]-glucose metabolism, which are mainly acetate and succinate, using proton nuclear magnetic resonance (¹H-NMR) spectroscopy [28]. The use of [¹³C]-enriched carbon source allows us to distinguish between the products excreted from metabolism of [U-¹³C]-glucose (doublet resonances labeled by red arrows in Fig 3C) and an uncharacterized internal carbon source (single resonances labeled by blue arrows in Fig 3C) [27]. As expected, the amounts of end products excreted from glycolysis are greatly reduced in the ^RNAiPFK.ni and Δfbpase/^RNAiPFK.ni cell lines, due to the RNAi leakage (Fig 3C). After tetracycline induction, the glycolytic activity is reduced by ~95% in the mutant cell lines, confirming that PFK activity is strongly affected but not abolished.

The role of PFK in gluconeogenesis was determined by IC–MS/MS analyses of the incorporation levels of ¹³C atoms from metabolism of [U-¹³C]-glycerol] into key gluconeogenic intermediates, in the parental, ∆fbpase, ^RNAiPFK (.ni and.i) and ∆fbpase/^RNAiPFK (.ni and.i) cell lines (Fig 3D). Interestingly, incorporation of [¹³C]-enriched atoms in H6P (F6P, M6P, G6P, and 6PG) was strongly reduced in the ^RNAiPFK.i mutant, compared to the parental and ^RNAiPFK.ni cells (Fig 3D). Unexpectedly, gluconeogenesis from glycerol was more affected in the ^RNAiPFK.i mutant than in the Δfbpase mutant. Indeed, the percentage of fully [¹³C]-enriched H6P (m6) from [U-¹³C]-glycerol was lower in the ^RNAiPFK.i mutant than in the Δfbpase mutant (44.6 ± 4.7% versus 67.5 ± 4.8%). In agreement with the involvement of PFK in gluconeogenesis, we also observed a synergistic effect in the Δfbpase/^RNAiPFK.i double mutants, since incorporation of [¹³C]-enriched atoms in H6P was further reduced in the double mutant compared to the single Δfbpase mutant (27.8 ± 13% versus 67.5 ± 4.8% of the H6P are fully [¹³C]-enriched, respectively). In glucose-depleted conditions (−Glc), the doubling time of the ^RNAiPFK.i cell line (18.4 h), which was identical to that of the Δfbpase cell line (18.4 h) [23], was increased compared to the parental cells (15.7 h) (Fig 3B). The doubling time was further increased in the Δfbpase/^RNAiPFK.i cell lines (32.7 h) (Fig 3B), which suggests that the absence of PFK affects growth in the absence of glucose. Taken together, these data clearly showed that PFK participates in gluconeogenesis in PCF trypanosomes incubated in the absence of glucose with a contribution at least as important as that of FBPase. The incorporation of [¹³C]-enriched atoms in H6P of the Δfbpase/^RNAiPFK.i cell line (Fig 3C) and the capacity of this mutant to grow in glucose-depleted conditions (Fig 3B), could be due to residual PFK expression. However, we cannot exclude that an additional enzyme, such as SBPase, is involved in gluconeogenesis.

SBPase is involved in gluconeogenesis with a minor contribution

To address the possible role of SBPase (Tb927.2.5800) in gluconeogenesis, either for its SBPase activity in the PPP or for its possible FBPase activity (see Fig 1A), SBPase alleles have been replaced by the blasticidin (BSD) and puromycin (PAC) resistance markers, respectively, to generate the Δsbpase (Δsbpase::BSD/Δsbpase::PAC) cell line, which was checked by PCR (S2 Fig) and western blotting (Fig 4A, line 3). The [¹³C]-incorporation profile after incubation with [U-¹³C]-glycerol was identical between the parental and Δsbpase cell lines, questioning an SBPase contribution to gluconeogenesis (Fig 4C). In addition, Incorporation of [¹³C] atoms in S7P, the product of SBPase activity, was not affected in the Δsbpase cell line (Fig 4C), suggesting that the putative cycle within the non-oxidative branch of the PPP involving SBPase is not significantly utilized in PCF trypanosomes for production of F6P, via the activity of transaldolase (TAL; see Fig 1A), under the conditions used.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. IC–MS/MS analysis of intracellular metabolites extracted from FBPase, SBPase and/or PFK mutant cell lines.

Panel A shows a western blot analysis of all nine mutants cell lines affecting expression of FBPase, SBPase, PFK, and/or HK using the anti-FBPase, anti-SBPase, anti-PFK, anti-HK, and anti-PFR (loading control) immune sera. For the Δsbpase/Δhk/Δpfk/^RNAiFBPase cell line, tetracycline-induced (.i) and non-induced (.ni) conditions have been analyzed. Lines through gels indicate where intervening lanes were cropped out of the image. Panel B shows growth curves of the tetracycline-induced and non-induced Δsbpase/Δhk/Δpfk/^RNAiFBPase cell lines and the PCF parental cells, in glucose-rich (+Glc) and glucose-depleted (−Glc) conditions. Cells were maintained in the exponential growth phase (between 10⁶ and 10⁷ cells ml⁻¹), and cumulative cell numbers reflect normalization for dilution during cultivation. Each value corresponds to the mean of three independent experiments (biological replicates). In panel C, the parasites were incubated for 2 h in PBS containing 2 mM [U-¹³C]-glycerol prior to metabolite extraction. The figure shows enrichment of key gluconeogenic intermediates at 0 to 7 carbon positions (m0 to m6, with a color code on the side) with [¹³C]-labeled atoms expressed as percentage of all corresponding molecules (MID; Mass Isotopomer Distribution). Each value corresponds to the mean of four independent experiments (biological replicates). The dotted arrows correspond to at least two enzymatic steps. For abbreviations, see Figs 1 and 2: P5P, Pentose 5-phosphate. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g004

