DTYMK-Related Neurodegeneration — Comprehensive Research Report

Preferred disease name: Childhood-onset neurodegeneration with progressive microcephaly (CONPM) Target MONDO ID: MONDO:0859241 Category: Mendelian, autosomal recessive, ultra-rare (≤5 reported patients worldwide as of 2025) Report date: 2026-08-01

⚠️ Curation Preflight Warnings (read before using this report)

1. Named-Entity / topic confusion risk is HIGH for this gene. A PubMed query for DTYMK returns ~58 records, of which the large majority are cancer-bioinformatics papers (hepatocellular carcinoma prognostic signatures, lung adenocarcinoma pan-cancer analyses, uveal melanoma, colorectal cancer). Only five primary records concern the human neurodegenerative disease. A deep-research tool asked about "DTYMK" will very plausibly return a coherent, well-cited, snippet-validating report about DTYMK as an oncology biomarker rather than about CONPM. Per the repo NEC preflight (CLAUDE.md §2b), the identity anchors for this entry are: gene DTYMK / HGNC:3061 / OMIM *188345, phenotype OMIM #619847, locus 2q37.3. Any report whose dominant subject is tumor prognosis is off-target.

2. Second NEC-adjacent risk: the first clinical report (Lam et al. 2019, PMID:31271740) framed DTYMK as a mitochondrial DNA depletion syndrome (MDDS) gene, not as a nuclear-genome-instability disorder. The two framings coexist in the literature and are not the same mechanistic claim. Do not blend them into a single unqualified causal chain — see §6.

3. Quote provenance. Abstract text reproduced below under "verbatim abstract" was fetched from NCBI E-utilities and is quotable. Sentences marked [full-text] were extracted from PMC full text by an intermediate summarizer and must be re-verified against the source before being used as an evidence snippet: — they are reported here as leads, not as validated quotes.

4. Ontology IDs suggested throughout are candidates. Every one must be verified with just validate-terms / OAK before commit.

1. Disease Information

Overview

CONPM is an ultra-rare autosomal recessive neurometabolic/neurodegenerative disorder caused by biallelic loss-of-function variants in DTYMK, which encodes deoxythymidylate kinase (thymidylate kinase, TMPK; EC 2.7.4.9). TMPK catalyses the penultimate step of dTTP biosynthesis (dTMP → dTDP). Because both the de novo (thymidylate synthase) and the salvage (thymidine kinase) routes to dTTP converge on dTMP upstream of TMPK, loss of TMPK constitutes a complete block of the canonical dTTP supply — a fact the discovery paper explicitly calls remarkable given that affected children are born alive.

Clinically the disorder presents as congenital-to-infantile onset, severe and progressive (postnatal-worsening) microcephaly, profound global developmental delay or frank developmental regression, early-onset seizures, spasticity with pyramidal signs, cortical blindness/absent visual tracking, feeding failure, and death in early childhood in the severe form, with a milder end of the spectrum (small head circumference, severe intellectual disability, hypotonia, poor speech, motor delay) documented in the sibling pair reported by Lam et al.

The neuroimaging signature is distinctive: profound generalized cerebral atrophy with severe ventricular and subarachnoid space enlargement, shrinkage/"disappearance" of the basal ganglia, and relative sparing of brainstem and cerebellum.

Key identifiers

Synonyms and alternative names

• Childhood-onset neurodegeneration with progressive microcephaly (CONPM) — OMIM/MONDO preferred
• DTYMK-related neurodegeneration
• DTYMK deficiency; thymidylate kinase (TMPK) deficiency; dTMP kinase deficiency
• Deoxythymidylate kinase deficiency
• DTYMK-related mitochondrial DNA depletion syndrome (Lam et al. framing; not an OMIM-recognized MDDS numbered subtype)

Provenance of information

Aggregated disease-level resources plus individual-patient case reports. There is no registry, no natural-history cohort, and no EHR-derived data for this disorder. All human phenotype information derives from five individually described patients in three publications. HPO annotations for OMIM:619847 are denominator-annotated over n=2 or n=4 individuals (see §3), which is itself a useful signal that all "frequencies" are case-count fractions, not population estimates.

2. Etiology

Disease causal factors

Purely genetic (monogenic, autosomal recessive). Biallelic loss-of-function variants in DTYMK (HGNC:3061). No environmental, infectious, or acquired etiology is known or postulated. No somatic mechanism in this disease (somatic DTYMK dysregulation is an oncology topic — see §12 note — and is mechanistically unrelated).

Primary quote (verbatim abstract, PMID:34918187):

"Here, we describe two unrelated children with bi-allelic variants in DTYMK, encoding dTMPK, which catalyzes the penultimate step in dTTP biosynthesis." "In summary, by combining genetic and biochemical approaches in multiple models we identified loss-of-function of DTYMK as the cause of a severe postnatal neurodegenerative disease and highlight the essential nature of dTTP synthesis in the maintenance of genome stability and neuronal survival."

Genetic risk factors

• Causal variants: biallelic (homozygous or compound heterozygous) DTYMK variants — see §4.
• Consanguinity is a documented risk contributor: Individual II of Vanoevelen et al. was born to consanguineous Egyptian parents and was homozygous for p.Pro81Leu [full-text]. The 2025 Mexican case was likewise homozygous for p.Pro81Leu (PMID:40696808).
• Carrier status: heterozygous carriers are unaffected. Mouse data support this directly — heterozygous Dtymk mice show ~3-fold reduced expression with no neural tube defects (PMID:38621447).
• No susceptibility loci, GWAS signals, or modifier genes have been reported. GWAS Catalog / PheGenI contain nothing relevant to CONPM.

Environmental risk factors

None identified. No toxin, exposure, lifestyle, occupational, maternal, or seasonal factor has been implicated. Age, sex, and parity are not risk factors (see §9 for the sex distribution caveat: n is too small to interpret).

One negative environmental experiment is informative: Tiani & Stover fed heterozygous Dtymk dams three different diets (folate-manipulated) and found no diet-dependent increase in neural tube defect risk in het offspring (PMID:38621447) — i.e., no demonstrated folate/one-carbon dietary modifier of Dtymk haploinsufficiency in mouse.

Protective factors

None identified. No protective allele, modifier, dietary, or lifestyle factor is documented.

However, there is one striking endogenous compensatory phenomenon that functions like a protective mechanism at the cellular level and is arguably the central unresolved question of this disease: non-neural tissues appear to be substantially spared, and patient fibroblasts proliferate normally despite undetectable TMPK activity. This is discussed as a mechanism/knowledge-gap in §6 rather than a protective factor, because the responsible enzyme is unidentified.

Verbatim abstract (PMID:34926941, ACS Omega 2021):

"In conclusion, TMPK mutations identified in patients represent loss of function mutations but surprisingly the proliferation rate of the patient-derived fibroblasts was normal, suggesting the existence of an alternative and hitherto unknown compensatory TMPK-like enzyme for dTTP synthesis."

