SEARCH

SEARCH BY CITATION

Keywords:

  • Genetics;
  • Neurology;
  • Peripheral nervous system disorders

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Background:

Although reporting the same clinical phenotype, inherited polyneuropathy in Leonberger dogs (ILPN) has been attributed to various modes of inheritance.

Hypothesis:

The ILPN is one disease with a major risk factor on chromosome X.

Animals:

Dogs affected by ILPN (n = 104).

Methods:

Pedigree analyses were performed by means of a case-control approach. Data were retrieved either from medical records of cases diagnosed by the first author (n = 13), from breeders (n = 18) or from different registries publishing data on affected dogs (n = 73). A comparison was made between the X-chromosome ancestry of fathers of affected male dogs and the ancestry of the X-chromosomes of mothers of affected dogs of either sex. A systematic random sample, obtained from an international database of registered Leonberger dogs, served as a reference population regarding ancestry.

Results:

Having one particular female, born 1943, in the X-chromosomal lineage is a major risk factor for developing ILPN. Sex distribution among affected dogs is in favor of a risk factor on the X-chromosome and contradicts a monogenic autosomal or mitochondrial inheritance.

Conclusions and Clinical Importance:

The ILPN is considered most likely to be one disease, and the inheritance of ILPN is best explained by an underlying X-linked mode of transmission for the phenotype. However, age at onset and severity of signs might be determined by contributing loci. This has consequences in molecular genetic studies and for breeding strategies aimed at eliminating this disease.

Abbreviations
ILD

International Leonberger Database

ILPN

inherited polyneuropathy in Leonberger dogs

SKK

Swedish Kennel Club

WILD

Worldwide Independent Leonberger Database

Inherited polyneuropathy in Leonberger dogs (ILPN) is a degenerative neurologic disease that has recently received increased attention, with dogs diagnosed all over North America and Europe, including Scandinavia.[1, 2] Case descriptions have been published on the internet, and the identity of a large population of affected dogs is available through this source. Two scientific publications report cases of ILPN with different modes of inheritance despite a similar clinical phenotype.[3, 4]

The major neurologic findings reported in affected dogs were distal muscle atrophy, a high-steppage pelvic-limb gait, reduced spinal reflexes and laryngeal paresis or paralysis.[3] Results from electrophysiology as well as muscle and nerve biopsy were consistent with a chronic distal symmetric polyneuropathy. Onset of clinical signs was observed at between 1 and 9 years of age. Clinically, electrophysiologically, and pathologically, this disease was reported to resemble Charcot-Marie-Tooth's syndrome, which is a group of heredodegenerative polyneuropathies in man.[2, 3, 5, 6] Subgroups of this human syndrome have been defined based on different modes of inheritance. Recessive and dominant autosomal as well as X-chromosome linked subgroups have been reported.[5, 6]

Of the 21 cases reported, 20 were males.[3] Pedigree analysis was performed on 15 of these affected dogs belonging to one six-generational family, and an X-linked recessive mode of inheritance was considered most likely for ILPN based on the results, with a probable partial or age-dependant penetrance.[3] An X-linked recessive mode of inheritance will generate a skewed sex distribution of affected individuals with predominantly males being affected. Males carrying the defect allele will display the diseased phenotype, whereas females can transfer the defect allele over generations as nondiseased carriers. Affected female offspring are expected only if a carrier (or affected) bitch is bred to an affected male. On the other hand, an affected male does not transmit the genetic defect to its male puppies.

There is a distal symmetric polyneuropathy in French and Belgian Leonberger dogs with neurologic signs identical to the dogs in the report earlier.[3, 4] Based on pedigree analysis of a family group consisting of 10 affected females and 6 affected males, an autosomal recessive mode of inheritance was suggested.[4] As an explanation for the suggested diverging patterns of inheritance, the condition in these European Leonberger dogs was presumed to be genetically different from the polyneuropathy in American Leonbergers reported previously.

A genetic test was recently launched for a severe early-onset form of ILPN termed LPN1.1 The identified mutation in homozygotes is reported to be responsible for approximately one-third of the cases of polyneuropathy in Leonbergers, and is introduced as one of several possible genetic risk factors. However, about 25–30% of young onset ILPN-affected dogs are actually clear of the identified mutation. According to a recent research abstract, 2 loci on autosomal chromosomes 16 and 7 are strongly associated with Leonberger polyneuropathy.2 Neither of the loci, in isolation or combination, appear to be shared by all Leonberger dogs affected by ILPN, but both loci might contribute to disease severity and pathogenesis.

