August 1, 2018 (Vol. 38, No. 14)
Forensic Science Use of Genetic Analysis Leads to Ground-Breaking Work
Television writers portray DNA evidence as a slam dunk, sealing the fate of many a villain in a fast-paced game of cat and mouse. The reality, however, is that a single DNA sample requires days to analyze, and many samples never get processed at all. DNA profiling has come a long way since its debut in 1986, but in many ways, it’s still in its infancy. Here are four ways researchers are breaking new ground with forensic uses of genetic analysis.
Whose Cell Is It Anyway?
A DNA sample can indeed put a criminal behind bars or exonerate an innocent suspect—but only if it gets tested. If it stays in a neglected rape kit, for example, DNA won’t serve the cause of justice. Thousands of rape kits stay on shelves for years simply because each kit takes so long to process. The first big hold-up occurs when the rape kit’s sperm cells are separated from the victim’s own cells. This process, which requires multiple rounds of washing and centrifugation, can last eight hours.
“We can reduce that multistep process to a one-step process that takes just an hour,” says Utkan Demirci, Ph.D., professor of radiology at Stanford University. Dr. Demirci worked with Fatih Inci, a research scientist in his lab, to develop a microfluidic chip that grabs the sperm cells while allowing everything else to be washed away, all in one go.
Previously, investigators tried using separation procedures incorporating sperm-binding antibodies, but these antibodies often come away empty-handed because they grasp at surface antigens that can begin to degrade within hours. Seeking more durable handholds, Dr. Demirci and his colleagues turned to lectins, highly stable proteins that bind sperm cells to oligosaccharides that coat egg cells.
The sperm cell’s surface is studded with many lectin molecules, which boost the sperm-capturing efficiency of the researchers’ oligosaccharide-covered chip surface. The researchers spent several years optimizing their chip’s surface chemistry and designing a flowthrough to retain the highest possible percentage of sperm cells.
They also teamed up with Lenny Klevan, Ph.D., a forensic expert, and George Duncan, Ph.D., a DNA supervisor at the Broward County Sheriff’s Office Forensic Laboratory, to test the chip under real-world conditions. Even when the chip was used on decade-old swabs, it successfully snagged sperm, demonstrating up to 92% capture efficiency.
The chip, Dr. Demirci asserts, is easy to use and less dependent on human labor than previous protocols. “This will be the advantage in the long run,” he adds, “when the technology is translated and commercialized.”
Smaller, Cheaper, faSTR
Once the relevant DNA is isolated, it is cut into fragments, which are analyzed to create a DNA profile. Over the last 10 years, innovators have created “rapid DNA” systems that have cut the time between cheek swab and DNA profile down to 90 minutes. Though technically portable, these machines still weigh a hefty 150 lb. “We refer to that as ‘two-man luggable,’” says James P. Landers, Ph.D., a professor at the University of Virginia (UVA). Besides maintaining UVA appointments in chemistry, mechanical engineering, and pathology, he holds executive posts at ZyGEM US and MicroGEM.
Challenged to design a system that could fit in a backpack, Dr. Landers and his team, aided by UVA’s Applied Research Institute, embarked on what Dr. Landers calls “a three-year trek to deliver something paradigm shifting.” The effort was sponsored by the Department of Defense.
By radically reimagining the mechanics that could be used in a rapid DNA system, the team created a device weighing just 10 pounds. “It’s roughly the size of two reams of paper,” Dr. Landers points out. In this device, which is called “faSTR,” a spinning microfluidic disc puts fluids into motion, making it unnecessary to resort to vacuum pumps or other bulky hardware. “The device is the size of an old-fashioned CD,” Dr. Landers notes. “But it has fairly complex fluidic architecture.”
The much smaller dimensions of the new system required some fine-tuning to optimize the biochemistry. According to Dr. Landers, the faSTR system can complete a short tandem repeat (STR) analysis from a swab in 35 minutes. Naturally, a system that is smaller and faster than its predecessor is going to be more expensive, right? Not necessarily.
“We’ve pioneered a method for making microfluidic discs out of overhead transparencies,” Dr. Landers states. A pattern of microfluidic channels is created by printing toner onto the disc everywhere except where the channels will be. Production of the discs can involve a dozen layered transparencies, with different channel patterns printed onto the different layers, allowing three-dimensional designs.
Handheld, while-you-wait DNA profiling, such as that provided by faSTR, could transform the forensic collection of genetic information. How useful that information will be in solving crimes depends, in some cases, on the availability of searchable databases of DNA profiles for comparison.
ZyGEM US, the microfluidics and forensics arm of MicroGEM, is developing the faSTR Profiling System, a field DNA analysis device that can complete a short tandem repeat (STR) analysis from a swab in 35 minutes. Currently focused on criminal justice and military applications, faSTR will eventually extend to clinical diagnostics, potentially as a home-based testing system.
