Progress of medicine has evolved at a fascinating pace. In the pre 20th century a medical diagnosis was focused on an organ or tissue and had to be based on the symptoms, medical history and physical examination, whereas in the mid-20th century diagnosis was possible on a cellular level with more and more laboratory investigations available. With biochemical and image-based technology the pathology could be made on the protein level and now we have entered the molecular technology and the power of gene-based diagnosis [Gerling et al., 2003]. The skin and its appendages are involved in more than one third of all monogenic diseases and therefore genodermatoses are excellent models for genetic research [Scriver et al., 2001]. Advances in genetics and genomics have revolutionized biomedical science and our understanding of diseases processes and pathophysiology. This new level of science will influence the individualized management of patients with genodermatoses and improve health and quality of life of our patients. Our understanding of diseases has received an increasing insight and with this knowledge the classification of diseases has to be adapted continuously.
THE ECTODERMAL DYSPLASIAS
Ectodermal dysplasias are a large group of heterogeneous heritable conditions characterized by congenital defects of one or more ectodermal structures and their appendages: hair (hypotrichosis, partial or total alopecia), nails (dystrophic, hypertrophic, abnormally keratinized), teeth (enamel defect or absent) and sweat glands (hypoplastic or aplastic). As a rule, ectodermal dysplasias are not considered as pure “one-layer diseases.” Mesodermal and occasionally endodermal dysplasias may coexist. Ectodermal dysplasia deals with a genetic missspelling in genes responsible for skin and appendage formation. Embryogenesis occurs in distinct tissue organizational fields and specific interactions among the germ layers that may lead to a wide range of ectodermal dysplasias exist when genes important for development are mutated or otherwise altered in expression. Ectodermal dysplasias mirror very impressively how medical progress and understanding has a very fast pace. In 1971 only eight forms of ectodermal dysplasias were known [Freire-Maia and Pinheiro, 1984]. Nowadays approximately 200 different ectodermal dysplasias have been delineated, about 30 have been identified at the molecular level with identification of the causative gene [Freire-Maia et al., 2001]. In a time of such a fast medical evolution it is difficult to decide what should be the criteria for a classification of ectodermal dysplasias. Itin and Fistarol 2004 emphasized, that rather commonly non fully expressed phenotypes exist, which make a clinical diagnosis more difficult. Freire-Maia and Pinheiro used the clinical aspects for their classification and Priolo integrated molecular genetic and clinical aspects for her scheme [Pinheiro and Freire-Maia, 1994; Priolo and Lagana, 2001]. Those two more historical classification schemes have the difficulty that when applied strictly, several additional groups of diseases should be integrated within the term of ectodermal dysplasias, for example, keratodermas with skin or hair alterations or the ichthyoses with associated abnormalities. Such consequent classification would lead to an endless list of diseases and would be useless for the practical work. Recent evidence implicates a genetic defect in different pathways orchestrating ectodermal organogenesis. Modern molecular genetics will increasingly elucidate the basic defects of the different syndromes and yield more insight into the regulatory mechanisms of embryology. In this way a reclassification of ectodermal dysplasias will be possible according to the function of their involved mutated genes and their pathways. Lamartine 2003 recently proposed a helpful classification according to the functions of the genes discovered in different types of ectodermal dysplasia. He attempted to separate ectodermal dysplasias into four major functional subgroups: cell–cell communication and signaling; adhesion; transcription regulation; and development.
Why do we create classifications? Clinical observations lead to hypotheses with general applicability. With increasing data the hypotheses have to be adapted. Enormous information about genes is available electronically from biomedical literature in the form of full texts and abstracts. In addition, various publicly available databases (such as GenBank, Gene Ontology, and Entrez) provide access to gene-related information at different levels of biological organization, granularity, and data format. This information is being used to assess and interpret new basic science knowledge [Natarajan and Ganapathy, 2007]. Classifications are just working tools and mirror the scientific time. In general, they are only useful for a limited time until science has new criteria for classification.
After 25 years of clinical work it came clear to me that patients do not follow the textbooks and their classifications. Not all individuals affected by ectodermal dysplasias will have physical findings that fit the description of a specific syndrome. There may be a great deal of variation in the clinical appearance of the same type of ectodermal dysplasia from one affected person to the next. There are more patients who do not have the fully expressed syndrome but have only very few signs and symptoms. However, together with clinical key findings and on a molecular basis the mutation can be found. With such an approach a new classification of ectodermal dysplasia has to include this fact. The following example should underline the above mentioned fact:
A 2-year-old boy was referred with hypotrichosis at the scalp. He had prolonged dermatitis and dry scaling on the vertex with no other complaints. Family history was unremarkable and physical examination confirmed that the child was otherwise healthy. Some superficial scars were visible. The hair was dry and curly. Nails had mild dysplasia but hand and feet were normal and teeth and mouth were also unremarkable (Figs. 1–3).