To determine whether SBPase contributes even modestly to gluconeogenesis, it was necessary to produce an FBPase/SBPase/PFK triple mutant. This task was complicated by the lethality of the ^RNAiPFK.i cell line in glucose-rich conditions (Fig 3B). The absence of PFK certainly induces the accumulation of G6P and F6P produced by HK and then glucose-6-phosphate isomerase (PGI), respectively, at the expense of glycosomal ATP, which would rapidly become limiting and cause cell death. We can therefore expect that the deletion of the PFK gene in the HK null background to be viable. The T. brucei genomes contain two tandemly arranged HK genes (HK1, Tb927.10.2020 and HK2, Tb927.10.2010), which are 98.3% identical with only 11 aa difference. The HK genes were inactivated in the EATRO1125.T7T background using the marker-free CRISPR-Cas9 approach we recently developed for trypanosomes, by introducing in all HK alleles a short fragment containing stop codons and a BamHI restriction site (Fig 5A) [29]. After transfection, single cells were isolated and clonal populations were screened for the insertion of repair template by PCR and BamHI digestion. The PCR fragment amplified from the selected Δhk cell line was fully cleaved by BamHI, indicating that all HK alleles were inactivated by the repair cassette (Fig 5B), as confirmed by western blot analyses (Fig 5C). Abolition of glycolysis in the Δhk cell line was monitored by quantifying the amounts of excreted [¹³C]-enriched and non-enriched end products from metabolism of [U-¹³C]-glucose and proline, respectively, using ¹H-NMR spectroscopy (Fig 5D). Indeed, [¹³C]-enriched acetate, succinate and alanine were no longer detectable in the Δhk spent medium, while excretion of non-enriched acetate, succinate and alanine from the metabolism of proline remained the same as in the parental cell line. Then, the PFK alleles were successfully inactivated using the same CRISPR-Cas9 approach in the Δhk background (Δhk/Δpfk), as confirmed by western blot analyses (Fig 5B and 5C). As expected, the doubling time of glycolysis-compromised Δhk and Δhk/Δpfk cell lines remained the same in the proline containing PCF culture medium with or without glucose, while the parental cell line grew faster in the presence of glucose (Fig 5E).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. HK is not essential for growth and protects against the absence of PFK.

Panel A shows a schematic representation of the strategy used to inactivate the HK genes by inserting a short sequence containing a succession of stop codons flanking a BamHI restriction site. The dot in the guide RNA sequence corresponds to the Cas9 cleavage site, indicated by a vertical bar at position 509 bp in the HK genes. The 50-bp flanking sequences allowing repair through homology-directed repair are shown in gray in the HK genes and the sequence of the repair cassette. Transfection with the ribonucleoprotein complex, composed of recombinant Streptococcus pyogenes Cas9 (SpCas9), the in vitro-synthesized guide RNA and the repair cassette, results in rapid and efficient genetic modifications of the parasites, in marker-free conditions [29]. After transfection, the cells were cloned using a cell sorter in 96 well plates and the growing clonal populations were screened by PCR. The resulting edited HK genes and the position of the oligonucleotides used to screen clones by PCR are shown underneath. Panel B shows the PCR confirmation of inactivation of all HK alleles in the selected Δhk and Δhk/Δpfk clones. PCR products were analyzed on agarose gel after digestion with BamHI, allowing easy discrimination of gene inactivation on all alleles, by the BamHI digestion of the 668-bp PCR fragment into the 359-bp and 309-bp fragments only in the Δhk cell lines. Confirmation of HK genes inactivation by western blot analysis with the anti-HK, anti-PFK, and anti-PFR (loading control) is shown in panel C. Panel D shows ¹H-NMR analyses of [¹³C]-enriched and non-enriched end products (succinate, acetate and alanine) excreted from the metabolism of [U-¹³C]-glucose and proline, respectively, by the parental (WT) and Δhk cell lines incubated in PBS containing 4 mM [U-¹³C]-glucose and 4 mM proline. A part of each spectrum ranging from 1.2 to 2.6 ppm is shown. The doublet resonances corresponding to [¹³C]-enriched molecules produced from [U-¹³C]-glucose and the single resonances corresponding to non-enriched molecules produced from proline are highlighted by red and blue arrows, respectively. Panel E shows growth curves of the PCF parental, Δhk and Δhk/Δpfk cell lines, in the proline containing PCF culture medium with also (+Glc) or without glucose (−Glc). Cells were maintained in the exponential growth phase (between 10⁶ and 10⁷ cells ml⁻¹), and cumulative cell numbers reflect normalization for dilution during cultivation. Each value corresponds to the mean of three independent experiments (biological replicates). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g005

To generate the SBPase/PFK, SBPase/FBPase and FBPase/SBPase/PFK mutants, the marker-free CRISPR-Cas9 gene inactivation and RNAi approaches were used. Briefly, the HK and/or PFK genes have been inactivated in the SBPase null background to generate the Δsbpase/Δhk and Δsbpase/Δhk/Δpfk cell lines, in which the pLew100-RNAi-FBPase construct was introduced to generate the Δsbpase/Δhk/^RNAiFBPase and Δsbpase/Δhk/Δpfk/^RNAiFBPase cell lines, respectively. The incorporation profile of [¹³C]-enriched atoms in G6P, M6P, S7P, and R5P is similar in the Δfbpase/^RNAiPFK.i and Δsbpase/Δhk/Δpfk/^RNAiFBPase.i cell lines (Fig 4C, lanes 8 and 10, respectively), confirming that the contribution of SBPase is minor or nonexistent.

