Abstract
Cancer is a complex disease involving genetic and phenotypic changes. Several single nucleotide polymorphisms (SNPs) have been associated with the risk of breast cancer development in women; however, little is known regarding their influence on canine mammary tumor risk. We assessed the influence of SNPs in genes related to human breast cancer susceptibility, with respect to the risk of development of mammary tumors in dogs. Sixty-seven canine SNPs in proto-oncogenes, tumor suppressor genes, genes involved in DNA repair, and in hormonal metabolism were evaluated in 212 bitches with mammary tumors and in 161 bitches free of mammary neoplasia. A significant association with mammary neoplasia risk was identified for 2 SNPs in RAD51 (rs23623251 and rs23642734) and one SNP in the STK11 gene (rs22928814). None of the other SNPs were related to the risk of mammary tumor development. The identification of genetic profiles associated with risk of mammary neoplasia is of great importance, supporting the implementation of specific clinical management strategies in high-risk animals.
Advances in molecular, genetic, and clinical research have markedly increased our knowledge of the basic biology of cancer. Cancer is clearly recognized as a complex disease involving a sequence of genetic, epigenetic, and phenotypic changes. The disease is characterized by aberrant gene structure or function that drive the progressive switch of normal cells into neoplastic ones.7,14 Exhaustive research in molecular genetics applied to oncology has led to the identification of genetic profiles associated with clinical and pathologic features of neoplasms, with the biological behavior of the tumors, and with their response to treatment.3,19 In humans, several single nucleotide polymorphisms (SNPs), mostly in proto-oncogenes, tumor suppressor genes, genes coding for proteins involved in DNA repair, and in sexual steroid hormone metabolism have been associated with an increased risk of cancer development, including breast cancer. 13
To date, only a few studies have reported an association between SNP and canine mammary tumor (CMT) risk. Some authors described significantly different allele frequencies for several SNPs when comparing high- and low-risk breeds for mammary tumors, underlining the importance of the genetic background in the susceptibility to mammary tumor development in dogs. 2 Other authors have recently associated SNPs in BRCA1 and BRCA2 genes with mammary neoplasm development in dogs.5,20 Furthermore, an association between SNP in the ESR1 gene and the risk for mammary tumors in English Springer Spaniels has also been described. 2
We assessed the influence of SNPs in genes known to be associated with human breast cancer susceptibility, in the risk of development of CMT. For that purpose, 67 canine SNPs in proto-oncogenes (HER-2, EGFR), tumor suppressor genes (TP53, STK11), and genes involved in DNA damage recognition and repair (BRCA1, BRCA2, RAD51, CHECK2, PTEN, BRIP1) and in hormonal metabolism (ESR1, PGR, PRLR, COMT) were selected (Table 1).
Genes and single nucleotide polymorphisms (SNPs) assessed for association to canine mammary tumor risk.
CFA = canine chromosome number; Dash (–) = allele undetected in the population.
SNP undetected.
A case-control study was conducted involving 373 dogs: 212 bitches with histologically confirmed mammary tumors (mean age: 10.0 y); and 161 bitches >7-y-old, without evidence of mammary neoplasia (mean age: 9.0 y). The breeds represented in both groups were very similar, with 45% and 46% of mixed-breed dogs included in case and control population, respectively. Furthermore, 32% and 22.4% of the bitches from the case and control group, respectively, belonged to a series of 5 more representative breeds: Poodle, Boxer, Labrador Retriever, Cocker Spaniel, and German Shepherd. All of the owners gave informed consent regarding the use of the material for research purposes. Our study was performed under an institutionally approved protocol (Approval P151/2016).
Genomic DNA was extracted from peripheral blood samples obtained by standard venipuncture (High Pure PCR template preparation kit, Roche, Mannheim, Germany). The DNA quality was evaluated by measuring the optical density, and the quantity was assessed (NanoDrop 1000 spectrophotometer, Thermo Fisher Scientific, Waltham, MA). SNP genotyping was performed (MassARRAY iPLEX Gold technology, Agena Bioscience, San Diego, CA) in the Genomics Unit of the genotyping service of the Gulbenkian Institute of Science (Unidade de Genómica/Serviço de Genotipagem do Instituto Gulbenkian de Ciência, Portugal).