Verbatim abstract (PMID:34994281, Nucleosides Nucleotides Nucleic Acids 2022):

"Deficiency in TMPK activity due to genetic alterations of DTYMK, i.e., the gene coding for TMPK, causes severe microcephaly in humans. However, no defects were observed in other tissues, suggesting the existence of a compensatory enzyme for dTTP synthesis." "…because of its low activity, isoform 6 is unlikely be able to compensate for the loss of TMPK activity caused by deletions and/or point mutations of the DTYMK gene."

Gene–environment interactions

None documented. CTD, PheGenI, and the primary literature contain no GxE evidence for DTYMK/CONPM. Theoretically plausible but untested interactions worth flagging as knowledge gaps: (i) thymidine/deoxynucleoside availability (relevant to the NCT04802707 trial, §12); (ii) antifolate or nucleoside-analogue drug exposure in carriers, given that the pathway is the target of 5-FU/methotrexate-class agents; (iii) genotoxic exposure (UV/ionizing radiation) given the demonstrated DNA-damage-response defect. None of these are supported by data — do not curate as evidence-backed.

3. Phenotypes

3.1 HPO annotations for OMIM:619847 (authoritative, with case-count frequencies)

Retrieved from the HPO annotation API (ontology.jax.org/api/network/annotation/OMIM:619847). Frequencies are n-of-N counts over the reported cases, not population frequencies. Denominators of 4 indicate annotation across both clinical reports (Vanoevelen n=2 + Lam n=2); denominators of 2 indicate a single report.

Phenotypes documented in the literature but NOT in the current HPO annotation set (candidate HPO-annotation gaps worth flagging upstream):

3.2 Phenotype characteristics by domain

Neurodevelopmental / cognitive - Type: clinical sign + developmental - Onset: congenital to infantile (HP:0003577 2/2; HP:0003593 2/2). Feeding problems from day 3 of life in Individual I [full-text]. - Severity: severe to profound in the Vanoevelen/Hernández-Carreto cases (essentially no milestones achieved); moderate-severe intellectual disability with poor speech in the Lam siblings. - Progression: progressive with frank regression — this is the defining feature separating CONPM from static primary microcephaly. Individual I was hypotonic at birth then "developed spasticity with opisthotonus within 1 year of age" [full-text]. - QoL: catastrophic. Total dependence for all activities of daily living; no communication; tube feeding; recurrent aspiration/respiratory illness. No formal QoL instrument (EQ-5D, PROMIS, PedsQL) has been applied — a genuine data gap.

Microcephaly - Type: physical manifestation / clinical sign - Onset: congenital but predominantly postnatal-progressive. Individual I had OFC 31 cm at birth (2nd centile — i.e., borderline, not markedly microcephalic) and reached −7.6 SD by 9 months; Individual II reached −7.4 SD by 26 months [full-text]. This trajectory (near-normal at birth → extreme by 1–2 years) is the single most curation-relevant temporal detail and should drive a clinical_course: PROGRESSIVE qualifier. - Severity: extreme (−7 SD or worse). - Frequency: 2/2 in the severe cases; small head circumference in the milder siblings.

Seizures / epilepsy - Type: clinical sign - Onset: 6 months (febrile seizures, Individual I); 15 months (myoclonic jerks, Individual II) [full-text]; epilepsy present in the 2-year-old Mexican case (PMID:40696808). - Semiology: febrile seizures, myoclonic jerks, status epilepticus. - EEG: Individual I had a flat-trace EEG treated with phenobarbital [full-text] — consistent with profound cortical loss rather than a primary channelopathy. - Progression: progressive; drug response poor.

Motor / tone - Biphasic pattern: neonatal hypotonia → evolving spastic hypertonia with hyperreflexia, clonus, extensor plantar responses, spastic tetraplegia, opisthotonus, limb dystonia. Individual II retained "good control of the head" while having increased distal tone, bilateral clonus, and positive Babinski signs [full-text].

Visual - Absent visual tracking/eye contact from the first months; annotated as cerebral (cortical) visual impairment (HP:0100704). "No eye contact was ever made" (Individual I) [full-text].

Growth / feeding - SGA (1/2), short stature (1/2), poor feeding from the neonatal period requiring nasogastric tube, dysphagia, laryngomalacia. Individual II birth weight 2250 g (−1.8 SD) [full-text].

Laboratory abnormalities — important because they are the metabolic-workup handles - Increased circulating lactate (HP:0002151) — 2/2. LOINC: 2524-7 (Lactate [Moles/volume] in Serum or Plasma), 32693-4 (Lactate, blood); CSF lactate LOINC 2519-7. - Hyperalaninemia (HP:0003348) — 1/2. LOINC 26603-3 (Alanine [Moles/volume] in Plasma). The lactate+alanine pattern is the classic mitochondrial-disease screen and explains why the Lam siblings were worked up as MDDS. - Microcytic hypochromic anemia and elevated liver enzymes (Individual II) [full-text]. - Fibroblast dTMPK enzyme activity — the diagnostic functional assay (see §10).

Urogenital - Cryptorchidism 2/4, micropenis 1/2 — notable as the only consistent extra-CNS structural finding, and worth flagging mechanistically since gonadal/germline tissue is highly proliferative.

Neuropathology - Gliosis (HP:0002171). "Pathology in individual I confirms massive neuronal dropout, only sparing the dentate nucleus and brain stem." [full-text]

4. Genetic / Molecular Information

Gene

Protein: 212-aa homodimeric P-loop kinase of the thymidylate kinase family. Reaction: dTMP + ATP → dTDP + ADP (Mg²⁺-dependent). Localizes to cytosol, nucleus, and mitochondrion (UniProt P23919). Human Protein Atlas reports low tissue specificity (tau 0.29), detected in all tissues, assigned to the "Bone marrow – Nuclear processes" expression cluster, with HPA IF subcellular localization to mitochondria and moderate, non-distinctive expression across brain regions. Curation note: the absence of brain-enriched expression means the brain-restricted phenotype is not explained by expression pattern — it must be explained by the proliferative/repair demands of neurodevelopment plus the postmitotic vulnerability of neurons.

Reported pathogenic variants (all germline; ClinVar-verified)

Genotype assignments by patient:

p.Pro81Leu is the recurrent allele, seen in 3 of 5 reported patients across Egyptian, Dutch, and Mexican ancestries — consistent with recurrent mutation rather than a founder effect (no shared haplotype reported; the three ancestries are unrelated).

Functional consequences — loss of function via loss of obligate dimerization

Verbatim abstract (PMID:34926941, ACS Omega 2021):

"Here we show that in fibroblasts derived from a patient, the P81L and D128N mutations led to a complete loss of TMPK activity in mitochondria and extremely low and unstable TMPK activity in cytosol. Despite the lack of TMPK activity, the patient-derived fibroblasts apparently grew normal. … The wild-type TMPK mainly exists as a dimer with high substrate binding affinity, that is, low KM value and high catalytic efficiency, that is, kcat/KM. In contrast, all mutants were present as monomers with dramatically reduced substrate binding affinity and catalytic efficiencies. Based on the human TMPK structure, none of the mutated amino acids interacted directly with the substrates."