Importantly, all our attempts to generate the Δsbpase/Δhk/Δpfk/Δfbpase quadruple mutant using the CRISPR approach failed, probably because this mutant is not viable whatever the growing conditions. Indeed, if FBPase, PFK and SBPase are the only ways to produce F6P from F1,6BP, the Δsbpase/Δhk/Δpfk/Δfbpase cell line should not be able to produce the essential metabolite G6P from glucose, glycerol or proline, the main carbon sources used by the parasite. To test this hypothesis, we developed an assay to quantify the efficiency of mutant generation of the FBPase gene, using a CRISPR-Cas9-based approach, in the parental EATRO1125.T7T, Δhk and Δsbpase/Δhk/Δpfk cell lines. After transfection with the ribonucleoprotein complex containing the guide RNA and the BleR repair cassette, which confers resistance to phleomycin, cells were cloned in four 96-well plates in glycerol-rich medium supplemented with 5 µg ml⁻¹ phleomycin and the growing clonal populations were all screened by PCR with primers flanking the FBPase gene (Fig 6A). As expected, inactivation of the FBPase gene in the parental and Δhk cell lines was highly effective, since of the 72 and 44 phleomycin-resistant clones selected, 67 and 42 had both FBPase alleles inactivated (homozygous clones), respectively (Fig 6B–6D). The absence of FBPase expression was confirmed by western blot analyses (Fig 6C). The few other selected clones are heterozygous, with a single inactivated FBPase allele, as confirmed by PCR using one primer in the BleR gene and the other flanking the FBPase gene (Fig 6A and 6B), and western blot analyses (Fig 6C). In contrast, only five clones have been selected from the transfected Δsbpase/Δhk/Δpfk cell line, all heterozygous, with FBPase still expressed from the active FBPase allele (Fig 6B–6D). Assuming that gluconeogenesis is essential in the HK null background, the failure to obtain homozygous Δsbpase/Δhk/Δpfk/Δfbpase cell lines confirmed that FBPase, PFK and also SBPase are all involved in gluconeogenesis, with no other alternative for F6P production from F1,6BP.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. SBPase is the only possible alternative to FBPase and PFK for gluconeogenesis.

Panel A shows a schematic representation of the strategy used to inactivate the FBPase alleles by inserting the phleomycin resistance marker (BleR) flanked by 5′ and 3′ untranslated regions in the FBPase alleles. The dot in the guide RNA sequence corresponds to the Cas9 cleavage site, indicated by a vertical bar at position 357 bp in the FBPase gene. After transfection of the parental EATRO1125.T7T, Δhk and Δsbpase/Δhk/Δpfk cell lines with the ribonucleoprotein complex, the cells were cloned using a cell sorter in six 96-well plates (384 wells) in the presence of 5 µg ml⁻¹ phleomycin and 10 mM glycerol before screening the growing clonal populations by PCR. The resulting edited FBPase gene and the position of the primers used to screen clones by PCR are shown underneath. 561-bp (1) and 1428-bp (2) PCR fragments were obtained, with primers flanking the insertion site, for the native and inactivated FBPase alleles, respectively. To control the correct insertion of the repair cassette, primers located in the BleR gene and downstream of the insertion site were used to generate an 880-bp PCR fragment (3), which is present in all the analyzed phleomycin-resistant clones (see panel B). Panel B shows the PCR analysis of all clones containing a single inactivated allele (heterozygotes), while only three Δsbpase and Δhk/Δsbpase clones are shown among the numerous homozygous ones (both alleles deleted) obtained for the parental and Δhk transfections. PCR products obtained with the primers flanking the insertion site (1 and 2 in panel A) and the PCR control (3) were analyzed on agarose gel. Panel C confirms FBPase gene inactivation on one or two alleles, in selected clones, by western blot analysis with the anti-HK, anti-SBPase, anti-PFK, anti-FBPase, and anti-PPDK (loading control) immune sera. Panel D recapitulates for each transfection the number of plated cells (Total), of phleomycin-resistant clones (BleR), of cells containing 0 (Homozygous) or 1 (Heterozygous) functional FBPase allele and the percentage of BleR clones containing 1 functional FBPase allele (% Hetero). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g006

FBPase expression is upregulated under glucose-depleted conditions and in the absence of PFK or SBPase

Since gluconeogenesis is required in the absence of glycolysis, the parental, Δfbpase, ^RNAiPFK.i, and Δsbpase cell lines have been grown under glucose-depleted and glucose-rich conditions to compare the expression of FBPase, PFK, and SBPase, quantified by western blot analyses (Fig 7). FBPase expression is significantly increased by 1.4-fold in glucose-depleted conditions compared with glucose-rich conditions in the parental PCF cell line. Interestingly, this upregulation in glucose-depleted conditions was further significantly increased in the ^RNAiPFK.i and Δsbpase mutants (2.9- and 1.8-fold compared to WT grown in glucose-rich conditions, respectively), which is in agreement with the contribution of PFK and SBPase in gluconeogenesis. In contrast, FBPase is not upregulated in the ^RNAiPFK.i and Δsbpase mutants grown in glucose-rich conditions and expression of PFK and SBPase was not affected under the different conditions and in the genetic backgrounds analyzed (Fig 7).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Expression of FBPase, PFK, and SBPase under different growth conditions and in different genetic backgrounds.

