Argentine veterinarian Guillermo Giovambattista remembers the day in 2015 when he received an unusual call in his laboratory: there had been a murder in his city, Buenos Aires. On the other end of the phone was a member of the Prosecutor's Office. The main suspect in the crime had been arrested and his shoe hid a possible clue: dog poop crushed on the sole. Giovambattista, director of the Veterinary Genetics Institute, had collaborated with authorities on hundreds of cattle theft cases, but this assignment was different. The resolution of a homicide was in his hands.
"The feces that the suspect had stepped on could be from the dead man's pet," recalls the veterinarian, from the National University of La Plata. His team took samples from the deceased's dog and compared the DNA with that of the excrement from the sole of the alleged killer. The results showed that they were 20 times more likely to belong to the same dog than to two different ones. It was not a definitive clue, but it was enough to add to others and corner the accused.
Veterinary forensic genetics has taken a leap since then, explains Christopher Phillips of the University of Santiago de Compostela. The team of this geneticist was key in solving the case of Eva Blanco, a 17-year-old girl who was raped and murdered on April 20, 1997 when she was returning to her house in the Madrid town of Algete. The DNA extracted from the semen did not match that of any suspect and the crime remained unpunished until 2015, when Phillips' group re-analyzed it and ruled that the anonymous genetic material was most likely from a North African man. The Civil Guard agents did the rest. They investigated all the North Africans who lived in Algete in 1997 and ended up arresting Ahmed Chelh, a Hispanic-Moroccan who committed suicide in January 2016 in the Madrid prison of Alcalá Meco.
Geneticists want to achieve with dogs what they have already begun to achieve with humans. "Now we can predict a dog's breed and color fairly accurately from a DNA sample," says Phillips, who collaborates with CaDNAP, a group of European geneticists dedicated to the analysis of canine genetic material. In the Buenos Aires murder it was easy to compare two DNA samples to see if they belonged to the same animal. The difficult thing is to make a robot portrait of an unknown dog from a piece of saliva, hair or excrement found at a crime scene.
Each cell of a dog contains in its nucleus a code of 3,000 million letters of DNA, with the instructions that allow it to bark, wag its tail or run for a stick. Scientists on the CaDNAP project look for small variations of a single letter—called single nucleotide polymorphisms—that are associated with the appearance of a dog. They have identified 43 of these genetic markers, which are used to predict, with a probability of 80%, the color of a dog or the shape of its skull, according to preliminary results presented at the last congress of the International Society for Forensic Genetics, organized in September in Prague (Czech Republic).
"The dog, as man's best friend, is going to be increasingly relevant in forensic investigations and will help identify the guilty and, therefore, reduce the number of crimes", says biologist Josephin Heinrich, researcher at the Medical University of Innsbruck (Austria) and the CaDNAP project. The group is now working to refine the method and also be able to predict body size, eye color, tail morphology and ear shape.
At the Prague conference, Heinrich discussed the effects that a single change in a dog's DNA letter can have. The biologist showed the photographs of four Dalmatians. The first, with the classic look. The second, with yellowish spots instead of black, due to a change in the MC1R gene, associated with red hair in humans. The third animal had brown spots due to a change in the TYRP1 gene. And the fourth Dalmatian was a long-haired variety, due to a mutation in its FGF5 gene. The appearance of the four dogs could easily be predicted from their droppings.
A year ago, the CaDNAP group tried to find out the breed of 392 dogs from their genetic material, in an experiment published in the specialized journal Forensic Science International: Genetics. Because most DNA is identical from one animal to another, and because it's very expensive to analyze everything, the scientists focused on 13 short, highly variable regions, called "short tandem repeats." Those 13 specific places in the genome were enough to hit the race in more than 97% of the cases.
"At first, the issue of domestic animals was of great interest due to the case of the Snowball cat in the US," recalls the seasoned geneticist Antonio Alonso, director of the National Institute of Toxicology and Forensic Sciences, in the Madrid town of Las Rozas. The specialist refers to the murder of Shirley Duguay, a 32-year-old woman, on October 3, 1994, on Prince Edward Island, in Canada. The main clue to the crime was a man's leather jacket, with Duguay's blood and eight white hairs attached.
A police inspector, Roger Savoie, tried to prove that those hairs belonged to the cat of the main suspect, Douglas Beamish, ex-partner of the murdered woman. It was the first time that animal DNA was used in a trial. The analyzes ruled that those eight hairs were indeed from Snowball, Beamish's cat. The defendant was sentenced to 18 years in prison. "But the reality is that this is not a highly required analysis in forensic laboratories," Alonso acknowledges.
