Before DNA: Blood Typing

In 1901, Karl Landsteiner discovered that blood from different people could be grouped into several types. When a blood transfusion was performed between people of different types, the red blood cells were quickly destroyed. But when a transfusion was performed between people of the same blood type, the donor cells would survive. The modern-day A-B-O blood typing system was born, enabling the first successful blood transfusions, and earning Landsteiner the Nobel Prize in 1930.

The discovery of blood typing had important implications in forensics as well as in medicine. In 1916, in the first case to rely on the new technology (see the section below), Leone Lattes applied a technique he had developed to identify the blood type of dried blood. For the first time, it was possible to know definitively that a given sample of blood could not have come from a specific person.

Blood typing was an extraordinarily useful forensic tool, and the technology underwent several improvements in the decades following its discovery. Additional typing methods besides the original A and B markers were discovered that allowed for further subtyping. Particularly important were the Rh factor (which most of us know as the + or – in our blood types) and the Barr body (present in females but not males – allowing for the determination of gender from a blood source). However, blood typing suffered from a number of shortcomings. (1) Many people share the same blood type, making high confidence matches difficult or impossible. (2) Blood typing required a bodily fluid or its residue, and so in cases with small amounts of material, non-fluid biological materials (such as hair or skin), or sample degradation, typing was impossible. Because of this, DNA based tests would come to replace blood typing as a forensic tool.

Case Example: A Bloody Nose or a Murderer?

When Leone Lattes, a professor at Italy’s University of Torino, developed a test for typing dried blood, he received immediate requests to apply his services. In one of the first cases, a man was accused of committing a murder. Among the evidence against him were a large number of blood stains found on his overcoat. The man claimed these stains were the result of a bloody nose, but the police suspected that they came from the victim. When Lattes tested the blood, he found that the victim had blood type O, while the blood on the overcoat was type A (the same as the suspect). This supported the suspect’s story and caused the police to change the focus of their investigation.

To read more about this case and others, consult Latte’s article: “Two Practical Cases of Individual Diagnosis of Human Blood,” which can be found in the following collection from the NCJRS (note: this is a very large PDF file).

RFLP: The First DNA Fingerprinting

In 1984, Alec Jeffreys developed the first DNA test capable of separating individuals, based on a technique called RFLP (Restriction Fragment Length Polymorphism). Together with Peter Gill, who was developing techniques for recovering DNA from blood stains and from semen collected after a rape, he subsequently demonstrated that RFLP could be applied successfully in a forensic setting. The RFLP method would go on to become the nearly universal DNA forensic tool until it was eventually overtaken by STR based methods in the 1990s.

The RFLP technique begins by using molecules called Restriction Enzymes to cut up DNA into specific pieces. While different versions of the technique exist, the most common version relies on special regions in the genome, called VNTRs (Variable Number Tandem Repeats). These regions are parts of the genome that have short DNA sequences repeated different numbers of times in different people. The special property that makes them ideal for use in a DNA test is their high variability. While one person might have a region containing 10 copies of a repeated sequence, another person might have 20 copies of the repeat, and another might have 25 copies. A number of independent VNTR regions exist in the genome, and the RFLP technique provided a way to measure the length of many of these regions in the same experiment (important in a forensic situation where material is often scarce). While two individuals might share the same length profiles for some of these regions, the chance of them sharing the same lengths at all of them is exceptionally small.

RFLP was a significant improvement over blood typing. Unlike blood typing, the technique could produce very high-confidence matches (the chance of an accidental match was less than one in a billion, and was subsequently made even smaller). And it was also applicable to older samples that had undergone too much degradation for blood typing to work. However, there were still a number of limitations that would be improved upon by later techniques. (1) A large amount of starting material was required for the technique. (2) When DNA in a sample degrades, it breaks up into progressively smaller pieces. The sizes of the sections of DNA measured in RFLP were relatively large, so if DNA was too degraded, the test couldn’t be performed. (3) There was some subjective interpretation required in examining the results that could result in different labs reporting different answers. The STR based method that came to replace RFLP offered significant improvements over all of these shortcomings.

Case Example: DNA’s First Exoneration

The first application of DNA fingerprinting in a criminal case turned out to be a powerful portent of the revolutionary role DNA would play in exonerating the innocent. In 1986, police were investigating the rape and murder of two girls, Lynda Mann and Dawn Ashworth. The police had a 17-year-old man, Richard Buckland, in custody. Buckland, who suffered from learning disabilities, had confessed to the murder of Dawn Ashworth after police questioning. But he denied murdering Lynda Mann. The police hoped to use DNA to link Buckland to the Mann murder. They called in Jeffreys and Gill to perform a test to match semen from the crime scenes to Buckland.

When the test results came in, they revealed that the same man had murdered both girls, but this man was not Buckland. Buckland had given a false confession and the real killer was still at large. Out of leads, the police ran a DNA screen of 4,000 men – nearly every young adult male in the village. The guilty man still didn’t turn up in the screen and the case was about to run cold, when a man from the village revealed an interesting story. A friend of his named Colin Pitchfork had paid the villager to take the DNA test for him. Pitchfork was subsequently arrested and proven to be the killer of both girls. Jeffreys’ test had both freed an innocent man and found the real murderer.

