HOW? DUNIT

DNA, or deoxyribonucleic acid, is a relative newcomer to the world of forensic science. It is also an incredibly useful tool for identification. In fact, only DNA and fingerprints (see Chapter Twelve: Fingerprints) are absolutely individualizing as no two people have ever been found to share the same DNA or fingerprints. Except for identical twins, that is. They possess identical DNA but different fingerprints. No one knows why their fingerprints are different, but they are.

Our understanding of DNA and its uses in both medicine and forensics is a rapidly evolving field. And like the blood chemistry discussed in the previous chapter, it is an extremely complex subject to grasp. In this chapter, I will attempt to simplify it a bit and give you some degree of understanding how this valuable and exciting field impacts the world of forensics.

Your body is made up of approximately sixty trillion cells. Each of these cells, with the exception of the red blood cells (RBCs) in your blood, contains a nucleus, and it is within the nucleus that your DNA resides.

The human body has cells of many types: heart cells, brain cells, blood cells, liver cells … you get the idea. And each of these cell types has a specific function and works in concert with one another to produce a functional human being. But how do all these cells know what they’re supposed to do? They have an instruction manual that tells them what type of cell they are and exactly what they are supposed to do. This instruction manual comes in the form of the DNA molecule.

Some of our DNA is packaged into units called genes. These are the basic units of heredity. The genes are in turn arranged along a long structure called a chromosome. Humans possess forty-six chromosomes arranged into twenty-three pairs. These pairs are numbered 1 through 22, with the final two being the sex chromosomes, called X and Y.

The DNA molecule is a polymer (long string of repeating components) of smaller molecules called purine bases. Though many different purine bases exist, only four are involved in the production of DNA: guanine, cytosine, thymine, and adenine. Scientists typically refer to these by their first letter: G, C, T, and A. All life is based on these four molecules.

The number of bases strung together in any given DNA strand can be in the millions or billions, and they can hook up in any conceivable order. The order in which these are linked determine the message contained within the DNA. The highly variable nature of this pattern is what makes DNA so useful for identification.

DNA is double-stranded, which means that it consists of paired strands (polymers) of these bases that are wound together in a double helix, a spiral-like structure. It looks like a twisted ladder (see Figure 10-1). When these bases pair up to form a double strand, each strand is a mirror image of its mate. The reason is that rules of base pairing dictate that C only binds with G, and A only with T. It has to do with the size and shape of these bases, but digging deeper into that is not necessary for you to understand how DNA is used in the forensic world.

The term genome is used to refer to the total DNA within a cell. Each person has approximately three billion base pairs in his DNA—that’s six billion bases in all. Since these bases can be put together in any order, the possible base sequences for any given DNA strand is literally astronomical. This is the basic reason that we are all different and the reason DNA typing (DNA fingerprinting) in the forensics lab is so accurate.

But, do we use all of our genome? The answer is yes, and no. Each DNA strand is made up of two different types of DNA: Genes, which make up about 5 percent of our DNA and determine our genetic characteristics and inheritance, and what is called non-encoded DNA, which makes up the other 95 percent. This non-encoded DNA is also affectionately called junk DNA.

This junk DNA supports and affects certain gene functions. And it is the DNA that is of most interest to the forensic scientist.

Genetic individuality is fixed at conception when a person receives half of his chromosomes, and thus his DNA, from each parent. The mother donates one chromosome from each of her twenty-three pairs to each egg she produces. Which member of each pair she donates is independent of which member of every other pair she donates. If we consider each member of the twenty-three pairs as either A or B, she would have a Chromosome 1A and 1B, Chromosome 2A and 2B, and so on. One egg could possess all A’s, another all B’s, another half A’s and half B’s, or any combination of A’s and B’s. For example, one egg could be ABBAABBA, etc., while another could be BBAABBAAA, etc. Since each chromosome has two choices (A or B) and since they are independent of one another, the possible combinations are 2 multiplied by itself 23 times. That is, 2 × 2 = 4; 4 × 2 = 8; 8 × 2 = 16, etc. Do this 23 times and you’ll see that the mother can produce 8,388,608 different types of eggs.

Figure 10-1: The DNA double helix. The rules of base pairing dictate that cytosine (C) must pair with guanine (G), and adenine (A) must pair with thymine (T). This pairing holds the “twisted ladder” together.

