以商用犬隻基因型鑑定套組進行台灣之比特類犬STR基因庫建立及基因多型性分析

因比特類犬隻攻擊並傷害犬隻及人類事件不斷增加，根據動物保護法第八條規定，行政院農業委員會於2021年10月26日禁止飼養、輸入或輸出美國比特鬥牛犬和美國史大佛夏牛頭犬。為幫助未來立法的實施，必須制定一個有效的品種鑑定標準。比特類犬包含了數個品種及該品種之混種犬。目測品種鑑定已被證實不準確，所以我們應考慮使用更客觀的方法，例如分子標記。研究顯示，使用核内DNA之單一核苷酸多型性（SNP）和短縱列重複序列（STR），可成功將動物品種分類並鑑別。品種鑑定之建立需要參考族群，但我們對於台灣的比特類犬族群所知甚少。因此，本研究目的有三：一、透過外觀比例，評估較爲客觀的目測品種鑑定；二、探討台灣比特類犬族群結構和其相關性；三、建構台灣比特類犬的STR基因庫，作為未來發展比特類犬隻品種分子鑑定技術之基礎。此研究對象共78隻犬隻（55隻比特類犬隻、16隻非比特類犬隻及7隻未確認品種之犬隻），材料爲犬隻唾液及口腔黏膜拭子。除提及之外，研究中的犬隻品種皆由Embark Veterinary Inc.先通過SNP品種鑑定服務，確認目標犬隻之品種後，再進行分組。首先，本研究根據1991年英國危險犬法（DDA）公佈的指引，透過50隻犬隻（9隻福爾摩沙高山犬FMD、25隻混種比特類犬MPT及6隻美國比特犬APT）進行目測品種鑑定評估。結果顯示APT組所獲得的平均分數顯著高於其他組，而爲了提高準確性，所有外觀測量比例標準一起作評估。我們也利用了商用犬隻基因型鑑定套組Canine Genotype Panels 1.1及2.1經篩選後所得之25個基因座，對42隻犬隻（11隻FMD、14隻MPT、17隻APT）進行基因多型性分析、族群結構性及遺傳分化差異性分析。APT組的基因多型性低於MPT組及FMD組，但相較於其餘文獻之結果之德國狼犬和黃金獵犬，具有相似或略高的基因多型性，而相較柯基犬具更高的基因多型性，與其他研究中的當地犬種相比，基因多型性則較低。各組別之間沒有顯著的遺傳分化差異，而相較於MPT和APT之間，FMD與MPT及FMD與APT之間的遺傳分化差異較大，MPT和APT之間的遺傳分化差異則小至可忽略。族群結構性分析則可見APT及MPT被分配到同一個集群中，而FMD 和其中一隻具有有趣品種成分的MPT 個體被分配到另一個集群中。PCoA分析則顯示MPT 和 APT 可以被分為兩個獨立但緊密且重疊的集群。與族群結構性分析結果相似，Neighbour-Joining (NJ) 算法可以將族群形成兩個獨立的群體，即比特類犬和非比特類犬。 NJ 方法和個體基因分配測試均取得了合理的分配成功率（56.2%-75.0%），而等位基因頻率可達到較高且較有效的品種分配。這兩個商用犬隻基因型鑑定套組亦顯示接近 0 的累積Probability of identity, P(ID) 值、累積 P(ID)sibs 值及高的Exclusion probability (EP)，有效於個體鑑定及親緣鑑定，因此可以應用於獸醫法醫中。總結來説，這項研究證明了分子標記對於準確的品種鑑定仍然更加可靠，而透過商用犬隻基因型鑑定套組，我們可建立台灣一小部分比特類犬遺傳多型性基因庫，同時建立一個基礎的 DNA 的品種鑑定系統。

關鍵字

比特類犬； STR基因庫；遺傳多型性分析；動物法醫學

並列摘要

Due to the rising cases regarding the aggression of pitbull-type dogs in Taiwan, the Council of Agriculture, Executive Yuan, has banned American Pit Bull Terrier and American Staffordshire Terrier from being owned, imported or exported under Animal Protection Law on the 26th of October 2021. Standards of breed identification must be formulated to aid the implementation of the legislation in the future. Pitbull-type dogs encompass purebred dogs of various breeds and dogs presumed to be mixes of those breeds. Visual breed identification among breeds is often shown to be inaccurate. Thus, a more objective method, such as molecular markers, should be considered. Single Nucleotide Polymorphism (SNP) and Short Tandem Repeat (STR) of nuclear DNA have been described in animal breed-related studies and have revealed promising results. To identify a breed, a reference population is required, and there is little knowledge of the pitbull-type dog population in Taiwan. Therefore, this study aims to assess visual breed identification via body measurement ratio, determine the population structure and relatedness, and construct STR profiles of pitbull-type dogs in Taiwan for future molecular technique establishment for canine breed identification. A total of 78 dogs (55 pitbull-type dogs, 16 non-pitbull-type dogs and 7 dogs with breeds unconfirmed) were included in the study, and saliva swabs were taken from each dog. The breeds of dogs in the current study are confirmed via SNP breed identification by Embark Veterinary Inc., except as mentioned. Firstly, visual breed identification was assessed through 50 dogs (9 Formosan Mountain Dogs, FMD, 25 mixed pitbull-type dogs, MPT and 16 American pitbull terrier dogs, APT) using a more objective body measurement ratios based on guidelines of The Dangerous Dogs Act (DDA) 1991 (UK). Results showed that the APT group obtained a significantly higher average score compared to the other groups and all body measurement ratio criteria should be evaluated together to increase accuracy. We also utilized commercial Canine Genotype Panels 1.1 and 2.1, which after selection, amount to 25 loci, to undergo a genetic polymorphism analysis, population substructure, genetic differentiation analysis on 42 dogs (11 FMD, 14 MPT, and 17 APT). APT was genetically less diverse than the MPT and the FMD group, while having a similar or slightly higher genetic diversity compared to German Shepherd and Golden retrievers, a higher genetic diversity than Pembroke Welsh Corgi, and less variable gene pool compared to the local dog breeds in other studies. There was no significant genetic differentiation among the groups, while there were higher genetic differences between FMD and MPT as well as FMD and APT compared to the negligible differences between MPT and APT. APT and MPT were assigned to a single common cluster, while FMD and a single MPT individual with an interesting breed profile were assigned into another in population structure. PCoA analysis showed that MPT and APT could be grouped into separate yet close and overlapping clusters. Similar to the population substructure results, two separate clans could be formed by the NJ algorithm, which is pitbull-type and non pitbull-type group. Both the Neighbour-Joining (NJ) method and individual assignment tests achieved a fair assignment success (56.2%-75.0%), with the allele frequency method having the highest assignment success and efficacy. The two kits also showed close to 0 cumulative Probability of identity, P(ID) value, cumulative P(ID)sibs value and high exclusion probability which is exemplary for individual identification and parentage analysis, thus will be useful in veterinary forensics application. In conclusion, this study suggested that molecular markers are still more reliable for an accurate breed identification and an informative genetic diversity profile of a small part of the pitbull-type dog population in Taiwan as well as DNA-based breed identification can be established and achieved using the commercial Canine Genotype kits.