RNA sequencing to support intronic variant interpretation: A case report of TRAPPC12‐related disorder

Intronic variants are recognized in the rare disease community as challenging to interpret and therefore not often associated with disease due to the limitations of prediction algorithms. TRAPPC12 is a component of the transport protein particle (TRAPP) complex, which has a role in the endoplasmic reticulum to Golgi trafficking (Scrivens et al., 2011). An autosomal recessive TRAPPC12-related disorder was first described in 2017 (Milev et al., 2017), characterized by progressive early-onset encephalopathy with brain atrophy and spasticity (MIM: 617669). An emerging spectrum of severity of TRAPPC12-related disorders has been suggested, with most variants reported being either missense or truncating (Aslanger et al., 2020; Gass et al., 2020). Here, we present an individual with epilepsy, intellectual disability, and brain anomalies in whom we identified TRAPPC12 (MIM: 614139) compound heterozygous intronic variants. RNA sequencing was key in variant interpretation, leading to a clinical diagnosis. Our patient is a female born to non-consanguineous French–Canadian parents. At 13 months of age, she was unable to roll over or pull to stand. She first walked at 4.5 years of age. She may have had her first seizure at 3–4 months of age and was diagnosed with epilepsy at age 2 years. Investigations at 4–5 years of age (2008–2009) included a microarray and FMR1, Rett syndrome, and Angelman syndrome testing; these were all normal. At age 14 years, she presented to the Genetics clinic with intellectual disability, epilepsy, repetitive arm and hand movements and a progressive scoliosis. Her overall functioning was deemed comparable to a 2-year old; she was ambulating independently with a broad-based gait, had five words and five signs used appropriately, was able to follow a one-step command and was daytime toilet trained. There were no hearing or vision concerns. At age 15 years, she had normal growth parameters, with a height at the 25th percentile, weight at the 50th percentile, and head circumference at the 75th percentile. She was noted to have distinctive features including a wide face, malar hypoplasia, a thin upper vermilion and a pointed chin (Figure 1a). No spasticity was noted. She had brain magnetic resonance imaging (MRI) at age 2 and 16 years; both revealed severe hypoplasia of the pons, a short corpus callosum, and bilateral dysplastic hippocampi, with no overall interval change (Figure 1b,c). Initial trio clinical exome sequencing (ES) in 2018 was nondiagnostic and the family was subsequently enrolled in the Care4Rare Canada research program. Study design approval was obtained from the institutional research ethics board (Children's Hospital of Eastern Ontario; CTO-1577) and the parents provided informed consent. Research re-analysis of the clinical ES data (1 year and 5 months after initial clinical ES reporting) identified two heterozygous TRAPPC12 intronic variants in trans: g.3461389C > A; c.1531-3C > A (before exon 7; absent in gnomAD as of January 2023) and g.3469469A > G; c.1776+3A > G (after exon 9; gnomAD allele frequency of 0.00001193 as of January 2023; chr2(GRCh37); NM_016030.5). Given that neither of these variants have been previously associated with a phenotype, their presence is either extremely low or absent in controls, and multiple splice variant predictors (SpliceSiteFinder-like, MaxEntScan, and GeneSplicer) anticipate a deleterious effect on splicing for each, both these variants would therefore be classified as a variant of uncertain significance (VUS) individually by ACMG criteria (PM2 and PP3). This VUS classification explains why neither of these variants were reported by the clinical laboratory as the report included additional comments that “variants of uncertain significance are only rarely reported at our discretion.” Research RNA sequencing (RNA-Seq) was performed to investigate these variants further using blood from the affected individual collected in PAXgene Blood RNA Tubes. Libraries were prepared using the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina with poly(A) enrichment, then sequenced on an Illumina NovaSeq 6000 using an SP flow cell to generate paired-end 2 × 150 bp reads, with a target depth of 100 million reads (acquired >110 million reads for this sample). Briefly, read alignment and junction detection were performed using STAR (v2.6.1c) in two-pass mode to the GRCh37 reference genome and Ensembl (release 87) annotations were used. RSEM (version 1.2.22) was utilized for gene and transcript expression level quantification. From these analyses, expression levels of the TRAPPC12 transcript were not significantly different from controls. In the affected individual, however, there were instances of exon 7 skipping, (p.(Ile511ProfsTer13)), absent in controls (z = inf), and exon 9 skipping, (p.