Of course, the father can produce the same number of different types of sperm. Add to this that any of these eight million sperm types can combine with any of the mother’s eight million egg types to produce a fertilized egg and the possibilities become huge. In fact, there are more than eight trillion possible combinations. No wonder you don’t look like your sister.

It’s this diversity that allows the forensic scientist to identify a perpetrator or exonerate a suspect with such a high degree of accuracy. This discriminatory power was first exposed to the public in the famous Colin Pitchfork case.

FORENSIC CASE FILES: THE COLIN PITCHFORK CASE

In 1983, fifteen-year-old Lynda Mann was brutally raped and murdered near the rural English town of Narborough. In 1986, Dawn Ashworth, also fifteen years old, met a similar fate, sending a cold panic through the community. When the police investigation hit a wall, local officials decided to try the new technique of DNA matching, which had just been developed by Dr. Alec Jeffreys at the University of Leicester. The police believed that the killer lived and worked in the area, so they asked all males in the area to submit a blood sample for testing. After screening several thousand samples, no match was made. Then, a man came forward and told police that a co-worker had persuaded him to give a blood sample in his place. The man’s name was Colin Pitchfork. In 1987, the police obtained a sample from Pitchfork, a match was made, and he confessed. In 1988 he was sentenced to life in prison. This was the first time that mass DNA screening had been used to solve a criminal case.

For many years forensic scientists searched for a method to absolutely identify an individual from materials left at a crime scene. Fingerprints were the first discovery that provided such positive proof. But fingerprints aren’t found at every crime scene. Some criminals wear gloves or wipe prints from any objects they might have touched.

However, DNA-containing materials are frequently left at crime scenes without the perpetrator’s knowledge. Since DNA is found in essentially every cell in the body, virtually any biological material from the criminal will reveal the perpetrator’s identity. Blood, semen, saliva, hair, skin, sweat, and tears can each contain DNA evidence.

Swiss biologist Friedrich Miescher (1844–1895) first discovered DNA in 1868, but it was many years before it was truly understood what DNA was and what it did. In 1943, while working with bacteria, Oswald Avery (1877–1955), Colin MacLeod (1909–1972), and Maclyn McCarty (1911–2005) discovered that DNA carried genetic information, and in 1953, James Watson (1928– ), Francis Crick (1916–2004), and Maurice Wilkins (1916–2004) elucidated the double-helical structure of the DNA molecule.

As scientists continued to analyze this molecule, it became apparent that all humans, and indeed all primates, share a large amount of the genome. This means that much of your DNA is exactly like everyone else’s and also identical to that of the chimpanzees in the local zoo. If this is the case, how can DNA be used to distinguish one person from another? The key is that we share a “large amount” of the genome, but not all.

In 1984, Alec Jeffreys (1950– ) and his associates at the University of Leicester discovered that each person’s DNA was actually unique. By using special restriction enzymes (more on these later) that cut DNA into shorter pieces, they found that certain areas of this long DNA molecule exhibited polymorphism (many different forms). It turns out that these variable areas are unique in each of us, and it is the analysis of these areas that allows discrimination of one individual from another. Shortly after discovering this polymorphism, Jeffreys developed a process for isolating and analyzing these areas of human DNA. He termed this analysis DNA fingerprinting. It is also called DNA typing.

DNA evidence first entered a U.S. courtroom in 1985 and the first conviction based on DNA evidence in the world was in 1987 (see “Forensic Case Files: The Colin Pitchfork Case”).

DNA polymorphism is found in non-encoded junk DNA. These areas are highly variable in length and base sequence, and this is what is important to forensic DNA typing. The reason is that it has been found that certain base sequences within the non-encoded DNA segments are constantly repeated. These repeating sequences are called satellites or, depending on their size, minisatellites or microsatellites. These satellite sequences repeat throughout a specific location (called a locus) within the strand. Since these segments are of variable length and repeat along the length of the DNA strand a variable number of times, they are called variable number tandem repeats (VNTRs).

Before we look at how these repetitive satellite sequences are used, I want to introduce a different type satellite sequence known as short tandem repeats (STRs), which have further increased the discriminatory power of DNA.

STRs are tandem repeats similar to VNTRs except that they are much shorter and repeat frequently throughout the DNA chain. Also, there are many more known STRs than there are VNTRs, which gives the forensic scientist more repeats to analyze.