Panel A shows a western blot analysis of the EATRO1125.T7T parental (WT), Δfbpase, ^RNAiPFK.i, and Δsbpase cell lines cultured under glucose-rich (+G) and glucose-depleted (−G) conditions during 3 days, using the anti-FBPase, anti-PFK, anti-SBPase, and anti-PFR (loading control) immune sera. In panel B, the corresponding specific signals after scanning the bands from 3 to 5 experiments are presented. The obtained quantitative values were normalized with the control values (PFR) and expressed as the level of expression relative to the parental cells grown under glucose-rich conditions (WT +G). Each value corresponds to the mean (±SD) of 3–5 independent experiments (biological replicates) and a Student t test was performed by assuming equal variances to determine the p-value (^*p < 0.1; ^**p < 0.05; ^***p < 0.01; ^****p < 0.005). nd: not detectable. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g007

Glycosomal metabolite content under different carbon source conditions

In this context, it is important to determine how the glycosomal environment under gluconeogenic conditions can be understood in view of the known kinetic properties of PFK. In other words, do the gluconeogenic products of PFK (F6P and ATP) significantly drop compared to its substrates (F1,6BP and ADP) in proline-rich/glucose-free conditions? Indeed, the kinetic properties of T. brucei PFK imply that, to catalyze its reverse reaction, the PFK gluconeogenic substrates (F1,6BP and ADP) have to be approximately 500-fold more abundant within the glycosomes than its products (F6P and ATP) [26]. To address this question, we determined the total intracellular concentrations of metabolites by IC-HRMS by adding an internal standard ([U-¹³C]-labeled Escherichia coli extract) into the T. brucei cell extracts, as described before [12]. The F1,6BP/F6P ratio was 2.5 times lower under proline conditions than under glucose conditions, whereas one would expect a high increase of this ratio to force PFK to work in the direction of gluconeogenesis in glucose-depleted conditions (Fig 8). We consider that these measured amounts reflect the glycosomal contents. Indeed, F1,6BP and F6P are primarily produced in the glycosomes and may equilibrate with the cytosolic pool through the glycosomal pores, but their possible conversion through the cytosolic glycolytic activity has probably no impact on the steady-state amounts of gluconeogenic intermediates, as previously estimated [30]. However, the observed unchanged ADP/ATP ratio is not informative. Indeed, the total intracellular amounts of ATP do not reflect the glycosomal situation, since the net production of ATP mainly occurs in the mitochondrion of the PCF trypanosomes and ATP/ADP exchange between the glycosomes and the other subcellular compartments is unlikely [30].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Analysis of intracellular metabolites.

Intracellular concentrations of metabolites in PCF trypanosomes grown in SDM79 medium containing 10 mM glucose (Glucose) or not (Proline). The relative amounts of intracellular metabolites in proline conditions are expressed as a percentage of that in glucose conditions fixed at 100%. The sequence of the metabolic reactions is represented in linear form underneath the figure, with the entry of glucose and proline indicated by arrows. The boxed and underlined metabolites are the gluconeogenic substrates and products of PFK, respectively. Each value corresponds to the mean of three independent biological replicates. For abbreviations see Fig 1A: Cit, citrate. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g008

To estimate the impact of different carbon sources on the glycosomal ATP levels, we used an ATP-specific fluorescence resonance energy transfer (FRET)-based indicator, named ATeam, that is composed of a bacterial F_oF₁-ATP synthase ε subunit sandwiched between a cyan- and a yellow-fluorescent protein (CFP and YFP, respectively) [31]. In the ATP-bound form, the ε subunit retracts to bring the two fluorescent proteins close to each other, which increases FRET efficiency and allows detection of changes in ATP level upon fluorescence quantification (Fig 9A). To focus on the glycosomal ATP levels, the ATeam cassette was fused to the N-terminal extremity of a Myc-tagged glycosomal protein containing a C-terminal peroxisomal targeting signal (PTS1), i.e., glycerol-3-phosphate dehydrogenase (GPDH; EC 1.1.1.8; Tb927.8.3530), which was recently also used to target a cytosolic protein exclusively into this organelle [32]. The resulting ATeam-Myc-GPDH protein was conditionally expressed in the presence of tetracycline [33]. Immunofluorescence analyses showed that the ATeam-Myc-GPDH recombinant protein was located in glycosomes (Fig 9B). We previously showed that FRET efficiency was significantly reduced and the CFP fluorescence lifetime significantly increased in the ^OEATeam-Myc-GPDH.i cells incubated in the presence of glycerol compared to glucose conditions, due to accumulation of Gly3P at the expense of glycosomal ATP [33] (Fig 9C and 9D). However, FRET and CFP fluorescence lifetime were similar in proline and glucose conditions (Fig 9C and 9D), suggesting that the glycosomal ATP level is not affected by glucose depletion.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. Estimation of glycosomal ATP levels.

