Blaschko‐Linear <scp>TGM1</scp> ‐ <scp>nEDD</scp> Associated With Mosaic Genome‐Wide Uniparental Isodisomy

Autosomal recessive congenital epidermal differentiation disorders (EDDs) (OMIM #190195), also known as the ichthyoses, are associated with biallelic pathogenic variants in genes critical for epidermal barrier function, including TGM1 and ABCA12 (Herman et al. 2009). Pathogenic variants in TGM1 lead to transglutaminase-1 deficiency, a hallmark of TGM1-nEDD, also known as lamellar ichthyosis, due to preventing the crosslinking of precursor proteins during the formation of the cornified cell envelope cytoplasmic layer, a structure essential for the skin's barrier function (Herman et al. 2009). In two previously reported patients with Blaschko-linear EDD, postzygotic second-hit variants in ABCA12 were found, leading to a linear ichthyosiform erythroderma phenotype (Bhatia et al. 2025; van Leersum et al. 2020). Here, we report the first documented patient with mosaic TGM1-nEDD associated with genome-wide uniparental isodisomy (UPD). A six-year-old girl was born at 39 weeks of gestation via spontaneous vaginal delivery following an uncomplicated pregnancy. Birth weight was 2784 g (5.7th percentile), length 44 cm (< 1st percentile), and head circumference 34 cm (51st percentile). Since birth, she has had persistent short stature and low weight (both < 1st percentile) with relative macrocephaly (10th percentile), despite adequate caloric intake and absence of gastrointestinal symptoms. Developmental milestones were achieved on time, and there are no neurologic abnormalities. At birth, she had adherent dark scale on the face and scalp and Blaschko-linear hyperpigmented scaling on the chest, back, and extremities, without a collodion membrane. Physical examination at age 6 demonstrated hyperkeratotic papules, hyperpigmented patches, and plate-like scales in linear patterns along the lines of Blaschko, involving the right trunk, back, axilla, legs, and arms (Figure 1a–e). There were no extracutaneous abnormalities, hemihypertrophy, or dysmorphic features. Biopsy of affected skin showed prominent orthokeratotic hyperkeratosis, a normal-to-focally diminished granular layer, minimal spongiosis, and a mild superficial dermal perivascular mononuclear infiltrate. Unaffected skin exhibited an unremarkable epidermis with a basket-weave stratum corneum and normal thickness of the granular, spinous, and basal layers (Figure 2a,b). This study was approved by the Yale University Institutional Review Board and was conducted in accordance with the Declaration of Helsinki. Written informed assent and consent were obtained from the patient and her parent. Paired exome sequencing (ES) of saliva and cultured keratinocytes from an affected skin sample was performed. ES of saliva identified a heterozygous pathogenic TGM1 (NM_000359) variant, c.425G>A (p.Arg142His). Affected keratinocytes demonstrated homozygosity for the same variant, along with genome-wide homozygosity across all high-quality variants (Figure 2c–e). Phasing analysis using SHAPEIT4, a population-based phasing tool (Delaneau et al. 2019), demonstrated that germline heterozygous variants from saliva segregated into parental haplotypes, whereas homozygous variants detected in affected keratinocytes uniformly corresponded to a single parental haplotype across 1327 independent phase blocks (Figure 2f). These findings strongly support mosaic genome-wide UPD as the genetic mechanism underlying the Blaschko-linear mosaicism observed in this patient. Mosaic presentations of autosomal recessive EDDs are rare. Cutaneous mosaicism arises from developmental genetic events that generate genetically distinct keratinocyte populations distributed along the lines of Blaschko, reflecting clonal expansion during early embryogenesis (Happle 1993; Kouzak et al. 2013; Paller 2001). When loss of heterozygosity (LOH) unmasks an autosomal recessive condition, the resulting segmental disease phenotype is termed recessive mosaicism. There is no consistent evidence that mutant or wild-type keratinocytes possess a selective advantage in mosaic skin disorders, suggesting that patterned involvement may instead reflect neutral clonal expansion during development (Paller 2001). UPD is a recognized mechanism by which both homologous chromosomes are inherited from a single parent and can result in disease by unmasking recessive variants through copy-neutral LOH or disrupting genomic imprinting (Bens et al. 2017; Engel 1980; Spence et al. 1988). UPD has been reported in several EDDs, including those caused by pathogenic variants in ALOX12B, SPINK5, ABCA12, and CERS3 (Castiglia et al. 2009; Numata et al. 2014; Polubothu et al. 2018; Zhang et al. 2022). However, these cases involve germline UPD of an individual chromosome. UPD has been identified across a range of conditions, including imprinting disorders, overgrowth syndromes, and malignancies (Inbar-Feigenberg et al. 2013; Makishima and Maciejewski 2011; Yamazawa et al. 2010). Mosaic genome-wide UPD represents the most extensive form of UPD and is exceptionally rare (Inbar-Feigenberg et al. 2013). Reports of mosaic genome-wide maternal UPD