Periodontal Ehlers–Danlos syndrome is associated with leukoencephalopathy

Individuals of two families clinically diagnosed with periodontal EDS were included in the study. All patients included in the study underwent general physical and neurological examination. The clinical investigation list and the questionnaire for reporting on periodontal EDS were previously published in Kapferer-Seebacher et al. [14] as supplements. Briefly, clinical investigation included assessment of joint and skin features and photographs. Oral investigation included periodontal investigations, orthopantomogram, and intraoral photographs. The diagnosis of early severe periodontitis was based on radiologic evidence of severe alveolar bone loss (≥ 50%) at an age of ≤ 25 years or history of complete tooth loss due to tooth mobility at an age of ≤ 30 years. The questionnaire includes questions on previous dental treatments and smoking, joint and skin features, and other features like recurrent infections, hoarse voice, organ ruptures, and other systemic conditions. Family A was investigated at the VU University Medical Center, Amsterdam, in 2017. Clinical and genetic data of Family B have been previously reported by Kapferer-Seebacher et al. [1]. Neurologic investigation of family B was conducted in 2017 and 2018 at the Medical University of Innsbruck, Austria.

All patients included in the study underwent MRI of the brain. We scored all MRIs according to a previously published standardized protocol [15].

Whole exome sequencing was performed on three affected individuals from family A. Genomic DNA was isolated from blood and 2.5 μg was sheared on a Covaris S2 instrument (Covaris, Woburn, MA). DNA libraries were prepared using Kapa Biosystems reagents (Kapa Biosystems, Wilmington, MA), and 1.0 μg of library was used for enrichment with Roche/NimbleGen SeqCap EZ MedExome (Roche, Basel, Switzerland) according to the manufacturer’s protocol. Sequencing was performed on an Illumina HiSeq 2500 platform (Illumina, San Diego, CA, USA) with 125 base pair paired-end reads. Over 92% of the capture region was covered ≥ 30× with a mean bait coverage of approximately 105× for each sample. Variant calling was performed using an in-house analysis pipeline. Alignment of sequence reads to the human genome (hg19) was performed with the Burrows–Wheeler Aligner tool (BWA-MEM v0.7.10) using default settings. Subsequently, Picard Tools (v1.111; http://​picard.​sourceforge.​net/) was used for sorting and marking duplicates. For local realignment and base quality score recalibration, we used the Genome Analysis Toolkit (GATK; v3.3-0; Broad Institute, Cambridge, MA) and for variant calling, we used the GATK HaplotypeCaller. Variants were filter tagged using the GATK VariantFiltration and annotated by snpEff (v4.0). Variant prioritization was performed using Cartagenia Bench Lab NGS (Agilent Technologies, Santa Clara, USA). In short, a classification tree was used to select for variants present in all three affected individuals and virtually absent in control cohorts dbSNP build 142 (http://​www.​ncbi.​nlm.​nih.​gov/​projects/​SNP), 1000 Genomes Phase 3 release v5.20130502, and ESP6500 (http://​evs.​gs.​washington.​edu/​EVS/​) as well as in-house controls. Prerequisite was that these variants had been genotyped in at least 200 alleles. Subsequently, the remaining variants were further prioritized based on literature, predicted (deleterious) effects on protein function by, e.g., truncating the protein, affecting splicing, amino acid change, and evolutionary conservation.

Family A is a Caucasian Dutch four-generation family (Fig. 1); individuals A-II-1, A-III-1, and A-III-2 were available for clinical and genetic investigations. The C1R mutation c.926G>T (NM_001733.4) resulting in p.(Cys309Phe) was identified in all of them. This cysteine at position 309 is located in the Sushi/SCR/CCP domain and is involved in disulfide bond formation that stabilizes the C1r CCP1 module [1]. In addition, c.927C>G results in an amino acid change of the same amino acid p.(Cys309Trp) and was previously identified in other periodontal EDS patients [1]. No known or likely pathogenic variants were found in 154 genes related to leukoencephalopathies, including vascular leukoencephalopathies. All individuals were clinically diagnosed with periodontal EDS (for details see, Table 1; Figs. 1 and 2).

At clinical investigation, the male individual A:II-1 was 58 years of age. In addition to signs of periodontal EDS, his medical history included diabetes type II, high blood pressure, hypercholesterolemia, and cardiac arrhythmias; he never smoked. From age 50 years he experienced slow and mild cognitive decline with memory, concentration and spatial orientation problems, dyspraxia, aggression, and depression. At age 57 years, he was admitted in a state of confusion during pneumonia, while in the hospital, he had a seizure lasting 20 min with a focal start in the right hand and secondary generalization. Neurological examination at age 57 years revealed mild tremor and slight cerebellar ataxia, no other abnormalities. Individual A:III-2 was 31 years of age at time of investigation. She never smoked, had a normal blood pressure, and had no complaints of cognitive decline. Neurological examination at age 31 years was normal. Individual A:III-1 was 8 years of age at time of investigation. He had mild learning problems, especially language problems, but followed normal level education. Neurological examination at age 8 years was normal.

