mlcD (Q7Z2B8) — Myosin-II / Cytokinesis Prediction Assessment OpenScientist openscientist-autonomous 3 citations 2 artifacts 2026-07-10T03:53:33.149597 citations file

OpenScientist final report (text/html)

OpenScientist final report (application/pdf)

mlcD (Q7Z2B8) — Myosin-II / Cytokinesis Prediction Assessment

Gene: mlcD (UniProt Q7Z2B8, MLCD_DICDI), Dictyostelium discoideum (NCBITaxon:44689)
Focus: computational_prediction — hypothesis prediction-myosin-cytokinesis
Terms tested: GO:0000281 mitotic cytokinesis; GO:0030837 negative regulation of actin filament polymerization; GO:0043520 regulation of myosin II filament assembly; GO:0071976 protein localization to cell division site.
Reference context: doi:10.64898/2026.03.19.712954

Executive Judgment

Verdict: REFUTED (over-annotation / paralog–class misassignment).

BioReason-Pro-SFT predicts myosin-II and cytokinesis regulatory roles for mlcD. Primary
biochemistry and the curated UniProt/dictyBase record independently establish that mlcD is
the dedicated essential light chain of MyoD, a single-headed class I (unconventional) myosin —
not a component of the conventional myosin-II (mhcA) contractile machinery. All four predicted
terms belong to the myosin-II system (heavy chain mhcA plus its regulatory light chain and
essential light chain), whose bipolar filaments drive cytokinesis. Class I myosins are monomeric,
membrane-associated motors that do not self-assemble into filaments, so mlcD cannot mediate or
regulate myosin-II filament assembly or contractile-ring–based cytokinesis.

The most important caveat: I could not access the specific reference doi:10.64898/2026.03.19.712954
programmatically, and I did not run a MyoD-null phenotype search successfully (PubMed queries for
class-I phenotypes returned no hits in this environment). The refutation nonetheless rests on
direct, unambiguous primary evidence (protein sequencing + complex purification) and on the
curated functional record.

Evidence Matrix

Citation	Evidence type	Direction	Claim tested	Key finding	Context	Confidence / limitations
PMID:12826013 (De La Roche, Lee, Côté 2003)	Direct assay (protein sequencing, ESI-MS, ITC, complex purification)	Refutes seed	Is mlcD a myosin-II light chain?	MlcD is a 16 kDa 4-EF-hand protein co-purifying with MyoD, a long-tailed class I myosin; FLAG-MlcD complexes with MyoD but not MyoB/MyoC; low Ca²⁺ affinity (cannot sense physiological Ca²⁺)	D. discoideum, native + FLAG-tagged expression	High. Directly defines class/partner identity.
PMID:21671662 (Crawley et al. 2011)	Interaction / biochemical mapping	Refutes seed	Light-chain assignments across Dictyostelium myosin-I	Each long-tailed myosin-I has a unique light chain: MyoB–MlcB, MyoC–MlcC, MyoD–MlcD; short-tailed MyoA/MyoE use calmodulin	D. discoideum	High. Confirms MyoD–MlcD pairing.
UniProt Q7Z2B8 (MLCD_DICDI)	Database / curated record	Refutes seed	Protein identity, family, oligomeric state	Recommended name "Myosin-ID light chain"; FUNCTION "light chain for myosin-D"; SUBUNIT "Myosin I… Inability to self-assemble into filaments… Interacts with myoD; does not interact with myoB or myoC"	Curated	High (review/database-level, but backed by the primary papers above).
dictyBase GO (via UniProt xrefs)	Database (experimental IDA/IPI)	Qualifies / competing	What processes/locations are experimentally supported?	Curated terms: myosin complex (IDA), myosin heavy chain binding→myoD (IPI), Ca²⁺ binding (IDA), actin wave (IDA), macropinocytic cup cytoskeleton (IDA). None of the 4 predicted myosin-II/cytokinesis terms present	D. discoideum	High. Absence of predicted terms in curated set is informative.
InterPro/PANTHER/Pfam (Q7Z2B8)	Structural/evolutionary	Qualifies	Domain architecture	147 aa, four EF-hands (EF-hands 2–4 degenerate), CALM/Myosin/TropC-like (IPR050230), calmodulin-like — consistent with an EF-hand myosin light chain, not a filament-assembly regulator	Sequence	High.
Iteration-2 NW/BLOSUM62 (this study)	Computational (sequence identity)	Qualifies/refutes	Which clade does MlcD belong to?	MlcD 45.6% id to calmodulin & 45.2% to MlcB (myosin-I LC) vs only 30.8%/31.2% to myosin-II mlcR/mlcE; reproduces published ~44% CaM benchmark	D. discoideum paralogs	Medium-high. Global alignment, single method; concordant with primary data.
PMID:10423462 (Dai et al. 1999)	Mutant phenotype (micropipette aspiration)	Qualifies (class context)	What do Dictyostelium myosin-I motors do?	Amoeboid myosin-I's drive pseudopod formation, macropinocytosis; double mutants lose ~50% cortical tension; required for migration — NOT cytokinesis	D. discoideum myosin-I mutants	Medium (about myosin-I class generally, not MyoD-specific).