Kinetics [full-text, Table 2 of PMID:34926941 — re-verify before use]:

Key mechanistic insight for the pathophysiology graph: the missense substitutions are not active-site contact residues. They act allosterically/structurally, converting the obligate homodimer into a catalytically crippled monomer. This is a dimerization-disruption LoF, not a substrate-binding LoF — a distinction worth capturing as a distinct MOLECULAR-scale pathophysiology node.

Cellular enzyme activity [full-text, PMID:34918187]: - Individual I fibroblasts: 0.62 pmol/min/mg protein (essentially undetectable) - Mother: 43.65 pmol/min/mg; Father: 31.08 pmol/min/mg (both normal) - Statistical significance: mother vs. proband p = 1.46 × 10⁻⁶; father vs. proband p = 5.77 × 10⁻⁶ - Compartment-specific finding: no detectable TMPK activity in mitochondria; low, unstable activity in cytosol (PMID:34926941 abstract, verbatim above).

Modifier genes

None identified. The most important open question in this space is not a modifier gene per se but the identity of the unknown compensatory TMPK-like enzyme postulated to sustain dTTP synthesis in non-neural tissue. Frisk et al. systematically excluded the five non-canonical DTYMK mRNA isoforms (isoforms 2–5 lack essential substrate-binding elements; isoform 6 retains intact catalytic centres but has <0.1% of isoform-1 activity) — PMID:34994281. Vanoevelen et al. considered and rejected CMPK2: "has no apparent capability to use dTMP as a substrate, it would appear unlikely that CMPK2 can fulfill this function. Thus, a compensatory pathway for dTTP generation remains to be proven." [full-text]

This is the single best KNOWLEDGE_GAP discussion item for the dismech entry.

Epigenetics

No DNA methylation, histone-modification, chromatin, or episignature data exist for CONPM. DTYMK is not on any published rare-disease episignature panel. Not applicable / no data.

Chromosomal abnormalities

None reported as a cause of CONPM. Note that DTYMK lies at 2q37.3, within the interval commonly deleted in 2q37 deletion / Brachydactyly–Mental Retardation syndrome (which dismech already curates as 2q37_Microdeletion_Syndrome.yaml, driven by HDAC4). There is no evidence that DTYMK haploinsufficiency contributes to the 2q37 deletion phenotype, and the mouse heterozygote data argue against it (PMID:38621447). Do not cross-link these two entries as mechanistically related without new evidence; a "same locus, different mechanism" note is appropriate if any link is made at all.

Population constraint

gnomAD constraint metrics (pLI / LOEUF / missense Z) for DTYMK could not be retrieved through available tooling in this session (the gnomAD browser is a JS app and the GraphQL endpoint requires POST). Do not assert a pLI or LOEUF value from memory. Fetch these directly before curating a constraint claim. The relevant biological constraint statement that IS evidenced: homozygous Dtymk knockout is embryonic lethal in mouse while heterozygotes are normal (PMID:38621447) — i.e., recessive essentiality, not haploinsufficiency.

5. Environmental Information

• Environmental factors: none. No entries in CTD linking environmental chemicals to CONPM. (CTD does contain DTYMK–chemical interactions from toxicogenomic screens, but these are gene-expression associations in unrelated experimental contexts and are not disease-etiologic — do not curate them as risk factors.)
• Lifestyle factors: not applicable (congenital-onset monogenic disorder).
• Infectious agents: not causal. Infection is however a major proximate cause of death: Individual I died at 18 months from cardiopulmonary arrest following a respiratory illness; Individual II died at 32 months from pneumonia and coma [full-text]. Curate these as complications/terminal events (§11), not as etiology.
• Fever as a trigger: Individual I had recurrent febrile seizures (HP:0002373) [full-text]. Whether fever is a genuine decompensation trigger (as in some intoxication-type IEMs) is unstudied — a legitimate hypothesis-flagged item, not an evidenced claim.

6. Mechanism / Pathophysiology

6.1 The causal chain (proposed pathograph, upstream → downstream)

[MOLECULAR] Biallelic DTYMK LoF variants (p.Pro81Leu / p.Asp128Asn / p.Ala99Thr / p.Asp96fs)
     │
     ▼
[MOLECULAR] Loss of TMPK homodimerization → monomeric, catalytically crippled enzyme
     │  (kcat/KM reduced 85–96%; D128N below detection)
     ▼
[MOLECULAR] Loss of dTMP → dTDP phosphorylation (EC 2.7.4.9); COMPLETE block of the
     │  canonical dTTP supply, because de novo (TYMS) and salvage (TK1/TK2) both
     │  converge on dTMP UPSTREAM of TMPK
     ▼
[CELLULAR] dTTP insufficiency in cells with high replicative/repair demand
     │        ├──► Impaired DNA replication; S-phase collapse (patient fibroblasts 2.8% S-phase
     │        │     vs. 23.65%/16.03% in parents) [full-text]
     │        ├──► Nucleotide-pool imbalance → RIBONUCLEOTIDE MISINCORPORATION into genomic DNA
     │        │     (dtymk⁻/⁻ zebrafish gDNA fragility comparable to Rnaseh2⁻/⁻ mouse) [full-text]
     │        └──► Impaired DNA damage response (persistent γH2AX 24 h post-UV) [full-text]
     ▼
[CELLULAR] Genome instability in the developing and mature CNS
     ▼
[CELLULAR] Neuronal apoptosis / neuronal dropout
     │  (dtymk mutant zebrafish: significantly more apoptotic cells in forebrain,
     │   p = 6.45 × 10⁻⁶ vs. wild type) [full-text]
     ▼
[TISSUE] Progressive cerebral and striatal atrophy with gliosis;
     │  relative sparing of brainstem, cerebellum, dentate nucleus
     ▼
[ORGANISM] Progressive microcephaly, developmental regression, epilepsy, spastic
    tetraplegia, cortical blindness → early childhood death

┌─ PARALLEL / CONTESTED ARM (Lam et al. 2019) ─────────────────────┐
│ Mitochondrial TMPK activity loss → mitochondrial dTTP pool       │
│ depletion → mtDNA replication failure → mtDNA DEPLETION →        │
│ OXPHOS deficiency → lactic acidemia (HP:0002151, 2/2) and        │
│ hyperalaninemia (HP:0003348, 1/2)                                │
│ STATUS: mtDNA depletion shown "in silico" only in one sibling;   │
│ mitochondrial TMPK activity loss confirmed biochemically         │
│ (PMID:34926941). Curate as an EMERGING mechanistic_hypothesis.   │
└──────────────────────────────────────────────────────────────────┘

┌─ UNRESOLVED COMPENSATION (the central paradox) ──────────────────┐
│ Unknown "TMPK-like" enzyme sustains dTTP in non-neural tissue →  │
│ near-normal bulk dNTP pools; normal fibroblast proliferation;    │
│ viability to birth despite complete canonical-pathway block.     │
│ Identity unknown. CMPK2 and DTYMK isoform 6 both excluded.       │
│ Curate as KNOWLEDGE_GAP.                                         │
└──────────────────────────────────────────────────────────────────┘

6.2 Molecular pathways

• Pyrimidine deoxyribonucleotide biosynthesis — KEGG hsa00240 (Pyrimidine metabolism). Reactome hosts the reaction under pyrimidine deoxyribonucleotide biosynthesis (the Reactome ContentService returned 403 in this session; fetch and verify the exact stable ID before curating — do not assert an R-HSA ID from memory).
• De novo arm: dUMP —(TYMS, folate-dependent)→ dTMP
• Salvage arm: thymidine —(TK1 cytosolic / TK2 mitochondrial)→ dTMP
• Convergence point: dTMP —(DTYMK/TMPK, blocked)→ dTDP —(NME/NDPK)→ dTTP
• Downstream: DNA replication, DNA repair, mtDNA replication.

