Incomplete tricarboxylic acid cycle and proton gradient in Pandoravirus massiliensis: is it still a virus?

Giant viruses of amoebae challenge traditional virus definitions due to their large particle sizes and genomes, the latter encoding functions once thought exclusive to cellular life (e.g., translation components, defense systems). Pandoraviruses are particularly remarkable with genomes rich in ORFans. The study asks whether Pandoravirus massiliensis exhibits a hallmark of autonomous life: the ability to produce or maintain energy gradients. Given that energy generation in biology is often associated with proton gradients across membranes, the authors investigated whether P. massiliensis virions possess a membrane potential and whether viral genes may encode components related to the tricarboxylic acid (TCA) cycle that could contribute to such a gradient. They report detection of a proton gradient in virions, modulation of this gradient by metabolic regulators, transcription of viral genes weakly similar to TCA enzymes, and functional demonstration of a viral isocitrate dehydrogenase, collectively challenging classical boundaries of virology.

The introduction synthesizes prior discoveries that eroded classic viral hallmarks: detection of giant viruses (Mimivirus, Pandoravirus) with micrometer-scale particles and megabase genomes; presence of translation-associated genes in Tupanvirus and Klosneuvirus; antiviral defense systems in Mimivirus (MIMIVIRE); viral cytochrome P450 monooxygenases; and widespread auxiliary metabolic genes in environmental NCLDV metagenomes (e.g., fermentation, sphingolipid, nitrogen metabolism). Prior studies suggest viruses can reprogram host metabolism (e.g., cyanophages carrying photosynthesis genes, phage-encoded electron transport proteins) and influence biogeochemical cycles. However, before this work there was no evidence that giant viruses utilize metabolic gene products for their own energy needs. This background motivates probing for energy generation, specifically proton gradients and TCA cycle components, in P. massiliensis.

- Immunofluorescence and mitochondrial dyes: Amoebae (Acanthamoeba castellanii) infected with P. massiliensis were stained with MitoTracker Deep Red 633 and TMRM to assess membrane potential during the replication cycle at multiple time points (H4–H16). Mouse polyclonal anti–P. massiliensis antibodies (FITC-labeled secondary) enabled virion-specific co-localization. Controls included S. aureus (positive) and cowpox virus supernatant (negative). Purified virions were obtained by filtration (5 µm), low-speed spin to remove debris, ultracentrifugation (80,000 g, 20 min), PBS washes, and resuspension. Purified particles were stained with MitoTracker (15 min, 37 °C) or TMRM (100 µM stock; 1 µL added to 1 mL particles; 30 min, 30 °C) and imaged by confocal microscopy. 3D analyses and SEM compared virions versus isolated amoebal mitochondria to exclude confounding mitochondrial debris. - Depolarization assay: Purified virions were incubated with CCCP (100–400 µM) at 35 °C overnight prior to TMRM staining and confocal imaging. Viral viability/quantity was assessed (TCID50) and replication impact tested by infecting A. castellanii, then treating with CCCP (100 µM), followed by immunofluorescence and qPCR (SYBR Green; designed primers; Ct < 35 positive). - Acetyl-CoA modulation: Purified virions were exposed to acetyl-CoA at low (100 nM) and high (0.8 mM) concentrations prior to TMRM staining to assess effects on membrane potential; S. aureus served as control. - Bioinformatics: P. massiliensis genome screened by BLASTp (nr; e ≤ 1e−3), DELTA-BLAST against CDD, PSI-BLAST, HMMER/Pfam, and PHYRE2 structural prediction to identify proteins with similarity to TCA cycle enzymes; orthologs assessed with ProOrtholog (≥30% identity, ≥50% coverage); COG annotation with focus on class C (energy metabolism; score > 50 considered significant). Multiple sequence alignments (MUSCLE) and phylogeny (FastTree ML; 100 bootstraps; iTOL formatting) were conducted where possible. - Transcriptomics: RNA-seq on infected amoebae sampled across the replication cycle (30 min to 8 h post-inoculation) and on freshly released virions (11 h post-infection) after removal of debris (5-µm filtration, 500 × g). qRT-PCR targeted predicted TCA-cycle ORFs across time points (H0 to H16), with DNase treatments to remove DNA, one-step SYBR RT-qPCR, triplicates, positivity as Ct < 35. - Proteomics: LC-MS/MS of proteins from purified virions and infected amoebae (H0–H16). Database: six-frame translated viral genome fragmented into 320 aa windows (30 aa step), tryptic constraints, plus host A. castellanii proteome. Viral hits compared to COG classes. - Cloning/expression: Seven predicted TCA-related ORFs (132, 181, 206, 577, 695, 768, 1245) were codon-optimized, N-terminal Strep-tagged, synthesized, and cloned into pET24b(+). Expression in E. coli BL21(DE3)-derived strain; induction at OD600 ~0.6, then 20 °C for 20 h. Lysis (lysozyme, DNase I, PMSF), sonication, and StrepTrap HP affinity purification (AKTA) with biotin elution. Identities confirmed by MALDI-TOF. - Enzyme assays: Commercial kits used for citrate synthase, aconitase, and isocitrate dehydrogenase (IDH) activities; reactions in 96-well plates, 100 µL, quadruplicates, monitored at 450 nm for NAD(P)H formation. Kinetics (Michaelis–Menten) derived using Gen5 and Prism; human IDH used as reference. IDH activity was also assayed directly on purified virions. - Statistics: Comparative analyses of fluorescence intensities and treatment effects; significance typically at p < 0.05.