Panel A shows a schematic drawing of the ATeam probe adapted from [31]. Variants of CFP (mseCFP) and YFP (cp173-mVenus) were connected by the ε subunit of Bacillus subtilis F_oF₁-ATP synthase. In the ATP-free form (top), extended and flexible conformations of the ε subunit separate the two fluorescent proteins, resulting in a low FRET efficiency. In the ATP-bound form, the ε subunit retracts to draw the two fluorescent proteins close to each other, which increases FRET efficiency. In panel B, the glycosomal localization of ATeam-Myc-GPDH was confirmed by immunofluorescence assays on the ^OEATeam-Myc-GPDH.i cell line by using an anti-aldolase immune serum as a glycosomal marker. The yellow CFP signal was converted to green in order to merge it with the red fluorescence corresponding to aldolase. Panel C shows a comparative analysis of the ratio of YFP emission (FRET) and CFP emission after excitation at 435 nm in the ^OEATeam-Myc-GPDH.i cell line incubated in the presence of 5 mM of glucose (Glc), glycerol (Glyc) or proline (Pro) (means ±SD, n = 2 independent experiments, ****p-value associated with the Student t test is < 0.0001). Panel D shows a comparative analysis of the CFP fluorescence lifetime of the same cell line incubated in the same conditions (means ±SD, n = 3 independent experiments, ****p-value associated with the Student t test is < 0.0001). Panel E shows a schematic drawing of the MaLion probe adapted from [34]. The ε subunit of B. subtilis F_oF₁-ATP synthase was inserted into the green fluorescent protein (GFP) and the red fluorescent mApple protein (RFP) to design the single fluorescence protein-based ATP indicators, named MaLionG and MaLionR, respectively. In the ATP-bound form, the ε subunit retracts and induces a conformational change of the GFP and RFP, which becomes fluorescent. Panels F and G show the glycosomal and cytosolic localization of MaLionR-Myc-GPDH and MaLionG by immunofluorescence assays on the ^OEMaLionR-Myc-GPDH.i and ^OEMaLionG cell lines, respectively, by using an anti-aldolase immune serum as a glycosomal marker. Expression of MaLionR-Myc-GPDH in the tetracycline-induced cell line (.i) was confirmed in Panel H by western blotting using the anti-GPDH immune serum. Panels I and J shows fluorescence analyses of the ^OEMaLionR-Myc-GPDH.i and ^OEMaLionG.i mutants (two cell lines for each), respectively, maintained during several days in glucose-depleted conditions, washed and then incubated during 30 min in PBS containing or not (No carbon source) 5 mM of glucose, glycerol or proline before microscopic analyses (means ± SD, n = 3 independent experiments and a Student t test was performed by assuming equal variances to determine the p-value, ^*p < 0.1; ^**p < 0.05; ^***p < 0.01; ^****p < 0.005; ns, not significant). The obtained quantitative values were expressed as the level of expression relative to the “No carbon source” condition. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.15148560.

https://doi.org/10.1371/journal.pbio.3002938.g009

To confirm these data, we used also another ATP sensor, i.e., MaLion, based on a single fluorescent protein exhibiting an increase in fluorescence emission at a single wavelength upon sensing ATP [34]. The MaLion sensor is composed of a bacterial ε subunit of F_oF₁-ATP synthase inserted into a fluorescent protein (Fig 9E). The conformational changes induced by the ATP-bound ε subunit increase GFP or RFP fluorescence as a function of ATP concentration. The MaLionR-Myc-GPDH recombinant protein and the MaLionG protein expressed in the ^OEMaLionR-Myc-GPDH.i and ^OEMaLionG.i cell lines were used to monitor glycosomal and cytosolic ATP, respectively. The glycosomal and cytosolic localization of MaLionR-Myc-GPDH and MaLionG, respectively, was confirmed by immunofluorescence analyses (Fig 9F and 9G) and tetracycline-dependent expression of MaLionR-Myc-GPDH was confirmed by western blotting using the anti-GPDH immune serum (Fig 9F). Fluorescence emission was measured in cells incubated in PBS containing glucose, glycerol, proline or no carbon source. The fluorescence intensity, which reflects the amounts of ATP, is not affected by the absence of carbon source compared with glucose conditions. This is probably due to the maintenance of the cellular ADP/ATP ratio when an unknown endogenous carbon source is used in the absence of external carbon sources [27], with a considerable reduction in metabolic fluxes. As observed with the ^OEATeam-Myc-GPDH.i cell line, two ^OEMaLionR-Myc-GPDH.i cell lines showed a strong reduction of glycosomal ATP under glycerol conditions, while incubation with proline induced a significant increase in glycosomal ATP compared with glucose conditions (Fig 9I). However, as expected, cytosolic ATP levels remained relatively stable whatever the incubation conditions (Fig 9J). Taken together, these surprising data suggest that PFK converts F1,6BP to F6P in the presence of relatively high levels of ATP, which is not in agreement with what is predicted to occur free in solution, with the known equilibrium constant of the PFK reaction and the enzyme’s known kinetic properties [35].

