A 15 Mb large paracentric chromosome 21 inversion identified in Czech population through a pair of flanking duplications
© Drabova et al.; licensee BioMed Central Ltd. 2014
Received: 9 April 2014
Accepted: 20 June 2014
Published: 19 August 2014
Inversions are balanced structural chromosome rearrangements, which can influence gene expression and the risk of unbalanced chromosome constitution in offspring. Many examples of inversion polymorphisms exist in human, affecting both heterochromatic regions and euchromatin.
We describe a novel, 15 Mb long paracentric inversion, inv(21)(q21.1q22.11), affecting more than a third of human 21q. Despite of its length, the inversion cannot be detected using karyotyping due to similar band patterns on the normal and inverted chromosomes, and is therefore likely to escape attention. Its identification was aided by the repeated observation of the same pair of 150 kb long duplications present in cis on chromosome 21 in three Czech families subjected to microarray analysis. The finding prompted us to hypothesise that this co-occurrence of two remote duplications could be associated with an inversion of the intervening segment, and this speculation turned out to be right. The inversion was confirmed in a series of FISH experiments which also showed that the second copy of each of the duplications was always located at the opposite end of the inversion. The presence of the same pair of duplications in additional individuals reported in public databases indicates that the inversion may also be present in other populations. Three out of the total of about 4000 chromosomes 21 examined in our sample carried the duplications and were inverted, corresponding to carrier frequency of about 1/660. Although the breakpoints affect protein-coding genes, the occurrence of the inversion in normal parents and siblings of our patients and the occurrence of the duplications in unaffected controls in databases indicate that this rare variant is rather non-pathogenic. The inverted segment carried an identical shared haplotype in the three families studied. The haplotypes, however, diverged very rapidly in the flanking regions, possibly pointing to an ancient founder event at the origin of the inversion.
The identification of inv(21)(q21.1q22.11) supports the notion that paracentric inversions are the most common form of chromosomal variation and that some of them may still remain undetected.
KeywordsInversion Duplication USP25 SNP arrays Founder effect
Since the identification of the cytogenetically visible heterochromatic inversion inv(9)(p11q12) almost half a century ago, many other polymorphic inversions involving both heterochromatin (1qh, 16qh and Yqh) and euchromatin have been described in human. The most common euchromatic inversions are the 22–25 Mb long inv(2)(p11.2q13), the 22.6 Mb long inv(10)(p11.2q21.2) and the very frequent 4.7 Mb long inv(8)(p23.1)[3–7]. Paired-end sequencing and other focused approaches have recently led to the identification of many smaller inversions. The Database of Genomic Variants currently lists 583 merged inversion regions, and the InvFEST database lists 85 well validated human inversions[8, 9].
Inversions often have no clinical significance, although opposite examples exist, like the recurrent inversion in the F8 gene causing haemophilia A. Therefore the identification of a rare inversion prenatally or in an affected child always possesses a question of a possible phenotypic influence. Inversion carriers also have a risk of unbalanced offspring; considering paracentric and pericentric inversions together, the risk is about 1%. Two mechanisms contribute to the formation of unbalanced gametes in inversion carriers. First, an odd number of recombinations within the inversion loop in an inversion heterozygote may lead to aberrant chromosomes (with a higher risk in pericentric inversion heterozygotes producing duplication-deletion chromosomes and a lower risk in paracentric inversion heterozygotes producing dicentric and acentric recombinant chromosomes expected to be lethal)[11, 12]. Second, due to a less favourable arrangement of low-copy repeats at the breakpoints of the inverted allele, several inversions have been described to predispose to rearrangements including those associated with some microdeletion syndromes[13, 14].
Two scenarios exist for the evolutionary origin of human polymorphic inversions: they can be either recurrent, or can descend from a single ancestral event. While the inv(2)(p11.2q13) is recurrent due to the local genome architecture, the inv(10)(p11.2q21.2) seems to have just one ancestral founder in Northern Europe, and similarly the inv(8)(p23.1) shows a strong founder effect. A large study of different inversions concluded that recurrence is rather exceptional and that most inversions are likely to be identical by descent.