The hypothesis behind the present study was that inherited polyneuropathy in the Leonberger breed is one disease with a recessive major risk factor on the X-chromosome, and that the relevance of this hypothesis could be tested primarily by a comparison between the X-chromosome ancestry of fathers of affected male dogs and the ancestry of the X-chromosomes of mothers of affected dogs of either sex. The rationale for choosing these 2 groups for comparison was that fathers do not contribute X-chromosomes to their sons, and if the risk factor is located on the X-chromosome, fathers of affected males can be assumed to be representative of the general population regarding this factor. On the other hand, mothers must be either carriers of the risk factor or affected by ILPN to produce affected offspring under this hypothesis. A larger group of dogs either diagnosed by the first author (KHJ) or reported from various sources to be affected by ILPN was therefore investigated by pedigree analyses. The study was conducted with the specific aim to establish if the occurrence of this hereditary disease in the breed could be explained by an underlying genetic defect in the X-chromosome with contributing loci, for example, the identified LPN1-mutation, determining, eg, severity and age of onset in subsets of affected dogs.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Affected Dogs

Leonberger dogs were classified as being affected by ILPN (n = 104) if they were either diagnosed by a board-certified neurologist in Sweden (KHJ) (n = 13), or published by the Leonberger Club of America Health, Research and Education Committee as diagnosed by clinical signs, nerve and muscle biopsies (n = 36), or presented in their cause-of-death registry as affected by polyneuropathy (n = 2).c Additional data on affected dogs were retrieved by personal communication from Swedish breeders (n = 18), from individual owners and breeders that publish details of affected dogs on web sites (n = 11),5,6,7,8 from health status as described in the International Leonberger Database (ILD) (n = 4),3 and from a German cause-of-death registry (n = 20).9 All these dogs were also classified as affected. Affected dogs were identified by year of birth, sex, country of origin and registered pedigree name and/or identification number.

Control Dogs and Reference Population

Parents of affected dogs were included for comparison of the ancestry of the X-chromosomes. These so-called control dogs were classified into 3 groups; fathers of affected male dogs (male-fathers), fathers of affected female dogs (female-fathers), and mothers of affected dogs of either sex (mothers). Whenever a father or a mother itself had been reported to be affected by ILPN, that parent was classified among the affected dogs and hence excluded as a control dog. To compare the X-chromosome ancestry of the included fathers and mothers to a reference population of dogs, a systematic random sample including every tenth dog born in 2008 and listed in the ILD was obtained, and the X-chromosome ancestry of these dogs was investigated. Within the reference population, the ancestry of the X-chromosomes of Leonbergers from France was compared with the ancestry of the X-chromosomes of Leonbergers from all other countries.

Pedigree Analyses

The pedigrees of affected dogs, control dogs, and dogs in the reference population were analyzed regarding their X-chromosomal lineages. Pedigrees were retrieved from the Swedish Kennel Club (SKK) registry,10 ILD, and the Worldwide Independent Leonberger Database (WILD).11 The completeness of the information included in the ILD and WILD was validated by comparing pedigree data for affected dogs registered in the SKK registry (n = 26) with information in ILD and WILD.

Each affected dog, control dog, and reference dog, was traced back to the earliest X-chromosome ancestor found in the registry of ILD. X-chromosomal lineages were depicted. The proportion of affected dogs, control dogs and reference dogs, respectively, with a specific X-chromosome ancestor is presented.