A detailed description of this case can be found on Qiagen’s website here.

PCR, STRs and the Onset of Modern DNA Testing

In 1983, Kary Mullis developed what was to become one of the most transformational techniques in modern science. Named PCR (Polymerase Chain Reaction), the technique would come to revolutionize the way biology, and along with it forensic DNA testing, was done. PCR allows any desired region of DNA to be amplified, turning trace amounts of starting DNA into amounts large enough for testing. At a basic level, PCR works by taking advantage of the DNA-copying machinery used by natural organisms. By using short DNA molecules that match a particular region, called PCR primers, a laboratory investigator can force this copying machinery to only copy that specific region. By repeating the procedure, the number of copies of the region can be repeatedly doubled, resulting in an exponential increase in copies of the desired DNA.

PCR was originally applied to DNA testing in the late 1980s, around the same time that Jeffreys’ RFLP technique was first developed. At first, PCR based testing involved examining a region of DNA called the HLA locus that is involved in immune function and is highly variable between people. However, this test couldn’t offer the level of specificity that the RFLP technique provided and, consequently, was only used in situations when RFLP didn’t work. A few years later, in the early 1990s, a technique was developed to use PCR on regions called STRs (Short Tandem Repeats). These are regions in the genome with repeated sequences, similar to the repeated regions used in the RFLP method, but significantly smaller. With this advance, PCR became superior to RFLP, and soon replaced it in most crime labs. Today, the use of PCR on STRs is the standard forensic DNA test in most countries, and data from this test forms the basis for CODIS, the FBI’s DNA database.

PCR has two main advantages over the earlier techniques. (1) PCR can be used to amplify even miniscule amounts of starting material, so even trace evidence left behind can often yield results. (2) The STR regions used in modern PCR testing are significantly shorter than those used in RFLP. Therefore, the PCR technique will frequently work even when the DNA is too degraded to be used in RFLP. Many of the modern advances in DNA testing have worked to push for even greater improvement in these two categories.

Case Example: Ronald Jones

In 1985, a woman was grabbed by a man known as “Bumpy,” who asked her for change to buy a drink. Later that day, she was found dead, and semen was found on her, indicating that she had been raped. Friends of the victim testified that Ronald Jones was “Bumpy,” the man who had grabbed the victim earlier that day. Both blood typing and DNA testing were attempted on the semen found on the victim. The blood typing results were presented as evidence of a possible match, though the test was flawed because the semen was mixed with the victim’s own cells, which prevented an accurate typing. An RFLP DNA test could not offer conclusive results. Jones was convicted in 1989 and sentenced to death. After years of appeals, in 1997 permission was granted to apply PCR based DNA testing to reanalyze the sample. This time the test was conclusive: there was no match between Jones and the semen found on the victim. Jones was finally freed in 1999 after 10 years in prison.

More information on Jones’ case can be found on the Innocence Project website.

Y Chromosome Based Tests

While the traditional STR based methodology is still the most common technique in use today, there are situations when specialized techniques are needed. One such technique is Y chromosome testing. In forensics, the most common use of this method is in rape cases. The biological samples recovered after a rape are a complex mix of cells from both the victim and the rapist. In order to get a clean DNA profile from the perpetrator alone, it is necessary to separate this mix. A standard technique in these situations is to use a chemical that digests most cells but has no effect on sperm cells. However, in some cases, such as when the rapist has had a vasectomy, or when very little sperm can be recovered, this isn’t possible. In the case of a female rape victim, a Y chromosome test can then be used. Because women lack the Y chromosome, the results of the test will only reflect the male DNA in the sample.

Y Chromosome based tests are very similar to normal STR tests (again short regions with repeats that vary highly between individuals are used). However, the Y chromosome based tests offer a greater challenge in interpretation. This is because the DNA profile generated for a male by such a test will be identical to his father’s profile (except in the case of a mutation in a single generation, which is relatively rare). This is different from the traditional STR test, where 50% of the DNA tested comes from the mother and 50% from the father, producing a unique profile in every individual. Because of this, it can be difficult or impossible to differentiate between close male relatives. Therefore, Y chromosome tests are usually only used when DNA cannot be recovered using traditional STR methods.

Case Example: Exoneration from an Old Sample

In 1981, a woman was raped in her apartment in Dallas by an unknown man. The woman only saw part of the man’s face briefly during the attack, but when presented with a photo lineup, she identified Charles Chatman as her rapist. A blood type test revealed that the rapist had type O blood, the same as Chatman’s (as well as 40% of the African American population). Chatman was found guilty and sentenced to 99 years in prison. In 2004, after spending 23 years in jail, he was granted a DNA test. Unfortunately, the test was unsuccessful and couldn’t produce a conclusive result. The Innocence Project of Texas then requested that a Y-STR test be conducted. This test proved that Chatman wasn’t the rapist, and he was released from prison in 2008.

More information on Chatman’s case can be found on the Innocence Project website.

