Mitochondria of a human skeletal muscle cell line associate with the parasitophorous vacuoles of three archetypal lineages of Toxoplasma gondii and influence fatty acid import

Abstract

Toxoplasma gondii, an obligate intracellular parasite, acquires host nutrients to sustain its intracellular replication. A key interaction involves host mitochondrial association (HMA) with the parasitophorous vacuole membrane, previously thought to be strain- and cell-type-dependent, and notably absent in type 2 strains in fibroblasts. Here, we report that in human skeletal muscle KD3 myotubes, all three archetypal T. gondii strains—including type 2—demonstrate significant HMA. This association was confirmed by mitotracker staining and transmission electron microscopy. Notably, HMA appears to correlate inversely with the parasite’s uptake of exogenous ¹³C-labeled fatty acids, suggesting a competitive nutrient environment shaped by host mitochondrial proximity. These findings highlight host cell-type specificity in mitochondrial interactions and suggest that HMA may function as a modulator of nutrient acquisition in a context-dependent manner. This work revises the understanding of strain-specific HMA and underscores the complexity of host-parasite metabolic interactions in muscle tissue, a physiologically relevant niche for chronic T. gondii infection.

This study reveals that all major Toxoplasma gondii strains associate with mitochondria in human muscle cells, challenging previous beliefs and suggesting this is a host-dependent strategy for acquiring nutrients.

Introduction

The obligate intracellular parasite Toxoplasma gondii uses host cell nutrients for growth and survival in virtually all nucleated cells of warm-blooded animals (Blume and Seeber 2018). During the active invasion process tachyzoites form their proliferation niche within a parasitophorous vacuole (PV) by invaginating the host cell plasma membrane but largely excluding host proteins. Properties of this PV membrane (PVM) are imparted by secretory parasite proteins and include permeability to small molecules (Schwab et al. 1994, Gold et al. 2015) and association with particular host organelles. The host endoplasmic reticulum, the Golgi apparatus, the centriole, and mitochondria have been shown to associate with the PV membrane (PVM) (Coppens and Romano 2018) (Romano et al. 2013). Host mitochondrial adhesion (HMA) to the PVM has been discovered originally in macrophages (Jones and Hirsch 1972) and further investigated in human foreskin fibroblasts. In these host cells HMA is parasite strain-dependent and considered to be absent in type 2 but to be present in type 1 and type 3 parasites (Dubey et al. 1998, Pernas et al. 2014). TgMAF1b is considered to be the parasites effector protein mediating HMA. It is mainly expressed in type 1 parasites but not type 2 tachyzoites when cultured in HFF cells. Consistently, absence of MAF1b abolished HMA in RH type 1 parasites and its heterologous expression in Me49 type 2 parasites conferred HMA (Pernas et al. 2014, Adomako-Ankomah et al. 2016). Its paralog MAF1a is expressed in all three isotypes but has an unknown function (Adomako-Ankomah et al. 2016). The corresponding host mitochondrial receptor proteins TOM70, the essential host mitochondrial chaperone HSPA9 and the GAP domain of Ral GTPase Accelerating Protein α1 were identified by a combination mutagenesis, proteomic, RNAi, and yeast two hybrid screening approaches (Blank et al. 2021, Powell et al. 2024).

The significance of HMA has been suggested to include cytokine modulation (Pernas et al. 2014), a growth advantage during acute toxoplasmosis (Adomako-Ankomah et al. 2016), and an increase in cyst burden during the chronic phase of the infection (Blank et al. 2021). Using a fluorescent fatty acid analog, mitochondrial fusion around the PV has also been suggested to limit parasite fatty acid uptake by nutrient competition and restricting growth of type 1 RH parasites (Pernas et al. 2018). Mitofusin 1 and 2 mediate mitochondrial fatty acid import and have been implicated in this nutritional host defense (Li et al. 2022). It remains unclear, if and how this mechanism applies to type 2 HMA-negative parasites and whether HMA is also strain-dependent in other host cell types.