Statistical Analyses

The relationship between sex and occurrence of ILPN was tested by means of a case-control approach by a two-sided Fisher's exact test. The proportion of all dogs in the reference population with female A1 in the X-chromosomal lineage was compared with the proportion among all male-fathers by means of the same test. Finally, the proportion with A1 in the X-chromosomal lineage was compared between the 3 groups of control dogs by two Fisher's exact tests (female-fathers versus male-fathers and mothers versus male-fathers). The statistical analysis was performed with the software package Stata version 11.12

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The ancestry of the X-chromosomes was documented for a total of 104 affected dogs, 20 females (19%), 83 males (81%), and 1 dog of unknown sex (Fig 1). The affected dogs were born between 1965 and 2007. The countries of origin for the affected dogs were Canada (n = 5), Czech Republic (n = 1), Denmark (n = 1), France (n = 1), Germany (n = 11), the Netherlands (n = 8), Norway (n = 3), Sweden (n = 26), Switzerland (n = 5), United Kingdom (n = 17), and the United States (n = 25). In addition, the ancestry of the X-chromosomes was documented for all male-fathers (n = 49), all female-fathers (n = 16), all mothers (n = 78), and a total of 436 different dogs in the reference population born in 2008. Nine of the affected dogs (4 females and 5 males) were reported to be parents to other affected dogs, but were included in the group of affected dogs in accordance with the inclusion/exclusion criteria.

image

Figure 1. X-chromosomal lineages depicted for 94 Leonberger dogs affected by inherited polyneuropathy with a common X-chromosome ancestor, female A1. Arrow points at female A1. The figure shows the widespread distribution of affected dogs through many generations.

Download figure to PowerPoint

The earliest X-chromosome ancestors in all analyzed pedigrees were 4 different females found in the registry, female A, B, C, and D (Table 1). Female A, born approximately in 1915, was represented in the investigated pedigrees through 3 different X-chromosomal lineages in the mid-1940s, female A1, A2, and A3 (Table 2). Female C, registered as a Leonberger, was reported to descend from a Newfoundland X-chromosomal lineage, whereas the X-chromosome ancestor of the female D-Leonberger was reported to be a “non-Leo, possible Ovtcharka.” No breed ancestry of females A and B was noted in the database.

Table 1. The 4 earliest X-chromosome ancestors (female A, B, C, and D) found in the pedigrees of 104 Leonberger dogs affected by inherited polyneuropathy, their parents (classified into mothers, fathers to affected females and fathers to affected males, respectively), and 436 reference dogs. Number and proportions of dogs in the respective groups with one or another of the X-chromosome ancestors are given.
DogBorn (approx.)AffectedMothersFemale-fathersMale-fathersReference Dogs
Female A191598/104 (94%)72/78 (92%)15/16 (94%)42/49 (86%)398/436 (92%)
Female B19450001/49 (2%)0
Female C19451/104 (1%)1/78 (1%)01/49 (2%)9/436 (2%)
Female D19705/104 (5%)5/78 (6%)1/16 (6%)5/49 (10%)29/436 (7%)
Table 2. The 3 different X-chromosomal lineages (female A1, A2, and A3) in the mid-1940s descending from female A (born approximately 1915), found in the pedigrees of 104 Leonberger dogs affected by inherited polyneuropathy, their parents (classified into mothers, fathers to affected females and fathers to affected males, respectively), and 436 reference dogs. Number and proportions of dogs in the respective groups with one or another of the X-chromosome ancestors are given.
DogBorn (approx.)AffectedMothersFemale-fathersMale-fathersReference Dogs
Female A1194594/104 (90%)68/78 (87%)15/16 (94%)29/49 (59%)286/436 (66%)
Female A219452/104 (2%)2/78 (3%) 09/49 (18%)58/436 (13%)
Female A319452/104 (2%)2/78 (3%) 04/49 (8%)54/436 (12%)

Among the affected dogs, the proportion of males to females differs significantly (P < .001) from that (50 : 50) expected by a monogenic autosomal inheritance or a mitochondrial inheritance. The total number and the proportion of affected dogs, female-fathers, male-fathers, mothers, and dogs in the reference population from the different X-chromosome ancestors, respectively, are presented in Table 1. In Table 2, the corresponding figures are shown for dogs from different bitches born in the 1940s in the X-chromosomal lineages from female A. Female A1 is represented in the X-chromosomal lineage of 29/49 (59%) of the male-fathers, which is not significantly distinguishable from the proportion (66%) in the reference population (P = .43). However, female A1 is represented in the X-chromosomal lineage of 68/78 (87%) of the mothers, and in the X-chromosomal lineage of 15/16 (94%) of the female-fathers. There was a significant difference in the proportion that had female A1 in the X-chromosome ancestry between female-fathers and male-fathers (P = .013), as well as between male-fathers and mothers (P = .005).