The length of the repeated sequence in a VNTR may be hundreds of base pairs long, but the repeating sequence in an STR is only three to seven bases long. And STRs repeat over segments of the DNA strand that is four hundred or less bases long. This means that by using STRs, even degraded or damaged DNA samples can be used for testing (discussed later in this chapter).

DNA fingerprinting or typing is complex and not easy to grasp. There are numerous techniques available and even more on the horizon as research into this field is ongoing. We’ll consider the basics of some of the more common techniques—one old, two current, and one future.

The oldest DNA analysis method still in use is called restriction fragment length polymorphism (RFLP). The major problem with RFLP is that it requires a rather large DNA sample that is of good quality. For this reason, many labs now use the combination of polymerase chain reaction (PCR) and short tandem repeats as their method for DNA analysis. As we will see, this process is automated in more sophisticated labs through a process known as multiplexing. Here several STRs are extracted and amplified at the same time. This allows the lab to define a number of different DNA markers in a very short period of time.

The future of DNA testing may lie in single nucleotide polymorphism (SNP), where the level of differentiation falls to a single base. This technique can be easily automated, which makes it a very efficient method.

This is an older, more expensive, more time-consuming, and less accurate method for DNA analysis, but it is still in use. The basic steps in RFLP are:

DNA EXTRACTION: Before the lab can analyze the DNA, it must be separated from the material that contains it. Since the DNA resides within the nuclei of the cells, it must be extracted from the cells without damaging the DNA itself. There are many methods for accomplishing this, and none are particularly better than any other. The type of tissue to be analyzed and the particular lab performing the procedure determine which method is used.

These procedures usually employ protein-destroying enzymes called proteases. These enzymes break down the proteins of the cell wall and other cellular structures, but do not harm the DNA, which is not a protein. The sample is mixed with a solution of salt, detergent, and a protease enzyme. The enzyme digests (breaks down) the proteins of the cells and releases the DNA into the solution. An organic solvent such as chloroform or phenol is added. The DNA is soluble in the water solution, while the protein fragments are soluble in the organic solvent. The two solvents separate—as does vinegar and oil salad dressing—with the water solvent, which contains the DNA, layering out over the denser organic one. Alcohol is then added to the DNA-water solution. This precipitates out the DNA, which is filtered. It is now ready to use.

DNA FRAGMENTATION AND AMPLIFICATION: The long DNA strands are then fragmented into smaller portions using a restriction enzyme, which cuts the strand at predictable locations. The location chosen is one not involved in the VNTR repeating pattern, since preservation of this pattern is necessary for determining the number of repeats. The original DNA molecule, which might be a million bases in length, is cut into fragments that might be one hundred to ten thousand bases long.

FRAGMENT SEPARATION: The fragments are separated using gel electrophoresis (see the appendix), which separates them according to size. The shorter the fragment, the faster and farther it will migrate through the gel. The reason is that the longer fragments meet more resistance as they move through the gel than the shorter ones do. In this way, the fragments are separated into groups, which will appear as bands according to length.

FRAGMENT TRANSFER: THE SOUTHERN BLOT: Now that the DNA fragments have been separated into bands so that they can be compared with other DNA samples, they have to be transferred to a medium where they can be handled. A gel simply won’t work. Ever tried to pick up Jell-O? It’s not an easy thing to do since gelatin isn’t very sturdy. So how can we store, preserve, and compare two samples of gel? Enter Edward Southern and his technique, which bears his name: the Southern blot.

In this process, which is similar to mopping up a spill with a paper towel, a sturdy nylon membrane is placed on top of the gel. The DNA bands on the gel transfer to the nylon, retaining their positions relative to one another so that the all-important pattern of the bands is unaltered.

FRAGMENT TAGGING AND VISUALIZATION: Now that the DNA has been moved to a sturdier environment, the bands must be made visible. This is accomplished using radioisotope probes, which are simply DNA fragments tagged with a radioactive isotope such as phosphorus 32 (P-32). After the probes are attached, the nylon membrane is placed between two sheets of X-ray film and an autoradiograph (autorad) is made.

An autorad makes use of a standard sheet of X-ray film, but unlike the X-ray you get at the doctor’s office, no external X-rays are needed. The radioactive isotopes in the tagged probe constantly release radiation that will expose the film just as X-rays will. The exposed film now reveals the band pattern of the DNA sample. This produces the familiar DNA fingerprint pattern (see Figure 10-2). It’s like a personal bar code.