Trypanosomes and cell cultures

The PCF of T. brucei EATRO1125.T7T (TetR-HYG T7RNAPOL-NEO), which constitutively expresses the T7 RNA polymerase gene and the tetracycline repressor under the control of a T7 RNA polymerase promoter for tetracycline inducible expression [56], was cultivated at 27 °C in the presence of 5% CO₂ in SDM79 medium containing 10% (v/v) heat-inactivated fetal calf serum and 3.5 mg ml⁻¹ hemin [57]. Alternatively, a glucose-depleted medium derived from SDM79, named SDM79-GlcFree, was developed to cultivate the procyclic cells [23]. This SDM79-GlcFree medium consists of glucose-depleted SDM79 medium, containing 20% (v/v) heat-inactivated fetal calf serum, in which the EATRO1125.T7T cells were cultured during 72 h to consume the glucose coming from the serum and then diluted with the same volume of glucose-depleted SDM79 medium without serum to finally obtain SDM79-GlcFree. Glucose depletion was verified by NMR spectrometry analyses (with a detection threshold ≤1 μM) and to prevent import of residual glucose, 50 mM N-acetyl-d-glucosamine, a non-metabolized glucose analogue inhibiting glucose transport, were added to the medium. The glucose-rich and glycerol-rich conditions were obtained by adding to SDM79-GlcFree 10 mM glucose (without N-acetyl-d-glucosamine) or 10 mM glycerol, respectively.

Gene knockout

Generation of the ∆fbpase (EC 3.1.3.11; Tb927.9.8720) null mutant and rescue ∆fbpase/FBPase cell line was described before [23]. Replacement of both alleles of the SBPase gene (EC 3.1.3.11; Tb927.2.5800) by the BSD and PAC resistance markers via homologous recombination was performed with DNA fragments containing a resistance marker gene flanked by the SBPase UTR sequences. Briefly, the BSD or PAC orf were PCR amplified with primers containing 50-bp upstream (5′ primer) or downstream (3′ primer) of the ATG or stop codons, respectively, of the SBPase gene. The EATRO1125.T7T parental cell line was used to generate the SBPase knockout cell line (Δsbpase::BSD/Δsbpase::PAC, called ∆sbpase). After transfection, selection was done in SDM79 medium containing hygromycin (25 μg ml⁻¹), neomycin (10 μg ml⁻¹), blasticidin (10 μg ml⁻¹), and/or puromycin (1 μg ml⁻¹).

Gene inactivation by CRISPR/Cas9 inactivation

Gene inactivation was achieved by inserting double-stranded DNA (repair cassette) corresponding either to a short sequence (TAGCTAAGTGAggatccTGAATAATTAG) containing six successive stop codons in the three reading phases (underlined in the short sequence above) and a BamHI restriction site (italic letters in the short sequence above) or to a resistance marker (phleomycin, BleR), as recently described [29]. The repair cassettes were also flanked by 50-bp homologous to the 5′ and 3′ sequences of the Cas9 cut site. To inactivate the HK (EC 2.7.1.1; Tb927.10.2010 and Tb927.10.2020), PFK and FBPase genes, the RNA guides used are GAGCAGACGAAGGTTAACCGTGG (ARNg-HK_134), GTTGATACTCTCGAGCGCCCGG (ARNg-PFK_94), and ATATTACTGCTGCCGTCCAGAGG (ARNg-FBPase-374_rev), respectively. Prior to transfection, equal quantities (0.4 µmol each) of TracrRNA and gRNA from Integrated DNA Technologies (IDT) were combined. The mixture was then heated at 95 °C for 5 min and gradually cooled to room temperature. Subsequently, this mixture was combined with 30 µg of SpCas9 sourced from IDT (Alt-RTM SpCas9 Nuclease 3NLS, Catalog #1074181). 5 × 10⁵ cells were respectively transfected using Amaxa nucleofector device with Basic Parasite Nucleofector Kit 2 (Lonza) (100 µl) with 1 µg of purified repair cassette (in a final volume of 10 µl) and 30 µg of Cas9 protein preloaded with the mixture of TracrRNA and gRNA. Cells were transfected using program X-001 and resuspended in 5 ml of medium containing or not 5 µg ml⁻¹ phleomycin. Following a 2-day culture period, a portion of the transfected cells underwent DNA extraction using the NucleoSpin Blood kit (Macherey-Nagel) followed by PCR amplification using primers that flank the Cas9 cleavage site. Gel analysis of the PCR products is expected to show two bands corresponding to the wild-type and modified gene. Cells were subsequently cloned using a cell sorter (TBM Core facility), and the selection of inactivated cells was performed by DNA extraction using the NucleoSpin Blood kit (Macherey-Nagel) followed by PCR amplification using primers flanking the Cas9 cleavage site as previously described. Guide RNAs were designed using EuPaGDT [58], from http://tritrypdb.org. Primers and guide RNAs used were synthesized by IDT. All genetic modifications were confirmed by sequencing the PCR product (Eurofins) generated from primers flanking the modifications.

Inhibition of gene expression by RNAi

RNAi-mediated inhibition of FBPase gene expression was performed in the TetR-HYG and T7RNAPOL-NEO background by expression of stem-loop “sense-antisense” RNA molecules of the targeted sequences [56,59] using the pLew100 expression vector, which contains the phleomycin resistance gene (kindly provided by E. Wirtz and G. Cross) [60]. Production of the pLew-FBPase-SAS plasmid and the ^RNAiFBPase cell lines was described before [23]. To generate the ∆sbpase/∆hk/^RNAiFBPase and ∆sbpase/∆hk/∆PFK/^RNAiFBPase mutants the ∆sbpase/∆hk and ∆sbpase/∆hk/∆pfk cell lines were transfected with the NotI-linearized pLew100-FBPase-SAS plasmid in 2-mm electroporation cuvettes with Basic Parasite Nucleofector Kit 2 on the Amaxa Nucleofector device (Lonza), using X-001 program. Selection of the mutant cell lines was performed in glucose-rich SDM79 medium containing hygromycin (25 µg ml⁻¹), neomycin (10 µg ml⁻¹), puromycin (1 µg ml⁻¹), blasticidin (10 µg ml⁻¹), and phleomycin (5 µg ml⁻¹).