Materials and methods

Host cells and parasite culture

All cultures were maintained in a 36.6°C humidified CO2 (10%) incubator as described previously (Christiansen et al. 2022).

Briefly, BJ-5ta human foreskin fibroblast (ATCC CRL-4001 HFF) monolayers were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco) supplemented with 25 mM glucose (Sigma–Aldrich), 4 mM l-glutamine (Thermo Fisher Scientific), 1 mM sodium pyruvate (Capricorn Scientific), 100 U/ml penicillin, 100 μg/ml streptomycin (Thermo Fisher Scientific), and 10% heat-inactivated bovine serum (Capricorn Scientific).

The immortalized human myoblast cell line KD3, initially derived from normal subcutaneous female muscle tissue, was cultured in DMEM supplemented with 25 mM glucose, 4 mM l-glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 μg/ml streptomycin, 2% Ultroser G (Cytogen GmbH), and 20% heat-inactivated fetal bovine serum (FBS) (Capricorn Scientific). The differentiation of myoblasts into myotubes was induced at 70% confluency by changing medium to DMEM supplemented with 25 mM glucose, 4 mM l-glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 μg/ml streptomycin, 2% horse serum (HOS) (Capricorn Scientific), 10 µg/ml human insulin (Sigma–Aldrich), 5 µg/ml human holo-transferrin (PAN Biotech), and 1.7 ng/µl sodium selenite (Sigma–Aldrich) for 5–7 days (Maus et al. 2024).

RH (Sabin 1941), Pru-Δhxgprt tdTomato (Pru-tdTomato) (Chtanova et al. 2008), and Type III NED (Darde et al. 1992) were maintained in vitro in HFF monolayers grown in DMEM with 25 mM glucose, 4 mM L-glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 μg/ml streptomycin, and 1% heat-inactivated FBS (tachyzoite medium) (Christiansen et al. 2022). Freshly egressed parasites were passaged by transfer to new HFF monolayers.

Thin-section electron microscopy

Confluent HFF cells and differentiated KD3 myotubes were infected with parasites for 24 h. Cells were fixed in a mixture of 1% paraformaldehyde (Sigma–Aldrich) and 2.5% glutaraldehyde (Sigma–Aldrich) in 0.05 M HEPES buffer (pH 7.4) for 3 h at room temperature and were embedded in plastic as described in (Christiansen et al. 2022). Briefly, cells were scraped from the flask or dish bottom, embedded in agarose, and post-fixed in osmium tetroxide, tannic acid, and uranyl acetate. After dehydration in ethanol, cells were embedded in Epon resin. Thin sections (60–70 nm) were contrasted and analyzed with a transmission electron microscope operated at 120 kV. Images were recorded with a 1k CCD camera (Megaview III, EMSIS, Germany) using the montaging option of the software (iTEM, EMSIS, Germany).

Immunofluorescence assay

Uninfected host cells were grown in monolayers on 12 mm round glass coverslips (TPP) and infected with T. gondii tachyzoites for 24 h. Mitotracker^TM Deep Red FM (Thermo Fisher Scientific) was applied for 30 min at 200 nM in D1 medium at 37°C in a humidified CO2 incubator. Cultures were chased for 15 min in culture media and washed three times with PBS, fixed with 4% paraformaldehyde, and mounted on microscopy slides (Thermo Fisher Scientific) in Fluoromount-G (Sigma–Aldrich) containing DAPI (1:3 000) (Thermo Fisher Scientific). Monochromatic images were recorded on a Zeiss Apotome Imager equipped with a Plan-Fluor 63x/1.45 oil M27 objective. For MAF1a and MAF1b stains, samples processed as described before (Christiansen et al. 2022). Briefly, fixed samples were neutralized in 100 mM glycine/PBS for 5 min, permeabilized for 20 min in 0.2% TX100/PBS, blocked in 2%BSA/0.2%TX100/PBS, and stained in blocking solution for 1 h at room temperature. Polyclonal mouse primary antibodies (αTgMAF1RHa1 and αTgMAF1RHb1 (Adomako-Ankomah et al. 2016) were used at 1:500 dilutions. Images were recorded on a Zeiss Observer using a Plan-Fluor 63x/1.4 oil M27 objective. Images other than MAF1a and MAF1b stains were imported into ImageJ version 1.52a for coloring, the generation of overlays, and quantification. MAF stains were processed in Zeiss Zen 3.4 blue edition. The quantification of HMA was done manually on blinded samples by counting the number of PVs with apparent HMA.