Mitochondrial Based Tests

Similar to Y chromosome based tests, testing based on mitochondrial DNA can be used in situations when traditional STR testing isn’t possible. Mitochondria are specialized compartments in our cells that produce the cells’ energy. They have their own DNA, separate from the 23 pairs of chromosomes in the nuclear genome. This DNA is passed directly from a mother to her children (both male and female) with no contribution from the father.

Mitochondrial based tests differ somewhat from the standard STR method, but are similar in principle. Typically, PCR is used to amplify two regions of the mitochondrial DNA that are highly variable between people. Samples are then matched by comparing the sequence in these two regions. The advantage of mitochondrial DNA is that it typically lasts longer than nuclear DNA. Most cells have hundreds or even thousands of copies of mitochondrial DNA (and only two copies of nuclear DNA), so the chance that some of the mitochondrial DNA will survive degradation is much higher.

The challenges of mitochondrial based tests are similar to those utilizing the Y chromosome. Because all of us get our mitochondrial DNA entirely from our mother, maternally related family members will have identical or nearly identical mitochondrial DNA. This makes distinguishing between individuals based on mitochondrial DNA more difficult and the level of certainty from a “match” far lower than under standard testing.

Case Example: A Hair but No Root

In 1996, there were a series of rapes in Indiana, and police suspected that a serial rapist was on the loose. Richard Alexander became a suspect and was arrested by the police. While one of the rape victims and her fiancé testified that Alexander was the woman’s rapist, there were a number of pieces of contradictory evidence. Even after Alexander was put in prison, a number of rapes occurred, in which Alexander could not possibly have been the perpetrator. However, in one of these rapes, Alexander’s photograph was accidentally shown in a photo array, and the victim identified him as her rapist.

Then in 2001, a man named Michael Murphy confessed to one of the rapes for which Alexander had been convicted. Hair samples had been collected from this rape, but it wasn’t possible to extract a DNA profile at the time. In the five years since Alexander’s conviction, the development of mitochondrial based tests had changed this. While a DNA profile cannot be recovered from hairs without roots using standard STR based methods, it can frequently be recovered using a mitochondrial based test. After Murphy’s confession, the hairs were tested. They were found to match Murphy and not Alexander. After five years in prison, Alexander was released at the end of 2001.

More information on Alexander’s case can be found on the Innocence Project website.

Non-Human DNA Forensics

The use of DNA testing is not limited to humans, and in recent years many of the forensic tests developed for people have been adapted to a number of other animals and even plants. These techniques have been applied in a number of criminal cases, particularly where pet material has been left behind at a crime scene. While most of the techniques are similar to those used for humans, the methodologies are less developed in most cases. Commercial tests are only available for a small number of animal species, and largely haven’t been standardized as they have for humans.

Case Example: Snowball

In 1994, in Prince Edward Island, Canada, a 32-year-old woman disappeared from her home. Weeks later, a man’s jacket was found in the woods near her house. The jacket contained stains of the woman’s blood along with hairs identified to be from a domestic cat. By 1995, the woman’s body had been found and a suspect in the case had emerged: the woman’s estranged husband. The man lived with a pet cat, Snowball, whose hair color matched the cat hairs found in the bloody jacket. An STR based DNA test was performed on the cat hairs from the jacket and on a sample taken from Snowball, and was found to be a match. This evidence helped link the husband to the crime, and he was convicted in 1996 of second degree murder.

A more detailed description of this case can be found in the follow article: Marilyn A. Menotti-Raymond, Victor A. David & Stephen J. O’Brien, Pet cat hair implicates murder suspect, 386 Nature 774 (1997). The following is a link to the article (note: may require a subscription to access).

Modern DNA Testing and the Future of DNA Forensics

While many of the techniques used today were developed in the 1990s, some (such as Y chromosome tests) weren’t in widespread use until well into the 2000s. More recent developments in DNA testing have focused on a number of areas. These include: (1) The recovery of DNA from increasingly trace amounts of biological evidence. For example, DNA can now frequently be recovered from a small number of skin cells left behind when someone touches an object. (2) Recovery of DNA from degraded samples. Tests based on alternative regions, such as Mini-STRs, have been developed and are starting to gain use. In these tests, the DNA examined is smaller than that used in standard STR tests, allowing for the use of more highly degraded DNA.

One of the critical lessons that has emerged over the history of DNA testing is the importance of retaining biological samples, even when they don’t provide useful results at the time of the investigation. In many cases, a conviction was overturned only when a new technique was developed that allowed for testing of a saved sample. In untold numbers of other cases, an innocent person has remained in prison because potentially exculpatory evidence was lost.

The future is likely to bring a number of important developments in DNA testing. Improvements are likely to continue in the recovery of trace and degraded evidence. Additionally, there will be a continued extension of forensic tests to non-human DNA evidence, including animal, plant, and microbial DNA. The future will also bring about new possibilities fraught with important legal and societal implications. Among these will be the use of DNA from a crime scene to create a physical profile of a suspect. Currently, this is far too difficult for all but the broadest of characteristics, but in the coming decades it will likely become increasingly practical.