- Detection of virion membrane potential: During infection and in purified preparations, a subset of P. massiliensis virions showed MitoTracker Deep Red 633 and TMRM staining co-localizing with anti-Pandoravirus antibodies. Approximately 20% of mature particles were MitoTracker-positive. TMRM staining of purified virions produced strong signals comparable to S. aureus (positive control), while cowpox virus supernatant produced none. - Specificity controls: 3D imaging and SEM distinguished virions from amoebal mitochondria; purified mitochondria displayed distinct morphology, reducing the likelihood of mitochondrial debris causing dye signals. - CCCP depolarization: Pre-incubation of virions with CCCP (100–400 µM) significantly reduced TMRM fluorescence versus untreated controls (p < 0.05), confirming dye uptake reflects membrane potential. Viral titers (TCID50) and early infection qPCR Ct shifts (ΔCt 1.85–2.57 at H0–H3) indicated no major loss of particle numbers; a nonsignificant reduction in antibody-labeled particles on amoebae at 400 µM CCCP suggested potential involvement of membrane potential in early infection steps. - Acetyl-CoA modulation: Low acetyl-CoA (100 nM) increased TMRM fluorescence in virions (p < 0.05), while high acetyl-CoA (0.8 mM) decreased it (p < 0.05), mirroring effects in S. aureus controls, consistent with regulation by a TCA-related metabolite. - Genomic signatures: Eight ORFs exhibited low sequence similarity to TCA enzymes: putative citrate synthase (ORF577), aconitase (ORF1245), isocitrate/isopropylmalate dehydrogenases (ORF132, ORF864), α-ketoglutarate-to-succinic semialdehyde route via α-ketoglutarate decarboxylase (ORF762), succinate dehydrogenase (ORF181), fumarase (ORF206), and upstream acetyl-CoA synthetase (ORF595). No ATP synthase or classical electron transport chain genes were detected. - Transcription: RNA-seq detected 6/8 predicted TCA-like ORFs transcribed mainly between 4–8 h post-infection. qRT-PCR indicated all targeted ORFs (including 595, 577, 1245, 132, 864, 762, 206) were transcribed during the productive phase, with lowest Ct values around H4–H6. - Proteomics: Two predicted TCA-related proteins were detected in mature virions: ORF762 (putative α-ketoglutarate decarboxylase) and ORF595 (putative acetyl-CoA synthetase). - Enzymology: Recombinant ORF132 showed isocitrate dehydrogenase activity with Michaelis–Menten kinetics (Km ≈ 6.8 × 10−5 M; kcat ≈ 13.1 s−1; R² = 0.993). Specific activity was reported as 4 (units as reported by authors). ORF864 showed no IDH activity. IDH activity measured in purified virions was similar to recombinant ORF132. For comparison, human IDH showed specific activity 6.3 per µg, kcat 16.3 s−1, Km 585.4 µM, kcat/Km 2.78 × 10−4 s−1 M−1 (R² = 0.997). - Evolutionary context: Phylogeny for ORF132 clustered pandoravirus orthologs (P. massiliensis ORF132, P. braziliensis ORF339, P. neocaledonicus ORF902) together, distinct from archaeal sequences; insufficient homology in the other seven ORFs precluded robust phylogenetic inference. - No ATP detection and absence of ATP synthase in the genome suggest any energy-related activity is partial or utilizes alternative mechanisms. Overall, evidence supports a membrane potential in virions and partial, functional TCA-like capacity via at least one enzyme (IDH).

