Detection of mutations within exons 4 to 8 of the p 53 tumor suppressor gene in canine mammary glands

Fifteen female canines with mammary tumors and 6 normal females were used to study mutations in exons 4 to 8 of the p53 gene. DNA samples from the tumors, respective adjacent normal mammary tissue and mammary glands from healthy animals were sequenced and analyzed for the presence of mutations. Mutations were found in 71.8% of the samples and the most frequent were missense mutations. The most attacked exons in the mammary tumor were 5, 7 and 8, with 23.4, 31.6 and 23.4% mutations, respectively. Canine mammary tumors are related to mutations in gene p53 and mutations mostly occur in the region of the protein that is linked to the DNA in the cell nucleus, which can change the functionality of the cell and propitiate tumor growth. Despite being macroscopically normal, the mammary tissue adjacent to the tumors has mutations that can lead to recurrence if not removed together with the tumor.

The p53 protein contains 393 amino acids and has been divided structurally and functionally into four domains.The sequence-specific DNA-binding domain of p53 is located between amino acid residues 102 and 292.This is a protease-resistant, independently folded domain containing a Zn 2+ ion that is required for its sequence-specific DNA-binding activity (Cavalcanti Junior et al., 2002).The p53 protein is a factor that enhances the transcription rate of a number of known genes that at least partially carry out p53-dependent functions in a cell (Levine, 1997).Cells lacking the p53 gene or containing mutant p53 are unable to arrest the cell cycle (Teifke and Lohr, 1996).The p53 protein plays a central role in the regulation of cell proliferation, genome stability and programmed cell death, repairing injured DNA and leading to a new transcription factor (Van Leeuwen et al., 1996;Walker e Rapley, 1999;Muto et al., 2000;Setoguchi et al., 2001).Activation of p53 often results in cell cycle arrest, presumably to allow for DNA repair before replication or mitosis.A primary mechanism by which p53 negatively controls the cell cycle is through the transcriptional activation of p21, which is a protein with the function of retaining the cell cycle at the R point, thereby allowing DNA repair.When mutations are numerous, p53 induces apoptosis and avoids the duplication of an injured cell (Albrechtsen et al., 1999).
Overexpression of the mutant p53 protein has also been observed in canine tumors of epithelial, mesenchymal and round cell origins (Sagartz et al., 1996, Wolf et al., 1997).The expression of p53 assessed by real-time polymerase chain reaction (PCR), however, may be heterogeneous in mammary tumors such as adenoma, adenosarcoma and their metastasis in lymphnodes (Klopfleisch and Gruber, 2009).
Mammary gland tumors are the most common neoplasms occurring in female dogs, representing 40-50% of malignant tumors (Davidson, 2003;Itoh et al. 2005;Queiroga et al., 2005).The histological tumor type is an important factor for the survival of affected animals (Sontas et al, 2009).Mammary carcinomas in dogs have similarities with breast cancer in humans, including the high prevalence of adenocarcinoma, frequency of metastasis and progressive disease (Sartin et al., 1992).Canine Tp53 is similar both in structure and function to human Tp53, therefore canine cancer may provide a useful clinical model in the search for effective anti-cancer therapies based on Tp53.
Although female canines commonly experience recurrences in other mammary tissues following the surgical excision of tumors, studies have been limited to evaluating the gene condition of the tumor tissue and few have addressed the remaining, apparently healthy tissue (Veldhoen et al., 1999).The possibility of early diagnosis from these tissues that may exhibit abnormalities expressed by genes even before the tumor develops can lead to more effective treatments prior to tumor development.
The present study was conducted with the purpose of investigating exons 4 through 8 of the Tp53 suppressor gene for frequency and mutation types in spontaneous canine mammary gland tumors and adjacent tissue without tumors, using the mammary glands of healthy female canines as the negative control.