Fatty acid uptake experiments

Host cells were grown in T60 dishes and infected with T. gondii parasites at MOI 2 for 24 h. Algal U-¹³C-fatty acids (Cambridge Isotope Laboratories) were coupled to BSA and added until host lysis was imminent. All parasites were labeled equally long. Monolayers were quenched in ice-cold PBS and washed three times, before syringe release and filtration through 3 µm filters. Total lipids were extracted as described previously (Blume et al. 2015) in biphasic chloroform, methanol, and water (Carl Roth), dried, and transmethylated in chloroform, methanol, and trifluoromethylphenyltrimethylammonium hydroxide (TCI international) in a ratio of t2:1:0.015. FAMES were analyzed on a GCMS with high-efficiency ion source (5977B, 7890B Agilent) equipped with a 10 m VF-5 ms guard column and 30 m VF-5 ms separation column. Data analysis and natural abundance correction were done in DExSI (Dagley and McConville 2018) and GraphPad PRISM 9.1.0.

Results

Mitotracker stain indicates accumulation of host mitochondria at the parasitophorous vacuoles of the three archetypal T. gondii strains in KD3 myotubes but not human fibroblasts

We recently established KD3 human skeletal muscle cells a host cell type that enables the long-term maturation of tissue cysts of the main lineages of T. gondii. To characterize the replication niche of tachyzoites in these myotubes, we infected them with type 1 RH, type 2 Pru, and type 3 NED tachyzoites and stained with mitotracker highlighting both, parasite and host mitochondria (Fig. 1a). Surprisingly, we observed that three strains exhibited host mitochondria in the immediate vicinity of their PVs, as indicated by prominent mitotracker stain, including to Pru parasites. To exclude parasite isolate-specific reasons for this unexpected behavior, we tested the HMA behavior under previously reported conditions and monitored mitochondria in infected human foreskin fibroblasts (Fig. 1b). Mitotracker staining reveals no apparent adhesion of host mitochondria to the PV of Pru type 2 parasites but to the PVs of type 1 RH and type 3 NED parasites (Fig. 1b). Manual quantification of PVs with associated host mitochondria revealed that virtually all vacuoles of type 1 and type 3 parasites and 70% of type 2 parasites exhibited HMA (Fig. 1c). This indicates that HMA of Pru parasites is not as extensive and may be qualitatively and functionally different. Subsequently, we investigated the expression of MAF1b and MAF1a in myotube-cultured Pru and RH parasites by indirect immunofluorescence microscopy (Supplementary Fig. S1). MAF1b was clearly expressed at the vacuole of RH parasites (Supplementary Fig. S1A) but absent from Pru parasites, as previously reported for HFF-grown type 2 parasites. MAF1a was expressed in both parasites in a granular and potentially secreted form (Supplementary Fig. S1B) consistent with previously described localization (Adomako-Ankomah et al. 2016). Together, these data suggest that factors other than MAF1b are responsible for the apparent association of the host mitochondria with the PVM in KD3 myotubes.

Figure 1.

Adhesion of KD3 myotube mitochondria to the vacuoles of T. gondii. (a) Merged images of KD3 myotubes infected for 24 h with RH (type 1), Pru (type 2), and NED (type 3) parasites. Mitochondria were stained with MitoTracker^TM Deep Red FM and DNA with DAPI. Fluorescent channels were merged with bright-field images. Scale bar indicates 10 µm. (b) Merged images of human foreskin fibroblasts infected for 24 h with RH (type 1), Pru (type 2), and NED (type 3) parasites. Mitochondria were stained with Mitotracker and DNA with DAPI. Fluorescent channels were merged with bright-field images. Scale bar indicates 10 µm. (c) Quantification of mitochondrial adhesions from three independent experiments with at least 15 parasitophorous vacuoles per experiment. Shown is the percentage of parasitophorous vacuoles (PV) exhibiting host mitochondrial adhesion. ***P < 0.001, Mann–Whitney test.

