Forensics 2.0

Something wasn’t adding up. Earlier that summer, Tony Wheatley, then a detective at the Morgan County Sheriff’s Office, went to a home just outside of Versailles, Missouri, to investigate a reported suicide attempt. Police knocked on the door of retired archaeologist Gary Rex Walters, but no one answered. So they entered to ensure the safety of any occupants. Wheatley says that he and his fellow officers did not find any people inside, but they did spy marijuana pipes in plain view. The officers left the premises and applied for a search warrant. When they later reentered the home, they found four open wooden coffins full of human remains—bones, teeth, and skulls.

Wheatley called the Morgan Country Coroner, who indicated that the remains shouldn’t have been stored in Walters’s home. The detective confiscated the coffins and their grisly contents. “We went ahead and secured them for safe keeping until we could figure out what was going on,” Wheatley says. “I’m not trained in anthropology or anything like that, so I didn’t know how old they were or what they were.”

Walters argued that he had permission to own the skeletons contained in the coffins, saying that he had excavated the remains near Iztapa, Guatemala, sometime in the 1970s. He even produced decades-old documentation from the Guatemalan government to prove the legality of his cache.

But Wheatley says he “still wasn’t convinced that those documents covered those remains that we had seized. . . . We wanted to make sure that they weren’t newer bones that [Walters] had come across locally.” He added that there are several unmarked Native American burial sites in central Missouri and that there was a black market for items recovered from those sites. “We didn’t want him trying to sell Native American bones,” Wheatley says.

Wheatley’s concerns only grew as he looked into the archaeologist’s background. In the 1990s, Walters had been accused of stealing human remains that he was supposed to help relocate from a historic African American cemetery in St. Louis to make room for transportation infrastructure in the city. That incident, which revolved around his insistence that he was owed money for his work on the cemetery relocation project, was apparently resolved when he returned 28 bodies and received $90,000 for his work.

As an anthropologist, our job mostly in the past was focused on creating what we call a biological profile from the skeleton to help identify them. As technology is changing, our rules are honestly shifting.—Lindsay Trammell,
St. Louis Medical Examiner's Office

The discovery of the human skeletal remains in Walters’s home reanimated old suspicions. “[The investigators] were just trying to verify whether he was telling the truth or whether these were skeletal remains of modern individuals that had gone missing or been killed,” Trammell says. “So that was one of the reasons they called me—to see if we could discern, based on looking at the remains, were they from Guatemala, were they from this cemetery, or were they from somewhere different altogether?”

But Trammell’s preliminary work left her puzzling over what seemed like a mixed bag of remains. So she called for backup. Trammell sent samples to Cris Hughes of the University of Illinois at Urbana-Champaign and Chelsey Juarez of North Carolina State University (NCSU), who performed genome sequencing and isotope analysis, respectively, to provide more detailed information on the skeletons that would either corroborate or contradict the archaeologist’s story.

“Really, as an anthropologist, our job mostly in the past was focused on creating what we call a biological profile from the skeleton to help identify them,” Trammell says. “As technology is changing, our rules are honestly shifting, because now you do have DNA and now you do have isotopes, where these types of tests can tell you quite a bit from the skeleton.”

The archaeologist’s bizarre case is one of only a handful of examples where such techniques have been applied to criminal investigations. While searching for DNA matches to established databases or suspect samples is common practice in many crime labs, genome sequencing to establish a body’s likely ancestry and isotopic analyses to suggest its possible geographic origins remain exceptions to the rule of standard forensic workups. But as technologies mature—from single-cell sequencing to epigenetic analyses—investigators are beginning to rely more and more on advanced forensic methods, yielding unprecedented insights into victims, perpetrators, and their crimes.

“It’s taken us a while to get to this point, truthfully,” says Seth Faith, a researcher at NCSU’s Forensic Science Institute. “But now we have this wonderful technology in front of us and people who are trained in how to use it and interpret that data.”

The devil’s in the DNA

© PHILLIP PSAILA/SCIENCE SOURCEBecause they had collaborated on cases before, Trammell knew that Hughes, who also serves as a deputy forensic anthropologist at the Champaign County Coroner’s Office, was clued in on the latest in genome sequencing as applied to forensics. Trammell was also aware that Hughes trained under University of Illinois ancient DNA researcher Ripan Malhi, who has used advanced sequencing techniques to recover and decode DNA from 6,000-year-old human remains. “[Hughes is] definitely the expert when it comes to DNA,” Trammell says. “I knew that she could test for mitochondrial haplogroups”—genetic signatures that are associated with geographic regions—“and [determine] whether or not they were consistent with an individual that was African or Native American.”

Trammell sent small bone and teeth samples from 10 of the crania found in the archaeologist’s wooden coffins. Of those, Hughes took subsamples from six and sequenced hypervariable region 1 (HVR1) of the mitochondrial DNA, which can help determine an individual’s haplogroup. “To jibe with [the archaeologist’s] story, I would have to see something that was a Native American haplogroup,” Hughes says. Of the four samples that yielded sequenceable DNA, three did have HVR1s consistent with Native American ancestry. However, one sample’s HVR1 placed the mitochondrial genome within West African haplogroups, suggesting that not all of the remains in the wooden coffins were from Guatemala after all. “Some of those materials in there were basically raising a red flag and were not consistent with his story,” Hughes says.