describe individuals with developmental abnormalities, intrauterine growth restriction, and relative macrocephaly (Bens et al. 2017). In contrast, mosaic genome-wide paternal UPD has been more frequently reported and is typically associated with features of Beckwith–Wiedemann syndrome, transient neonatal diabetes, and Angelman syndrome (Gogiel et al. 2013; Inbar-Feigenberg et al. 2013; Wilson et al. 2008). Non-mosaic genome-wide UPD is incompatible with life due to global disruption of genomic imprinting (Gogiel et al. 2013), whereas mosaic forms permit survival and yield highly variable phenotypes depending on the distribution of uniparental cell lineages (Inbar-Feigenberg et al. 2013; Wilson et al. 2008). Clinical features reflect both genome-wide homozygosity and dysregulation of imprinted gene expression across multiple loci. Intrauterine and postnatal growth restriction are common in mosaic genome-wide maternal UPD, likely reflecting the influence of maternally expressed imprinted genes and placental dysfunction, whereas mosaic paternal genome-wide UPD has been associated with placental abnormalities (i.e., cystic placenta), overgrowth phenotypes, and a significantly increased risk of malignancy (Bens et al. 2017; Gogiel et al. 2013; Wilson et al. 2008; Yamazawa et al. 2010). Most reported cases of mosaic genome-wide UPD have been identified through phenotypes associated with imprinting (Bens et al. 2017), and lineage-restricted manifestations confined to cutaneous tissues remain exceedingly uncommon. Although parental origin could not be determined in this case due to a lack of parental samples and the absence of methylation analysis, the patient's clinical features, including small size for gestational age, persistent growth restriction, and relative macrocephaly, are more consistent with patterns described in mosaic maternal genome-wide UPD (Bens et al. 2017). However, this interpretation remains speculative, and parental DNA would be necessary to confirm our interpretation. Mosaic genome-wide UPD is thought to arise from errors occurring at or shortly after fertilization that disrupt normal biparental genomic contribution. These include abnormalities in pronuclear development, such as failure of one parental pronucleus to replicate or contribute, retention of extra maternal genetic material (i.e., polar body retention), or duplication of a single parental haploid genome (endoreduplication), resulting in a diploid cell population derived from one parent. When such events occur early in embryogenesis, they can generate a mosaic blastocyst composed of both biparental and uniparental cell lineages, which subsequently contribute to different tissues during development. The Blaschko-linear distribution in our patient supports expansion of a progenitor keratinocyte population derived from a uniparental lineage during early embryogenesis. Given that TGM1 deficiency impairs epidermal barrier function, affected keratinocytes would not be expected to confer a biological advantage. However, we cannot exclude the possibility that differential survival or subtle developmental dynamics influenced the persistence of the UPD cell population. The apparent confinement of the phenotype to skin does not exclude the presence of low-level mosaicism in other tissues. Prior reports of mosaic genome-wide UPD have demonstrated variable tissue distribution. Because mosaic genome-wide paternal UPD reports have demonstrated an increased risk of malignancy (Gogiel et al. 2013; Inbar-Feigenberg et al. 2013), affected individuals may warrant consideration of longitudinal clinical surveillance, although the precise risk in cases without systemic manifestations remains uncertain. This patient expands the mechanistic spectrum of mosaic EDDs by demonstrating that mosaic genome-wide UPD can underlie Blaschko-linear disease. Recognition of this genetic mechanism has important implications for diagnosis, genetic counseling, and clinical surveillance. First-degree relatives may remain carriers of the germline variant. Ultimately, genetic testing of affected tissue is critical for diagnosis. Awareness of recessive mosaicism caused by mosaic genome-wide UPD when confronted with Blaschko-linear EDD patterns in early childhood is critical for accurate diagnosis, appropriate patient counseling, and informed molecular testing strategies. This research received funding from the National Institutes of Health (R01AR068392, 1S10OD030363-01A1). A.J.L. is supported by the 2025 Research Fellowship from the Pediatric Dermatology Research Alliance. Reviewed and approved by the Yale IRB (#0809004252). The patient and their parent in this manuscript have given written informed consent to the publication of their details. K.A.C. has served as an investigator for Abbvie, Biocryst, Boehringer-Ingelheim, Janssen, and Regeneron. He has served as a consultant for Biocryst, Mitsubishi Tanabe, and Palvella Therapeutics, and as a scientific advisor for Resvita. The data that support the findings of this study are openly available in ClinVar (accession number: SCV007541352) and in dbGAP at https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs004172.v2.p1.