MATERIAL AND METHODS
Fifteen canine females with malignant mammary tumors (aged between five and fifteen years) and six females without tumors submitted to elective ovary-hysterectomy (aged between 1.5 and six years, control group) of various breeds were studied.Representative sections of the mammary tissue (tumor, normal tissue of mammary gland adjacent to the tumor and mammary tissue from normal animals) were submitted to routine histopathological evaluation carried out by three pathologists to characterize tumor type alteration (or its absence) in affected and control mammary glands (Misdorp et al., 1999).
Genomic DNA from 0.5g of mammary tissue, frozen in liquid nitrogen, was extracted using the phenol:chloroform technique.Each sample was mixed with 100l TE (Tris 10 mM -Invitrogen Life Technologies and l mM EDTA, pH 8.0 -Sigma Chemical Co.) and 100l phenol (Merck), pH 8.0.After homogenized and centrifuged at 18.000g, for 5min at 4°C, 100l supernatant were collected and added in l00l of phenolchloroform (1:1) (Maniatis et al., 1998), and centrifuged at 18.000g for 5min.The supernatant was collected and 100l chloroform (Merck) was added, mixed for 1 min and centrifuged at 18.000g for 5min.In a new tube, 10l ammonium acetate 3M (Cromato Produtos Químicos LTDA) were added followed by 100l of the supernatant of tube one and 100l isopropanol (Merck).This mixture was homogenized for 1min and incubated for 30min in a freezer at 20°C -after that, it was centrifuged at 18.000g for 15min.The supernatant was then removed and the pellet washed with 500l 70% ethanol by centrifugation at 18.000g for 5min at 4ºC.Thereafter, ethanol was removed and the contents held at room temperature for 30min.At the end of this period, it was diluted with pure water.The extracted DNA was submitted to 1% electrophoresis agarose gel (Invitrogen Life Technologies) to determine its quality.
PCR oligonucleotides for amplification of the Tp53 fragments were designed on the basis of previously published sequencing data (Chu et al., 1998;Muto et al., 2000).Conventional PCR reactions were carried out for a final volume of 20μl, containing 50ng of genomic DNA, 20mM Tris-HCl, pH 8.4; 50mM KCl; 2mM MgCl 2 , 0.2mM dNTP mix, 10 pmoles of each primer (Table 2) and 2.5 U Taq DNA-polymerase (Invitrogen Life Technologies Ltda).Each PCR amplification consisted of an initial denaturing reaction performed at 94°C for 2 minutes, followed by 35 cycles of denaturing at 94°C for 1min, annealing at 58°C for 1 minute, polymerization 72°C for 1 min and a final elongation reaction at 72°C for 3 minutes, performed in a thermocycler (Eppendorf Personal Mastercycler).
The PCR products were purified using the PureLink PCR Purification Kit (Invitrogen, USA) and submitted to nucleotide sequence analysis using the MegaBACE 1000 automatic sequencer with the DYEnamic ET Dye Terminator Kit (Thermo Sequenase™ II DNA Polimerase).Analyses of nucleotide sequences from PCR products, which were obtained from the sequencing analyzer, were carried out using the GenBank search program (www.ncbi.nlm.gov/BLAST).

