How are mitochondrial ATT (Ile) start codons translated as Methionine?

Question

In some vertebrate species, some mtDNA start codon sequences are ATT but these are translated as Methionine rather than Isoleucine. What is the mechanism for this non-standard translation?

The main example below is in a bird, blue crane (NCBI Genbank excerpt below). The NCBI data give the standard note on completion of the stop codon but there is no note about modification of the start codon. We have confirmed this ATT start codon from PCR-Sanger sequencing an independent sample. There are similar cases in other vertebrates (Aves, Squamata, Amphibia) such as ND2 in Goggia geckos (example 2 below, also independently confirmed by Sanger sequencing. In this genus the start codon DNA varies from ATT to ATA or ATC: 50/59, 6/59, 3/59). I think I have seen this in other mtDNA genes. In many other birds and reptiles ND2 starts with ATG.

Are these modifed post-transcription to ATG? If so how? Could they reflect a variant of the mtDNA code where AUU = Met? Or where any AUX codon is recognized as Met.

Example 1. ND2 from Blue Crane (NCBI Genbank refseq excerpt)

LOCUS NC_020572 16696 bp DNA circular VRT 18-APR-2013 DEFINITION Anthropoides paradiseus mitochondrion, complete genome. ACCESSION NC_020572 VERSION NC_020572.1 DBLINK Project: 193816 BioProject: PRJNA193816 KEYWORDS RefSeq. SOURCE mitochondrion Anthropoides paradiseus (blue crane) ORGANISM Anthropoides paradiseus

 gene            3970..5008
                 /gene="ND2"
                 /db_xref="GeneID:15088361"
 CDS             3970..5008
                 /gene="ND2"
                 /note="TAA stop codon is completed by the addition of 3' A
                 residues to the mRNA"
                 /codon_start=1
                 /transl_except=(pos:5008,aa:TERM)
                 /transl_table=2
                 /product="NADH dehydrogenase subunit 2"
                 /protein_id="YP_007624320.1"
                 /db_xref="GeneID:15088361"
                 /translation="MNPHAKLIFLTSLVLGTTITISSNHWMSAWAGLEINTLAIIPLI
                 SKSHHPRAIEAAIKYFLVQATASALVLFSSMINAWSTGQWDITQLNQPMPCLLLTTAI
                 AMKLGLVPFHFWFPEVLQGSSLTTALLLSTVMKFPPITILFLTSHSLNPALLTSMAVA
                 SAALGGWMGLNQTQIRKILAFSSISHLGWMTIIIMYSPKLTLLTFYLYSLMTITVFLT
                 LNTTKALKLSTMMITWTKIPTLNATLMLTLLSLAGLPPLTGFLPKWLIIQELTKQEMT
                 TAATIITMLSLLGLFFYLRLAYYSTITLPPNSTNHMKQWHTDKSTNTLIAIFTSLSAL
                 LLPLSPMILTII"

 3901 agtagagtca gctaaaaaag ctatcgggcc cataccccga aaatgatggt ttaacccctt
 3961 cctctacta**a tt**aacccaca tgcaaaacta atcttcctca caagcctagt cctagggaca
 4021 accattacaa tctcaagcaa ccattgaata tcagcctgag caggcctaga aatcaatact
 4081 ctcgccatta tccccctcat ttcaaaatcc caccacccgc gagccatcga agccgcaatc
 4141 aaatatttcc tagtacaggc aaccgcttca gcactagtcc tcttctcaag tataatcaac
 4201 gcatgatcta caggacaatg agatattacc caattaaacc aaccaatacc atgcctttta

Example 2. ND2 in Goggia (Gekkota, squamata) NCBI nucleotide locus ref.

LOCUS KF666817 1424 bp DNA linear VRT 09-SEP-2014 DEFINITION Goggia lineata voucher MCZ R184778 NADH dehydrogenase subunit 2

Answer 1

The key point in understanding non-AUG start sites in mRNAs is the nature and role of the initiator tRNA.

There is a special initiator met-tRNA which differs in structure from all the elongator tRNAs in such a way that it — and only it — is recognized by the IF-2 type initiation factors (and not by the EF-Tu/EF-1 elongation factors). In a complex with the initiation factor and GTP it can bind to the ribosome at the P-site. Normally the ribosome will also be interacting with an AUG codon at the P-site, but in some cases it is not, so that initiation may still occur.

Thus, the initiating amino acid is always methionine, regardless of the codon, because methionine is the only amino acid with which the initiator tRNA can be charged.

In the case of bacterial mRNAs, it has long been known that about 10% start with other codons, predominantly GUG. This can be explained by their relative position to the ribosomal binding site on the mRNA — the Shine and Dalgarno sequence. I imagine this is what happens in mitochondria.

In the case of eukaryotic cytoplasmic mRNAs, an initiator met-tRNA complex normally scans from the 5ʹ-end, initiating at the first AUG or first the first AUG in an ‘appropriate context’. The precise definition of an appropriate context is unclear, and on can imagine that in some cases such a context pauses the ribosome at a non-AUG codon, at which initiation begins.

References

“The Biochemistry of the Nucleic Acids”, 11th edition, Chapter 12 (1992) RLP Adams et al. — this has all the older original references, but not the recent ones.

Wikipedia entry on Shine and Dalgarno sequence.

Eukaryotic scanning model: Hinnebush review (2011)

Initiation in prokaryotes and eukaryotes Londi review (2015)

Related SE Biology Answer on initiation and elongation met-tRNAs (by me)