In this report we describe a serendipitous identification of inv(21)(q21.1q22.11), a large, 15 Mb long paracentric inversion affecting more than a third of the long arm of human chromosome 21 (chr21), which cannot, however, be identified using karyotyping. Repeated identification of the same pair of small duplications in cis on chr21 in three Czech families using single nucleotide polymorphism (SNP) arrays prompted us to hypothesise that this remarkable co-occurrence of two remote copy number variants (CNVs) could be specific for an unusual rare structural variant of chr21, for example one with an inversion of the DNA segment flanked by the duplications, and this scenario was confirmed by multiple experiments. The inversion may not be restricted to the three families reported, as suggested by the presence of the same pair of duplications in additional individuals reported in public CNV databases.
FISH probes used in the study
Probe specification (Supplier)
chr21 centromeric probe (proximal to both duplications)
D21Z1 (Cytocell, Cambridge, UK)
probe overlapping the proximal duplication
RP11-482O14 (BlueGnome, Cambridge, UK)
probe at the proximal end of the segment between the duplications
RP11-45 M19 (Source BioScience, Nottingham, UK)*
probe at the distal end of the segment between the duplications
RP11-64I20 (Source BioScience)*
probe overlapping the distal duplication
RP11-134 K3 (Source BioScience)*
DSCR1 (or chr21qter subtelomere) probe (distal to both duplications)
PN21 (Kreatech Diagnostics, Amsterdam, The Netherlands) (or D21S1575 (Cytocell))
We identified a pair of identical duplications on three out of a total of about 4000 unrelated human chr21 tested, and we showed that the 15 Mb segment flanked by the duplications forms a yet unknown large paracentric inversion of chr21. A thorough database and literature search revealed the same combination of similar duplications in one patient with developmental delay (individual 9885009, chr21:16,956,885-17,141,961 (nssv1172163) and chr21:32,098,070-32,217,892 (nssv1172207)) and in one unaffected control (individual 1780854261_A, chr21:16,956,866-17,145,011 (nssv1151408) and chr21:32,086,323-32,224,934 (nssv1151611)) from one study, and in one unaffected control (individual SW_0033, chr21:16,948,663-17,145,770 (essv7004290) and chr21:32,082,962-32,242,849 (essv7004391)) from another study. Several other individuals were reported to carry just one of the duplications; taking into account the relatively small size of the segments, it is possible that some array platforms or CNV calling algorithms could overlook one duplication from the pair. By analogy with our results, it could be expected that these chromosomes also carry the inversion. This together with our estimate of carrier frequency of about 1/660 may indicate that the inversion could be more common, and that it may not be limited to the Czech population.
This notion can be further supported by the observation of chr21 haplotypes rapidly diverging outside of the inverted region, confirming that the Czech inversion carriers studied were not closely related. At the same time the shared identical haplotype within the inverted segment pointed to the common descent of the three inverted chr21 from a single ancestral event, a scenario observed in most inversions. Although the very large size of the inversion may have allowed for double crossing over, thus limiting linkage disequilibrium only to regions adjacent to the breakpoints of the inverted segment, the haplotypes observed in our families were strictly uniform throughout the whole segment. This may support the view that recombination is actively suppressed in inverted regions[18, 19]. The inverted segment contains hot spots of variable recombination rate which can potentially contribute to this phenomenon.
The inheritance of the rearranged chr21 from healthy parents in Families A-C and its presence in an unaffected child from Family C together with the likely occurrence of the inversion in unaffected controls from public databases indicate that this rearrangement is unlikely to be a rare pathogenic variant responsible for the abnormal phenotype in patients from our Families A and B or in individual 9885009. The repeated identification of the inversion in affected persons may simply reflect the proportion of affected patients tested using microarrays among all individuals analysed, at least in the Czech cohorts. However, it cannot be excluded that this rare variant could represent a predisposing factor with incomplete penetrance and/or variable expressivity possibly contributing to neurodevelopmental disorders in conjunction with other genetic, epigenetic or environmental factors.