6.3 Cellular processes and GO term suggestions

All GO IDs above are candidates — verify each with uv run runoak -i sqlite:obo:go info GO:XXXXXXX -O obo before commit.

6.4 Protein dysfunction

Loss of function via failure of obligate homodimerization. Not misfolding-aggregation, not gain of function, not dominant-negative (heterozygous parents and heterozygous mice are unaffected). Structural rationale: the substituted residues (P81, A99, D128) do not contact substrate; structural modelling explains how each substitution destabilizes the dimer interface / catalytic architecture (PMID:34926941). Structure resources: PDB entries for human TMPK exist (search "human thymidylate kinase" in PDB); AlphaFold model AF-P23919.

6.5 Metabolic changes

• Profound reduction of dTDP/dTTP synthetic flux through the canonical route.
• Paradox to curate carefully: measured steady-state dNTP pools in patient fibroblasts and in zebrafish mutant larvae "resemble those of normal controls" [full-text]. The disease is therefore best modeled as a flux/compartment/demand-limited defect, not a bulk pool-depletion defect. This distinction matters for any downstream biomarker claim — a normal fibroblast dNTP panel does not exclude CONPM.
• Secondary systemic markers: elevated lactate (2/2) and hyperalaninemia (1/2) → a mitochondrial-disease-like biochemical signature that will route these patients into an MDDS workup.
• CHEBI candidates (verify with OAK): dTMP CHEBI:17013, dTDP CHEBI:58369, dTTP CHEBI:37568, ATP CHEBI:30616, thymidine CHEBI:17748, deoxycytidine CHEBI:15698, L-lactate CHEBI:16651, L-alanine CHEBI:16977.

6.6 Immune system involvement

None known. No autoimmunity, immunodeficiency, or interferonopathy has been reported. This is worth an explicit negative note because the mechanism — ribonucleotide misincorporation into genomic DNA, explicitly benchmarked in the paper against Rnaseh2-null mouse DNA — is the exact molecular lesion of Aicardi–Goutières syndrome type 4 (RNASEH2B/C/A), which is a type I interferonopathy. Whether CONPM has an unrecognized cGAS-STING/interferon component is an explicitly attractive, entirely untested hypothesis and an excellent KNOWLEDGE_GAP / proposed_experiments item (measure interferon signature in patient fibroblasts and dtymk zebrafish).

6.7 Tissue damage mechanisms

Neuronal apoptosis and neuronal dropout with reactive gliosis; not oxidative stress, ischemia, or fibrosis. Zebrafish histology showed "empty spaces, indicative of neurodegeneration" in brain [full-text]. Human neuropathology: "massive neuronal dropout, only sparing the dentate nucleus and brain stem" [full-text].

6.8 Biochemical abnormalities

Enzyme deficiency (EC 2.7.4.9). Diagnostic assay: dTMPK activity in cultured fibroblasts (see §10). No receptor or ion-channel defect.

6.9 Epigenetic changes

No data.

6.10 Molecular profiling

• Transcriptomics: no patient RNA-seq published for CONPM. GTEx/HPA show broad, non-tissue-specific DTYMK expression.
• Proteomics / metabolomics / lipidomics: no disease-specific studies. No MetaboLights or PRIDE datasets for CONPM.
• Single-cell / spatial: none.
• Functional genomics screens: DTYMK is a common essential gene in DepMap CRISPR screens (consistent with the mouse embryonic lethality) — verify the current DepMap common-essential call before citing. The uveal-melanoma study demonstrates pharmacological DTYMK inhibition (YMU1) synergizing with PARP1 inhibition (pamiparib) (PMID:39195238), which is independent orthogonal support for the "DTYMK loss → DNA repair burden" arm of the mechanism, in a cancer rather than neuronal context.

Verbatim abstract (PMID:39195238, Cells 2024):

"Our hypothesis of the double hit into tumoral DNA metabolism as a possible therapeutic option in uveal melanoma was confirmed since combined targeting of DTYMK and PARP1 affected all tested cytophysiological parameters with the highest efficiency."

7. Anatomical Structures Affected

Organ level

• Primary: brain (central nervous system). UBERON:0000955 (brain).
• Secondary: none primarily; respiratory (aspiration pneumonia, respiratory failure, laryngomalacia), GI (dysphagia, feeding failure), haematological (microcytic anemia, 1 patient), hepatic (transaminase elevation, 1 patient), urogenital (cryptorchidism, micropenis).
• Body systems: nervous (dominant); with secondary respiratory, digestive, and reproductive involvement.

Regional CNS involvement — the sparing pattern is diagnostically important

• Lateralization: bilateral and symmetric (generalized atrophy). No asymmetry reported.

Tissue and cell level

Key modeling caution: the phenotype has two temporally distinct cellular substrates — (a) a proliferative phase defect (neural progenitors, replication/S-phase) explaining microcephaly, and (b) a postmitotic defect (mature neurons, DNA repair burden) explaining progressive degeneration. These should be separate pathophysiology nodes with distinct biological_scale: CELLULAR tags, not bundled.

Subcellular level

8. Temporal Development

Onset

• Congenital onset (HP:0003577) 2/2 and infantile onset (HP:0003593) 2/2 — both annotated, reflecting that prenatal/neonatal features (SGA, polyhydramnios, breech, neonatal feeding failure, neonatal hypotonia) precede the overt neurological syndrome.
• Prenatal signals are nonspecific: polyhydramnios 1/2, breech presentation 1/2, SGA 1/2. Head circumference is near-normal at birth (Individual I: 31 cm, 2nd centile) — CONPM is not reliably detectable by prenatal or newborn OFC.
• Onset pattern: insidious/chronic-progressive with a subacute regression phase in infancy.