GO Curation Implications (leads — require curator verification)

Do NOT add GO:0000281 (mitotic cytokinesis), GO:0030837 (neg. reg. actin filament polymerization),
GO:0043520 (regulation of myosin II filament assembly), or GO:0071976 (protein localization to cell
division site) to mlcD. These are conventional-myosin-II / contractile-ring functions and are
class-inappropriate for a myosin-I light chain.
GO:0043520 specifically: refuted. The correct effectors of Dictyostelium myosin-II filament
assembly are the mhcA heavy chain (regulated by myosin heavy-chain kinases) and its regulatory /
essential light chains — not MlcD. If any myosin light chain warrants GO:0043520, it is the
myosin-II RLC/ELC paralog, not mlcD.
Retain / support instead (MF/CC): MF GO:0032036 myosin heavy chain binding (IPI to myoD) and
the calmodulin-like light-chain role; CC GO:0016459 myosin complex. Prefer these informative terms
over generic "protein binding."
Class-appropriate BP/CC leads (verify against MyoD phenotype literature): class-I myosin
processes at the plasma membrane — macropinocytosis/endocytosis and cortical actin dynamics
(consistent with curated macropinocytic cup cytoskeleton and actin wave localizations).

Mechanistic Scope

Direct molecular function of the gene product: MlcD is a calmodulin-like EF-hand essential light
chain that binds the IQ motifs in the neck of the class I myosin heavy chain MyoD, stabilizing
its lever arm. Its Ca²⁺ affinity is low, so it is not a Ca²⁺ sensor; De La Roche et al. propose it
confers Ca²⁺-insensitive regulatory properties distinguishing MyoD from calmodulin-bearing myosin-I.
The predicted terms describe downstream cellular processes of a different motor system (myosin-II
contractile ring); they are neither the immediate activity of MlcD nor a documented phenotype of it.

Conflicts and Alternatives

Paralog/class confusion (most likely source of the prediction): "mlc*" myosin light-chain genes
form a paralog pool spanning classes. The myosin-II regulatory (mlcR) and essential (mlcE) light
chains legitimately carry cytokinesis/filament-assembly context; a sequence/name-similarity or
frequency-biased transfer plausibly propagated those terms onto mlcD.
Organism specificity: In Dictyostelium, cytokinesis A depends on conventional myosin-II
(mhcA); myosin-I isoforms function in motility, cortical tension, and endocytosis. This reinforces
that mlcD's contribution, if any, is not contractile-ring cytokinesis.
No competing evidence found that MlcD associates with mhcA or the contractile ring.