Transmission electron microscopy confirms association of host mitochondria with the parasitophorous vacuole membrane in KD3 myotubes

To confirm the putative associations between KD3 mitochondria and the PVM of type 2 parasites, we performed thin-section electron microscopy of Pru parasite-infected KD3 myotubes and used RH parasites as controls (Fig. 2a). Indeed, the analysis revealed a close contact between the membranes of the PVMs and host mitochondria. In contrast, in infected human fibroblasts HMA was restricted to RH parasites (Fig. 2b).

Figure 2.

Thin-section electron microscopy indicates host mitochondrial adhesion of type 2 Pru parasites. Sections through (a) KD3 myotubes and (b) human foreskin fibroblasts infected with type 1 RH or type 2 Pru parasites for 24 h. While in the KD3 myotubes, both parasite strains induce an association of mitochondria with the membrane of the parasitophorous vacuole, in human foreskin fibroblast, only the RH parasite induce this association. Scale bars indicate 3 µm. Arrows indicate host mitochondria and squares indicate enlarged area shown below.

Fatty acid uptake correlates with host mitochondria association status in HFF and KD3 cells

The adhesion of host mitochondria has been suggested to influence the uptake of fatty acids (Pernas et al. 2018). To test whether the HMA status of the three T. gondii lineages in HFF and KD3 myotube cells correlated with fatty acid import, we infected both cell types with parasites of the three isotypes for 24 h and exposed intracellular parasites to BSA-coupled universally labeled algal ¹³C-fatty acids (Fig. 3) until just before egress. GCMS analysis of FAMES from total lipids detected major fatty acids and significant label incorporation in all three parasite strains. Interestingly, all fatty acids except FA20:0 and FA20:4 exhibited higher labeling in Me49 type 2 parasites in HFF cells (Fig. 3a), while in KD3 myotubes all parasite strains showed comparable label incorporation (Fig. 3b). We also observe a generally higher label incorporation into myotube-grown parasites. The data are consistent with a negative effect of HMA on the parasites access to exogenous fatty acids through the host cells and highlight host cell type-specific import activity of exogenous fatty acids.

Figure 3.

Fatty acid uptake. Uptake of U-¹³C fatty acids into T. gondii tachyzoite strains grown in (a) human foreskin fibroblasts or (b) KD3 myotubes during 24 h as measured by GCMS. Shown is the % of ¹³C label incorporation into pools of fatty acid molecules. In HFF cells fatty acid labeling of Me49 parasites differs significantly from RH and NED parasites except for FA20:0 and FA20:4 according to a Mann–Whitney test. *, ** and *** indicate at least P < 0.05, P < 0.005, and P < 0.0005, in respectively, in the pairwise comparisons between Me49 with RH and NED strains. Data represent the mean with S.E.M. of two independent experiments with 4 biological replicate cultures in total.