Progression

• Rate: rapid by neurodegenerative standards. OFC crosses from ~2nd centile to −7.6 SD within 9 months (Individual I) and to −7.4 SD by 26 months (Individual II) [full-text].
• Course: relentlessly progressive; no remission, no relapsing-remitting pattern, no plateau documented in the severe form.
• Stages (proposed, for curation):
• Neonatal (0–3 mo): hypotonia, feeding failure, NG-tube dependence, absent visual fixation. OFC near-normal.
• Infantile decline (3–15 mo): rapid OFC deceleration, spasticity replacing hypotonia, opisthotonus, seizure onset, no milestones acquired/loss of acquired milestones.
• Advanced (15 mo–death): spastic tetraplegia, myoclonus, status epilepticus, flat-trace EEG, complete cortical/striatal atrophy on imaging, respiratory and swallowing failure.
• Terminal: intercurrent respiratory infection → pneumonia/cardiopulmonary arrest.
• Duration: lifelong; life-limiting. Death in childhood 2/2 (HP:0003819) — ages 18 and 32 months in the two severe cases.
• Milder end of spectrum: the Lam siblings survived to allow assessment of intellectual disability and speech, indicating a genuinely broader survival range. Precise ages/outcomes require the Clin Chim Acta full text.

Patterns

• Remission: none, spontaneous or treatment-induced.
• Critical periods: the first 12 months — the window during which OFC collapses and the atrophy becomes established. Any disease-modifying intervention would almost certainly need to act prenatally or in the first months. Curate this as the therapeutic window, and note that no such intervention exists.
• Fever/illness as a decompensation trigger: suggested by febrile seizures and by the fact that both deaths followed intercurrent respiratory infection, but not established as a metabolic-decompensation mechanism.

9. Inheritance and Population

Epidemiology

• Prevalence: not documented. No ORPHA prevalence class is available (no Orphanet entry found). For a dismech Prevalence record the honest structured encoding is:
• measure_type: CASES_IN_LITERATURE
• prevalence_class: ULTRA_RARE (or NOT_YET_DOCUMENTED)
• notes: "Five patients reported worldwide as of 2025 (Vanoevelen 2022 n=2; Lam 2019 n=2; Hernández-Carreto 2025 n=1)."
• Do not populate rate_per_100000.
• Incidence: unknown.
• Verbatim abstract (PMID:40696808): "Only four cases have been reported in the literature to date. This paper's objective is to describe the fifth globally reported case of CONPM and the first documented in a Mexican patient…"

Genetic epidemiology

Population demographics

• Affected populations: Dutch, Egyptian, Chinese (Hong Kong), Mexican — i.e., no ethnic clustering; the disorder is pan-ethnic and its rarity is a function of allele rarity, not population structure.
• Geographic distribution: none (case reports from Europe, North Africa, East Asia, Latin America).
• Variant geography: p.Pro81Leu reported in Netherlands, Egypt, Mexico; p.Asp128Asn in the Netherlands; p.Ala99Thr and p.Asp96fs in Hong Kong.
• Sex ratio: among the fully characterized cases, 1 female (Individual I) and 2 males (Individual II, Mexican case); the Lam siblings' sexes require the full text. n = 5 is far too small to infer a sex ratio — curate as "no sex bias reported" rather than a ratio.
• Age distribution: all affected individuals identified in infancy/early childhood; no adult-onset or adult-diagnosed cases.

10. Diagnostics

Clinical / laboratory tests

Imaging

Brain MRI is the highest-yield diagnostic modality. Expected findings: - Profound generalized cerebral atrophy (HP:0002059) - Marked lateral ventricular dilatation (HP:0006956) and widened subarachnoid spaces (HP:0012704) - Small/"disappeared" basal ganglia — the most distinctive feature - Normal-sized thalamus, brainstem, and (usually) cerebellum - Serial imaging showing progression is more informative than a single study.

Verbatim (PMID:34918187 abstract): "Brain imaging revealed severe cerebral atrophy and disappearance of the basal ganglia."

Electrophysiology

• EEG: abnormal; flat trace documented in the advanced stage of Individual I [full-text]. No pathognomonic pattern.
• EMG/NCS/ECG: no reported abnormalities; peripheral nerve involvement not described.

Biopsy / pathology

• Neuropathology (post-mortem, Individual I): massive neuronal dropout with sparing of dentate nucleus and brainstem; gliosis [full-text].
• Skin biopsy for fibroblast culture is the key diagnostic specimen (enables the enzyme assay).
• Muscle biopsy: no characteristic ragged-red/COX-negative findings reported; if the MDDS hypothesis is pursued, muscle mtDNA quantification would be the test.

Genetic testing

• Recommended first-tier approach: whole-exome sequencing (WES) or whole-genome sequencing (WGS) with trio analysis. All five reported patients were diagnosed by exome sequencing. WES is what identified the Mexican case ("confirmed through whole-exome sequencing (WES)", PMID:40696808 verbatim) and the Lam siblings ("whole exome sequencing is often needed for their diagnoses", PMID:31271740 verbatim).
• Gene panels: DTYMK should be — and in some laboratories now is — included on (a) progressive/primary microcephaly panels, (b) neurodegeneration-in-childhood panels, and (c) mitochondrial DNA depletion syndrome panels. Its inclusion is inconsistent across vendors; verify panel content in GTR for any specific lab before recommending. Confirm current listings at https://www.ncbi.nlm.nih.gov/gtr/.
• Single-gene testing: appropriate only for targeted familial-variant testing / cascade screening after a proband diagnosis.
• Sanger confirmation + segregation in both parents is essential to establish biallelic status and phase (as done in PMID:40696808).
• CMA / karyotype / FISH: low yield for CONPM itself. CMA remains reasonable as part of a general microcephaly workup and would detect a 2q37.3 deletion contributing one allele in a compound-heterozygous configuration — worth explicitly considering, since a whole-gene deletion in trans with a point variant would be missed by exome-only analysis.
• mtDNA testing: mtDNA sequencing and copy-number quantification are indicated given the MDDS differential and the lactate/alanine profile.
• Repeat expansion testing: not indicated.
• RNA-seq: useful as a second-tier tool to resolve the deep-intronic VUS (c.239+1045_239+1050del, VCV003065236) or other candidate splice-affecting alleles.

Clinical criteria and differential diagnosis

No consensus diagnostic criteria exist (too few patients). Diagnosis = compatible phenotype + biallelic DTYMK variants (± functional confirmation).

Differential diagnosis — the most useful section for a curator, because CONPM's imaging and biochemical profile overlaps several well-known entities:

Screening

• Newborn screening: not applicable. There is no analyte biomarker, no treatment, and head circumference is near-normal at birth. CONPM fails standard Wilson–Jungner criteria.
• Carrier screening: DTYMK is not on standard expanded carrier screening panels. Its inclusion would be defensible only in the context of a known family.
• Cascade screening: targeted variant testing of at-risk relatives and reproductive partners after a proband diagnosis — this is the highest-value screening application.

11. Outcome / Prognosis

Survival and mortality

• Death in childhood (HP:0003819) in 2/2 of the severely affected annotated individuals.
• Documented ages at death: 18 months (Individual I, cardiopulmonary arrest following respiratory illness) and 32 months (Individual II, pneumonia and coma) [full-text].
• The milder Lam siblings survived long enough for formal intellectual assessment, establishing that survival beyond early childhood occurs at the mild end. Precise survival data require the primary full text.
• No 5-/10-year survival statistics, no life-expectancy estimate, no mortality rate — n is too small. For dismech, curate "death in early childhood in the severe form; survival documented in the milder form" with the two specific ages, rather than any derived rate.
• Disease-specific mortality: deaths are attributable to the disease via its complications (respiratory infection, aspiration, respiratory failure) rather than to a single organ failure.