RESULTS AND DISCUSSION
Malignant mammary tumors ranged in size from 0.8 to 22cm at the largest diameter and exhibited various histological types (Table 1).The adjacent mammary gland tissue exhibited no histological abnormalities; the same was true for the animals without tumors.The amplified samples had a homology rate of 86.3% with the canine p53 protein (GenBank -AAB42022.1).For the analysis of the sequences, samples of each exon from the mammary gland or the control animals were aligned in order to obtain the most conserved segment of each sequence.These segments were the reference for detecting abnormalities (mutations) in the samples from the animals with tumors, which minimized the effect of polymorphisms.
Analyses were carried out on 96 nucleotide sequences from different tissues (tumor and adjacent mammary gland tissue) and exons.A total of 71.9% of these sequences had at least one mutation in some exon.The most frequent mutations in all exons were missense mutations in both groups (Table 3 and 4).Only 27 (28.1%)sequences had no mutation in the conserved region of each exon.The missense mutation type has been reported to have a high frequency, such as 20.0% in only three exons (5, 7 and 8) (Muto et al., 2000;Wong et al., 1999).In the present study, five exons were studied (4 to 8), allowing the observation of a greater number of mutations.The occurrence of multiple mutations in Tp53 has previously been observed in human (Glebov et al., 1994) and canine tumors (Veldhoen et al., 1999), with a high frequency of silent mutations (Strauss, 1997).In the present study, silent mutations accounted for 30% of the mutations in the tumor tissue.
Although the high number of mutations may be considered polymorphism, it is difficult to affirm this in the present study, since these regions were more conserved in healthy females (control), which were used for comparison.One way to demonstrate that the mutations in the tumors are not polymorphisms is to compare the tumor tissue to non-tumor tissue in the same animal, where there should be no mutations (Strauss, 1997).Curiously, the normal mammary tissue adjacent to the tumors exhibited missense mutations comparable in number to the tumors studied.Moreover, despite their lower frequency, nonsense mutations were found in the normal mammary tissue of the animals bearing malignant tumors.However, these mutations were not coincident in the same animal and the notion of polymorphism can be discarded.The presence of these mutations may represent molecular alterations in the normal mammary tissue and a high risk of these mammary glands developing tumors in the future.
Mutations in the normal tissue of animals bearing tumors could indicate a germ line origin of these Tp53 abnormalities (Veldhoen et al., 1999).Unfortunately, no other tissue from these animals was available for a comparison of the observed abnormalities.
Tp53 mutations are related to a poor prognosis and shorter overall survival rates in human breast cancer (Béroud and Soussi, 1998).Indeed, 53.8% of the animals in the present study experienced a recurrence of tumors within 1.5 years and death ensued in 71.4% of these cases.
In the mammary tumors, exons 5, 7 and 8 were the most altered, with 23.4,31.6 and 23.4% mutations, respectively (Tab.4).Although these exons exhibited a higher degree of mutations, the results of the present study reveal a relevant mutation rate in exon 4 (18.3%).In the adjacent mammary tissue, exons 4, 5 and 8 were the most frequently altered (30.6, 25.6 and 35.3%, respectively) (Tab.4).The high frequency of tumor samples that harbor Tp53 mutations, compared with the results of other authors, may be due to the size of the tumors, which were large and allowed collection of tumor material without contamination from the DNA of normal tissue (Chu et al., 1998;Muto et al., 2000).The mutations in the normal mammary glands could mean that alterations may be present early in tissues, before they are detected by cytological and histopathologic exams.
Among the more frequent mutations reported here were G (17.5%) and C (13.3%) deletions and A (15.4%) and T (14.7%) insertions.When the insertion or deletion was in the third base and did not change the amino acid, the mutation was considered silent (28.7%).If the mutation modified the reading of the codon sequence, it was considered a missense mutation.The deletion of one base can result in an altered sequence of amino acids and the production of a defected protein (Setoguchi et al., 2001).The observed rates of transversion (16.8%) and transition (7.7%) were lower than the 16.66% transitions observed by other authors (Lee and Kweon, 2002).A variable rate of mutations in Tp53 is found in mammary tumors.Some studies report no Tp53 mutations in human tumors (Kanaya et al., 2002).Others identify at least 15.0% mutations in primary mammary carcinomas (Chu et al., 1998).Other authors have also identified mutations in exons 5 and 8, which are considered mutation hot spots (Chu et al., 1998;Muto et al., 2000).On the other hand, exon 1 has not demonstrated any mutation, based on the evidence of hot spots (Mayr et al., 2000).Exons 4 to 8 used in the present study have also been studied in canine mammary tumors by other authors.Forty-nine mutations have been identified, with 75.5% located in the central domains of canine Tp53 (Setoguchi et al., 2001).
Although few mutations were repeated in the same codon, variation was found in the codons with alterations.It has been reported that mutations may be distributed in 200 or more p53 codons (Hainaut e Hollstein, 2000).This variation in the mutation position in the gene results in different expression patterns of the p53 protein.et al., 1999).These mutations are especially related to codons 248 and 273, which code arginine, an amino acid responsible for the protein binding to DNA.Codon 175 (arginine) has been related to the protein conformation for this binding (Levine, 1997).
The homology between the p53 gene in humans and canines was obtained from the alignment of GenBank sequences and was 86%, which is higher than previously reported rates of 81% (Chu et al., 1998) and 79.7% (Setoguchi et al., 2001).The Tp53 sequence is composed of 828 and 883 nucleotides in canines and humans, respectively.However, only 626 nucleotides have identity between the two sequences.This difference in the sequence of these species could be due to deletions in the proline-rich regions in the canine Tp53, which developed through evolution (Kraegel et al., 1995;Veldhoen and Milner, 1998).The conserved region of the nucleotide sequences allow comparative studies on human breast cancer using the canine model.However, it was not possible to identify a single nucleotide polymorphism that repeated in all tumors and could be considered a diagnostic factor, such as the example of codon 72 for breast cancer in humans (Lång et al., 2009).
This homology allowed the analysis of codons related to the binding of DNA to the protein in canine species, including the codons responsible for the structural conformation of the protein, which were all in the central domain of DNA binding.Although few samples exhibited mutations in these codons, the majority of the alterations were located near them.Insertions and deletions of a nucleotide, which predominate among missense mutations, result in an incorrect reading of the sequence and defected structure and function of the peptide (Walker and Rapley, 1999;Kreuzer and Massey, 2002).

CONCLUSION
The results of the present study indicate that abnormalities in the Tp53 gene are involved in the genesis of canine mammary tumors.Moreover, these abnormalities may be present early in normal tissue, as the mutations were detected in the macroscopically and histologically normal mammary tissue adjacent to the tumors.The complete surgical excision of the total mammary chain on the affected side could be justified in order to avoid the appearance of further tumors.Knowledge on the molecular events of tumor genesis in canine species will contribute toward a better understanding of cancer in these animals.

Table 2 .
Sequences of specific sense and antisense oligonucleotide primer pairs used for the amplification of exons 4 to 8 of the canine p53 gene using polymerase chain reaction (PCR)

Table 3 .
Mutation types observed in the amplified sequences of 5 exons from the canine p53 gene in individual mammary tumors and normal glands from female canines Anima l

Table 4 .
Mutation types observed in the amplified sequences of 5 exons from the canine p53 gene in mammary tissues (tumor and adjacent mammary tissue) from female canines