The proximal duplication involves USP25, a ubiquitously expressed gene coding ubiquitin specific peptidase 25, a protease which breaks polyubiquitin chains on degraded proteins and is thus involved in ubiquitin metabolism and recycling. Defects in other ubiquitin specific peptidases, for example USP15, have been associated with autism. It is unclear if the rearranged chr21 carries one intact copy of USP25, and if the expression of this copy is influenced by a possible positional effect associated with the rearrangement. The distal duplication harbours five members of the KRTAP gene family coding keratin associated proteins present in the interfilamentous matrix of the hair cortex. The five small genes are fully embedded in the duplication and the influence of the rearrangement on their expression or the expression of other neighbouring members of the KRTAP family is also unclear. In addition to genes directly affected by the breakpoints or by the duplications, other genes from the inversion or its flanking regions can be influenced, as described in other inversions.
There was no remarkable history of infertility, spontaneous abortions or developmental defects in extended pedigrees of inv(21)(q21.1q22.11) carrier parents from Families A-C (multiple abortions in Family A were reported in the maternal branch which, however, did not carry the inversion). This is in accord with the generally expected low risk of abortions in inversion carriers or of children carrying recombinant chromosomes, which can in turn be associated with the suppression of recombination within the inverted segment[18, 19]. Nevertheless, evidence exists that some types of recombination (the U-type exchange within the inversion loop) or the breakage of dicentric recombinant chromosomes can produce abnormal gametes capable of giving rise to affected offspring in paracentric inversion carriers[4, 23]. The literature search has not revealed any inv dup del(21q) which could be associated with inv(21)(q21.1q22.11); however, due to the position of the inversion, the inv dup del arrangement would have a very large terminal deletion of 21q and its carriership may not be compatible with life. A patient has been described with a combination of symptoms of proximal 21q monosomy and some symptoms of Down syndrome who carried an analphoid marker chr21 consisting of inverted duplication of 21q21.1qter with a neocentromere. A similar marker chr21 could originate as an acentric recombination product in an inv(21)(q21.1q22.11) heterozygote; however, identical microsatellite alleles on both segments pointed in this case to an intrachromosomal origin of the rearrangement, or to an additional recombination immediately distal to the distal inversion breakpoint. Another case with a de novo 9-break complex chr21 rearrangement had breakpoints in band 21q22.11; however, the resolution of the study did not allow deciding if the breakpoints were identical with those on the inverted chr21 described here.
Although the inversion affects 15 Mb of DNA and more than a third of the long arm of chr21, due to no observable difference in the banding pattern of the inverted chr21 this rearrangement is likely to escape attention during karyotyping. There are several reports in the literature of rare paracentric inversions in 21q identified cytogenetically [reviewed in, but they seem to be different, and their identification in the light microscope in fact rather argues against them being exactly identical with the current inversion.
The research was in compliance with the Helsinki Declaration and was approved by the local ethics committee (University Hospital Motol) under reference number EK 597/09. Informed consent with participation in the study was obtained from all subjects or their legal guardians.
SNP array analysis was performed in three Czech patient cohorts. The first cohort consisted predominantly of children with a broad range of phenotypes including developmental delay, ID, autism, seizures, and congenital and growth anomalies, and involved 1298 unrelated patients representing 2596 chr21. In 88 of these patients both parents and in 33 patients one parent were also analysed on the array, thus adding 209 unrelated chr21 and totalling the number of chr21 in this cohort to 2805. The second cohort consisted of 477 foetuses referred for prenatal SNP array testing, with both parents or one parent analysed in 27 or 11 cases, respectively, totalling the number of chr21 to 1019. The third cohort consisted of 28 couples analysed as a part of pre-conception care, who harboured 112 chr21. The three cohorts thus yielded a grand total of 3936 chr21.
Family A was ascertained through a female patient with moderate ID and hyperactivity. At the age of 9.5 years she had a long face, frontal bossing, long nose with a broad tip and large nares, epicanthal folds, upslanted palpebral fissures, big open mouth with a narrow upper lip and prominent lower lip, sparse teeth and small low-set dysplastic protruding ears. Her face resembled that of patients with Angelman syndrome, but testing for this condition was negative. She also showed wide stance with bent knees, long fingers on both hands and feet, and pedes planovalgi. Her growth was normal. The patient had an older step brother and two younger brothers who, however, were not available for analysis. The youngest brother suffered from developmental delay and hyperactivity. Both parents of the patient finished primary education. The mother had a history of one spontaneous abortion and there were multiple occurrences of abortions in the maternal branch of the family. SNP array testing was performed in the patient, and later also the parents were analysed.