Morbidity and function

• Profound, permanent, global disability: no independent sitting, rolling, vocalizing, or smiling in the severe form ("he did not roll over, sit, vocalize, or smile") [full-text]; total care dependence; enteral feeding; refractory epilepsy; cortical blindness; spastic tetraplegia.
• No quality-of-life instrument has been applied. No EQ-5D, PedsQL, PROMIS, or CPCHILD data exist. Do not populate QoL scores.
• GBD/WHO carry no disease-specific burden estimate.

Complications

Aspiration and recurrent respiratory infection → pneumonia (a documented cause of death); respiratory failure (HP:0002878); laryngomalacia (HP:0001601); dysphagia (HP:0002015) with tube-feeding dependence (HP:0033454); status epilepticus (HP:0002133); failure to thrive/short stature; contractures secondary to spastic tetraplegia (expected, not explicitly reported); microcytic anemia; transaminase elevation.

Recovery potential

None. Neuronal loss is irreversible; no disease-modifying therapy exists; no recovery or plateau has been documented in the severe form.

Prognostic factors

No validated prognostic model. Observationally suggestive (all low-confidence, n=5): - Rate of OFC decline in the first year is the most face-valid clinical prognostic index. - Age at seizure onset (6 months vs. 15 months, both severe). - Residual enzyme activity: biochemically plausible as the primary determinant (p.Ala99Thr retains the highest residual kcat/KM of the characterized missense alleles, and the Lam siblings carrying A99T are at the milder end — an appealing but unproven genotype–phenotype correlation with n=2; explicitly flag as a hypothesis, not a finding). - Prognostic biomarkers: none validated.

12. Treatment

There is no disease-specific or disease-modifying therapy. Management is entirely supportive.

Pharmacotherapy

Pharmacogenomics: no CPIC/PharmGKB guidance specific to DTYMK. A worth-noting theoretical consideration (unstudied — do not curate as evidence): the affected pathway is the target of thymidylate-synthase inhibitors (5-FU, capecitabine) and antifolates (methotrexate); whether DTYMK carriers or patients have altered sensitivity is unknown.

Advanced therapeutics

• Gene therapy: none. No preclinical AAV or gene-replacement program identified. DTYMK is small (212 aa CDS, well within AAV capacity) and the disease is recessive LoF — i.e., theoretically tractable — but the postnatal-degeneration timeline and the near-complete atrophy by 12–24 months make the therapeutic window extremely narrow, and no program exists.
• Gene editing / cell therapy / mRNA / siRNA / ASO: none. ASO is mechanistically inapplicable (missense LoF, not a splice or knockdown target — with the possible exception of the deep-intronic VUS if it proves splice-altering).
• Enzyme replacement: not feasible — TMPK is an intracellular, nuclear/cytosolic/mitochondrial kinase acting on a phosphorylated, membrane-impermeant substrate.

Substrate/nucleoside supplementation — the one active clinical-trial handle, with a critical mechanistic caveat

NCT04802707 — "Deoxynucleosides Pyrimidines as Treatment for Mitochondrial Depletion Syndrome", Phase II, open-label, single-centre, recruiting; deoxycytidine (dC) + deoxythymidine (dT) orally, escalating over 22 days then maintained at 400 mg/kg; ages 0–60; up to ~200 participants. This trial explicitly lists DTYMK among its eligible genotypes.

The record is already present in the dismech reference cache (references_cache/clinicaltrials_NCT04802707.md), and the following is a verified quotable snippet from that cached file:

"The subjects included are children (0-18Y), with positive MDS diagnosis and express mutations in one of the following genes: POLG, POLG2, C10orf2, RRM2B, MPV17, SUCLA2, SUCLG1, FBXL4, DTYMK."

⚠️ Mechanistic caveat that MUST accompany this trial in the knowledge base. Deoxynucleoside substrate-enhancement therapy works in TK2 deficiency because supplying dThd/dCtd bypasses a kinase-limited first salvage step. In DTYMK deficiency the block is at dTMP → dTDP, i.e., downstream of where supplemental thymidine enters the pathway (thymidine → TK1/TK2 → dTMP → [BLOCK]). Supplying more dThd therefore increases the substrate that is already accumulating proximal to the block and has no obvious mechanistic route to restoring dTTP. Supplemental dCtd could in principle relieve dNTP-pool imbalance on the pyrimidine side, but that is speculative.

Recommended curation: record NCT04802707 as a clinical trial whose eligibility includes DTYMK, and attach an explicit discussions entry of kind: KNOWLEDGE_GAP (or a mechanistic_hypotheses entry with status: EMERGING) stating that the substrate-bypass rationale is not established for DTYMK given the position of the enzymatic block, with proposed_experiments = measure dTTP pools and mtDNA copy number in DTYMK-deficient cells ± dT/dC. Do not curate dC/dT as an evidenced treatment for CONPM.

Surgical and interventional

• Gastrostomy (PEG) for tube-feeding dependence — clinically standard for this level of dysphagia (NG tube documented; PEG not explicitly reported). NCIT:C15329 Surgical Procedure; consider a gastrostomy-specific NCIT term.
• Orchidopexy for cryptorchidism, where clinically indicated. NCIT:C16186 Orthopedic Surgical Procedure is not correct here — look up a urological/orchidopexy NCIT term.
• Airway management for laryngomalacia and respiratory failure.

Supportive and rehabilitative

All NCIT IDs must be verified with uv run runoak -i sqlite:obo:ncit info NCIT:Cxxxxx -O obo.

Treatment outcomes, adverse events, algorithms

• No response-rate data (no disease-specific therapy has been trialled in CONPM).
• No CONPM-specific adverse-event data. FAERS contains nothing indexed to this disease.
• No treatment algorithm, guideline, NCCN/society pathway, or GeneReviews chapter exists for CONPM. Management follows generic severe-neurodegenerative-encephalopathy principles.
• Personalized medicine: the only genotype-guided element currently available is reproductive (see §13).

13. Prevention

Primary prevention

Not achievable for an affected fetus/child — the disorder is determined at conception. Prevention is entirely reproductive: - Genetic counselling for at-risk couples: 25% recurrence risk per pregnancy, 50% carrier risk for unaffected sibs, standard AR counselling. NCIT:C15240. - Carrier testing of the proband's parents and extended family (particularly relevant in consanguineous kindreds). - Prenatal diagnosis by CVS or amniocentesis with targeted testing of the known familial variants. - Preimplantation genetic testing for monogenic disease (PGT-M) — technically straightforward once the familial variants are known. - Consanguinity counselling in populations where first-cousin union is common.