Hughes knew that Iztapa was abandoned by natives in 1350 CE, before Europeans arrived in the area, meaning that “if we’re looking at the DNA, and we’re looking at maternal ancestry, you should expect to only see Native American links,” she says. “We shouldn’t see anything that’s European or African, because there was no contact. They hadn’t come over yet.”

But Hughes knew of another analysis that could provide further evidence of the skeletons’ origins, so she suggested Trammell contact Juarez to analyze the relative abundances and ratios of isotopic forms of common elements, which can yield clues about where the deceased individual lived from birth to death.

Isotopes tell most

Juarez also received bone and tooth samples from the same 10 crania that Trammell had sampled for Hughes. Using mass spectrometry and focusing on the crown of the first permanent molar—which starts forming in utero and finishes developing by age 3—she performed analyses that quantified the amounts of isotopic strontium in the skeletal materials. She then compared those levels with maps of known isotope distributions in both coastal Guatemala and the St. Louis region. “We did not find overlap in the Latin American region where [the archaeologist] was claiming that the bones were from,” Juarez says. “We did find overlap in the St. Louis area.”

ISOMAP: This map shows estimates of oxygen isotope ratios  (essentially 18O:16O) expected in the hair of people living throughout the lower 48 United States. The values are derived from observed isotope ratios in tap water and hair samples. Researchers can create similar maps for isotopes of a handful of other elements, such as hydrogen and strontium, using values from human tissues, soil, or rainwater. Such maps are being used by criminal investigators to reconstruct the geographic origin or travel history of victims or suspects.ANNU REV EARTH PLANET SCI, 38:161-87, 2010

Juarez’s results provided further support that the bones in the archaeologist’s skeletal stockpile were not all from ancient Guatemala, as he had contended. “Based on the results that we were getting, my first thought was that these remains were coming from multiple different archaeological sites from various regions,” says Trammell. “To me it seemed like this was something that he was possibly collecting throughout his career. But without testing every single element, there’s no way to prove that.”

Hughes and Juarez submitted their data and reports on their analyses to Trammell, who passed those findings onto investigators in Morgan County.

The reach of forensics

One of the best and broadest examples of applying next-gen tools and methodologies to forensic sequencing can be seen in a project at the Armed Forces DNA Identification Laboratory (AFDIL). Researchers there are using next-gen sequencing to sift through the genomes of fallen service members whose remains were recovered on foreign battlefields from World War II, the Korean War, and Vietnam.

Charla Marshall, chief of the emerging technologies section at AFDIL, headed up the validation of a sequencing protocol that investigators are now using to repatriate the remains of more than 800 Korean War veterans. The skeletal remains were buried in the National Memorial Cemetery of the Pacific in Hawaii (colloquially referred to as the “Punchbowl”) in 1953 after being disinterred from their original burial grounds near Korean battlefields, shipped to Japan, preserved in formalin, and shipped to Hawaii. In the late 1990s, the AFDIL retrieved the bodies from the Punchbowl to begin identifying the soldiers. But the bodies had been fixed in formalin, which induces DNA crosslinks that disrupt sequencing reads; so the researchers were unable to extract DNA fragments long enough to sequence with the Sanger method.

We are using the very sensitive methods of next-gen sequencing to recover 55- to 75-base-pair fragments, which are not amenable to Sanger sequencing strategies that require 20-base-pair primers on either end.—Charla Marshall,
Armed Forces DNA Identification Laboratory

“Right now we are using the very sensitive methods of [next-gen sequencing] to recover 55- to 75-base-pair fragments, which are not amenable to Sanger sequencing strategies that require 20-base-pair primers on either end,” Marshall says.

In July of this year, the Defense POW/MIA Accounting Agency (DPAA) announced that researchers at AFDIL had identified US Army Corporal Charles White, who died fighting in North Korea in 1951 at the age of 20. Using Marshall’s protocol to sequence Cpl. White’s mitochondrial genome, the team matched it to mtDNA provided by a niece, a nephew, and a sister. Last summer, his remains were returned to his family in Lexington, Ohio, and he was buried with full military honors. And Cpl. White is just one of many. “Between March of 2016 and September 30 of 2016, we have processed 80 samples through the [sequencing] procedure,” says Tim McMahon, chief of AFDIL. “And we’re running at about a 45 percent success rate,” in terms of pulling useable sequences from the chemically degraded DNA.

Isotope analyses, which can narrow a person’s travel and life history down to a set of geographic locations, are also becoming more common in forensics labs. In 2015, scientists working on the murder of 2-year-old Bella Bond, whose virtually unidentifiable body was found in a trash bag near Boston Harbor, used isotopes to help determine where she may have lived during her tragically short life. By comparing oxygen isotopes in her hair and teeth to known values of oxygen isotopes in drinking water throughout the country, investigators suspected that she had spent most of her life in the New England area, information that helped police narrow the search for her murderers and arrest the girl’s mother and her boyfriend for the crime in September 2015.

