What about Leoma? Could a new DNA test indicate, beyond a reasonable doubt—in fact, with virtual certainty—whether the remains I had exhumed, not once but twice, were Leoma or some other murder victim? No one was on trial for murder in this case—Leoma’s great-nephew had long ago confessed to killing her, and served six years for manslaughter (though in 2006 he recanted). In this case it wasn’t a jury I was testifying before, but a family, one whose doubts I hoped to lay to rest forever.
In the O.J. Simpson case, the LAPD had been able to gather a fresh sample, within hours after the victims’ death. In the Leoma Patterson case, unfortunately, the samples were more than twenty-five years old, and that was worrisome. DNA is immense, but it’s also relatively fragile. Over time, it can break down or be destroyed by bacteria, acids, or other chemicals in the body or in the environment. Would enough DNA have survived with the tough, enameled capsules of the two molars nestled once more in my pocket at 34,000 feet as I streaked toward San Antonio? Please, I prayed silently.
SAN ANTONIO’S RIVERWALK, packed cheek-by-jowl with restaurants, bars, hotels, and tourists, is like a miniaturized, Disneyfied version of the canals of Venice or Amsterdam. Sprawling alongside the Riverwalk, the Henry B. Gonzales Convention Center encompasses 1.3 million square feet: roughly thirty acres, which is bigger than many family farms in East Tennessee. Toward the end of February 2007, those 1.3 million square feet were swarming with thousands of forensic scientists, converging on San Antonio for the annual convention of the American Academy of Forensic Sciences. The AAFS boasts more than six thousand members, in fields ranging from A (anthropology) to Z (zoarchaeology: “Doc, are those bones animal or human?”). The convention center’s cavernous exhibit hall—bigger than the football field in Neyland Stadium—was jammed with hundreds of vendors’ booths hawking a staggering array of goods and services: DNA analysis. Fingerprint scanners. Toxicology screenings. Disposable biohazard suits. Rape-test kits.
At Booth 117—sandwiched between two software companies—I met Jason Eshleman, a slight, soft-spoken, but swift-talking scientist. Jason is one of an emerging crop of twenty-first-century anthropologists inhabiting a niche I could not have imagined a half century ago: molecular anthropology.
Jason earned a Ph.D. at the University of California at Davis, under the tutelage of David Glenn Smith, who helped pioneer the application of modern DNA testing to ancient human skeletons. Smith enticed Jason to UC-Davis for his graduate work by posing an intriguing anthropological question: Were the dominant Native American tribes of the American Southwest—the Apaches and Comanches—the region’s first human inhabitants, with a claim on the land dating back some ten thousand years? Or were the Apaches and Comanches interlopers, invaders who had swooped down from Canada only recently—“only” being about four thousand years ago? Smith told Jason that if he followed the trail of DNA evidence, he could answer that question. What he didn’t tell Jason was that extracting and purifying ancient DNA was a nightmarishly difficult process. “I wouldn’t wish it on my worst enemy,” Jason laughed as he told me the story of his grueling research.
Trace Genetics, the DNA lab Jason co-founded after earning his Ph.D., had recently been acquired by DNA Print, a much larger firm, largely because of Jason’s expertise in detecting and analyzing ancient DNA. One of his most notable successes had been finding DNA in the teeth of a ten-thousand-year-old skeleton from the Queen Charlotte Islands, located off the coast of British Columbia (an archipelago that bills itself as the “Galapagos of Canada”). Encouraged by this evidence of his skill, I handed over the envelope containing two molars I had pulled from the jaws of Maybe-Leoma.
Some scientists thrive on the conceptual; their minds can envision particles that the most powerful microscopes can’t show us; processes that can’t be directly observed, but only inferred, guessed at, by interpreting a stew of complex biochemical by-products. I am not one of these scientists. I need bones and teeth—things that I can see with my eyes and grasp with my hands. Jason Eshleman, on the other hand, can see with his mind’s eye, grasping the complex interactions of the most complex molecules in the body, DNA.