The presence of a membrane potential in P. massiliensis virions, its sensitivity to CCCP, and modulation by acetyl-CoA collectively indicate that virions maintain an electrochemical gradient influenced by central metabolic intermediates. The identification, transcription, and partial protein-level detection of viral ORFs weakly similar to TCA enzymes, together with functional IDH activity (ORF132) in vitro and in virions, suggest a partial TCA-like pathway in the viral context. This may contribute to early infection by facilitating processes such as genome delivery, analogous to osmotic and ion channel effects observed in other viruses. From an ecological and evolutionary perspective, these findings extend the concept of auxiliary metabolic capacity in large DNA viruses, implying deeper viral participation in energy flux and metabolic reprogramming of hosts. Although classical ATP synthase and respiratory chain components were not detected, alternative or incomplete cycles (as seen in cyanobacteria with modified TCA pathways) could be operating, or these enzymes may serve roles in redirecting host carbon metabolism upon infection. The phylogenetic distinctness of ORF132 and lack of clear homology for other ORFs raise questions about origins (e.g., horizontal transfer, divergence) and functions beyond canonical TCA roles. Overall, the data blur boundaries between viral and cellular life by demonstrating energy-related features at the virion level.

This study provides the first evidence of a membrane potential in Pandoravirus massiliensis virions and identifies eight viral ORFs with low similarity to TCA cycle enzymes, all of which are transcribed during replication. Two of these proteins are present in mature virions, and one (ORF132) encodes a functional isocitrate dehydrogenase with measurable kinetics. Membrane potential is sensitive to CCCP and is modulated by acetyl-CoA, supporting a link to central metabolism. These findings challenge traditional definitions of viruses and suggest that pandoraviruses may possess partial, virus-encoded metabolic capabilities that influence infection and potentially host metabolism. Future work should include structural characterization of candidate enzymes (especially IDH), genetic and complementation assays to confirm functions in vivo, comprehensive searches for additional metabolic partners, and precise quantification of energy production and flux during infection.

- Sequence similarity of the eight ORFs to canonical TCA enzymes is low, limiting confidence in functional assignments beyond ORF132 and precluding robust phylogenetic inference for most candidates. - No genes for ATP synthase or classical electron transport chain were identified, and ATP production was not detected, indicating any metabolic pathway is incomplete or unconventional. - Only a subset of virions (~20%) stained with mitochondrial dyes; functional heterogeneity among particles is possible. - CCCP reduced membrane potential but had limited measurable impact on virion numbers and early infection by qPCR/immunofluorescence, complicating causal links to infectivity. - Proteomics detected only two of the predicted TCA-related proteins in mature particles, and several recombinant proteins (e.g., citrate synthase, aconitase) showed no activity under assay conditions. - Potential confounding by residual host mitochondrial material was addressed by imaging and controls but cannot be absolutely excluded. - Exact publication date not provided; some methodological descriptions contain typographical inconsistencies that may affect precise replication without consulting supplementary materials.