Lesley Chesson, president of Salt Lake City–based IsoForensics, says that her company gets requests from investigators all over the country for isotope analyses like the one used in the Bella Bond case. “This application for modern casework has really been taking off in maybe the last 10 or 15 years,” she says. “We work with folks from all over the United States—police departments, state bureaus of investigation, and sheriff’s departments.”

IsoForensics was spun out of the University of Utah, where geochemist Gabe Bowen conducts research on the cutting edge of isotope analyses with an approach called isomapping or isoscapes. Isomaps are predictive maps that display the distribution of isotopes of a particular element that can be compared with isotope ratios in bones, teeth, hair, or other tissues to estimate the geographic origins, travel histories, diets, or date of death of a person.

“Isoscapes are trying to take our first-principles—understanding of the physical, biological, and chemical systems that control isotope variation—and turn those into predictive maps that provide a fingerprint for interpreting forensic data,” Bowen says. “It would be like if you wanted to know where a letter came from in the mail. It’s got a ZIP code on it. You need a map of the ZIP codes across the U.S. to interpret that to figure out where it came from.”

But building isomaps is a slow, ongoing process. “These days, we’ve got massive [DNA] databases against which you can compare a sample,” Bowen says. “We’re not at that point yet [with isomaps]. And, unfortunately, there hasn’t really been a concerted, community-wide effort to build the necessary databases.”

A lack of funding, Bowen adds, makes this a difficult proposition. “So we’re left in a situation where most applications, when they come up, involve making some measurements and then scraping to fund samples for comparison or pull together data from published sources that can provide comparators,” he says. “So that means there’s not a plug-and-play application in most cases. Each application still, at some level, involves a little bit of research. And when you start getting into operational forensic work, that’s not very attractive.”

Outside the chalk outline

© ISTOCK.COM/SHERALEESIn addition to their use in missing-person cases, next-gen sequencing approaches may extend beyond the identification of complete or partial human genomes. In 2015, Faith’s colleagues at NCSU published a paper in PLOS ONE that reported the sequencing of fungal DNA from nearly 1,000 dust samples collected from indoor environments across the U.S. (10:e0122605). The researchers identified as many as 40,000 fungal taxa in these samples, and developed an algorithm that can place a given sample of dust within a couple hundred kilometers of its geographic origin based on the signature of fungal species it contains.

“Being able to predict within 200 kilometers where something’s come from in the United States is pretty impressive,” Faith says. “There’s no other technology to date that can get that type of precision.” High global diversity and their resistance to desiccation make fungal taxa capable geographic identifiers. Faith adds that his group has launched a larger project in concert with the US Department of Defense to create a world map of dust fungal taxa. This, Faith says, may help investigators predict the origin and travel history of samples involved in smuggling cases, port investigations, or hazardous material trafficking, among other scenarios.

Faith has also illustrated the utility of next-gen sequencing in the context of forensic epigenetics to identify individuals (Electrophoresis, 35:3096-101, 2014). “Here in the lab, we’ve done some work to look at tissue sores to establish where the DNAs come from based on the epigenetic attributes, like methylation patterns,” he says. “We’re also trying to develop an approach to differentiate identical twins based on epigenetic patterns.”

Isotope analyses are also being developed for use in food safety, wild­life forensics, poaching investigations, and in African ivory smuggling cases.

And the same mass spectrometry technology that fuels isomapping and other isotope analyses is also applied to characterizing the composition of other molecules hidden away in crime scene samples. “We’re seeing folks using mass spec to identify the proteins in hairs, for example, and trying to make those attributable to individuals,” Faith explains.

Bowen says that isotope analyses are also being developed for use in food safety, wildlife forensics, poaching investigations, and in African ivory smuggling cases. “There’s a lot of different directions that people are going with it.”

The mystery lingers

Back at the St. Louis Medical Examiner’s Office, Lindsay Trammell is still housing piles of bones recovered from the archeologist’s home in central Missouri. No charges were filed in the case, and the wooden coffins, along with other, nonhuman artifacts contained in them, were returned to the man. According to an official with the Morgan County Sheriff’s Office, however, “there are still some people in St. Louis that are trying to trace [the skeletal remains] back to some victims.” For that reason, the official says, “we’re not at liberty to discuss it yet.”

While Trammell’s traditional biological workup and the extra analyses performed by Hughes and Juarez made a compelling case that the archaeologist wasn’t completely truthful about the bones found in his house, the data generated by these forensic techniques is not yet admissible as evidence in court. (See “Challenges” below.)

“It’s a really interesting case, with three anthropologists in the youngest field of anthropology with different specialties,” Hughes says. “Then again, it’s kind of sad. No charges were filed. This is the hard part about the work. When there’s no closure or there’s no proper resolution for those individuals whose family members are kept in this guy’s house.”