Exploring genotype–phenotype correlations in CREBBP: comment on the literature and description of an additional patient with an atypical outcome

Rubinstein-Taybi syndrome (RTS) is a rare, mainly sporadic, autosomal dominant genetic disorder characterized by distinctive facial features, broad thumbs and toes, postnatal growth retardation, microcephaly, and a degree of intellectual disability [1]. Average IQ typically ranges between 35 and 50; however, cognitive outcomes are highly variable and rare affected individuals have been reported with intellect within the normal range [1]. Genetic variation in the cAMP response element-binding protein (CREB)-Binding Protein (CREBBP) gene is implicated in ~55–75% of RTS patients [2]. CREBBP encodes a transcription co-activator, which interacts with a large number of proteins. Frameshift, nonsense and missense, and splicing variants (listed in order of prevalence) have been described throughout the CREBBP gene and represent the majority of genetic variation responsible for RTS [2]. Approximately 17% of RTS patients with CREBBP variation have intragenic CREBBP deletions or whole gene deletions encompassing CREBBP [2].

To date, over 250 RTS-causing genetic variants have been identified in this gene (www.lovd.nl/CREBBP) and almost all of these variants are private [3]. Though rare, descriptions of patients with recurrent variants are a key tool to assess for genotype–phenotype correlations. Multiple CREBBP RTS cohorts have been characterized to identify possible genotype–phenotype correlations, but no statistically significant trends have been recognized as of yet [3–7]. A more severe lethal phenotype was associated with CREBBP deletion as part of a contiguous gene deletion syndrome [8], though this has not been consistently recapitulated in additional cohorts [4–6, 9, 10]. On the other hand, milder phenotypes have been reported when a CREBBP microdeletion exists in mosaic form and is also postulated to exist with missense variation falling outside of the histone acetyltransferase domain of CREBBP [6, 11]. However, a larger cohort of RTS patients did not identify any differences in prevalence or severity of clinical features for truncating versus single amino acid substitutions, nor for specific exonic locations [4]. This is considered separately from the known association between a phenotypically distinct condition, Menke–Hennekam syndrome (MKS) (OMIM #618332), with missense variation in exon 30 and 31 of CREBBP.

MKS is a recently described condition consisting of core features of developmental delay, autistic behavior, short stature, microcephaly, and vision and hearing impairments [12–14]. These patients present without the characteristic broad thumbs of RTS and have dysmorphology that is quite distinct from RTS. Given that all variants attributed to MKS are missense variants and the dysmorphology of these patients resembles that of 16p13.3 duplication syndrome (the locus of CREBBP), the existing hypothesis is that MKS arises from CREBBP gain-of-function (versus haploinsufficiency for RTS). Functional studies are required to further clarify this potential mechanism, especially in the context of the recent description of three patients with MKS and nonsense variants in exon 31 [15].

In RTS, the spectrum of cognitive outcomes is highly variable, though it is most often described in the moderate to severe intellectual disability range. A potential cognitive genotype–phenotype correlation was identified in 2015 by Spena et al., who described a second individual with a CREBBP p.Y1175C variant (NM_004380.2:c.3524A>G) and intellect in the normal range [6, 16]. It was postulated that this specific missense variant may cause only partial loss of function, thus resulting in a milder phenotype than the more prevalent RTS nonsense/frameshift variants. No further individuals have been described with recurrent variants and an absence of intellectual disability.

In 2019, Perez-Grijalba et al. described a female with typical RTS facial dysmorphisms and microcephaly [4]. Interestingly, she was not growth delayed, nor did she experience psychomotor or language delay, behavioral problems, or intellectual disability. She was found to have a nonsense variant in exon 5 of CREBBP: NM_004380.2:c.1318C>T, p.(Arg440Ter).

Herein we describe a second individual with the same exon 5 genetic variant. Our proband is the first child of healthy non-consanguineous parents of British Isles and Indian ancestry. Both parents have bachelor degree educational attainment. The proband was assessed by a clinical geneticist at eight days of age and clinically diagnosed with RTS. Clinical features were found to be in keeping with RTS and included: downslanting palpebral fissures, columella that extends beyond the level of the alae nasi, left cryptorchidism, broad and radially deviated thumbs, and broad halluces.

His early motor and speech development was only mildly delayed. He has had three psycho-educational assessments over the course of his life: at 5, 11, and 17 years of age. The first two assessments found his overall aptitudes to be at the lower end of the average range of cognitive function, though they were unevenly developed and he had specific learning disabilities. He received strong educational support through a private school setting for most of his elementary grades, as well as considerable parental support in his education. He was transitioned back to a public setting for his high school years. His latest psycho-educational assessment performance (with measures that included WAIS-IV [Wechsler Adult Intelligence Scale-Fourth Edition; Canadian Norms] and the Vineland Adaptive Behaviour Scales), at 17 years of age met the criteria for a diagnosis of a mild intellectual disability with both global cognitive aptitudes (5th percentile) and adaptive functioning (1st percentile) falling below the average range. He completed high school and is currently pursuing post-secondary certification at a Canadian college.

Thus, this patient represents a second individual with CREBBP exon 5 R440* variant RTS and comparatively good neurodevelopmental outcomes. Interestingly, we note that the two predominate isoforms of CREBBP differ only in their retention or skipping of in-frame exon 5: isoform a (ENST00000262367.10), the canonical isoform, includes exon 5, whereas isoform b (ENST00000382070.7) skips exon 5. Full-length isoform transcriptome analysis of the developing human brain has shown that isoform b has higher expression in the prenatal brain compared to the postnatal brain (log fold change in pre versus postnatal human brain = −3.05, p = −2.8e–15, FDR-corrected p = 5e–14 [17]) and thus we posit that the existence of a “normal”/non-variant CREBBP isoform (isoform b) in the developing brain in patients with R440* variants may somewhat preserve neurodevelopmental outcomes. However, we do note that three other individuals have been reported with different exon 5 nonsense variants and were described as having an intellectual disability, though there was no characterization of their degree of impairment [4]. In addition, cognition is a highly polygenic trait and there undoubtedly exists many other genetic and environmental factors that may modify this phenotype. Further detailed clinical descriptions of additional patients with exon 5 variants would be needed to fully explore this relationship.

In conclusion, this work summarizes genotype–phenotype correlations in CREBBP RTS and describes an additional patient with R440* RTS and an atypical neurodevelopmental outcome. Sharing information about the natural history of RTS patients with recurrent variants may help clinical geneticists more accurately prognosticate and optimally manage RTS patients in the future.

Author contributions

GEG planned, supervised the study, and performed clinical work. KMV drafted the manuscript and all authors read and approved the final version of the manuscript.

Funding

No financial assistance was received in support of this study.

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Informed consent was obtained from the patient and his parents. Institutional ethics approval is not required due to the descriptive nature of the case report.