By J. Thomas McClintock

A strand of hair, a drop of blood, a trace of saliva on a bottle, semen isolated from a rape victim. These bits of evidence from a crime scene can influence criminal investigations as well as the outcome of a trial. Instead of relying on fingerprints from a crime scene to identify suspects, law enforcement officers are learning to collect evidence that may contain DNA to identify individuals with virtual certainty. Violent crimes and sexual abuse cases that once might have gone unsolved due to insufficient evidence are now being solved based on the collection and analysis of such biological material.

DNA, or deoxyribonucleic acid, is the basic molecule of life that is imparted to us by our parents. It provides the genetic makeup that dictates the specific arrangement of building blocks that determines a person's individual characteristics. Because the order of these building blocks, or nucleotides, varies from person to person, scientists can compare DNA patterns or profiles from individuals and either link or eliminate a suspect to the evidence, in a manner similar to the use of fingerprints. Thus, DNA profiles have become powerful tools in the identification of individuals in criminal and paternity cases.

DNA testing, first introduced in the late 1980s, has helped solve many cases where other investigative leads failed. For example, from November 1997 to July 1998, eight rapes occurred in Southeast Washington, DC that confounded detectives because the victims gave investigators varying descriptions of their attackers. In December 1999, Leon Dundas of Jacksonville, Florida, was questioned in connection with three rapes in Jacksonville but released due to insufficient evidence. Jacksonville police could not take a blood sample because they lacked probable cause. However, the course of the investigation changed drastically when Dundas was found shot to death in March 2000. DNA profiles generated from blood samples taken from his body linked him to the three Jacksonville rapes as well as the eight District attacks. Consequently, DNA analysis helped solve this case that had no other investigative leads.

An increasing number of criminal convictions have been overturned by DNA test results. Such was the case when a local Maryland resident, accused of killing his wife in 1999, was freed from jail after DNA test results implicated another man now in custody in the District on an unrelated sexual assault. In another case, a developmentally disabled custodian from Manassas, Virginia was coerced into pleading guilty to a 1984 rape and murder he did not commit. In 1987, the real murderer finished serving a prison sentence (for burglary), and over the next three months he assaulted and killed three women in Richmond, Virginia and later raped and strangled a fourth woman in Arlington, Virginia. Police used DNA evidence to link the killer to the four killings and the 1984 rape. The Governor pardoned the disabled custodian in 1989 based on this DNA evidence.

The first widespread use of DNA tests involved RFLP (restriction fragment length polymorphism) analysis, a test designed to detect variations in the DNA from different individuals. In the RFLP method, DNA is isolated from a biological specimen (e.g., blood, semen, vaginal swabs) and cut by an enzyme into fragments. The DNA fragments are separated by size into discrete bands by gel electrophoresis, transferred onto a membrane, and identified using probes (known DNA sequences that are "tagged" with a chemical tracer). The resulting DNA profile, which resembles a simplified supermarket bar code, is visualized by exposing the membrane to a piece of x-ray film which allows the scientist to determine which specific fragments the probe identified among the thousands in a sample of human DNA. A "match" is made when similar DNA profiles are observed between an evidentiary sample and those from a suspect's DNA. A determination is then made as to the probability that a person selected at random from a given population would match the evidence sample as well as the suspect.

If the evidentiary sample contains an insufficient quantity of DNA for RFLP testing or if the DNA is degraded, a PCR (polymerase chain reaction)-based test may be used to obtain a DNA profile. The PCR-based tests provide rapid results and can serve as an alternative or as a complement to RFLP testing. As in RFLP analysis, DNA is first isolated from a biological specimen. Next, the PCR amplification technique is used to produce millions of copies of a specific portion of a targeted DNA segment. The PCR amplification procedure can be likened to a molecular xeroxing machine. The amplified PCR products are then identified by the addition of known DNA probes or separated by gel electrophoresis followed by chemical staining. The first commercial and validated PCR-based typing tests available were the HLA (human leukocyte antigen) DQ alpha system, now called DQA1, which can distinguish twenty-eight DQA1 types and the Polymarker (PM) system, which allows the forensic analyst to type five additional genetic markers. Such detection procedures reduce the analysis time from several weeks to twenty-four to forty-eight hours. As in RFLP testing, a "match" is made by comparing profiles from evidentiary samples to those from a suspect's DNA followed by probability calculations.

The latest method of DNA typing, called STR (short tandem repeat) analysis, has emerged as the most successful and widely used DNA typing procedure. STRs are sites on the chromosomes that contain short sequences that repeat themselves within the DNA. These elements serve as helpful markers for identification because of their abundance in the human genome. Following DNA extraction and PCR amplification, the DNA fragments or alleles are identified by capillary electrophoresis, a process that separates the alleles based on size. The resulting alleles are displayed graphically as "peaks." The STR process reduces the amount of time to obtain results and requires a sample size smaller than that needed for RFLP typing. A higher degree of discrimination and even individualization can be attained by analyzing a combination of STRs in a process referred to as multiplexing. Commercial kits (e.g., Profiler Plus and Cofiler) are available that utilize thirteen standardized STRs.

