Metro: enzyme-derived guanylate-kinase-like scaffold domain

The selected Metro protein A1Z8G0 has an altered nucleotide pocket consistent
with a catalytically inactive MAGUK scaffold. Its electronic kinase and
transferase annotations are not supported by retention of the guanylate-kinase
fold.
This conclusion combines target-sequence observations with experimental
work on the enzyme-to-scaffold transition; it is not a direct Metro enzyme assay.

Sequence observations

The target is the exact 595-aa benchmark protein. Its UniProt-annotated GUK-like
domain spans residues 389–580. Both MAFFT L-INS-i and G-INS-i give the same
correspondences for every inspected reference site. The independent run using
human GUK1 as the positional reference reproduces the key mapping.

Reference feature Yeast active GUK1 P15454 Human active GUK1 Q16774 Metro A1Z8G0 Rat PSD-95 P31016 Human MPP1 Q00013
ATP-loop lysine Lys15 Lys17 Arg402 Lys544 Arg295
Following ATP-loop residue Ser16 Ser18 Asn403 Asp545 Ser296
GMP recognition site Ser35 Ser37 Pro422 Pro564 Pro315
GMP recognition site Asp101 Asp103 His488 Ser631 Glu381

The complete target segment corresponding to yeast residues 9–16 is
GAPGVGRN (Metro396–403), compared with GPSGTGKS in yeast. A Lys-to-Arg
substitution alone would not establish inactivity; the broader pocket changes
and the experimentally characterized Ser-to-Pro correspondence are more
informative. Several other GMP-contacting positions remain conserved, including
the equivalents of yeast Arg39, Arg42, Tyr51 and Glu70. The domain has retained
substantial recognition-site structure rather than losing every ligand contact.

Direct outputs: site mappings,
local-pair alignment,
global-pair alignment, and
human-reference control.
Full accession records and their SHA-256 checksums are in
sources.json; methods and reproduction.

Experimental grounding and scope

Olsen and Bredt, Functional analysis of the nucleotide binding domain of
membrane-associated guanylate kinases
(2003),
PMID:12482754(https://pubmed.ncbi.nlm.nih.gov/12482754/),
DOI:10.1074/jbc.M210165200,
experimentally connected altered GMP-pocket residues to deficient nucleotide
recognition. Reciprocal substitutions restored GMP binding to PSD-95 without
restoring its catalysis. The reproduced PSD-95 positions Pro564/Ser631 provide a
control on this analysis; active yeast/human enzymes retain Ser/Asp at the
corresponding sites. The cached publication is abstract-only; the sequence
coordinates themselves are independently calculated here.

Zhu et al., Guanylate kinase domains of the MAGUK family scaffold proteins as
specific phospho-protein-binding modules
(2011),
PMID:22117215(https://pubmed.ncbi.nlm.nih.gov/22117215/),
DOI:10.1038/emboj.2011.428,
provides full-text structural and biochemical evidence for this functional
repurposing. Its comparison includes MPP-family domains, making it relevant to
Metro's p55/MPP architecture. The paper is cached with full text. This supports
the family-level interpretation but does not identify a specific phosphopeptide
partner of the Metro GUK domain.

The direct Metro study,
Bachmann et al., PMID:20427642,
DOI:10.1523/JNEUROSCI.0778-10.2010,
establishes an L27-dependent synaptic scaffold role. Together with the observed
pocket substitutions, it supports rejecting the electronic catalytic transfer.
The present data do not establish complete absence of ATP/GMP binding,
exclude all possible residual chemistry, or justify annotating Metro with a
particular phosphopeptide-binding specificity. Generic nucleotide-binding
annotations can remain unresolved separately from the catalytic claim.

Verification checklist