Secondary prevention (early detection)

• No population screening is justified or available. Newborn screening is not applicable (no analyte, no treatment, normal birth OFC).
• The practical "early detection" measure is serial head-circumference monitoring in infancy, which will flag the OFC deceleration and trigger a workup — but this detects the disease after neuronal loss has begun and does not alter outcome.
• Cascade testing within a known family is the only genuinely effective early-detection route.

Tertiary prevention (preventing complications) — this is where real clinical benefit lies

• Aspiration prevention: early swallow assessment, thickened feeds, timely conversion from NG to gastrostomy. Directly targets the documented cause of death (pneumonia).
• Respiratory infection prophylaxis: routine immunizations, influenza and RSV prophylaxis, pneumococcal vaccination, chest physiotherapy, prompt treatment of intercurrent infection.
• Seizure control to prevent status epilepticus.
• Contracture and positioning management for spastic tetraplegia.
• Nutritional optimization to prevent iron-deficiency/microcytic anemia and further growth failure.

Immunization

No disease-specific vaccine. Routine childhood immunization plus respiratory-pathogen prophylaxis is a high-value intervention given that both documented deaths followed respiratory infection. NCIT:C15346 Vaccination.

Public health and environmental interventions

Not applicable. No environmental modifiable risk exists.

14. Other Species / Natural Disease

Orthologs (Alliance of Genome Resources, all high-confidence "best score")

The yeast ortholog name CDC8 (cell division cycle 8) — which is also the human gene's legacy alias — is a direct historical signal of the gene's core cell-cycle function and of deep evolutionary conservation.

Naturally occurring disease in other species

None known. A search of OMIA and the veterinary literature identified no naturally occurring DTYMK-related disease in companion animals, livestock, or wildlife. No breed-associated variant; no VBO breed identifier applicable. No zoonotic potential and no cross-species transmission (this is a Mendelian metabolic disorder, not a transmissible condition).

Comparative biology

• Evolutionary conservation is exceptionally deep — a functional TMPK ortholog is present from yeast through humans, consistent with dTTP synthesis being a universal requirement.
• Comparative pathology diverges in an interesting and curation-relevant way:
• Mouse: homozygous null is embryonic lethal (PMID:38621447) — mouse does NOT recapitulate the human "viable but neurodegenerative" phenotype.
• Zebrafish: homozygous mutant is viable through early larval stages with microcephaly and neurodegeneration, then lethal — this is the closest phenocopy.
• Human: viable to birth with normal-sized head, then postnatal neurodegeneration.
• This species gradient (mouse lethal < zebrafish larval-lethal < human postnatal-degenerative) is itself evidence that the hypothesized compensatory dTTP pathway differs in capacity across species — an elegant framing for the HUMAN_MODEL_MISMATCH discussion.
• Enzymatic substrate specificity diverges too: Drosophila TMPK phosphorylates dTMP, dUMP, and also dGMP and dIMP (at low efficiency), unlike human TMPK (PMID:38518117 abstract, verbatim below) — relevant if a fly model is used, because compensation logic may differ.

Verbatim abstract (PMID:38518117):

"Unlike human TMPK, DmTMPK phosphorylated not only dTMP and dUMP but also dGMP and dIMP although with low efficiency. ATP and dATP are the most efficient phosphate donor but at higher concentration (>1 mM) ATP inhibited DmTMPK activity."

15. Model Organisms

15.1 Zebrafish — the flagship model (best phenocopy)

Allele: CRISPR-generated 5-bp deletion in exon 4, producing a premature stop codon 19 amino acids downstream [full-text]. Source: Vanoevelen et al., PMID:34918187. Should be registered/lookup-able at ZFIN (ZDB-GENE-990603-11).

Phenotype recapitulation — strong:

Limitations: (i) the fish shows non-CNS malformations (cardiac/intestinal edema, absent Meckel's cartilage, microphthalmia) that are not part of the human phenotype, suggesting a broader requirement in fish; (ii) the compressed larval timeline cannot model postnatal progressive degeneration over months–years; (iii) the two mechanistically most novel findings — ribonucleotide misincorporation and defective DDR — are demonstrated only in fish, never in human tissue. → This is a textbook HUMAN_MODEL_MISMATCH discussion item (evidence exists in a model; human translational validity is the open question), not a plain KNOWLEDGE_GAP.

Verbatim abstract (PMID:34918187):

"In addition, we generated dtymk mutant zebrafish that replicate this phenotype of microcephaly, neuronal cell death and early lethality. An increase of ribonucleotide incorporation in the genome as well as impaired responses to DNA damage were observed in dtymk mutant zebrafish, providing novel pathophysiological insights."

Supporting zebrafish developmental work (PMID:35346037, BMC Neuroscience 2022) — verbatim abstract excerpts:

"Our findings reveal that maternal-stored dNTPs are only sufficient for 6 cell division cycles, and the levels of dNTPs are inversely correlated to cell cycle length during early embryogenesis. TMPK and TK activities are prominent in the cytosol of embryos, larvae and adult fish and brain contains the highest TMPK activity." "Our results suggest that active dNTP synthesis in early embryogenesis is vital and that Dtymk is essential for neurodevelopment, which is supported by a recent study of dtymk knockout zebrafish with neurological disorder and lethal outcomes. Furthermore, there is a novel TMPK-like enzyme expressed at later stages of development."

Note the last sentence — it is independent, orthogonal support for the "unknown compensatory TMPK-like enzyme," this time from a developmental-stage expression study rather than from patient cells.

15.2 Mouse — essentiality established, disease NOT modeled

Tiani & Stover, Arch Biochem Biophys 2024, PMID:38621447, DOI 10.1016/j.abb.2024.109991.

• Homozygous Dtymk knockout is embryonic lethal.
• Heterozygotes across three dietary conditions showed no open neural tube defects, despite ~3-fold reduced dTYMK expression.
• Implication: the standard mouse null is not a usable CONPM model. A viable model would require a hypomorphic knock-in (e.g., the human p.Pro81Leu or p.Ala99Thr allele) or a conditional/neural-specific conditional knockout (Nestin-Cre, Emx1-Cre). No such mouse has been reported — this is the single most valuable missing reagent in the field and a strong proposed_experiments entry.
• Resources to check for existing alleles: MGI (MGI:108396), IMPC, KOMP/EuMMCR, IMSR.

15.3 Drosophila — characterized enzyme, no disease model yet

PMID:38518117 (Hu Frisk & Wang 2024) cloned, expressed, purified, and kinetically characterized DmTMPK, explicitly as groundwork for a fly disease model: "Drosophila has been used as an animal model to study pathogenic mechanism of neurological disorders… This study has laid a solid foundation for future study of TMPK function in Drosophila." No Dtymk mutant fly phenotype has been published. Caveat: broader substrate specificity than human TMPK (see §14).