Figure 10-2: The DNA fingerprint. Electrophoresis separates DNA sample fragments according to size, which results in columns of bands that can be compared with other samples.

The autorad gives you a picture of the DNA fragment pattern at a particular locus. This alone will not give you conclusive match, since some people share similar patterns at a single locus. It may, however, exclude a suspect. When comparing a known and an unknown DNA sample, if any band in the RFLP autorad doesn’t match, the two samples in question did not come from the same person. To make a match, multiple loci must be examined. We’ll look more deeply into why multiple loci are needed later.

When the forensic lab must match one or more samples, the electrophoretic gel is divided into several parallel columns called lanes. A DNA sample is placed at the beginning of each lane. When the electrophoresis process begins, the fragments in each lane move and separate independent of each other. This causes each sample to separate into a series of bands determined by the size of the various fragments in each sample. A match is made when the bands in the known and unknown columns match. This comparison can be done visually or with the aid of a computer.

The samples placed at the head of each column vary. Some may be control materials, such as DNA from bacteria, viruses, or lab-synthesized DNA. These are DNA samples with fragments of known size and can be used to estimate the size of fragments in any unknown samples. More importantly, one column contains the crime scene specimen and other columns contain samples from any suspects.

Figure 10-3: DNA matching. This DNA fingerprint compares an “unknown” crime scene sample with “known” samples taken from suspect A and suspect B. It should be apparent that the crime scene DNA did not come from suspect A, but could have come from suspect B.

For example, let’s say that a perpetrator shed blood at the scene of a homicide and the crime lab found and collected it. This constitutes an “unknown” sample since its origin is not known. If the suspect list includes two people, the lab takes samples from each. These are “known” samples since their origin is known. DNA fingerprinting is then used to compare the suspects’ DNA with the “unknown” crime scene DNA (see Figure 10-3).

By comparing this one locus, suspect A can be excluded. This is not his DNA. A single mismatch at any locus is exclusionary. But what of suspect B? The pattern matches his DNA profile for that locus; therefore, the sample obtained at the crime scene could be his. Testing more loci will prove or disprove this.

But RFLP is on the way out as a DNA test and has yielded to the combination of polymerase chain reaction and short tandem repeat.

The polymerase chain reaction (PCR) arrived on the forensic scene in 1992. This technique allows for repeatedly copying the DNA in a sample so that a larger quantity of identical DNA can be made. This process is called amplification. It requires as little as a billionth of a gram of DNA material.

PCR takes advantage of the method by which double-stranded DNA replicates itself in nature. Let’s say that the crime scene DNA sample is a single hair follicle. Let’s represent the DNA extracted from the cells of the follicle as the following:

Obviously, the actual strand would be much longer than this, but we’ll use this shorter segment since, for our purposes, it is more easily visualized. PCR involves several steps: denaturing, annealing, extending, and repeating.

Denaturing is the separation of the double-stranded DNA into its two component strands. Before each strand can be copied, it must be separated from its mate. This is accomplished by heating the sample to 94ºC to 96ºC. The result looks like this:

Annealing is the process of “priming” the copying process. Basically it’s a jump-start. Short primer DNA sequences are added to the DNA sample and it is heated to 55ºC to 72ºC. This initiates, or primes, the duplication reaction. After attachment of the primer segments, the strands might look like this:

Extending is the completion of the duplication process. Each strand is induced to manufacture its complementary strand by use of a DNA polymerase enzyme. This is the same type of enzyme that occurs naturally in the body. As in the body, each DNA strand serves as the template for synthesizing its complementary strand, resulting in two identical double-stranded DNA molecules.

You can see that the original single molecule of double-stranded DNA is now two identical copies.

Repeating the above three steps over and over rapidly multiplies the number of strands available for testing: 2 become 4, then 8, then 16, 32, 64, … etc.

Very quickly the original DNA sample grows into a more usable amount. This allows for DNA testing of even extremely small samples. In many modern labs this process is automated and relatively quick.

This technique was introduced in 1994. Short tandem repeats (STRs) are repeating microsatellites of DNA that are most often only four base pairs long, though they can range from three to seven. Their short sequence, multiple polymorphic types, and frequent repetition make them highly discriminatory and useful when the DNA sample of partially degraded or fragmented.

