Genetics of inherited skin disorders in dogs The Veterinary Journal

Canine genodermatoses represent a broad spectrum of diseases with diverse phenotypes. Modern genetic tech- nology including whole genome sequencing has expedited the identification of novel genes and greatly simplified the establishment of genetic diagnoses in such heterogeneous disorders. The precise genetic diagnosis of a heritable skin disorder is essential for the appropriate counselling of owners regarding the course of the disease, prognosis and implications for breeding. Understanding the underlying pathophysiology is a prerequisite to developing specific, targeted or individualized therapeutic approaches. This review aims to create a comprehensive overview of canine genodermatoses and their respective genetic aetiology known to date. Raising awareness of genodermatoses in dogs is important and this review may help clinicians to apply modern genetics in daily clinical practice, so that a precise diagnoses can be established in suspected genodermatoses.


Introduction
Currently, 6000-8000 rare diseases are known in humans, most of them inherited as monogenic traits. The causative genetic variants have been identified in more than 5600 monogenic diseases. 1 , 2 While each of these diseases is rare, it is currently thought that up to 10% of the human population in the United States may be affected by at least one rare disease. 3 More than 500 heritable diseases of the skin (genodermatoses) are known in humans (Leech and Moss, 2007;Feramisco et al., 2009;Lemke et al., 2014). Nonetheless, new patients with yet undescribed genodermatoses continue to be identified. Compelling arguments for the growing importance of genetic analyses in clinical practice were provided in an excellent review on human genodermatoses (Lemke et al., 2014).
In dogs, far fewer heritable skin diseases with known molecular aetiology are known. Five years ago, we compiled 19 canine genodermatoses (Leeb et al., 2017). Thanks to collaborations between private veterinarians, veterinary dermatology specialists and veterinary dermatopathologists and due to advances in genetics and sequencing technology, this number is growing rapidly. In this review, we will summarize data on 36 canine genodermatoses and their underlying causative genes. We searched OMIA 4 and PubMed 5 for the relevant information on canine genodermatoses and their respective genetic aetiology known to date.
Dogs and humans are closely related and share the vast majority of their ~20,000 protein coding genes. Not surprisingly, many genodermatoses are homologous between dogs and humans. Their comparative investigation in related species may be considered a One Health approach and is of mutual benefit to human and veterinary medicine. While the majority of newly described canine genodermatoses has already a well characterized homologous disease in humans, there are several examples, for which the causative gene was initially discovered in dogs. This suggests that more genes with important roles for the development and maintenance of a healthy skin are still likely to be discovered.
This review focuses on skin disorders with a monogenic inheritance or with clearly identified major genetic risk factors. We did not include normal hair morphology traits (e.g. long vs. short hair, curly vs. straight hair) or normal coat colour variation, except in those cases where a specific coat colour is clearly associated with a skin disease (e.g. coat colour dilution alopecia). A few syndromic diseases with major skin involvement were also included (e.g. acral mutilation syndrome, a sensory neuropathy, in which the skin lesions are a secondary consequence of the primary disease). The review will not deal with multifactorial diseases such as atopic dermatitis that are probably influenced by (many) genetic risk factors with small effects. We provide lists of the currently known canine genodermatoses and their underlying causative genetic variants. The review will also briefly discuss the use of genetic analyses as a complementary approach to clinical and histopathological examinations and its potential impact for precision medicine.

Canine genodermatoses with known genetic aetiology
A total of 36 genes related to heritable skin disorders were identified.
To facilitate the use for clinicians, we grouped these genes according to phenotype/disease. We classified the resulting diseases into ichthyoses and other disorders of keratinization (  Fig. 6). Our grouping is not mutually exclusive. For example, the epidermolytic hyperkeratosis caused by KRT10 variants is listed in the ichthyoses and the epidermolyses groups. The vast majority of canine genodermatoses is caused by relatively young deleterious alleles, which have arisen by spontaneous mutation events during the last decades. This explains why most genodermatoses are restricted to a single dog breed or a group of closely related breeds that may have had some recent genetic exchange. Notable exceptions are the ancient FOXI3 allele causing ectodermal dysplasia in several hairless dog breeds (Drögemüller et al., 2008) and the various MLPH alleles for coat colour dilution that segregate in many dog breeds and Mixed breed dogs (Drögemüller et al., 2007;Bauer et al., 2018c;Van Buren et al., 2020).