(Asp560_Gln592del)), observed in only 1 of our 263 controls at a lower frequency (accounting for 2.4% of splicing in this region in the single control sample compared to >8% in the affected; z = 53.6) (Figure 2a). Exon 9 skipping was anticipated to be observed in a small number of our control samples as the corresponding allele is observed in low levels in gnomAD. These aberrant transcripts were confirmed in fibroblast cells derived from the affected individual (Figure 2b). Western blot analysis of whole fibroblast cell extracts demonstrated less than half the amount of TRAPPC12 protein in the affected cells compared to controls (<43%; Figure 2c). Overall, our patient's presentation has significant overlap with other reported cases of TRAPPC12-related disorders. As with the three patients described by Milev et al. (2017), our patient had severe pons hypoplasia and corpus callosum anomalies. Notably, our patient is at the milder end of the phenotypic spectrum, as microcephaly, spasticity, and hearing or vision impairment were absent; however, one of the patients described in Aslanger et al. (2020) did not have microcephaly, and neither of the described patients with homozygous missense variants had visual and hearing impairment. The less severe presentation in the patient we describe is likely due to wild-type protein still being present at lower levels (<43% of the protein was detected by western blot in fibroblast cells from the affected individual compared to controls, Figure 2c), in contrast to the other liveborn affected individuals with biallelic missense and/or nonsense variants (Aslanger et al., 2020; Milev et al., 2017). As TRAPPC12-related disorders were relatively recently first described and have been associated with six variants to date (Aslanger et al., 2020; Gass et al., 2020; Milev et al., 2017) we likely have more to learn about the phenotypic spectrum and possible genotype–phenotype correlations. The compatibility with the autosomal recessive inheritance pattern, clinical findings, and in silico predictions suggest that the c.1531-3C > A and c.1776+3A > G TRAPPC12 variants are disease-causing. Furthermore, RNA-Seq established the damaging effect of these variants on the resulting transcripts and immunoblotting confirmed a significant decrease in protein levels. Nonetheless, these variants are still classified as VUSs by ACMG criteria as TRAPPC12 is not currently known to be haploinsufficient and it is also not known if ~40% of TRAPPC12 protein level is sufficiently decreased to result in disease (previous PM2 and PP3 criteria identified, and PS3_moderate now demonstrated). We do anticipate that TRAPPC12 levels are adequately decreased to lead to changes in Golgi apparatus morphology (Scrivens et al., 2011) and result in mitotic abnormalities (Milev et al., 2015), ultimately impacting its critical functions in neurons, but this requires further investigation. Overall, this study reveals the power of RNA-Seq to investigate potential splice-disrupting variants in intronic regions. Although ES and more recently genome sequencing (GS) are widely used for rare disease diagnosis, there are still many patient investigations that remain unsolved. One main challenge of these DNA-based assays is the lack of ability to decipher the functional and clinical impact of many variants, particularly those in noncoding regions. RNA-Seq is valuable for variant interpretation and its adaptation as a second-tier clinical test should be considered for patients without a molecular diagnosis following exome and/or GS, especially those with identified variants that require further investigations. This is the first known case of compound heterozygous splice site variants associated with a TRAPPC12-related disorder and an important demonstration of the power of RNA sequencing for variant interpretation. Priya T. Bhola and Aren E. Marshall drafted the article. Kym M. Boycott and Kristin D. Kernohan revised the article for intellectual content. Xueqi Wang, and Chantal F. Morel had major roles in acquisition of the data. Aren E. Marshall, Yijing Liang, Madeline Couse, and Elka Miller analyzed the data. Aren E. Marshall, Elka Miller, Kym M. Boycott, and Kristin D. Kernohan interpreted the data. The authors thank the patient and her parents, who provided consent for a case report publication. We would also like to thank the clinical laboratory, GeneDX, for their support of this research and for providing the ES data. This study was performed under the Care4Rare Canada Consortium funded by Genome Canada and the Ontario Genomics Institute (OGI-147), the Canadian Institutes of Health Research, Ontario Research Fund, Genome Alberta, Genome British Columbia, Genome Quebec, and Children's Hospital of Eastern Ontario Foundation. Aren E. Marshall was supported by a Canadian Institutes of Health Research (CIHR) fellowship award (MFE-176616). Kym M. Boycott was supported by a CIHR Foundation Grant (FDN-154279) and a Tier 1 Canada Research Chair in Rare Disease Precision Health. The authors have no conflict of interest to declare. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.