Page 33 - Withrow and MacEwen's Small Animal Clinical Oncology, 6th Edition
P. 33

12    PART I    The Biology and Pathogenesis of Cancer


         human tumors. The prototypical example was a structural aber-  from the GWAS suggested a role for CDKN2 and the associated
         ration resulting from a balanced chromosomal translocation that   cycle regulatory processes in OSA risk; the significance of these
                                                               pathways was independently confirmed as a key indicator of the
         creates a fusion gene composed of most of the BCR gene (located
  VetBooks.ir  on chromosome 22 in humans and on chromosome 26 in dogs)   prognosis in canine OSA using gene expression studies. 41,137,138  In
                                                               the case of HSA and B-cell lymphomas, two histologically differ-
         and a truncated form of the ABL gene (coincidentally located on
         chromosome 9 in both humans and dogs) in CML. 123  Both trans-  ent tumors that occur commonly in golden retrievers, the GWAS
         locations give rise to a derivative chromosome, the Philadelphia   identified two shared risk loci on chromosome 5. 133  This observa-
         (Ph) chromosome in humans and the “Raleigh” chromosome in   tion suggests that HSAs and B-cell lymphomas might originate
         dogs. Certain numeric aberrations (changes in DNA copy num-  from a common lineage of hematopoietic progenitor cells. 119,139
         ber) are similarly conserved in both species in a variety of can-  Such meticulous work has improved our understanding of how
         cers, including lymphoma, soft tissue sarcomas, OSA, and brain   distinct heritable traits segregate with cancer phenotypes in dogs,
         tumors. 124–129                                       although it remains to be seen if these traits will be shared between
            The development of specific tumors from cells harboring   closely related breeds or whether they contribute to risk indepen-
         such shared mutations is not surprising, but why would homol-  dently among different breeds. 127
         ogous,  highly  conserved  pathologic  rearrangements,  deletions,   Next-generation sequencing of tumor and normal exomes
         or amplifications occur in cells from distinct organisms? One   has been completed for canine lymphoma (Fig. 1.8), HSA, and
         possibility is selection for phenotype; in other words, the genetic   OSA. 41,59,140–142  In the case of B-cell lymphomas, 64 exomes were
         change “freezes” development at a particular stage for that cell   sequenced from golden retrievers and cocker spaniels. The tumors
         lineage and enhances growth and survival compared with nor-  had an average of 500 somatic mutations each, of which about 20
         mal cells in the niche. But it is also possible that these mutations   were nonsilent coding mutations. The most common recurrently
         are evolutionarily related on a mechanistic basis. For example,   mutated genes were TRAF3, FBXW7, and POT1, which regulate
         rearrangements of the immunoglobulin heavy chain locus and   signaling pathways of telomerase and autoimmunity.  FBXW7
         the MYC locus are thought to be due to recognition of MYC   mutations, in particular, were found in 15 of 54 (28%) golden
         flanking sequences by the recombinase enzyme system. 123  No   retriever samples and in 1 in 10 (10%) cocker spaniel samples. The
         such mechanism is known to be operative for other defined sites,   two most common, nonsilent mutations in FBXW7 were mutu-
         so these other mutational events could occur stochastically, with   ally exclusive and created arginine to leucine amino acid substitu-
         their recurrent characterization across multiple species being the   tion at positions 470 and 484, respectively. These residues are also
         result of the selective advantage provided by the acquired gene to   hotspots of FBXW7 mutation in human cancers.
         a cell of a highly specific lineage under highly specific conditions.   The exomes in T-cell lymphomas were more heterogeneous;
         Another possibility is that they are related to the nuclear anat-  still, 7 of 16 peripheral T-cell lymphomas from boxers harbored
         omy of the cell and specifically caused by proximity of chromo-  mutations in the  PTEN-mTOR pathways. The data, however,
         somal regions, cellular stress, inappropriate DNA repair (or, as   is not sufficient to distinguish whether this is due to a genetic
         mentioned previously, recombination), and DNA sequence and   propensity for mutations of PTEN and other functionally related
         chromatin features, such as reuse breakpoints. 130  A third intrigu-  genes in boxers, or to an association in which such mutations
         ing possibility is that cellular genomes are reverting to a confor-  favor selection for ontogenetically related tumors. Because genes
         mation that was found in a common ancestor (thus the high   in the PTEN-mTOR axis are also common targets of mutation in
         affinity and specificity between the rearranged chromosomal seg-  human cancers, studies of canine T-cell lymphoma might help to
         ments lead to the same recurrent event in many patients) but lost   unravel their contribution to the origin and progression of cancer
         during the process of chromosomal reorganization in evolution,   in a vulnerable breed and provide a discovery platform to test the
         or that these sites represent targets for gene deletions or duplica-  safety and efficacy of drugs that target these pathways.
         tions that have been repeatedly advantageous to species under   In the case of HSA, independent reports by two groups doc-
         conditions of natural selection and so have become embedded in   umented recurrent somatic mutations of TP53 and PIK3CA in
         their contemporary descendants.                       more than 50% of tumors within a background of many “private”
            In dogs and other domestic animals, the coexistence of genetic   mutations (i.e., unique to individual tumors). And in the case of
         isolates in closed populations we call “breeds,” along with animals   OSA, two independent reports identified somatic mutations of
         of mixed breeding, lends itself to the study of how a relatively   TP53 as the most common genetic abnormality in this tumor.
         homogeneous background influences cancer in outbred popula-  Exome  sequencing  also  confirmed  that  chaotic  OSA  and  HSA
         tions. Dogs were the first species in which genetic background   genomes had extensive, albeit minimally overlapping, copy num-
         was shown to mold tumor genomes and tumor gene expression   ber aberrations, which were far in excess of those found in lym-
         profiles, highlighting the utility of comparative approaches to   phomas 124,125,128 ; predictably, translocations resulting in de novo
         understanding cancer genetics. 55,128,131  Over the past decade,   fusion genes also were more frequent in HSA and OSA than they
         technologic advances in next-generation sequencing and bioinfor-  were in lymphoma. 42
         matics have improved our understanding of these relationships;   Most of the coding mutations identified by exome sequencing
         GWASs have identified disease-specific risk alleles in histiocytic   were confirmed in RNAseq data sets from the same cases, allow-
         sarcoma, 132  squamous cell carcinoma of the digit,  OSA,  canine   ing us to extend the observations to additional dogs and other
                                               28
                                                     14
                        27
         mammary tumors,  B-cell lymphomas and HSA, 133,134  mast cell   breeds and then to conclude that common somatic mutations in
         tumors, 135  and brain tumors 136  in susceptible dog breeds. In the   lymphoma, HSA, and OSA are not associated with breed, age, sex,
         case of OSA, the GWAS revealed 33 unique risk loci in three   or hormonal status (intact or neutered).
         breeds, and although none obviously overlapped, the existence of   Genomic studies, including GWAS, comparative genomic
         shared risk alleles might have been masked because they reside   hybridization, and next-generation sequencing, are expected to
         within fixed regions in some breeds (Fig. 1.7).  Convergent data   proceed in cats with the advent of the feline genome sequence. 143,144  
                                             14
   28   29   30   31   32   33   34   35   36   37   38