Purebred vs. Mixed breed dogs
Mixed breed dogs have a similar population structure as most human populations. Consequently, monogenic heritable diseases in Mixed breed dogs are relatively rare overall. However, dominant diseases affect Mixed breed and purebred dogs equally likely.
Purebred dogs are bred in strictly isolated populations. This requires a certain degree of inbreeding and promotes the expression of recessive alleles. These breeding practices have enabled to create ~400 phenotypically highly diverse dog breeds with relatively little intra-breed phenotypic variability. However, the inbreeding has also negative consequences. Inbred dogs have an increased probability to inherit identical genome segments from their parents. Consequently, recessive diseases also occur more frequently in purebred than in Mixed breed dogs. This partly resembles the situation in some isolated human populations where consanguineous marriages are common. In purebred dogs, the problem is further exacerbated by the excessive use of elite breeding animals ('popular sire effect'). Every dog carries a small number of recessive deleterious alleles. As long as the effective population size is reasonably large and the breeding dogs are used with roughly equal intensity, the population will contain many deleterious alleles with very small individual frequencies. The risk that any of these deleterious alleles will come together in a homozygous state in an offspring is not negligible due to the inbreeding, but is still limited. The excessive use of very few breeding animals may rapidly shift this towards a situation, in which some of the deleterious alleles reach very high frequencies. Therefore, without proper breeding programs, monogenic recessive diseases in purebred dogs may quickly evolve from rare to common diseases.
The most striking real example of these theoretical considerations can be seen in the recessive Golden retriever ichthyosis, which is caused by a PNPLA1 variant (Fig. 1). Due to the intensive use of a carrier as breeding sire in the 1990s, the pathogenic allele had reached the alarmingly high frequency of more than 50% in the breeding population at the time when the causative variant was finally identified (Grall et al., 2012). The corresponding genotype frequencies amounted to roughly 32% affected dogs, 49% heterozygous carriers and only 20% clear dogs (Owczarek-Lipska et al., 2011).

Targeted genetic testing for breeders and veterinarians
For many of the known genodermatoses, commercial genetic testing is available. Such a test will only interrogate a single genetic variant in the genome. An overview on different flavours of targeted genetic tests is provided in Supplementary Table S1 (see Appendix A: Supplementary material). These tests are very helpful, if a particular recessive disease has spread in a dog breed. When a recessive disease allele has a frequency of more than 1%-5% in a population, it is strongly recommended that breeders test their breeding animals to identify heterozygous carriers and to avoid carrier x carrier matings.
These targeted genetic tests can also be valuable for veterinarians to confirm a presumptive diagnosis, avoid unnecessary or invasive diagnostics and prevent futile therapies. A precise genetic diagnosis can be of great help in counselling a dog owner regarding the prognosis, required therapeutic management or humane euthanasia. For example, dogs with inherited footpad hyperkeratosis have a normal life expectancy but an impaired quality of life as walking is painful and their footpads require continuous lifelong management (Drögemüller et al., 2014;Plassais et al., 2015;Backel et al., 2020). Thus, having a dog with a genodermatosis can be a non-negligible financial and emotional burden for an owner (Fig. 4). If a dog with a suspected genodermatosis tests positive, the genetic test will definitively establish the suspected diagnosis. However, a negative genetic test result must be considered inconclusive. If the tested disease allele is not present, it is still possible that an independent mutation event, perhaps in the same gene, has happened. Targeted tests interrogate only a single position in the genome, they do not comprehensively test an entire gene or the entire genome.
There are numerous laboratories providing single targeted genetic tests and panels of targeted tests. Good overviews and databases on these laboratories are provided by the WSAVA 6 and the IPFD initiative. 7 Thus, the use of a targeted genetic test to confirm a suspected diagnosis is primarily recommended for recessive diseases and in purebred dogs of the same or at least a closely related breed, in which the causative genetic variant was originally reported. In all other constellations, it is unlikely that a targeted test will come up positive and then it may be better to consider a more comprehensive analysis, such as the sequence analysis of entire candidate genes or whole genome sequencing (WGS). Such analyses have become routine diagnostic procedures in human medicine. In veterinary medicine, they are technically feasible, but are not (yet) considered routine diagnostics and mostly undertaken in the framework of research projects.