Family B was ascertained through a currently 13-year-old boy with moderate to severe ID and autism with no language (he used only three words). The boy showed attacks of anger, aggression and self-aggression. His facial features included a long face, frontal bossing, distinct arched eyebrows, upslanted palpebral fissures, depressed nasal bridge, upturned nasal tip, big mouth with a thin upper lip, prominent incisors and small low-set ears. He also showed poor posture, wide stance with bent knees, genua valga and macrogenitalism. His mother was a high school graduate and worked as a librarian. His father was trained as a motor mechanic. The boy had one healthy older brother. Family B had no history of infertility or spontaneous abortions. A paternal aunt of the father suffered from epilepsy. Other genealogical data were unremarkable. Microarray analysis was performed in the patient and his parents.
Family C was identified through a currently 13-year-old boy with autism, unsteady gait and symptoms suggestive of neurofibromatosis including eye symptoms and cafe-au-lait spots. Parallel to the microarray analysis, a probably pathogenic missense NF1 mutation was identified in the patient (de novo or paternal, the father was not available). The patient had two apparently healthy siblings, a 15-year-old brother and a 3-year-old step brother. The family history was unremarkable and there was no history of infertility or spontaneous abortions. The family insisted on microarray analysis of the patient and his younger step brother. Subsequently also the mother of both boys was analysed.
Karyotyping of peripheral blood lymphocytes was performed using standard methods. Genomic DNA of individuals from Families A-C was tested using Human CytoSNP-12 BeadChips (~300 K; Illumina, San Diego, CA, USA). Array data were analysed using GenomeStudio (Illumina) and QuantiSNP. FISH employed standard protocols and various combinations of probes listed in Table 1. The study used genome build hg19.
Written informed consent was obtained from the patients' parents for the publication of this report and any accompanying images.
Copy number variant
Fluorescence in situ hybridisation
Single nucleotide polymorphism.
We thank the families of the patients for their cooperation, and Tereza Jancuskova and Sona Pekova for the preparation of FISH probes. The study was supported by grants CHERISH 223692, COST BM1004 and CZ.2.16/3.1.00/24022 from the European Commission, and NT/14200, DRO UH Motol 00064203 and LD11028 from the Czech Ministries of Health and Education.
- Walzer S, Breau G, Gerald PS: A chromosome survey of 2,400 normal newborn infants. J Pediatr 1969, 74: 438–448. 10.1016/S0022-3476(69)80202-7View ArticlePubMedGoogle Scholar
- Feuk L: Inversion variants in the human genome: role in disease and genome architecture. Genome Med 2010, 2: 11. 10.1186/gm132PubMed CentralView ArticlePubMedGoogle Scholar
- Salm MP, Horswell SD, Hutchison CE, Speedy HE, Yang X, Liang L, Schadt EE, Cookson WO, Wierzbicki AS, Naoumova RP, Shoulders CC: The origin, global distribution, and functional impact of the human 8p23 inversion polymorphism. Genome Res 2012, 22: 1144–1153. 10.1101/gr.126037.111PubMed CentralView ArticlePubMedGoogle Scholar
- Giglio S, Broman KW, Matsumoto N, Calvari V, Gimelli G, Neumann T, Ohashi H, Voullaire L, Larizza D, Giorda R, Weber JL, Ledbetter DH, Zuffardi O: Olfactory receptor-gene clusters, genomic-inversion polymorphisms, and common chromosome rearrangements. Am J Hum Genet 2001, 68: 874–883. 10.1086/319506PubMed CentralView ArticlePubMedGoogle Scholar