Family B is a five-generation Caucasian Austrian family, whose clinical and genetic data were previously reported (Family 1 in Kapferer-Seebacher et al.) [1]. The identified C1R mutation c.149_150TC>AT (p.Val50Asp) segregated with 15 individuals clinically affected by periodontal EDS in this family; in Fig. 1, only individuals relevant for the present study are depicted, with the same numbers as used in Kapferer-Seebacher et al. [1]. Individuals B:III-2, B:III-10, B:IV-1, B:IV-2, and B:IV-10 were available for the present neurologic analysis. All investigated individuals of family B had never smoked or were light smokers (≤ 5 cigarettes per day), and had no hypertension or hypercholesterinemia. All affected individuals in this family showed normal intellectual performance without evidence of cognitive decline (Fig. 3).

In male individual B:III-10, neurological examination at age 68 years was normal. The female individual B:III-2 reported on frequent headaches; neurological examination at age 56 years was normal. In female individual B:IV-10, neurological examination at age 43 years was normal. The female individual B:IV-1 suffered from depression. Neurological examination at age 34 years was normal. The female individual B:IV-2 was not available for clinical neurologic investigation, but reported frequent fainting spells and dizziness, and a history of frequent headaches in childhood.

In family A, individual A:II-1 underwent MRI at age 57 years (Fig. 4a (a–f)). It showed no cerebral or cerebellar atrophy. Almost diffuse, homogeneous cerebral white matter signal abnormalities were present, with an anterior-posterior gradient, in which the posterior cerebral white matter was somewhat better preserved. A thin rim of directly subcortical white matter was relatively, but not entirely spared. The anterior part of the temporal lobe and external and extreme capsules were affected. The corpus callosum and internal capsule were spared. Within the basal nuclei and thalami numerous small foci of high T2 signal were seen, which in part had a low signal on FLAIR, indicative of enlarged perivascular spaces. Two foci in the putamen and head of the caudate nucleus on the left were larger and cystic, indicative of lacunar infarctions. Within the brain stem, a lacunar infarction was present in the pons. Inhomogeneous T2 hyperintensities were present in the pons, the lower part of the midbrain and the cerebellar white matter. Gradient echo images showed evidence of a few microbleeds in the basal nuclei on the left and the dentate nucleus on both sides. No areas of diffusion restriction or abnormal contrast enhancement were seen. Individual A:III-2 underwent MRI at 31 years (Fig. 4a (g and h)). It showed signal abnormalities in the deep parietal and periventricular frontal and occipital white matter of limited extent. No signal abnormalities elsewhere, enlarged perivascular spaces, lacunar infarcts, areas of diffusion restriction, or microbleeds were visible. Individual A:III-1 underwent MRI at 7 years. It was normal.

In family B, individual B:III-10 underwent MRI at age 68 years (Fig. 4b (a–d)). It showed slight generalized cerebral atrophy with enlarged lateral ventricles and subarachnoid spaces. Extensive, homogeneous cerebral white matter signal abnormalities were present, without anterior–posterior gradient. The subcortical white matter was preserved in all areas and the anterior part of the temporal lobe and external and extreme capsules were not affected. The corpus callosum, internal capsule, brain stem, and cerebellum were spared. Numerous enlarged perivascular spaces were present spread over the cerebral hemispheres, involving white matter, basal nuclei and thalami. No lacunar infarcts, areas of diffusion restriction or abnormal enhancement after contrast were present. MR angiography and venography did not reveal vessel abnormalities. Individual B:III-2 underwent MRI at age 56 years. It showed numerous enlarged perivascular spaces. FLAIR showed a thin, continuous periventricular rim of abnormal signal and numerous small and larger spots of abnormal signal spread over the deep cerebral white matter. No lacunar infarcts, areas of diffusion restriction, or abnormal contrast enhancement were present. MR angiography and venography did not reveal vessel abnormalities. Individual B:IV-10 underwent MRI at age 43 years. The abnormalities were similar to those of individual B:III-2, but less extensive. Individual B:IV-1 underwent MRI at age 35 years. It only showed a very thin, continuous periventricular rim and a few small spots of abnormal signal spread over the deep cerebral white matter on FLAIR. MRI of individual B:IV-2 was obtained at age 33 years because of the neurological complaints described above. It revealed numerous enlarged perivascular spaces in the cerebral white matter. Additionally, a thin, continuous periventricular rim of abnormal signal and multiple small focal lesions were present in the deep cerebral white matter. No areas of diffusion restriction, microbleeds or abnormal contrast enhancement were observed.