The registries ILD and WILD were both in complete agreement with the SKK registry regarding ancestry of the X-chromosome of the affected dogs with a Swedish identity number. However, it was only possible to trace X-chromosomal lineages back to the beginning of the 1960s in the SKK database.

The investigated dogs in the reference population (n = 436) originated from 22 different European countries, Australia, New Zealand, 2 countries in North America, and 2 in Asia. The largest numbers of dogs were registered from France (n = 77), Germany (n = 67), United Kingdom (n = 44), Sweden (n = 43), and the United States (n = 32). Regarding the French reference dogs, 64% had female A1 in the X-chromosomal lineage, whereas 19, 15, and 1% had female A2, A3, and C in the X-chromosomal lineage, respectively. These figures are not significantly different (P > .08 to P > .6) from the investigated reference population of Leonbergers from all other countries (excluding French dogs).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The proportion of ILPN-affected males to females contradicts a monogenic autosomal or mitochondrial inheritance and instead suggests an X-linked mode of transmission. Pedigree analysis suggests that a commonly occurring ancestor, female A1, is the founder of X-chromosome linked ILPN in modern Leonbergers across the world. Although this could not be confirmed with certainty, the current analysis in conjunction with information on the genetic test1 and the recent research abstract2 indicates that ILPN is a polygenic disorder with a major risk factor on chromosome X. In the following, the argument for ILPN being X-linked and being one single disease is discussed in more detail. Furthermore, potential biases in the analysis as well as practical implications of the current findings are presented.

By comparing the X-chromosome ancestry of mothers and female-fathers to male-fathers, we compared obligate carriers of the hypothesized X-linked defect with dogs that were expected to be representative of the general Leonberger population. We did not use the group of affected dogs in these statistical calculations to avoid the bias from having more than one affected dog in some litters.

In all affected dogs (n = 10) descending from other X-chromosome ancestors than female A1 (and in the female-father with other X-chromosome ancestry) there is a theoretic possibility that an affected sire descending from female A1 in previous generations had transferred a defect X-chromosome to a carrier daughter. Fathers in these lineages with female A1 as the X-chromosome ancestor are depicted in Figure 2. Therefore, the affected dogs in this study are related to each other in a mode such that a possible genetic defect in an X-chromosome of female A1 might have been transmitted to all the affected dogs.

image

Figure 2. X-chromosomal lineages depicted for 10 Leonberger dogs affected by inherited polyneuropathy, where female A1 was not in the X-chromosome ancestry of the affected dog. In addition, one father of an affected female dog is inserted in the pedigree. Arrows point at female dogs that have been sired by stud dogs with female A1 in the X-chromosomal lineage, and thus potentially could have introduced a mutated X-chromosome in the respective X-chromosomal lineage.

Download figure to PowerPoint

The frequencies of males and females expressing a recessive X-linked condition should be f versus f2, respectively, in a population. From the material in this study, true frequencies for ILPN in males and females could not be calculated. However, counting backwards, with a sex distribution of 83 affected males and 20 affected females in this group of dogs, frequencies consisting of approximately 24% affected males and 6% affected females in the Leonberger population should be consistent with a nonfatal X-linked recessive trait. Neither calculations of prevalence and incidence of this disease nor a segregation analysis can be performed based on the material from this study.

Based on differing sex distribution, it is suggested that ILPN is a genetically different disease in Europe and America.[4] In the present study, the countries of origin of the affected dogs are not completely in agreement with the dogs in the reference population. This is partially because of the authors’ own case material originating from Sweden together with a considerable amount of affected dogs published by the Leonberger Club of America Health, Research and Education Committee. However, with the narrow genetic base for the breed and the mixture of European and American ancestors in the pedigrees of dogs from all included groups, it is unlikely that the breed is divided into 2 genetic pools. The authors of the present study find it most plausible that ILPN in Europe and America is the same disease, genetically determined by a defect in the X-chromosome, but phenotypically modified by gene variants on autosomal chromosomes. It cannot be excluded that the penetrance might also be modified by sex hormones or environmental factors.[3]

In human X-linked Charcot-Marie-Tooth's syndrome, males carrying the defect allele are moderately to severely affected, whereas heterozygous females tend to be mildly affected or completely asymptomatic.[7] The disease is considered a dominant disease because of indications for a toxic gain of function from the defect gene, and the variation in phenotypes between males and females is suggested to be caused by X-inactivation. A similar mechanism explaining the variation in phenotypes among ILPN-affected Leonberger dogs cannot be excluded.