But what do the DNA test results mean? The power of DNA evidence lies in the statistics. With RFLP, five markers or sites along a strand of DNA are typically analyzed yielding a probability of randomly selecting an individual from a given population ranging from one person in 100,000 to one in a million (1,2). With the PCR-based systems, specifically the HLA DQA1 and PM test, the probabilities may range from one person in 10,000 to one in 20,000. With STR analysis and when used in combination with all STR systems the power of discrimination may exceed 3 x 10 to the eleventh power, or the number of people on earth. Clearly, such strong evidence can have a powerful effect in courtroom proceedings. As the National Research Council states in The Evaluation of Forensic DNA Evidence (1996), "The state of the profiling technology and the methods for estimating frequencies and related statistics have progressed to the point where the admissibility of properly collected and analyzed DNA data should not be in doubt." (3).

However, a major point of controversy often lies in the interpretation of the DNA test results when samples contain material or DNA from more than one person. For a large variety of crimes, such as rape, the evidentiary samples will contain DNA from more than one contributor. Consequently, the evaluation of such mixtures is complex and must be interpreted carefully. For example, DNA markers (or fragments) from a sample containing a mixture originating from two individuals can be separated into major and minor components. However, even then a mixture can only be identified if the DNA markers of the minor component are above the "background noise." Moreover, a mixture may not always be evident by the presence of multiple bands (i.e., STR analysis) where the contributors actually share markers at a particular site on the DNA molecule. Fortunately, a case will usually comprise several stains or evidentiary samples which will reveal only one DNA contributor. These analyses, when compared to the other mixed and known samples in a case, may eliminate anyambiguities.

Several approaches have been used to assess the significance of an inclusion or match when samples containing DNA from more than one source have been detected in evidentiary samples. In STR analysis, one method involves the assignment of genotypes based upon peak height ratios, followed by the standard probability calculations. A recent inter-laboratory mixture study, evaluating the reliability of "peak heights ratios" generated by STR analyses, demonstrated the difficulties of determining alleles in mixtures (4). All participants were able to identify the alleles of the mixtures with the exception of some minor peaks. One laboratory reported a "stutter" (i.e., an artifact) as an allele, while two laboratories did not attempt to distinguish the genotypes of the contributors of the mixtures. The results demonstrated that the alleles were identified from the vast majority of the mixtures tested; however, the ability to determine the individual components of the mixture depended on the laboratory.

In a separate mixture study, sponsored by the National Institute of Standards and Technology, forty-five local, state, federal and commercial forensic laboratories were requested to specify all contributors in each sample mixture, provide STR profiles, and estimate the amount of DNA in the samples as well as the amount of recoverable DNA per sample (5). No participant in the study mistyped the single contributor sample. However, many laboratories did not attempt to fully type the contributors profile or they provided incorrect genotype assignments. The inability to correctly assign the proper genotype to a contributor was attributed to multiple shared alleles. Further investigations will clarify these results.

Perhaps the most significant investigative tool will allow crime laboratories to compare DNA types recovered from crime-scene evidence to those of convicted sex offenders and other criminals. This capability will be of tremendous value to investigators in cases where police have been unable to identify a suspect. In 1994, Congress enacted the Violent Crime Control and Law Enforcement Act, which included provisions for the FBI to establish a national DNA data bank, called CODIS (Combined DNA Index System), to allow crime laboratories to compare DNA information electronically. As of June 1998, all 50 states and the District of Columbia have passed legislation requiring the collection of DNA samples from offenders convicted of certain crimes (i.e., sex offenders, assault, murder, manslaughter, and endangering children). The resulting DNA profiles, which are unique to each individual, are entered into the convicted offender index of CODIS; whereas, the DNA profiles developed from crime scene samples are entered in the forensic index of CODIS. The CODIS software searches the two indexes for matching DNA profiles. This technology has revolutionized crime scene investigations and has provided a method to solve cases where no other investigative leads existed.

DNA profiles, generated from bits of evidence from a crime scene, are recognized at least as reliable and probative as fingerprints, if not more so. Instead of relying on fingerprints from a crime scene to identify suspects, law enforcement officers are learning to collect evidence that may contain DNA to identify individuals with virtual certainty. Violent crimes and sexual abuse cases that once might have gone unsolved due to insufficient evidence are now being solved based on the collection and analysis of such biological material.

References

1. Committee on DNA Technology in Forensic Science, Board on Biology, Commission on Life Sciences, National Research Council. 1992. DNA technology in forensic science. National Academy Press, Washington, D.C.
2. Lander, E.S., and B. Budowle. 1994. DNA fingerprinting dispute laid to rest. Nature 371: 735-738.
3. Committee on DNA Forensic Science: An Update, Commission on DNA Forensic Science: An Update, National Research Council. 1996. The evaluation of forensic DNA evidence. National Academy Press, Washington, D.C.
4. Ladd, C., N. C. S. Yang, and H. C. Lee. 2001. STR inter-laboratory mixture study. Proceedings of the American Academy of Sciences. Annual Meeting, Seattle, Washington.
5. Kline, M. C., J. W. Redman, D. L. Duewer, and D. J. Reeder. Results from the 1999 NIST mixed-stain study #2: DNA quantitation, differential extraction, and identification of the unknown contributors. National Institute of Standards and Technology Publication- Chemical Science and Technology Laboratory. Gaithersburg, MD.

J. Thomas McClintock, Ph.D.
DNA Diagnostics, Inc.
PO Box 4544 Crofton, MD 21114-4544
www.DNADiagnosticsinc.com
703-495-9090; 703-728-9090