The combination of PCR and STR has become the standard in most labs. The advantages are many. It requires much smaller samples and is faster and more reliable than RFLP. It can more easily be automated so that many samples can be done in a very short period of time. Using PCR and STR analysis allows samples to be analyzed in a couple of days as opposed to a month or more using RFLP.

The process of PCR-STR shares many of the steps we saw with RFLP. After the DNA has been extracted and fragmented, it is amplified by PCR in a thermal cycler, an instrument that varies the temperature throughout the repeating cycles. The fragments are then separated by gel or capillary electrophoresis, the latter making use of tiny capillary tubes. The electrophoretic device is attached to a computer, which analyzes the results and prints out the DNA profile.

Though the old bar code-looking profile can be generated by this method, the more automated STR analytic systems produce a printout that looks a bit different. Here the computer displays the STR peaks on a graph. The process of comparison is similar in that if all the peaks of the graph obtained from the analysis of two different DNA samples are identical, then the two samples share a common source (see Figure 10-4).

This process is almost fully automated now. Machines such as Applied Biosystems’ 3730 DNA Analyzer has a sample capacity of up to 384, making the analysis of many samples at once a reality.

Single nucleotide polymorphism (SNP) is a new technique that will likely see increased use in the future. The major problem at present is that it is expensive.

We saw that RFLP fragments were fairly long, a drawback that lessens their value in degraded or damaged samples (discussed later). This problem was circumvented by the discovery of STRs, which are very short fragments. But, what if the DNA examiner could use single nucleotide bases as the standard for matching? This would increase the discriminatory power of DNA even further. This is what SNP does.

If we searched for an “ATTA” STR repeat, these two strands would be indistinguishable since both have two ATTA repeats. But, with single nucleotide analysis the strands differ by a single base: The ninth base in the first sequence is guanine (G), while it is adenine (A) in the second one.

SNP can be used with restriction enzymes in the RFLP technique, or with PCR, where it can be easily automated. Theoretically, this will allow for discriminating two DNA samples based on a single nucleotide difference.

So, how do STRs and VNTRs work? We’ll use STRs in our example since they are shorter, more discriminatory, and have replaced VNTRs in most DNA labs.

By now, you know that we each have unique DNA and we receive our DNA from each of our parents. Remember that our chromosomes are paired, one coming from each parent. This means that our DNA, which makes up our chromosomes, is also paired. And since each of these DNA strands possesses STRs, we receive STRs from both parents. Earlier we saw that there were over eight trillion possible chromosomal combinations for the child of any two parents. The same goes for STR patterns.

This means that each of us will have a variable number of STRs in any given locus of our DNA. Since the number of any given STR at any given locus can be determined, and since the number of STRs at that locus varies from person to person, we can use these facts to determine if any two DNA samples share a common source. That is, did they come from the came person?

In addition, if we know how often a given number of STR repeats is found at a locus in the general population, we can use this information to calculate the odds that the two DNA samples came from the same person. This is similar to what we saw with ABO blood typing (see Chapter Nine) where blood type AB eliminated 97 percent of the population. A single locus of STR analysis can do the same.

But, how conclusive is a match from a single locus? Not very, but if the test is repeated from several locations, the odds add up quickly. Most labs use thirteen distinct loci in their analysis.

Let’s say we are dealing with a four-base-long STR such as CCTA. Let’s also say that by searching a certain locus on your DNA we find that you received six repeats of this particular STR from one parent (one DNA strand of the pair) and eleven repeats from the other (the other DNA strand of the pair). If we checked the same locus on my DNA, we might find that I received five repeats of this STR from one parent and twenty-one from another. Our DNA would be very different.

But would our DNA be different from everyone else’s on Earth? We couldn’t tell by looking at just this one locus. There may be other people who also received six and eleven, or five and twenty-one repeats. But what if we looked at a dozen loci? What are the odds that two people would have received the exact number of repeats from each parent at each of these loci? That would happen in only one of several hundred trillion conceptions. This means that no two people have the same pattern of STR repeats and thus no two people possess identical DNA.