One of the most famous, most challenging, and most controversial samples Jason had worked with was Kennewick Man, a male skeleton that a college student stumbled upon in the shallows of the Columbia River near Kennewick, Washington, in 1996. Initially believed to be the remains of a white homesteader, the bones were sent to a carbon dating lab just to be sure. The results—which put the age of the bones at 9,200 to 9,600 years old—ignited a fierce controversy. On one side were Native American tribes who wanted to rebury the remains as quickly and reverently as possible; on the other were scientists who wanted to study the bones for more clues about who first settled the Americas, where they came from, and how they got here. Several of my former Ph.D. students got caught up in the case, including George Gill, who teaches at the University of Wyoming; Doug Owsley, who heads the physical anthropology division at the Smithsonian Institution (and who got his bachelor’s degree at Wyoming, under George Gill); and Richard Jantz, who directs the Forensic Center at UT (and who developed ForDisc).
Just months after the discovery of the bones—while Jason was still working in Smith’s laboratory at UC-Davis—the lab received a small sample from the bones of Kennewick Man. Smith, Jason, and the other scientists were thrilled by the opportunity to do DNA analysis on the ancient remains, but before they had a chance to run the sample, the FBI came calling and took it. A consortium of Native American tribes had filed a legal motion to stop the analysis and recover the skeletal material, and a federal judge had granted an injunction and ordered the sample confiscated. Smith’s lab complied, but not before putting the sample in a vial with a tamper-evident seal.
Nine years later, after scientists finally won the right to make a thorough study of Kennewick Man, Jason got the sample back, still sealed in the vial. To his surprise and disappointment, the only DNA he found in the sample proved to be quite a bit younger than nine thousand years, and it wasn’t Native American. The DNA wasn’t from Kennewick Man’s, he finally discovered, but from another graduate student who had worked in the lab back in 1996. Jason didn’t learn the genetic makeup of Kennewick Man, but he did learn a powerful lesson about how easily samples can be contaminated. It was a lesson underscored on other occasions by a mysterious, persistent contaminator. “There was a period when I kept seeing the same DNA sequence over and over in my samples,” he recalls. “It wasn’t from me, and it wasn’t from anyone else who worked in the lab—we had everyone’s sequence on file, so we knew it wasn’t any of us.” Eventually, the culprit emerged: Jason’s girlfriend, who never set foot in the lab, but who shed enough stray skin cells to make her presence known, even though Jason always scrubbed up and suited up, much like a surgeon, before entering the lab.
As we talked amid the thicket of vendor booths in the cavernous exhibition hall in San Antonio, Jason explained how he would seek out and extract whatever DNA the teeth contained. He would start with just one of the teeth, keeping the other in reserve. His first step—and the reason he hadn’t worried about my contaminating the teeth with my own DNA—would be to scrub the tooth with a solution of bleach to remove any surface dirt or other residue. Next, he would soak it in the bleach solution another five minutes. The bleach bath would destroy any DNA on the outside of the tooth, but five minutes wouldn’t be enough time to allow the solution to penetrate the tooth’s enamel and permeate the dentin, where the DNA would be ensconced. Next, he’d rinse the tooth with sterile water, then dry it under an ultraviolet lamp (another step designed to destroy any contamination on the surface). To gain access to the DNA within the molar, he would crack the tooth into smaller fragments and soak them in a solution that would dissolve the calcium, break down the proteins, and free the DNA from the cells. After the pieces were dissolved—a process that would take about a we
ek, he said—he’d bind the DNA to silica beads, extract the digested proteins and other grunge, and finally wash the beads with an alcohol solution. “Just as alcohol washes away people’s inhibitions,” he explained, “it relaxes DNA’s inhibitions, causing it to release from the silica beads.”
