2.9 Mosaicism, X-Inactivation & Epigenetic Modifiers of Phenotype

2.9 Mosaicism, X-Inactivation & Epigenetic Modifiers of Phenotype

Quick Answer: Mosaicism = post-zygotic variant, so timing sets tissue distribution and blood may be falsely negative. Germline mosaicism keeps recurrence above zero after an "apparently de novo" dominant diagnosis. Skewed X-inactivation produces manifesting female carriers. Uniparental disomy usually arises by trisomy rescue, explains PWS/AS/Russell-Silver/Beckwith-Wiedemann subsets, and can unmask a recessive condition from a single carrier parent.

Subdomain 2B carries 15 scored items, and its second listed sub-topic is factors that influence phenotype and genotype. Penetrance, expressivity, anticipation, and heterogeneity are covered earlier in this chapter; this section handles the mechanisms that make a genotype inconsistent across cells or inconsistent with parental origin.

Mosaicism: one person, two or more cell lines

Mosaicism results from a post-zygotic mutational or segregation event. The earlier it occurs, the more tissues are involved.

Germline mosaicism and the recurrence conversation

This is the highest-yield mosaicism concept on the exam. When a child has an apparently de novo autosomal dominant condition and both parents test negative in blood, recurrence is low but not zero. Commonly quoted teaching figures include roughly 10–15% for Duchenne muscular dystrophy and around 6% for lethal (type II) osteogenesis imperfecta; many other dominant conditions are counseled at about 1% empirically. The counseling line is that "de novo" describes what was found in the parents' blood, not a guarantee about their gametes — and that prenatal or preimplantation testing remains available for that residual risk.

Detecting mosaicism

• Variant allele fraction (VAF) on next-generation sequencing is the practical signal: a heterozygous germline variant sits near 50%, while a mosaic variant may appear at 5–30% or lower.
• Read depth matters. Low-depth sequencing and Sanger confirmation both miss low-level mosaicism; the lab may need deep targeted sequencing or droplet digital PCR.
• Tissue choice matters. For a segmental skin phenotype, affected skin fibroblasts may be positive when blood is negative. For suspected mosaic aneuploidy, blood plus a second tissue may be needed.
• Classic somatic-mosaic disorders exist only in the mosaic state because the constitutional form is lethal: Proteus syndrome (AKT1), McCune-Albright syndrome (GNAS), and segmental NF1 are the standard examples.

X-inactivation and manifesting carriers

X-inactivation (lyonization) silences one X chromosome in each somatic cell of a 46,XX individual. It occurs early in embryogenesis, is usually random, and is clonally maintained — so all descendants of a cell keep the same inactive X.

Three consequences drive exam items:

1. Skewed X-inactivation produces manifesting carriers. If the X carrying the pathogenic variant is preferentially active, a heterozygous female can show symptoms. Classic manifesting-carrier scenarios include Duchenne muscular dystrophy (elevated CK, cardiomyopathy), Fabry disease, X-linked adrenoleukodystrophy, and hemophilia A/B with clinically relevant factor levels. A negative family history plus a symptomatic female does not exclude an X-linked condition.
2. Some genes escape inactivation. Roughly 15–25% of X-linked genes escape silencing to some degree, mostly in the pseudoautosomal regions and scattered loci. Escape explains why 45,X (Turner syndrome) has a phenotype at all — haploinsufficiency for escape genes such as SHOX — and why supernumerary X material in 47,XXY and 47,XXX is not phenotypically silent.
3. Discordant monozygotic twins. Female MZ twins can be discordant for an X-linked phenotype because inactivation was skewed differently in each twin — a favorite distractor-buster for "identical twins must be identical."

Uniparental disomy

Uniparental disomy (UPD) is inheritance of both homologues of a chromosome pair from one parent.

Mechanisms

• Trisomy rescue is the commonest route: a trisomic conceptus loses one chromosome; one-third of the time the retained pair is from a single parent. This is also why confined placental mosaicism and UPD travel together — the trisomic line persists in the placenta while the fetus is disomic.
• Monosomy rescue duplicates the single inherited chromosome, producing complete isodisomy.
• Gamete complementation and post-zygotic errors account for the remainder.

Imprinted-region UPD you must recognize

The single-carrier-parent trap. A child has a classic autosomal recessive condition, and testing shows the child is homozygous for a pathogenic variant — but only the father carries it and maternity is confirmed. The correct interpretation is paternal isodisomy for that chromosome, not non-paternity and not a lab error. Recurrence for future siblings is very low, because the event was a sporadic segregation accident rather than two carrier parents.

The epigenetic layer

Imprinting is maintained by DNA methylation at imprinting control regions, so imprinting disorders can arise by four distinct routes: deletion, UPD, imprinting-centre defect, or (for Angelman) a UBE3A sequence variant. This is why methylation-specific testing — methylation-specific MLPA or methylation-specific PCR — is the correct first-tier test for PWS/AS: it detects the abnormal methylation pattern regardless of which of the first three mechanisms produced it, and follow-up testing then distinguishes deletion from UPD from imprinting defect. Sequencing alone will miss all three.

Assisted reproductive technologies have been associated in the literature with a modest increase in imprinting disorders such as Beckwith-Wiedemann; counsel this as an association with small absolute risk rather than as an established causal quantity.

Common traps

• Reporting a negative blood test as excluding mosaicism when the phenotype is segmental.
• Telling parents of a child with an apparently de novo dominant condition that recurrence is zero.
• Concluding non-paternity when a child is homozygous for a variant carried by only one parent, instead of considering isodisomy.
• Assuming a symptomatic female excludes X-linked inheritance.
• Ordering sequencing alone for suspected Prader-Willi or Angelman syndrome instead of methylation analysis.
• Treating heterodisomy and isodisomy as interchangeable when reasoning about recessive unmasking.