Let’s look at another example. Say we analyze the STRs of a crime scene sample at five different loci and find the repeats at these loci as follows:

Now let’s say we know that the occurrence of each of these STR repeat patterns at these loci in the general population is 1 percent, 3 percent, 2 percent, 1 percent, and 2 percent, respectively. This means that one in one hundred people share this same repeat pattern at locus 1, three in one hundred share this same repeat pattern at locus 2, and so on. If a suspect’s DNA and DNA obtained at the crime scene show the exact same repeat patterns at all five loci, what are the odds that the DNA found at the scene came from someone other than the suspect? Since the inheritance of the STR patterns at each locus is independent of any other locus, the percentages (fractions) must be multiplied by each other. Like this:

1/100 x 3/100 x 2/100 x 1/100 x 2/100 = 12/10,000,000,000 or 12 out of 10 billion

This means that there are only twelve chances out of ten billion, or roughly one in a billion, that the DNA found at the crime scene came from someone other than the suspect. And this was using only five loci. As we will see later, the FBI database uses thirteen loci. Now, imagine if the suspect’s DNA matched the crime scene sample at thirteen loci. We would be looking at odds in the one per trillions.

Or put another way, if the STR count at all thirteen loci in a crime scene sample match the count at the same thirteen loci of the suspect sample, what are the odds that the crime scene sample came from someone other than the suspect? Astronomical would be the word.

So, DNA is a numbers game. The more loci used, the greater the odds that two matched samples share the same source.

Earlier, I mentioned degraded DNA. This is simply DNA that has been damaged and broken by heat, chemicals, decay, or some other process. The more degraded the sample, the more it is fragmented. Since DNA fingerprinting depends on counting the number of repeated sequences in a given locus of the DNA strand, if the DNA is already broken up, such a count becomes impossible. You can’t simply put the strand back together and then count. And indeed, severely degraded DNA, which has been broken into small fragments, is of little value. But what if it is only partially degraded and the surviving fragments are fairly long?

STR analysis can still be used in many such situations. Since STRs are much shorter than VNTRs and require less lengthy DNA segments for their location and counting, the likelihood that the pattern will be disrupted is much less when STRs are used. It is for this reason that STR analysis is becoming the norm for DNA fingerprinting, and why SNP analysis might soon become the standard.

Still, if the sample is severely degraded and the lab only has a pile of very short fragments or single bases to work with, no typing can be done. Not even STR. It would be like trying to read a book in which all the sentences had been reduced to fragments and single words. For Whom the Bell Tolls might be indistinguishable from The Cat in the Hat. However, if the book were only torn into chapters, we would have little trouble distinguishing between the two. A partially degraded DNA sample would be the latter situation, while a severely degraded sample would be the former.

But with good quality DNA samples, DNA typing is highly accurate. And when analyzed properly, its discriminatory power is absolute. It will not give false results. It will give either a match or no match, but it will not point the finger of suspicion in the wrong direction.

The first step in using DNA as a forensic tool is to locate the DNA. Without a usable sample, the crime lab will have nothing to work with, so a diligent search for DNA at the crime scene or on the victim or the suspect is critical.

DNA can be found in virtually every tissue and fluid in the human body, many of which are shed at crime scenes. Blood is the most common biological material encountered, but semen, saliva, tears, urine, bone, teeth, hair, and skin are often found at the scene, each of which can yield enough usable DNA for testing using modern techniques. Let’s look at these common sources in more detail.

TISSUES: The cells of skin and other tissues contain DNA within their nuclei.

BLOOD: The red blood cells of the blood have no nuclei, so they have no DNA, but the white blood cells do. When the lab extracts DNA from blood, it is the white blood cell DNA that is isolated for testing.

SEMEN: Semen has DNA within the spermatozoa. But if the person is azoospermic (produces no sperm) or has had a vasectomy—no sperm, no DNA. The epithelial cells that line the urethra do contain DNA. The urethra is the channel that connects the bladder to the outside; as the ejaculate moves along the urethra, it collects some of the urethral cells. The DNA in these cells can often be used to develop a DNA fingerprint.

SALIVA: Saliva itself contains no cells, but it collects the DNA-containing epithelial cells of the salivary ducts as it passes from the salivary glands to the mouth.

TEARS: Like saliva, tears contain no cells, but the epithelial cells that line the tear ducts do. These cells are carried out with the tears and can be a source of DNA.

HAIR: Hair itself contains no nuclear DNA, but the follicle cells do. Hair that has been cut or has fallen out naturally does not typically have follicular material attached and is not likely to possess nuclear DNA. But hair that has been yanked out often carries follicular material with it, and this can serve as a source for nuclear DNA. Still, the hair shaft itself contains a useful, special type of DNA called mitochondrial DNA, which we’ll look at shortly.