That’s when the real biochemical chemical wizardry would begin. Heating the solution almost to boiling causes the two legs of the DNA ladder to peel apart—like a zipper unzipping. When that happens, short (twenty-rung) pieces of complementary DNA called “primers” latch onto the long, unzipped halves of the mtDNA ladder, wherever their own sequences of A’s, G’s, C’s, and T’s mesh with the mtDNA’s—a step called “annealing.” The primers keep the ladder from zipping back together again as the solution is cooled slightly. Then, with the DNA unzipped, an enzyme called Taq (rhymes with “gack”)—extracted from organisms that live in sulfur hot springs and hydrothermal ocean vents—moves in and builds a complementary new ladder-leg on each of the long unzipped legs . . . transforming one unzipped ladder into two ladders, and thereby creating two DNA strands out of one. That entire cycle, called a “polymerase chain reaction,” or PCR, takes anywhere from twenty seconds to two minutes. At that point the entire solution is reheated to near-boiling again, the two ladders are unzipped, and the replication is repeated. It doesn’t take many PCR cycles of unzipping and replicating—doubling the number of copies each time—to turn a few strands of DNA into many. Very, very many: In an hour or less, a single strand can be transformed into billions of copies.
By the time I finished talking to Jason, my head was spinning. I was in biochemical and genetic overload, but I felt sure the work and the teeth were in good hands. I wished Jason good luck in his search for DNA in the teeth.
My molar handoff wasn’t the only casework I did while I was in San Antonio and surrounded by forensic experts. I took advantage of a book signing—Jon and I were signing copies of one of our Body Farm novels—to conduct an informal poll. We showed the 106 people in our book-signing line two images—Leoma Patterson’s photo, and Joanna’s facial reconstruction—and asked, “Are these the same woman, or not the same woman?” Of the 106 people polled, seventy-two said, “Yes, the same”; thirty-four said, “No, not the same.” It wasn’t a rigorous experiment by any means, but it sure was interesting to see people struggle to decide, and to listen in as they debated aloud with themselves or colleagues before voting.
I also did one other bit of research before leaving San Antonio. In the exhibit hall, I bumped into Murray Marks—one of my former Ph.D. students, whom I’d hired to join the forensic faculty at UT. Murray has used video superimposition in forensic cases, and he’s also done research on computerized facial reconstruction, so I was particularly interested in his opinion on the case. He studied the facial superimposition video—not just the freeze-frame, but the footage itself, as the image slowly dissolved from the photo to the skull, then back to the photo again. When I put the question to him, he looked up from the computer screen, his expression halfway between amusement and exasperation. “What, are you retarded?” he said. “Of course it’s her.” I hoped Leoma Patterson’s relatives would be as easy to convince. Convincing them would be easy, of course, if the DNA in the teeth confirmed that the bones were Leoma’s after all.
I didn’t realize what a big “if” that would turn out to be.
THREE WEEKS AFTER I handed the teeth to Jason, he phoned with discouraging news. “I’m not getting any DNA from the sample,” he said. The reason wasn’t clear. “It might be that I’m getting a lot of interference from humic material”—chemicals from the dirt, he explained, junking up the reaction—“or it might be that there just isn’t any DNA left.” It was possible, he went on, that acids from the soil, mold from the damp coffin, or bacteria had gradually invaded the tooth and destroyed the genetic material.
I was dumbfounded; how was it possible for DNA to survive for ten thousand years in teeth in the Pacific Northwest, but not for thirty years in East Tennessee?
“I don’t know,” Jason admitted. “There’s not a lot of data from that part of the country.” So we were providing new, discouraging data? I failed to find much comfort in that. “There’s an extraction I can perform to remove the humic acids,” Jason offered. “It might also remove some of the DNA—if there is DNA—but it might leave enough behind for me to get a sequence.” At this point we seemed to have nothing to lose by trying.
A nail-biting week passed. “I’m seeing some DNA,” Jason finally reported. Hallelujah! My excitement was short-lived, though: He had identified some pieces of mtDNA, but not enough yet to stitch together an entire sequence. “I’m going to process the second tooth now,” he said, “and I’m hoping that will give me enough for a whole sequence.” I hoped so, too.
But my hopes were in vain. The second DNA extraction failed to yield a complete mtDNA sequence. So did a third, a fourth, and a fifth. I had sent Jason more teeth; I sent him a four-inch section of femur. I sent him a cheek swab from Leoma’s granddaughter Michelle, for comparison. But there was nothing in the skeletal material to compare with Michelle’s DNA. Why not—where had the DNA gone? The answer came during a phone call to Dr. Cleland Blake, the medical examiner who had recovered and examined the remains back in 1979.