Knowledge Gaps

MyoD-null / MlcD-null cellular phenotype. Partially addressed: MyoD-specific knockout papers
were not retrievable here, but class-level phenotype data (Dai et al. 1999, PMID:10423462) show
Dictyostelium myosin-I motors drive pseudopod formation, macropinocytosis, and cortical tension —
not cytokinesis. A positive class-I process annotation for mlcD (e.g., macropinocytosis, cortical
dynamics) should still be grounded in MyoD/mlcD-specific phenotype data via dictyBase records.
Reference doi:10.64898/2026.03.19.712954. Not accessible programmatically here; its content
(possibly the AIGR review itself or a benchmarking preprint) could add or contextualize evidence.
Exact myosin-II light-chain paralog identities (mlcE vs mlcR). Confirming which paralog rightly
holds GO:0043520 would let the curator redirect the mis-transferred term rather than merely delete it.

Discriminating Tests

Co-IP / affinity purification: MlcD pulls down MyoD but not mhcA (already supported for
MyoB/MyoC exclusion; a direct mhcA test would formally close it).
Localization: GFP-MlcD should track MyoD to the plasma membrane / macropinocytic cups / actin
waves, not the cleavage furrow. Curated IDA localizations already favor the membrane pattern.
Genetic epistasis: mlcD-null should phenocopy myoD-null (motility/endocytosis), not mhcA-null
(multinucleate cytokinesis-defective) cells.
Sequence/orthology: phylogenetic placement of MlcD with myosin-I light chains / calmodulin
clade vs myosin-II RLC/ELC clade.

Curation Leads (require curator verification)

Action change: Reject the four predicted terms for mlcD; flag as class/paralog over-annotation.
Candidate references to verify: PMID:12826013 (identity, MyoD specificity, Ca²⁺ properties);
PMID:21671662 (MyoD–MlcD light-chain assignment). Snippets provided in the recorded findings.
Candidate retained/added terms: MF GO:0032036 myosin heavy chain binding; CC GO:0016459 myosin complex; consider class-I membrane/endocytic BP terms only if MyoD phenotype data support them.
Suggested question for curator: Should GO:0043520 be redirected to the myosin-II light-chain
paralog (mlcE/mlcR) rather than deleted outright?
Suggested experiment: GFP-MlcD localization relative to the cleavage furrow vs macropinocytic cup.

Computed Provenance — Pairwise Sequence Identity (Iteration 2)

Global Needleman–Wunsch alignment (BLOSUM62, gap = −8), MlcD (Q7Z2B8) vs D. discoideum paralogs.
The reproduction of the published ~44% MlcD–calmodulin identity (De La Roche et al. 2003, PMID:12826013)
serves as a method-validity check.

Comparison (vs MlcD Q7Z2B8)	UniProt	% identity	Aligned cols	Class
Calmodulin (calA)	P02599	45.6	147	Ca²⁺ sensor / myosin-I LC clade
MlcB — Myosin-IB light chain	Q54GL7	45.2	73*	Class I myosin LC
mlcR — Myosin-II regulatory LC	P13833	30.8	146	Myosin-II LC
mlcE — Myosin-II essential LC	P09402	31.2	141	Myosin-II LC

*MlcB record is only 73 aa (partial), so its alignment spans fewer columns.

Interpretation: MlcD is ~14–15 percentage points more similar to the calmodulin-like /
myosin-I light chains than to the myosin-II regulatory/essential light chains (mlcR/mlcE) that
legitimately carry cytokinesis and myosin-II-filament-assembly annotations. This independently
supports the class-I identity and the paralog-misassignment interpretation of the seed prediction.
(Computed value 45.6% ≈ published 44% → analysis validated.)

Limitations

Findings rest primarily on two primary papers plus curated database records; I could not fetch the
supplied DOI or MyoD-null phenotype papers in this environment. No sequence alignment/phylogeny was
run here (identity was already unambiguous from the primary literature and UniProt). Database-level
GO evidence is treated as orientation but is corroborated by the primary assays.