BONE: Bones have cells called osteocytes that contain DNA. DNA can be extracted from bones, sometimes even from those that are thousands of years old.

TEETH: Teeth are very hardy and are the last part of the body to dissolve away. The enamel is hard and contains no cells, but the pulp does. These pulp cells can survive for a very long time under some fairly adverse conditions. Drilling into the teeth of even very old skeletal remains can sometimes yield usable DNA.

Since DNA resides within biological materials, it is subject to the same putrefaction process that eventually destroys all human tissues. Since bacterial growth and putrefaction progresses more rapidly in warm moist environments, the best DNA samples are those that have been adequately dried and stored in a protective container. If drying is not feasible, wet samples are frozen until analyzed. If not properly collected and protected, DNA can degrade and be unusable.

How much DNA is needed? The simple answer is the more the better. However, with the use of the PCR technique, even very small samples can yield enough DNA for typing and matching. Usable DNA can come from a single hair with a follicle from an old hairbrush; a single tear or drop of blood; saliva in a bite mark or on a toothbrush, postage stamp or envelope, food, soda cans, telephones, pens and pencils, the face-side of the perpetrator’s mask; or even a tooth from a one thousand-year-old mummy. A January 2004 article in the Journal of Forensic Science suggested that human DNA could be extracted from maggots found on a decaying corpse up to four months after death.

The case of the famous Green River Killer shows how very small and very old DNA samples can be useful.

FORENSIC CASE FILES: THE GREEN RIVER KILLER

One of the most notorious and frightening serial killers in history was known as the Green River Killer. The moniker arose because the killer dumped his victims along the Green River near Seattle, Washington. Between 1982 and 1991, nearly fifty murders were attributed to the Green River Killer. The suspect list developed by the task force assigned to the cases was nearly as long.

In April 1987, police executed a search warrant on the premises of one of the suspects, Gary Ridgway. After obtaining evidence items from his house, they requested that he undergo a polygraph, but Ridgway refused. They then asked for a saliva sample and Ridgway complied by biting on a small square of surgical gauze. Unfortunately, the semen samples taken from many of the victims were too small for current testing procedures, so the samples, as well as Ridgway’s saliva, were stored. In the mid-1990s, the combination of STR and PCR analysis appeared.

Then in 2001, the lab tested Ridgway’s saliva sample obtained in 1987 along with semen samples taken from Opal Mills, Marcia Chapman, Cynthia Hinds, and Carol Christensen, all killed in 1982 or 1983. Using the new techniques of PCR and STR, the samples were amplified and compared. A match was made and Gary Ridgway was arrested and charged with four of the Green River killings. However, this case took a dramatic and controversial turn on November 5, 2003, when Ridgway plead guilty to forty-eight murders in exchange for a sentence of life without the possibility of parole, thus sparing himself a possible death sentence.

This case shows that if DNA samples are properly collected and stored, they can remain useful for decades.

Even though this absolute individuality is well established, some clever criminals still attempt to deny this fact in the hopes of winning an acquittal or overturning a conviction. Anthony Harold Turner of Milwaukee is such an example.

FORENSIC CASE FILES: THE FAKE RAPE

In 1999, Anthony Harold Turner was convicted of rape after DNA obtained from three victims matched his DNA with a probability of three trillion to one. Turner was somewhat of a self-educated DNA expert, and denied that the DNA was his. He stated that it must have come from someone with the exact same DNA. Since Turner did not have a twin brother, he was convicted. But, as he was awaiting sentencing a woman came forward saying that she had been raped. Imagine the prosecutors’ surprise when the DNA obtained from this victim also matched Turner, who was safely tucked away in jail. How could this be?

It turned out that some members of Turner’s family had paid the woman fifty dollars to claim that she had been raped. Where did the semen used to stage the fake rape come from? Turner managed to smuggle it from jail in a small ketchup packet.

But let’s confuse the situation a bit further. There are some people walking around with another person’s DNA in their blood. This was pointed out in a case worked by Abirami Chidambaram of the Alaska State Scientific Crime Detection Laboratory in Anchorage. Semen obtained from a rape victim matched that of a man who was in jail on another charge at the time of the alleged rape. Further investigation revealed that years earlier the man had received a bone marrow transplant from his brother. This meant that he and his brother, who was not an identical twin, now shared the same DNA in their blood. This also meant that the finger of suspicion for the rape was now directed at the brother. How is it possible for these two nonidentical twin brothers to share the same DNA in their blood cells?