Discussion

The recruitment of host mitochondria to the replication niche of is a feature of many intracellular pathogens (Maurice and Sadikot 2023), including T. gondii (Pernas et al. 2018; Li et al. 2022). While this feature was previously, thought to be absent in type 2 parasites (Pernas et al. 2014), we show here that in human skeletal KD3 myotubes type 2 Pru parasites indeed exhibit HMA. Muscle cells are a natural host cell type of T. gondii, in which differentiates into its bradyzoite form to persist in various host organisms (Weiss and Kim 2013). In contrast to HFF cells, HMA in myotubes does not occur on all vacuoles, instead we only find 70% of Pru PVs associated with host mitochondria. This is consistent with the presence of a distinct underlying mechanism. Further, HMA in RH strain-infected cells appears to occur around most of the circumference of the PV by long mitochondria. In contrast, host mitochondria of Pru strain-infected cells do appear shorter in TME analysis and cover smaller surface of the PV as revealed by fluorescence microscopy. Although we did not quantify this observation, it is again consistent with the presence of an alternative adherence mechanism. Toxoplasma gondii gene expression differs strongly across different host cells, including skeletal muscle cells (Swierzy et al. 2017), and hence additional parasite proteins that are absent in fibroblast cultures tachyzoites may be involved. Absence of TgMAF1b is considered the underlying reason for HMA-incompetence of type 2 parasites (Adomako-Ankomah et al. 2016). We tested for its expression in myotube-cultured Pru parasites by indirect immunofluorescence using TgMAF1b-specific antibodies but found it only to be expressed in RH parasites (Adomako-Ankomah et al. 2016). We also tested whether MAF1a, that does not confer HMA in HFF cells is expressed in myotube-grown parasites. Indeed, we find this isoform to be expressed in both RH and Pru parasites, and note expression patterns that are consistent with a previous report (Andomako-Ankomah et al. 2016). Interestingly, MAF1a appears to line the PV membrane of Pru parasites. Whether this indicates a tethering activity, remains to be investigated by reverse genetics and more comprehensive imaging. Other alternative HMA-conferring factors include TgMAF1RHb0 or ROP2. The latter is known to adhere to both the host ER and the mitochondria outer membrane when heterologously expressed in CHO cells (Sinai and Joiner 2001).

We further show that HMA correlates with fatty acid import. In particular, HMA appears to suppress fatty acid uptake in HFF cells in type 1 and type 3 parasites. However, we cannot exclude that different proliferation rates of the tested strains also impacts fatty acid uptake. We estimate that this likely represents a minor influence, as the rapidly replicating RH strain does not exhibit the highest ¹³C incorpration. Vice versa, it remains to be seen how this apparent impairment of nutrient uptake impacts parasite proliferation. To delineate potential non-HMA factors such as proliferation rate and host cell fatty acid metabolism, fatty acid import will need to be compared between non-HMA mutants and corresponding wild-type parasites.

The parasite also harbors fatty acid synthesis capabilities, that we did not monitor here, but that might complement their import to ensure adequate supply in different environments (Walsh et al. 2022). These anabolic pathways include the FAS2 pathway (Mazumdar et al. 2006; Ramakrishnan et al. 2012) and fatty acid elongases (Ramakrishnan et al. 2015). Interestingly, it has been suggested that HMA serves as part of an innate immune response suppressing tachyzoite growth by limiting fatty acid uptake (Pernas et al. 2018). In contrast, exogenous oleic acid that is readily imported into all three strains through both HFF and KD3 myotubes has been shown to also exert antiproliferative effects when available in high doses (Nolan et al. 2018). Another important fatty acid, on which we, however, did not focus on here, is lipoic acid (LA). Imported LA may be sourced from the host mitochondrion (Crawford et al. 2006) and is an essential cofactor for the parasites mitochondrial branched-chain amino acid dehydrogenase (BCKDH) in the mitochondrion (Oppenheim et al. 2014), while the apicoplast harboring the functionally largely equivalent pyruvate dehydrogenase is supplied with de novo synthesized fatty acid (Crawford et al. 2006).

Together, our findings underscore the importance of the host cell context in shaping parasite-host interactions, revealing that mitochondrial association is not a fixed trait of parasite genotype but a dynamic interface influenced by host cell type with implications for nutrient acquisition. Future studies may explore how HMA influences host immune responses, delineate the molecular mechanisms governing strain- and cell-type-specific mitochondrial recruitment, and clarify the extent to which this interaction modulates parasite nutrient uptake and proliferation dynamics across different host environments.

Conflict of interest

None declared.

Funding

M.B. and C.C. were funded by the Federal Ministry of Education and Research (BMBF) under project number 01KI1715 as part of the “Research Network Zoonotic Infectious Diseases.” M.B. and M.L. receive internal support from the Robert Koch Institute.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.