To down-regulate expression of the PFK gene (Tb927.3.3270) by RNAi, we produced the pLew-PFK-SAS plasmid, which is the pLew100 vector containing a sense and antisense version of the targeted 736-bp fragment (from position 578 to 1314 of the PFK gene), separated by a 67-bp fragment, under the control of a PARP promoter linked to a prokaryotic tetracycline operator. Briefly, a PCR-amplified 775-bp fragment, containing the antisense PFK sequence with restriction sites added to the primers, was inserted into the HindIII and BamHI restriction sites of the pLew100 plasmid. A 736-bp PCR-amplified fragment containing the sense PFK sequence was then inserted upstream of the antisense sequence, using HindIII and XhoI restriction sites (XhoI was introduced at the 3′-extremity of the antisense PCR fragment). After transfection of the EATRO1125.T7T and ∆fbpase cell lines with the NotI-linearized pLew-PFK-SAS plasmid, as described above, selection of the ^RNAiPFK and ∆fbpase/^RNAiPFK mutants was performed in glucose-rich SDM79 medium containing hygromycin (25 µg ml⁻¹), neomycin (10 µg ml⁻¹), and phleomycin (5 µg ml⁻¹), plus puromycin (1 µg ml⁻¹) and blasticidin (10 µg ml⁻¹) for the ∆fbpase/^RNAiPFK cell lines. Aliquots of all selected mutant cell lines were frozen in liquid nitrogen to provide stocks of each line that had not been cultivated long term in medium. Induction of RNAi cell lines was performed by addition of 10 µg ml⁻¹ tetracycline.

Expression of ATeam and MaLion recombinant proteins

To express in glycosomes the ATeam cassette, composed of the ε subunit of the F_oF₁-ATP synthase sandwiched by the CFP and YFP [31], the full-length GPDH gene (Tb927.8.3530) encoding the PTS1-containing glycosomal glycerol-3-phosphate dehydrogenase was inserted downstream of the ATeam cassette to generate a gene encoding the glycosomal ATeam-Myc-GPDH recombinant protein, as described before [33]. Briefly, a PCR-amplified 1833-bp fragment containing the ATeam sequence flanked by the HindIII and NdeI restriction sites, respectively, was inserted into the HindIII and NdeI restriction sites of the pLew100 vector containing the GPDH gene preceded by 10 Myc tag sequences. The resulting pLew-ATeam-Myc-GPDH plasmid was introduced in the EATRO1125.T7T parental cell line and clones were selected with phleomycin (5 µg ml⁻¹), in addition to hygromycin and neomycin.

To generate the pLew-MaLionR-Myc-GPDH plasmid, the ATeam cassette (HindIII-NdeI fragment) of the pLew-ATeam-Myc-GPDH plasmid was replaced by a PCR fragment containing MaLionR gene flanked by the HindIII and NdeI restriction sites, PCR amplified from the MaLionR plasmid (#113908, Addgene). To generate the pLew-MaLionG plasmid, a PCR fragment containing MaLionG gene (#113906, Addgene) flanked by the HindIII and BamHI restriction sites was inserted in the HindIII/BamHI sites of pLew100. The resulting pLew-MaLionR-Myc-GPDH and pLew-MaLionG plasmids were introduced in the EATRO1125.T7T parental cell line and clones were selected with phleomycin (5 µg ml⁻¹), in addition to hygromycin and neomycin.

Western blot analyses

Total protein extracts of PCF of T. brucei (5 × 10⁶ cells) were separated by SDS-PAGE (10%) and immunoblotted on TransBlot Turbo Midi-size PVFD Membranes (BioRad) [61]. Immunodetection was performed as described [61,62] using as primary antibodies, the rabbit anti-FBPase (1:1000), the rabbit anti-SBPase (1:250), the rabbit anti-PFK (1:200), the rabbit anti-PFR (1:10,000), the rabbit anti-HK (1:5000), the rabbit anti-GPDH (1:1000) [63], the rabbit anti-PPDK (1:1000) [64], and mouse anti-Myc 9E10 (αMyc; 1:100; gift from K. Ersfeld, Hull, UK). Anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase (Bio-Rad, 1:5000 dilution) was used as secondary antibody. Revelation was performed using the Clarity Western ECL Substrate as described by the manufacturer (Bio-Rad). Images were acquired and analyzed with the ImageQuant LAS 4000 luminescent image analyzer.

NMR spectrometry experiments

2 × 10⁷ procyclic T. brucei were centrifuged at 1400g for 10 min, then the pellet was washed twice with PBS and the cells were incubated for 6 h at 27 °C in 1 ml of incubation buffer (PBS supplemented with 5 g l⁻¹ NaHCO₃, pH 7.4) with 4 mM [U-¹³C]-glucose and with or without 4 mM of non-enriched proline. The viability of the cells during the incubation was checked by microscopic observation. At the end of the incubation, 500 µl supernatant were collected and 50 μl of maleate solution in D₂O (10 mM) was added as internal reference. ¹H-NMR spectra were performed at 500.19 MHz on a Bruker Avance III 500 HD spectrometer equipped with a 5 mm cryoprobe Prodigy. Measurements were recorded at 25 °C. Acquisition conditions were as follows: 90° flip angle, 5,000 Hz spectral width, 32 K memory size, and 9.3 s total recycle time. Measurements were performed with 64 scans for a total time close to 10 min 30 s. Resonances of the obtained spectra were integrated and metabolites concentrations were calculated using the ERETIC2 NMR quantification Bruker program [65]. To quantify glucose-derived acetate and succinate, we used the following procedure. Protons linked to acetate carbon C2 generates by ¹H-NMR five resonances, a single peak ([¹²C]-acetate) flanked by two doublets ([¹³C]-acetate). When [U-¹³C]-glucose is the only carbon source, the central resonance (1.85 ppm) corresponding to [¹²C]-acetate, probably derived from an unknown internal carbon source. As a consequence, it is not included in the glucose-derived acetate quantitative analyses.