Genetic analyses in a research context
In this chapter, we will briefly summarize the available methods to identify new pathogenic variants in dogs with suspected genodermatoses. Specific details on the methods are shown in Supplementary Table S1 (see Appendix A: Supplementary material). Gene panel sequencing and whole exome sequencing (WES), two currently popular approaches in human genetics, are not practical in dogs or any other domestic animal species due to cost considerations and a lack of commercially available reagents.
At the time of writing this review, only two approaches are routinely used in veterinary genetics. The first of these comprises PCR amplification followed by conventional Sanger sequencing of all exons of a given candidate gene. This approach has been in use for more than 20 years with little variations. It is most attractive for highly characteristic phenotypes, for which only a single candidate gene with few exons is known (e.g. verrucous epidermal keratinocytic nevi in female dogs due to heterozygous NSDHL variants; Leuthard et al., 2019). This approach requires a laboratory with experienced personnel and specifically adapted protocols (primers) for each disease and each individual candidate gene. It will also only detect relatively small coding variants and has little sensitivity for large structural variants. The main advantage of this method is that it requires very little data analysis and can provide results within 1-2 weeks.
The second commonly used method for genetic variant identification in canine genetics is whole genome sequencing (WGS). WGS has become increasingly popular during the last 10 years and is expected to become the primary approach for the analysis of single gene disorders in humans The phenotype of this genodermatosis is quite variable and often so mild that owners do not notice any problem. The role of the PNPLA1 gene for skin integrity was initially discovered in Golden retrievers. Only later on, human patients with a related homologous ichthyosis were identified (Grall et al., 2012). (B) Histologically, the epidermis is covered by dense laminar orthokeratotic keratin, which is exfoliating in layers. (C) Healthy control skin. Note the basket-weave orthokeratosis in the stratum corneum. . (Lionel et al., 2018). The laboratory costs to sequence a dog genome are less than $ 800. Unfortunately, these costs cover only the generation of the raw sequencing data and not (yet) the data analysis, which currently constitutes the major bottleneck for this technology. WGS has many advantages: It provides a nearly complete list of all genetic variants in a given dog's genome. Thus, WGS is the method of choice, if more than one candidate gene needs to be considered or if the candidate gene has many exons. The lab work for WGS is highly standardized and can be outsourced to commercial service providers. WGS has a higher sensitivity to detect structural variants than amplicon-based approaches. The only major drawback of WGS are the large data amounts necessitating complex bioinformatic analyses. Currently, this requires highly qualified personnel and considerable computational hardware. In the environment of the authors, WGS-based variant discovery in dogs still takes a few months. It is expected that WGS based approaches will further evolve and become more automated. In the not too distant future, it seems feasible that a WGS experiment of a single dog including data analysis may be completed in two weeks for approximately $1000. In specialized centres for human newborns, WGS can already be completed in less than a week to enable rapid therapeutic intervention (Hayeems et al., 2020). A hypothetical future diagnostic WGS analysis in a dog should provide a clear and concise report that is of use to veterinary clinicians. The report will have to be limited to all pathogenic and likely pathogenic genetic variants in clinically relevant genes with known functions. With these pending developments, it is a matter of time when the genetic analysis of canine patients with known genodermatoses transitions from research to standard diagnostics, similar to what has happened in human genetics (Lemke et al., 2014;Richards et al., 2015;Hayeems et al., 2020). However, there is always the possibility that spontaneous mutation events result in the emergence of truly new heritable diseases that have never been seen before. New heritable diseases will remain the primary domain of veterinary genetics research. For almost a third of the 36 genes listed in this review, their significance for dermatology has initially been discovered in dogs, prior to any insights from humans, mice or any other species (Figs. 1, 3, and 4). The population structures of dogs in general and purebred dogs in particular are ideally suited for the elucidation of causative genetic variants in new genodermatoses and other heritable phenotypes, when no functional candidate genes are available. In these situations, hypothesis-free unbiased approaches have proven stunningly successful in veterinary genetics Parker et al., 2010), although it is beyond the scope of this review to explain all these techniques in detail. The advances in sequencing technology and decreases in costs enable very direct experiments, in which complete genome sequences of single or multiple cases are directly compared to hundreds or thousands of control genome sequences. It is then straightforward to identify the so-called private variants that exclusively occur in an affected dog. If sufficiently large numbers of control genomes from genetically diverse dogs are available, these lists of private variants will become very short and may allow the direct identification of likely candidate causative genetic variants.