- Gilling M, Dullinger JS, Gesk S, Metzke-Heidemann S, Siebert R, Meyer T, Brondum-Nielsen K, Tommerup N, Ropers HH, Tumer Z, Kalscheuer VM, Thomas NS: Breakpoint cloning and haplotype analysis indicate a single origin of the common Inv(10)(p11.2q21.2) mutation among northern Europeans. Am J Hum Genet 2006, 78: 878–883. 10.1086/503632PubMed CentralView ArticlePubMedGoogle Scholar
- Fickelscher I, Liehr T, Watts K, Bryant V, Barber JC, Heidemann S, Siebert R, Hertz JM, Tumer Z, Simon Thomas N: The variant inv(2)(p11.2q13) is a genuinely recurrent rearrangement but displays some breakpoint heterogeneity. Am J Hum Genet 2007, 81: 847–856. 10.1086/521226PubMed CentralView ArticlePubMedGoogle Scholar
- Youings S, Ellis K, Ennis S, Barber J, Jacobs P: A study of reciprocal translocations and inversions detected by light microscopy with special reference to origin, segregation, and recurrent abnormalities. Am J Med Genet A 2004, 126A: 46–60. 10.1002/ajmg.a.20553View ArticlePubMedGoogle Scholar
- Martinez-Fundichely A, Casillas S, Egea R, Ramia M, Barbadilla A, Pantano L, Puig M, Caceres M: InvFEST, a database integrating information of polymorphic inversions in the human genome. Nucleic Acids Res 2014, 42: D1027-D1032. 10.1093/nar/gkt1122PubMed CentralView ArticlePubMedGoogle Scholar
- Macdonald JR, Ziman R, Yuen RK, Feuk L, Scherer SW: The Database of Genomic Variants: a curated collection of structural variation in the human genome. Nucleic Acids Res 2014, 42: D986-D992. 10.1093/nar/gkt958PubMed CentralView ArticlePubMedGoogle Scholar
- Lakich D, Kazazian HH Jr, Antonarakis SE, Gitschier J: Inversions disrupting the factor VIII gene are a common cause of severe haemophilia A. Nat Genet 1993, 5: 236–241. 10.1038/ng1193-236View ArticlePubMedGoogle Scholar
- Kleczkowska A, Fryns JP, Van den Berghe H: Pericentric inversions in man: personal experience and review of the literature. Hum Genet 1987, 75: 333–338. 10.1007/BF00284103View ArticlePubMedGoogle Scholar
- Madan K: Paracentric inversions: a review. Hum Genet 1995, 96: 503–515.View ArticlePubMedGoogle Scholar
- Osborne LR, Li M, Pober B, Chitayat D, Bodurtha J, Mandel A, Costa T, Grebe T, Cox S, Tsui LC, Scherer SW: A 1.5 million-base pair inversion polymorphism in families with Williams-Beuren syndrome. Nat Genet 2001, 29: 321–325. 10.1038/ng753PubMed CentralView ArticlePubMedGoogle Scholar
- Antonacci F, Kidd JM, Marques-Bonet T, Ventura M, Siswara P, Jiang Z, Eichler EE: Characterization of six human disease-associated inversion polymorphisms. Hum Mol Genet 2009, 18: 2555–2566. 10.1093/hmg/ddp187PubMed CentralView ArticlePubMedGoogle Scholar
- Thomas NS, Bryant V, Maloney V, Cockwell AE, Jacobs PA: Investigation of the origins of human autosomal inversions. Hum Genet 2008, 123: 607–616. 10.1007/s00439-008-0510-zView ArticlePubMedGoogle Scholar
- Cooper GM, Coe BP, Girirajan S, Rosenfeld JA, Vu TH, Baker C, Williams C, Stalker H, Hamid R, Hannig V, Abdel-Hamid H, Bader P, McCracken E, Niyazov D, Leppig K, Thiese H, Hummel M, Alexander N, Gorski J, Kussmann J, Shashi V, Johnson K, Rehder C, Ballif BC, Shaffer LG, Eichler EE: A copy number variation morbidity map of developmental delay. Nat Genet 2011, 43: 838–846. 10.1038/ng.909PubMed CentralView ArticlePubMedGoogle Scholar
- Vogler C, Gschwind L, Rothlisberger B, Huber A, Filges I, Miny P, Auschra B, Stetak A, Demougin P, Vukojevic V, Kolassa IT, Elbert T, de Quervain DJ, Papassotiropoulos A: Microarray-based maps of copy-number variant regions in European and sub-Saharan populations. PLoS One 2010, 5: e15246. 10.1371/journal.pone.0015246PubMed CentralView ArticlePubMedGoogle Scholar