Statistical analysis of data was performed (SPSS for Windows v.18, SPSS, Chicago, IL). Chi-square analysis was used to compare categorical variables. A 5% level was considered to define statistical significance. The observed number of each genotype was compared with that expected for a population in the Hardy–Weinberg equilibrium by using a goodness-of-fit chi-square test. Linkage disequilibrium (LD) analysis was performed for SNPs in genes in the same chromosome. Data were calculated and LD maps were constructed (Haploview v.4.2, https://www.broadinstitute.org/haploview/haploview).
A significant association with CMT risk was identified for 3 SNPs: rs23623251 and rs23642734 in RAD51, and rs22928814 in the STK11 gene (p = 0.004; p = 0.037; p = 0.008, respectively). For RAD51 SNP rs23623251, TT animals are more likely to develop mammary neoplasia than C allele carriers (odds ratio [OR] = 1.86; confidence interval [CI] = 1.22–2.83). In fact, TT animals correspond to 50.9% of all bitches with mammary tumors and only up to 35.4% of the animals from the control group. Furthermore, for RAD51 SNP rs23642734, the TT genotype represents 42.9% of animals with mammary tumors and 31.9% of the animals from the control population. Our data demonstrate that TT animals have a higher risk of developing mammary tumors than C allele carriers (OR = 1.58; CI = 1.03–2.43). LD analysis was performed for SNPs rs23623251 and rs23642734, and demonstrated that, in our population, these loci are in LD (r2 = 0.75; Fig. 1). For STK11 SNP rs22928814, T allele carriers have a higher probability of developing CMT than CC animals (OR = 1.82; CI = 1.17–2.84). Indeed, T carriers represent 40.3% of bitches with mammary neoplasia and only 26.7% of the animals from the control group. No significant association was found between the other SNPs included in our study and the risk of CMT development. The LD plots of all studied SNPs, including SNPs found to be associated with CMT in the different populations (control and disease groups), are presented as supporting information (Fig. 1). In 21 SNPs, LD could not be estimated because most of them were fixed SNPs (a homozygous SNP that has spread throughout an entire population) and some were not in agreement with Hardy–Weinberg equilibrium.
Linkage disequilibrium (LD) maps of genotyped single nucleotide polymorphisms (SNPs), including controls and canine mammary tumor cases. The dark gray-to-white gradient reflects higher to lower LD values. Dark gray diamonds without a number indicate complete LD (r2 = 1).
In a subsequent analysis, the association between RAD51 rs23623251 and rs23642734 and STK11 rs22928814 and the tumor type (benign or malignant) was assessed. No statistically significant association was found for any of the SNPs evaluated. However, for STK11 SNP rs22928814, CC carriers revealed a tendency to be associated with the development of benign tumors (p = 0.067). In fact, 40% of CC carriers developed benign neoplasms, whereas only 27.5% of the other genotypes developed this type of lesion. The association between RAD51 rs23623251 and rs23642734 and STK11 rs22928814 and the histologic classification of malignant tumors was also evaluated. For that purpose, malignant neoplasms were divided into 3 groups: A = complex carcinomas + carcinoma in benign tumors; B = simple carcinomas; C = carcinosarcomas + special type carcinomas + sarcomas. No significant relationship was observed between the animals’ genetic profile and the histologic subtype of the malignant mammary tumors for any of the SNPs assessed.