The remains still had bits of soft tissue on them, so Dr. Blake cleaned them. He did this by simmering the bones—“for a day or two,” he said—in water containing detergent and bleach. Lots of bleach. Jason applied bleach, briefly, to destroy any DNA contaminating the samples’ surfaces; Dr. Blake, on the other hand, had stewed the bones in it, and the combined assault of heat and chemicals had almost certainly nuked the DNA. Dr. Blake hadn’t known any better—remember, the bones were found many years before forensic DNA testing was available—but the odds that Jason would succeed suddenly looked very slim.
Finally, in late May—three months after I’d optimistically handed over those first two teeth in San Antonio—Jason called. It was time to pull the plug, he said. He’d done seven extractions, all without finding a usable DNA sequence. It was a bitter blow. Abandoning the DNA quest meant abandoning our hope of making a positive identification. Ironically, after nearly two years, countless hours, and the best forensic techniques we could apply—ForDisc, DNA, a clay facial reconstruction, an experimental computerized reconstruction, and a video superimposition—we were coming full circle, ending up right back where we began: in uncertainty and ambiguity.
I scheduled a meeting with Leoma’s relatives for June 1, 2007, to brief them on our efforts, our difficulties, and our confidence that—despite the GenQuest report, and the lack of sufficient DNA to refute it—the original identification had been correct after all. Personally, I felt sure that 05-01 was Leoma Patterson—the facial superimposition had convinced me—but I knew my belief wouldn’t satisfy some of the family members, and I was dreading the meeting.
Then, at ten o’clock the night before the meeting, I received some astonishing news from David Ray, the original TBI investigator. David had long since traded his TBI badge for a sheriff’s badge—he’d been elected sheriff of Claiborne County, up near the Kentucky border—but he’d heard about the twists and turns the Patterson case had taken. Intrigued by our difficulty finding DNA, David had rummaged around in the dusty files stored in his basement. There, he’d found his old TBI file on Leoma Patterson, and—wonder of wonders—in the file, he found the hair mat that had been recovered at the death scene back in 1979. The hair, along with bits of dried scalp, was sealed in a TBI evidence bag, the seals intact. My heart began to race. Unlike the bones, the hair and scalp had not been simmered in detergent and bleach. We were back in the game.
The following afternoon I briefed the family on all we’d done, and on how everything we’d done since the GenQuest test—the flawed, contaminated, water-muddling GenQuest test—
had supported the original identification of the remains as Leoma’s. The family appreciated all the efforts we’d made, but they were understandably disappointed that we couldn’t offer certainty. When I showed them the hair and scalp, though—which David Ray had brought down from Claiborne County that morning—their hopes soared again.
As they gathered around and watched, I slit the TBI evidence seal and opened the bag containing the hair and scalp. Snipping off a hank of hair, I sealed it in a Ziploc plastic bag, along with a bit of dried scalp, and overnighted the bag to Trace Genetics. Jason planned to divide the samples; he would analyze one-half, he told me, while his senior technician—working independently, in a separate lab—analyzed the other.
On June 18, 2007, Jason reported that both he and his technician had found plenty of DNA in the hair and scalp. For simplicity, they were looking only at mitochondrial DNA—a simpler, more durable form of genetic material than nuclear DNA. “What we don’t know yet is whether it matches the granddaughter’s,” he said.
Eight days later—on June 26, 2007—he knew: The DNA in the hair and scalp, and the DNA in Michelle’s cheek swab, were identical. What’s more, it was an unusual variety of mitochondrial DNA, one distinguished by two mutations. Statistically, the chances of a random match were extremely low, said Jason, just one in fifteen thousand. Turned around the other way, that meant the odds that Michelle was indeed the dead woman’s granddaughter—that the dead grandmother was Leoma Patterson and not some Jane Doe—were fifteen thousand to one: for all practical purposes, one hundred percent. After nearly two years and a wild forensic roller-coaster ride, we had identified Leoma Patterson at last, conclusively and positively. Maybe now Leoma—that is to say, Leoma’s family—could finally rest in peace.