Benefits of genetic analyses
Increasing numbers of known genodermatoses in dogs and the substantial variability of their clinical and histopathological phenotypes pose significant challenges for the clinician to reach the correct diagnosis. Genetic analyses have matured into an important part of the diagnostic portfolio (Fig. 7). Genetic analysis is ideally performed on an EDTA blood sample and does not require highly invasive sample procurement. The costs for targeted genetic tests are comparable to many routine laboratory tests. The costs for experiments to detect new genetic variation are admittedly higher, but still in the range of commonly used diagnostic tests like MRI. Moreover, if a new heritable disease is suspected, the costs for genetic analyses can often be covered by research funds. The genome sequence or any experimentally determined genotype will remain stable and need to be determined only once throughout the life of a dog.
Knowing the causative genetic defect of a genodermatosis enables reliable predictions about the mode of inheritance, which is an essential prerequisite for rational breeding decisions. For example, the discovery that a new ichthyosis in a young German shepherd was due to a spontaneously arisen dominant allele was very good news for the breed as only the affected dog and its parents had to be excluded from breeding in order to reliably eradicate this specific disorder (Bauer et al., 2017a). Knowledge on the molecular pathogenesis furthermore directed the therapeutic management.
In some instances, the genetic diagnosis sheds light onto different aspects of a heritable disease that are not immediately obvious from the clinical presentation. A Rottweiler with pronounced footpad hyperkeratosis developed severe allergies at a few months of age (Backel et al., 2020). While the allergies were initially considered as a separate disease, the genetic analysis revealed a genetic variant in the DSG1 gene, which leads to a barrier defect and causes a syndromic genodermatosis that is characterized by a combination of footpad hyperkeratosis and multiple allergies (Samuelov et al., 2013;Has et al., 2015;Backel et al., 2020).
In very rare instances, a precise diagnosis may even open options for a targeted therapy. Female dogs with heterozygous NSDHL loss-offunction variants develop verrucous epidermal keratinocytic nevi that   Jagannathan et al. (2013)). The SUV39H2 gene encodes a histone 3 lysine 9 methyltransferase, which mediates transcriptional silencing of chromatin. Research into this disease helped to better understand the role of epigenetic processes during differentiation of keratinocytes (Balmer et al., 2021).  follow Blaschko's lines and are often arranged in linear patterns (Bauer et al., 2017b;Leuthard et al., 2019;Christen et al., 2020). The NSDHL gene encodes NAD(P) dependent steroid dehydrogenase-like, an enyzme required for cholesterol biosynthesis. The striking pattern of the skin lesions is due to the random X-chromosome inactivation in females. Cholesterol biosynthesis is blocked in skin areas, in which the X-chromosome with the defective NSDHL allele is expressed. In these skin areas, the lack of cholesterol and the accumulation of toxic metabolic intermediates cause the skin lesions. If the NSDHL defect is diagnosed in time, a topical therapy containing cholesterol and lovostatin to prevent the further production of toxic metabolites may lead to a marked improvement of the lesions (Leuthard et al., 2019).

Conclusions
Genetic analyses provide a valuable complement to clinical and histopathological examinations. They often represent the fastest and least invasive approach to enable a precise diagnosis. Knowledge on the causative genetic defect in a genodermatosis is essential to give rational breeding recommendations and may even open up options for targeted therapy. The investigation of new genodermatoses has the potential to inform human medicine and to unravel new gene functions that are required for the development and maintenance of healthy skin und thus contribute to our mechanistic understanding of skin homeostasis.
Interested clinicians and pathologists are encouraged to consult with veterinary geneticists if they suspect a new genodermatosis.

Conflict of interest statement
The University of Bern holds a patent on genetic testing for hereditary nasal parakeratosis in Labrador retrievers. Royalties from this patent are paid to the University of Bern and Tosso Leeb.   7. Key requirements for the genetic analysis of a suspected new genodermatosis. If an inherited disease is suspected, it is essential to document a family history and to take appropriate samples suitable for DNA isolation. Obtaining additional samples from the parents (first priority) and/or littermates (second priority) increases the chances of success.