RAD51 is a protein involved in DNA repair pathways by homologous recombination. 1 Proper function of DNA repair enzymes is crucial for maintaining genome integrity. Failure of these proteins can lead to an exponential increase in DNA mutations, resulting in genomic instability and, eventually, in higher cancer susceptibility. 14 RAD51 also interacts with BRCA2, a well-recognized high-penetrance breast cancer susceptibility gene. 18 As in humans, changes in RAD51 messenger RNA and protein expression were described in CMTs, and an interaction between both BRCA2 and RAD51 canine genes has also been documented.10,11,15,16 STK11 is a serine/threonine protein kinase that plays several functions as a tumoral suppressor through control of cell growth and polarity, regulation of AMPK and other kinases, and cell cycle governance. 8 In humans, STK11 mutations are associated with predisposition to Peutz–Jeghers syndrome, a condition characterized by gastrointestinal polyposis, mucocutaneous hyperpigmentation, and increased cancer susceptibility, namely breast cancer. 12 There are also some analogies in the STK expression pattern between human breast cancer and feline mammary tumors. 4 Some genetic variations in RAD51 and STK11 genes were found in several types of human neoplasia, including breast cancer, hence they are both considered breast cancer susceptibility genes. 9 Comparative oncology is a rapidly growing and promising field of research. Merging scientific data from human and veterinary oncologic investigation clearly benefits knowledge of cancer development, progression, and management. 22 Canine and feline mammary tumors share several epidemiologic, clinical, and pathologic features with breast cancer, which makes them appropriate animal models for the translational study of this disease. 17 Our findings suggest that RAD51 and STK11 may play a similar role in mammary carcinogenesis in both humans and companion animals, thus constituting a valuable target for comparative oncology research.
Despite the fact that these particular SNPs in canine RAD51 (rs23623251 and rs23642734) and STK11 (rs22928814) genes are located in introns (non-coding DNA regions), they may have a functional significance in mammary tumor risk through several pathways, namely by coding for small regulatory elements that control gene expression or by affecting alternative splicing. Additionally, they may be linked to other functional SNPs from the same gene that indeed have impact on gene expression and/or function.14,21
Data from our investigation do not entirely match previous studies, which associated SNPs in ESR1, BRCA1, and BRCA2 genes with increased risk of CMT.2,5,20 This dissimilarity may be explained by differences in the canine population studied (such as sample size and composition), by the inclusion criteria, or by the laboratory methodologies employed.
Costs associated with genetic testing make inclusion of genetic tests difficult in routine veterinary oncology protocols. However, in an era characterized by growing concerns about animal health and welfare, it is reasonable to assume that an increasing number of pet owners will pursue genetic tests. This type of methodology allows for the identification of genetic profiles associated with a higher risk of mammary neoplasia, which may be reflected in long-term savings. The development of specific surveillance and clinical monitoring protocols in a group of high-risk animals could significantly reduce costs associated with the clinical management of the disease and also improve the animals’ quality of life. Furthermore, such development supports the implementation of preventive strategies aiming to reduce the exposure to other known risk factors (e.g., early spaying recommendation and avoidance of hormonal anti-contraceptive therapy).6,23
It is noteworthy that although most of the SNPs assessed in our study are not significantly associated with CMT risk, they may indeed influence the clinicopathologic features and biological behavior of the neoplastic disease, which are areas that deserve evaluation in future investigations.
Footnotes
Acknowledgements
We thank Dra. Bárbara Guerra Leal for her contribution to the statistical analysis of data. We thank the Liga Portuguesa Contra o Cancro—Núcleo Regional do Norte (LPCC-NRN; Portuguese League Against Cancer) for their support. We also thank Drs. Miguel França and Jorge Ribeiro from the UPVet Veterinary Hospital from ICBAS–Universidade do Porto, Dra. Cristina Pereira from Clínica Veterinária PetzAnimal, Dra. Raquel Vilaça from Clínica Veterinária de Custóias, Dra. Maria João Silva from Clínica Veterinária de Matosinhos, Dra. Flora Tinoco from Clínica Veterinária Dra. Flora Tinoco and Dr. Hugo Vilhena from Hospital Veterinário do Baixo Vouga, who contributed the biological material used for this work.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This work was partially supported by the institutional project CI-IPOP-22-2015 from the Research Center of the Portuguese Institute of Oncology of Porto, Portugal.