A Method for In Situ Reverse Genetic Analysis of Proteins Involved mtDNA Replication

Abstract

The unavailability of tractable reverse genetic analysis approaches represents an obstacle to a better understanding of mitochondrial DNA replication. Here, we used CRISPR-Cas9 mediated gene editing to establish the conditional viability of knockouts in the key proteins involved in mtDNA replication. This observation prompted us to develop a set of tools for reverse genetic analysis in situ, which we called the GeneSwap approach. The technique was validated by identifying 730 amino acid (aa) substitutions in the mature human TFAM that are conditionally permissive for mtDNA replication. We established that HMG domains of TFAM are functionally independent, which opens opportunities for engineering chimeric TFAMs with customized properties for studies on mtDNA replication, mitochondrial transcription, and respiratory chain function. Finally, we present evidence that the HMG2 domain plays the leading role in TFAM species-specificity, thus indicating a potential pathway for TFAM-mtDNA evolutionary co-adaptations.

TFAM’s Contributions to mtDNA Replication and OXPHOS Biogenesis Are Genetically Separable

“…Synthetic chTFAM variants ( Table S1 ) were generated with either native or human MTS by Twist Bioscience (South San Francisco, CA, USA). chTFAM MTS is functional in human cells [ 10 ]. Synthetic chTFAMs were cloned into pMA4659 (Addgene, Watertown, MA cat.…”

“…TFAM’s contribution to mtDNA replication is less well understood. However, the consensus is that a fraction of transcripts from the mitochondrial light strand promoter (LSP, identical in sequence to mtDNA heavy strand) are prematurely terminated at the guanine-rich conserved sequence block II (CSBII) to generate primers for mitochondrial heavy strand replication [ 9 , 10 , 11 ]. This model intimately links TFAM’s contributions to mtDNA replication and OXPHOS biogenesis.…”

“…Our understanding of mtDNA replication in human cells remains limited, in part due to the lack of tractable in vitro or in situ models to study this process. Recently, we developed the GeneSwap reverse genetic analysis method to study mtDNA replication in situ and made a striking observation that TFAMs from the ‘living fossil’ fish coelacanth ( Latimeria chalumnae ) and a frog ( Xenopus laevis ) can support mtDNA replication in human cells, whereas TFAM from phylogenetically less distant chicken ( Gallus gallus ) cannot [ 10 ]. This apparent paradox could not be explained by the extent of primary amino acid sequence similarity between proteins because chicken TFAM (chTFAM) is more similar to the human ortholog than TFAMs from either coelacanths or frogs.…”

“…This apparent paradox could not be explained by the extent of primary amino acid sequence similarity between proteins because chicken TFAM (chTFAM) is more similar to the human ortholog than TFAMs from either coelacanths or frogs. However, chimeric TFAM composed of either hTFAM N-terminal domain (NTD) and chTFAM C-terminal domain (CTD) or chTFAM NTD and hTFAM CTD supports hmtDNA replication [ 10 ]. This observation suggests that the inability of chTFAM to support hmtDNA replication is a cumulative effect of 100 aa residues that are at variance between mature forms of human and chicken proteins ( Figure 2 ) and that it cannot be attributed to any of these substitutions individually.…”

“…Therefore, each aa substitution in chTFAM is conditionally permissive (the condition here being the context of other substitutions in the same domain). Moreover, neither 67 aa that are at variance in the NTD nor 33 aa that are at variance in the CTD, collectively, can account for the chTFAM deficiency in hmtDNA replication [ 10 ].…”