Mass spectrometry analyses of [¹³C]-incorporation into cellular metabolites

For analysis of [¹³C]-incorporation into intracellular metabolites, EATRO1125.T7T parental and mutant cell lines grown in SDM79 medium were washed twice with PBS and resuspended in incubation solution (PBS containing 2 mM [U-¹³C]-glycerol or 2 mM [U-¹³C]-proline). The cells were incubated for 2 h at 27 °C before being collected on filters by fast filtration preparation for mass spectrometry analysis, as described before [12]. Total sampling time was below 8 s and the extraction of intracellular metabolites was carried out by transferring the filters containing the pellets into 5 ml of boiling water for 30 s. The extracts were briefly vortexed (~2 s), immediately filtered (0.2 µm), and chilled with liquid nitrogen. After lyophilization, the dried extracts were resuspended in 200 µl Milli-Q water prior to analysis. Three replicates were taken from each culture media, sampled and analyzed separately. Metabolites were analyzed by ionic-exchange chromatography coupled with tandem mass spectrometry (IC–MS/MS) using the method described by Bolten and colleagues [66]. Retention time on the column and multiple reactions monitoring transition of each analyzed metabolite are shown in Table 1 of [67]. The ¹³C mass isotopomer distribution of intracellular metabolites was determined from relevant isotopic clusters in the IC–MS/MS analysis, according to Kiefer and colleagues [68]. ¹³C mass isotopomer distribution measurements were performed using a triple quadrupole mass spectrometer (4000Qtrap, Applied Biosystems). To obtain ¹³C-labeling patterns (¹³C isotopologues), isotopic clusters were corrected for the natural abundance of isotopes other than ¹³C, using the in-house software IsoCor (available at MetaSys) [69].

Determination of intracellular metabolite concentrations by IC-HRMS and enzymatic assays

For the quantification of intracellular metabolites, the T. brucei PCF EATRO1125.T7T cell line maintained 2 h in PBS containing 10 mM glucose or 10 mM proline was sampled by fast filtration, as described above. The extraction of intracellular metabolites was performed as mentioned above before adding 200 µl of a uniformly [¹³C]-labeled E. coli cell extract as internal quantification standard [70]. Then, metabolites were analyzed by IC-HRMS as described above with measured concentrations of metabolites expressed as a total cellular concentration assuming a volume of 10⁸ cells being equal to 5.8 µl [37].

Immunofluorescence analyses

Cells were washed twice with PBS and fixed with 4% paraformaldehyde for 10 min at room temperature, spread on slides and permeabilised with 0.05% Triton X-100. After incubation in PBS containing 4% BSA overnight, cells were incubated for 45 min with anti-aldolase rabbit serum (Aldo; 1:1000). After washing with PBS, samples were incubated for 45 min with a secondary anti-rabbit IgG antibody conjugated to Alexa Fluor 594 (ThermoFisher). Slides were washed and mounted with SlowFade Gold (Molecular Probes). Images were acquired with the Metamorph software on a Zeiss Imager Z1 or an Axioplan 2 microscope as previously described [71].

Fluorescence intensity ratio measurements

The microscopy experiments were performed at the Bordeaux Imaging Center (BIC) on an inverted Leica DMI 6000 microscope (Leica Microsystems, Wetzlar, Germany) equipped with a resolutive HQ2 camera (Photometrics, Tucson, USA). The illumination system used was a Lumencor spectra 7 (Lumencor, Beaverton, USA). The objective used was a HCX PL APO CS 63X oil 1.32 NA. This system was controlled by MetaMorph software (Molecular Devices, Sunnyvale, USA). Cells were both observed in differential interference contrast mode by transmission and fluorescence microscopy (CFP excitation/CFP emission for donor acquisition, CFP excitation/YFP emission for FRET acquisition). We quantified YFP (FRET) and CFP emission using the ImageJ software (NIH, USA).

Fluorescence lifetime imaging microscopy (FLIM)

The FLIM measurements were performed at the BIC with the Lambert Instrument FLIM Attachment (LIFA, Lambert Instrument, Roden, the Netherlands) that allows the generation of lifetime images by using the frequency domain method. This system consists of a modulated intensified CCD camera Li² CAM MD, a modulated light excitation light source and a modulated GenIII image intensifier. For widefield epi-illumination, modulated LED (Light-Emitted-Diode) was used at 451 nm (3 W) for CFP excitation. Both the LED and the intensifier were modulated at frequency up to 100 MHz. A series of 12 images was recorded for each sample. By varying the phase shifts (12 times) between the illuminator and the intensifier modulation we calculated the phase and modulation for each pixel of the image. Consequently, we determined the sample fluorescence lifetime image using the manufacturer’s software LI-FLIM software. Lifetimes were referenced to a solution of erythrosin B (1 mg ml⁻¹) set at 0.086 ns [72].