15.4 In vitro / cellular models

• Patient-derived dermal fibroblasts (Individual I) — the workhorse. Available from the Vanoevelen/Bierau group (Maastricht). Used for enzyme assay, subcellular fractionation, S-phase analysis, dNTP pools. Key limitation: fibroblasts proliferate normally and thus do not phenocopy the disease — they are a biochemical readout, not a disease model.
• Recombinant TMPK (WT, P81L, A99T, D128N) expressed and purified for kinetics and size-exclusion dimerization analysis (PMID:34926941).
• iPSC / neural organoids: none reported. Given that (a) the phenotype is neuron-specific, (b) fibroblasts are spared, and (c) the compensating enzyme is unknown, patient-iPSC-derived cortical neurons and cerebral organoids are the obvious highest-value missing model. This is also directly relevant to MorPhiC-style cellular-phenotype curation (§CLAUDE.md MorPhiC pattern): a DTYMK-null iPSC line with category: Cellular phenotypes and evidence_source: IN_VITRO would be a natural future annotation — but note that DTYMK is not among the current MorPhiC anchor genes (ISL1, EOMES, GCM1, NKX2-1), so no MorPhiC data exist today.
• Cancer cell lines: MP41/MP46 uveal melanoma with the TMPK inhibitor YMU1 (PMID:39195238) — a pharmacological loss-of-function system, useful as orthogonal mechanistic support but not a neurodegeneration model.

15.5 Yeast

S. cerevisiae CDC8 — the classical cell-division-cycle mutant. Historically the source of the "TMPK is required for cell-cycle progression" understanding. Not used as a CONPM model but valuable for conservation arguments.

Appendix A — Master citation list

Database records: OMIM #619847; OMIM *188345; MONDO:0859241; MedGen C5676972 / UID 1801540; HGNC:3061; UniProt P23919; ClinVar VCV001686902/3/4/5, VCV004277573, VCV003065236; Human Protein Atlas ENSG00000168393; Alliance of Genome Resources HGNC:3061 orthology; HPO annotation set for OMIM:619847.

Appendix B — Curation checklist and flagged gaps for the dismech entry

Verification required before commit 1. Run just fetch-reference for every PMID above; verify every snippet is an exact substring. All [full-text]-marked sentences in this report came through an intermediate summarizer and must be re-verified against the PMC source (PMC8742820, PMC8679000) before use as evidence. 2. Run just validate-terms on every HP/GO/CL/UBERON/CHEBI/NCIT ID suggested here — none have been OAK-verified in this session. 3. Fetch gnomAD constraint values directly; do not assert pLI/LOEUF from this report (retrieval failed). 4. Verify the Reactome stable ID for pyrimidine deoxyribonucleotide biosynthesis (ContentService returned 403). 5. Obtain the Lam 2019 full text to confirm (a) the phase/pairing of A99T + D96fs, (b) the siblings' sexes and ages, (c) whether mtDNA depletion was measured or only inferred.

Recommended discussions entries - kind: KNOWLEDGE_GAP — identity of the compensatory TMPK-like enzyme. Attaches to the dTTP-biosynthesis-block node. Supported by three independent papers (PMID:34926941, 34994281, 35346037). Proposed experiments: unbiased biochemical purification of dTMP-kinase activity from patient fibroblasts; CRISPR screen for synthetic lethality with DTYMK loss. - kind: HUMAN_MODEL_MISMATCH — ribonucleotide misincorporation and DDR failure are shown only in zebrafish, never in human neurons or patient tissue; and mouse null is embryonic lethal while humans reach term. Proposed experiments: patient-iPSC cortical neurons/organoids; alkaline-gel and γH2AX assays in human cells; humanized hypomorphic knock-in mouse. - kind: KNOWLEDGE_GAP — interferon signature untested, despite the mechanistic parallel to RNASEH2-deficient AGS explicitly drawn in the primary paper. - mechanistic_hypotheses (status: EMERGING) — mtDNA depletion arm (Lam 2019, mitochondrial TMPK activity loss in PMID:34926941). Edges from the mitochondrial-TMPK-loss node should opt into this hypothesis group. Flag that the mtDNA depletion itself was shown "in silico" only. - kind: KNOWLEDGE_GAP — deoxynucleoside (dC/dT) substrate therapy rationale is not established for DTYMK, because the enzymatic block lies downstream of thymidine entry; yet NCT04802707 lists DTYMK as eligible. - Inter-patient discrepancy: cerebellar sparing (Vanoevelen) vs. cerebellar atrophy (Hernández-Carreto 2025).

Suggested biological_scale tags for pathophysiology nodes MOLECULAR — DTYMK LoF variants; loss of homodimerization; loss of dTMP kinase activity; dTTP biosynthesis block. CELLULAR — impaired DNA replication / S-phase collapse; ribonucleotide misincorporation; impaired DNA damage response; neuronal apoptosis. TISSUE — cerebral and striatal atrophy with gliosis. ORGANISM — progressive microcephaly; developmental regression; lactic acidemia; early childhood death.

Module conformance candidates: none of the existing dismech modules is a clean fit. genome_instability_mutation covers a genome-maintenance-defect → mutator-phenotype chain but is scoped to oncogenesis, and CONPM's output is neuronal apoptosis rather than clonal evolution — do not force conformance. If a module is warranted later, the natural one would be a new "nucleotide-pool-imbalance genome instability" or "replication-stress neurodegeneration" module shared with AGS/RNASEH2, PNKP, and the MDDS nucleotide-salvage disorders.

Sources

• DTYMK is essential for genome integrity and neuronal survival — Acta Neuropathologica (PMC8742820)
• Biochemical Characterizations of Human TMPK Mutations Identified in Patients with Severe Microcephaly — ACS Omega (PMC8679000)
• Deoxythymidylate kinase, DTYMK, is a novel gene for mitochondrial DNA depletion syndrome — PubMed 31271740
• Childhood-Onset Neurodegeneration With Progressive Microcephaly (CONPM) due to a DTYMK Homozygous Pathogenic Variant — Am J Med Genet A (PMID 40696808)
• Differential expression of enzymes in thymidylate biosynthesis in zebrafish — BMC Neuroscience
• Structural and functional analysis of human thymidylate kinase isoforms — PubMed 34994281
• DTYMK is an essential gene in mice and heterozygosity does not cause neural tube defects — PubMed 38621447
• Molecular characterization of Drosophila melanogaster thymidylate kinase — PubMed 38518117
• Co-Targeting of DTYMK and PARP1 as a Potential Therapeutic Approach in Uveal Melanoma — Cells 2024
• OMIM #619847 — NEURODEGENERATION, CHILDHOOD-ONSET, WITH PROGRESSIVE MICROCEPHALY; CONPM
• OMIM *188345 — DEOXYTHYMIDYLATE KINASE; DTYMK
• MedGen — Childhood-onset neurodegeneration with progressive microcephaly (UID 1801540)
• HPO annotations for OMIM:619847 — Human Phenotype Ontology
• MONDO:0859241 — EBI Ontology Lookup Service
• HGNC:3061 — DTYMK
• UniProt P23919 — Thymidylate kinase
• Human Protein Atlas — DTYMK tissue expression
• Alliance of Genome Resources — DTYMK orthologs
• ClinVar — DTYMK variants
• NCT04802707 — Deoxynucleosides Pyrimidines as Treatment for Mitochondrial Depletion Syndrome