The certainty of a correct diagnosis was high in about half of the affected dogs, namely those diagnosed by the first author (KHJ) and those published by the Leonberger Club of America Health, Research and Education Committee as diagnosed by clinical signs, nerve and muscle biopsies.4 For the remainder of the dogs, the diagnostic criteria were not provided, and the possibility of false positives must be considered. If there are false positives among the affected dogs, these misdiagnosed dogs would in reality have suffered from other diseases and as such should be more similar to the reference population regarding X-chromosome ancestry. If this so-called misclassification bias is nondifferential or independent of the exposure (in this case having female A1 in the maternal ancestry), it might have biased the effect estimates toward the null. Even with that possible scenario in mind, the statistical significances for a difference in X-chromosome ancestry are convincing. Regarding, on the other hand, a possibility for differential misclassification (dependent of the exposure) in our material, we have no reason to believe this has occurred. The affected dogs were reported directly by owners and breeders or found through various websites, and were probably not communicated with the investigation of the impact of female A1 in mind.

Status regarding ILPN was unknown for control dogs and dogs in the reference population, and false negatives are expected among these dogs. These groups of dogs may include both unreported cases and undetected or misdiagnosed cases.[1] Furthermore, only 4 different affected fathers to diseased daughters are known, even though all 20 affected females must have had a father with a mutation in the X-chromosome, if this disease is truly recessively X-linked. However, among the 104 affected Leonberger dogs in our material, as many as 26 dogs (18 males and 8 females) have offspring listed in ILD. Most probably, a majority of these dogs were bred without the breeder recognizing any clinical signs of a polyneuropathy at the time-point of mating. It seems plausible that there could have been affected parents that were never detected or reported to be affected by ILPN. Another reason to assume the presence of false negatives is the age of onset, reported to be as late as at 9 years of age[3] and 6 years of age[4] in some affected dogs. A substantial number of Leonberger dogs will die because of other reasons before they reach an age where signs of ILPN should have occurred in a genetically affected dog. In a study of mortality in over 350,000 insured Swedish dogs, the probability of death by 8 years of age was 41% for Leonberger dogs, and by 10 years of age, it was 74%.[8] The anticipated inclusion of false negatives in this study has probably biased the effect estimates toward the null.

The Leonberger registries ILD3 and WILD11 that were used in the pedigree analyses are both handled by end users, are available free on the internet, and have approximately 105,000 entries each. No fundamental differences were found in the quality of data from ILD and WILD regarding its use in this study. The completeness of the registries was validated with excellent results against the SKK registry.10 The SKK registry was considered the “gold standard” in these comparisons, because it is detailed and contains reliable pedigree data. It has been computerized since 1976 and, furthermore, identification of individual dogs by tattoo or microchip has been mandatory for dogs registered in the SKK database since 1997.

In conclusion, this study shows with statistically significant evidence that having female A1 as ancestor in the X-chromosomal lineage is a risk factor for developing ILPN. Together with the skewed sex distribution among affected dogs, the inheritance of this disease is believed to be best explained by an underlying X-linked mode of transmission for the phenotype. However, age at onset and severity of signs might be determined by modifying genes. With an anticipation of false negatives in the population, it is difficult to claim a given dog as genetically unaffected. This has consequences in molecular genetic studies where healthy controls are needed, and also results in difficulties for breeding strategies aimed at eliminating this disease. Defining the molecular genetic basis for all cases of this disease would be of help to the Leonberger breed. For this purpose, further studies are needed.

Footnotes
  1. 1
  2. 2

    Ekenstedt KJ, Drögemüller C, Minor KM, et al. Whole-genome association analysis reveals two loci strongly associated with Leonberger polyneuropathy. Fifth International Conference: Advances in Canine and Feline Genomics and Inherited Diseases. Baltimore, MD, 2010 (abstract)

  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12

    Statacorp., College Station, TX

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The authors thank Dr Helga R. Høgåsen for her help with French translation and Prof Leif Andersson for his views on the manuscript. The authors are also grateful to the contributing dog owners and breeders.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References