- Morel F, Laudier B, Guerif F, Couet ML, Royere D, Roux C, Bresson JL, Amice V, De Braekeleer M, Douet-Guilbert N: Meiotic segregation analysis in spermatozoa of pericentric inversion carriers using fluorescence in-situ hybridization. Hum Reprod 2007, 22: 136–141.View ArticlePubMedGoogle Scholar
- Jaarola M, Martin RH, Ashley T: Direct evidence for suppression of recombination within two pericentric inversions in humans: a new sperm-FISH technique. Am J Hum Genet 1998, 63: 218–224. 10.1086/301900PubMed CentralView ArticlePubMedGoogle Scholar
- Bhatt SS, Manvelyan M, Moradkhani K, Hunstig F, Mrasek K, Puechberty J, Lefort G, Sarda P, Weise A, Liehr T, Pellestor F: Inverted segment size and the presence of recombination hot spot clusters matter in sperm segregation analysis. Cytogenet Genome Res 2014, 142: 145–149. 10.1159/000356142View ArticlePubMedGoogle Scholar
- O'Roak BJ, Vives L, Fu W, Egertson JD, Stanaway IB, Phelps IG, Carvill G, Kumar A, Lee C, Ankenman K, Munson J, Hiatt JB, Turner EH, Levy R, O'Day DR, Krumm N, Coe BP, Martin BK, Borenstein E, Nickerson DA, Mefford HC, Doherty D, Akey JM, Bernier R, Eichler EE, Shendure J: Multiplex targeted sequencing identifies recurrently mutated genes in autism spectrum disorders. Science 2012, 338: 1619–1622. 10.1126/science.1227764PubMed CentralView ArticlePubMedGoogle Scholar
- de Jong S, Chepelev I, Janson E, Strengman E, van den Berg LH, Veldink JH, Ophoff RA: Common inversion polymorphism at 17q21.31 affects expression of multiple genes in tissue-specific manner. BMC Genomics 2012, 13: 458. 10.1186/1471-2164-13-458PubMed CentralView ArticlePubMedGoogle Scholar
- Courtens W, Grossman D, Van Roy N, Messiaen L, Vamos E, Toppet V, Haumont D, Streydio C, Jauch A, Vermeesch JR, Speleman F: Noonan-like phenotype in monozygotic twins with a duplication-deficiency of the long arm of chromosome 18 resulting from a maternal paracentric inversion. Hum Genet 1998, 103: 497–505. 10.1007/s004390050857View ArticlePubMedGoogle Scholar
- Barbi G, Kennerknecht I, Wohr G, Avramopoulos D, Karadima G, Petersen MB: Mirror-symmetric duplicated chromosome 21q with minor proximal deletion, and with neocentromere in a child without the classical Down syndrome phenotype. Am J Med Genet 2000, 91: 116–122. 10.1002/(SICI)1096-8628(20000313)91:2<116::AID-AJMG7>3.0.CO;2-QView ArticlePubMedGoogle Scholar
- Weise A, Rittinger O, Starke H, Ziegler M, Claussen U, Liehr T: De novo 9-break-event in one chromosome 21 combined with a microdeletion in 21q22.11 in a mentally retarded boy with short stature. Cytogenet Genome Res 2003, 103: 14–16. 10.1159/000076284View ArticlePubMedGoogle Scholar
- Mau UA, Petruch UR, Kaiser P, Eggermann T: Familial robertsonian translocation 15;21 and rare paracentric inv(21): unexpected re-inversion in a child with translocation trisomy 21. Eur J Hum Genet 2000, 8: 815–819. 10.1038/sj.ejhg.5200544View ArticlePubMedGoogle Scholar
- Entesarian M, Carlsson B, Mansouri MR, Stattin EL, Holmberg E, Golovleva I, Stefansson H, Klar J, Dahl N: A chromosome 10 variant with a 12 Mb inversion [inv(10)(q11.22q21.1)] identical by descent and frequent in the Swedish population. Am J Med Genet A 2009, 149A: 380–386. 10.1002/ajmg.a.32663View ArticlePubMedGoogle Scholar
- Colella S, Yau C, Taylor JM, Mirza G, Butler H, Clouston P, Bassett AS, Seller A, Holmes CC, Ragoussis J: QuantiSNP: an Objective Bayes Hidden-Markov Model to detect and accurately map copy number variation using SNP genotyping data. Nucleic Acids Res 2007, 35: 2013–2025. 10.1093/nar/gkm076PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.