Fighting Crime with Genetic Genealogy
Once DNA has been collected from a crime scene and analyzed, it can be compared to DNA taken directly from a suspect. If no suspect is available, authorities can search databases full of DNA “fingerprints” in hopes of finding a fingerprint that matches the DNA from the crime scene.
Law-enforcement databases typically contain DNA fingerprints taken from prior offenders, so if these databases fail to produce a match, investigators may need to widen their search. They may even consider delving into ancestry databases, which represent new troves of DNA information.
A new genetic genealogy service has been unveiled by Parabon NanoLabs. Investigators can work with the company to check their sample against DNA profiles that have been uploaded to a public database such as GEDmatch. Some DNA profiles may closely match the DNA from a crime scene. The more closely two profiles match, the more likely they come from closely related individuals, and perfectly matching profiles may even point to the same individual.
Ellen Greytak, Ph.D., director of bioinformatics at Parabon, emphasizes that not all DNA genealogy databases are fair game for law enforcement searches. “There are a number of companies that let you use their DNA-testing services,” she says. “These companies maintain their own private databases. Law enforcement has no access to those databases at all.”
Company-specific database restrictions also affect customers. For example, if two family members get tested by two different companies, they can’t compare their results. Enter GEDmatch. The company doesn’t do its own DNA testing. It just accepts files, which customers upload so they can compare their information with other customers’ information. Customers must expressly allow their GEDmatch files to be publicly searchable; only then can these files be searched by companies such as Parabon.
Starting with a DNA sample left at a crime scene by an unknown suspect—or by an unidentified victim—Parabon searches the database for relatives as far removed as a third cousin or a shared great-great-grandparent. The company has partnered with genetic genealogist CeCe Moore, best known for her work finding the biological relatives of adoptees. Together, the partners use genetic ties to build a family tree that will, if all goes well, lead back to the individual whose DNA has been obtained.
Recently, the team used this method to solve a 31-year-old murder case in Washington state. Two persons were in the database who were both second cousins to the suspect but genetically unrelated to each other. Using genealogical records, Moore constructed both family trees and found where they intersected by marriage, leading to a positive identification of the suspect.
When attempting to identify an unknown person from a DNA sample, Parabon NanoLab’s genetic genealogists first find a distant genetic match in a public genetic genealogy database (Match #1). They then use public records to build a family tree back to the most recent common ancestors of the unknown and the match (orange), afaater which they identify all descendants of those ancestors (blue). The unknown is generally one of the descendants, although it is possible that he or she is not listed in the genealogy tree because of adoption or misattributed parentage.
Proteins Do What DNA Can’t
Finally, when DNA is hard to come by, researchers can turn to protein. “Protein is intrinsically more stable,” says Glendon Parker, Ph.D., founder and CEO of Protein-Based Identification Technologies. Over time, DNA degrades until the remaining segments are too short to be amplified by PCR. Some forensic samples, such as fingerprint residues and hair strands, may yield too little DNA for analysis.
Still, DNA has become the go-to molecule for forensic identification because protein assays require a certain amount of finesse. Proteins denature and lose their three-dimensional structure, making enzyme assay results inconsistent. Protocols for analyzing DNA, on the other hand, became more streamlined and reproducible. “The whole field moved over to DNA and for very good reasons,” Dr. Parker says. “In the meantime, the field of proteomics has gone through a revolution.”
Protein-based identification employs the same principle at work in DNA profiling. Essentially, the technique detects single-amino-acid changes, brought about by single-nucleotide changes in the DNA. A protein containing single-amino-acid polymorphisms (SAPs) is digested with an enzyme, creating a set of peptides roughly 8–30 amino acids long. Then, mass spectrometry is used to determine the weights of the various peptides. Finally, the pattern of fragment sizes reveals changes in the amino acid sequence. “The pattern is complex, but it’s a direct function of the sequence of amino acids,” Dr. Parker explains.
Investigators can create a unique personal protein signature if they test enough of these genetically variable peptides. “It’s another way of reading what’s in their DNA,” Dr. Parker declares. He and his team found that testing around 60 different peptides gives a power of discrimination of one in a billion. So far, he says, over 300 different genetically variable peptides have been catalogued.
Considering that humans shed hair constantly, should people worry that their hair will reveal their personal genetic footprint? Not yet, says Dr. Parker. “At this point,” he notes, “we don’t have the databases we would need.” Protein-based identification can’t be used to search for a completely unknown suspect, only to compare the crime scene sample to a given individual. Traditional detective work won’t be supplanted anytime soon, it seems.