Our clinical diagnosis favored a mutation in P63 although only erosive dermatitis and alopecia was present and all the other manifestations were lacking. However mutation analysis confirmed a mutation in P63 at position 1700 where adenine was deleted.
A new classification should weight the importance of the clinical findings: There are key findings in the dermatologic examination which allow suggesting a complex syndrome even in absence of the other findings.
ONE GENE—DIFFERENT PHENOTYPES
How should a classification be made? If different phenotypes result from mutations in a single gene, the probable explanation is that a mutation disrupts different functions that are encoded by that gene. Examples are the DNA repair diseases such as the xeroderma pigmentosum (XP) group. The XPD gene encodes a subunit of the basal transcription factor TFIIH. The TFIIH complex has two main functions: Initiation of basal transcription and DNA repair [Itin et al., 2001]. The site of the mutation is important for the phenotype. Therefore a pure classification according to the gene mutation can not be the final solution. Genotypes are embedded in genomes. Individuality in phenotypes is embedded in components of the phenome which includes transcriptome, metabolome, proteome, etc. [Scriver, 2004]. The phenome, its layers, and its nodes, links and networks requires elucidation. Syndrome families might help to create functional genomics [Scriver, 2004]. Phenotype relationships might give clues for pathways and involved genes. Phenotype relationships together with gene functions will help in classifying ectodermal dysplasias. A strict molecular classification would be wrong because mutations in different genes can cause the same phenotypes and one gene can lead to several syndromes. There is a logic linkage between protein complexes and genetic syndrome families. Several diseases with overlapping clinical manifestations are caused by mutations in different genes but they are part of the same functional module.
How can we find these links? The following databases are very helpful:
London Dysmorphology Database LDDB
Pictures of standardized syndromes and undiagnosed Malformations (POSSUM)
How can we use databases for classification? Medical subject headings (MeSH) were used for text mining to analyze OMIM disease records [Brunner and Van Driel, 2004; Natarajan and Ganapathy, 2007]. This is a hierarchical tree in which for example nail is part of finger and finger is part of hand and hand part of limb. In this way it is possible to correct for differences in the level detail of description. To allow this process to be automated the keyword frequencies were represented as vectors. Starting from interesting phenotypic differences and then comparing the mutations that underlie them would probably be more effective. We could define terms, genes, proteins and functions which might help to classify ectodermal dysplasias. Using the keyword-vector method for Stickler syndrome, several related syndromes were recovered: notably the Marshall and oto-spondylo-megaepiphyseal dysplasia (OSMED) syndromes have similar facial and skeletal changes as Stickler syndrome but differ in other characteristics such as deafness and blindness. Molecular investigations have shown that COL2A1, COL11A2, and COL11A1 are involved in these syndromes [Brunner and Van Driel, 2004].
Genetic heterogeneity might result from different interactions at the protein level, such as receptor-ligand interaction. Such interactions have two important implications: once the first gene is found the other related genes are much easier and more effectively found. In addition genes that are involved in the same phenotype will ultimately be shown to have a functional relationship. This example highlights the concept that clinical classification will always precede molecular verification. Similarities in phenotype, biochemistry, genes and protein function can be found by computer assisted text mining and prediction on the simple assumption that mutations in different members of a protein complex may lead to comparable phenotypes can be made [Lage et al., 2007].
An example of genetic heterogeneity to molecular interactions and pathways can be given with Fanconi anemia:
9 loci have been found
Functions were unknown
All involved genes encoding DNA-repair
The same phenotype can evolve by mutations for the ligand or the receptor
The new classification scheme should work for clinicians (horizontal, associative thinking) but should also work for basic scientists (vertical, analytic thinking). Ectodermal dysplasias might be grouped clinically by: Clinical key findings (phenotypes) and syndrome families. In addition, ectodermal dysplasias might be grouped by genes and their functions:
Genes for cell–cell communication
Genes for adhesion
Genes for transcription
Genes for development
Genes for aging/DNA-repair
Genes for structure proteins
A further classification concept could include genes and their gene products (proteins):
What should the new diagnostic criteria for ectodermal dysplasia include now and how has the classification of ectodermal dysplasia changed? In a new concept grouping of clinical findings are made and gene pathways and functions as well as syndrome families are considered and clinical key symptoms are included.
What will help the new classification? Clinicians do not need to know all details of all entities which constitute ectodermal syndromes but they should recognize key findings. With these key findings they can put a patient in a syndrome family. Specification can then be done later by molecular analysis.