Bone marrow transplants are typically done in patients’ suffering from certain types of leukemia or some other blood disease. The patient is given chemotherapeutic agents that kill off all his native bone marrow cells and then bone marrow from a compatible donor is infused into the patient’s vein. The bone marrow material migrates to the patient’s bone marrow, sets up housekeeping, and begins cranking out blood cells. This means that the circulating blood cells now have the DNA of the donor’s marrow and not that of the patient. DNA testing of the blood will thus match both the bone marrow donor and recipient, a situation that is seen naturally only in identical twins.

How can the forensic DNA examiner get around this? Test other cells from the patient. Buccal cells, or cells from any other tissue in his body, will reflect his native DNA and will not match the DNA profile of his own blood. A bone marrow transplant does not change the DNA in all the recipient’s cells, only those of his bone marrow and blood.

To confuse matters even further, more recently bone marrow transplants are also done in patients in whom their native bone marrow is not completely destroyed by chemotherapy before the infusion. This means that their bone marrow will be a combination of their own and that of the donor and their blood DNA will reflect this combination of native DNA and donor DNA.

Another confusing situation arises from a rare genetic condition called chimerism. In Greek mythology, the Chimera was composed of parts from various animals. Descriptions of this creature vary but an example would be one with a lion’s head, goat’s body, and a snake’s tail. In humans, a chimera results from the abnormal combination of two or more fertilized (and at times non-fertilized) eggs. Let’s look a bit of basic genetics.

Fraternal twins come from two separate eggs and sperm cells. They are as different as they would be if they had been born years apart. They are twins only because they shared the same womb at the same time. Identical twins come from a single egg and sperm. After the egg is fertilized, it begins to divide to produce more identical cells. After the first division, if the two daughter cells pull apart and then each goes on to develop a separate fetus, the two fetuses will have the exact same DNA and will thus be identical twins.

A chimera is formed when two fertilized eggs (each egg different and each fertilized by a different sperm cell as in fraternal twins) join together and go on to produce a single fetus. Here, since the fetus is the result of two eggs and two sperm cells, the child will have two different types of DNA. It’s as if two fraternal twins were blended together into one person, which is essentially what happens. As you might guess, the chimeric individual would have two distinct DNA patterns. This could greatly confuse DNA testing.

ABO blood typing can be used to exclude paternity, but cannot absolutely state that the man in question is the father of the child. To establish paternity, DNA is used.

The first step in this process is to profile (fingerprint) the DNA of the mother and the child. These are then compared to the profile of the suspected father. We receive all of our DNA from our parents. We get no DNA from any other source, so the child’s DNA pattern should be a combination of those of the mother’s and the father’s. This doesn’t mean the child will have every fingerprint band that each parent possesses, but it does mean that the child cannot have a band that neither parent has. Where would it come from?

In paternity testing, if the child possesses a DNA fragment that is not present in either the mother or the suspected father, then the man is not the child’s parent. This fragment must have come from someone else (the real father) and paternity for the suspect father is excluded (see Figure 10-5).

Figure 10-5: DNA paternity testing. Every band in the child’s DNA fingerprint must match a band from one or the other parent. If the child possesses bands that did not come from either the mother or the suspect father, the suspect father cannot be the child’s parent and paternity is excluded. In this example, the child possesses a band that is not found in either the mother or the suspect father, so the suspect did not father the child.

FORENSIC CASE FILES: THE IAN SIMMS CASE

In this unusual case, paternity testing was used to obtain a criminal conviction. On February 9, 1988, Helen McCourt, a twenty-two-year-old insurance clerk in the small village of Billinge in northwest England, disappeared. She had stopped at the George and Dragon, a local pub owned by Ian Simms, to whom she may have had a romantic link. Witnesses reported hearing screams from the pub and when police confronted Simms he had several scratches on his face.

The police found a great deal of evidence: hair matching McCourt’s and one of her bloody earrings in Simms’s car; bloodstains on the stairway and the bedroom floor of Simms’s home; and, in various areas of the county, McCourt’s bloodstained coat and clothing, which also held hairs from Simms’s dog and fibers from his carpet, and a length of electrical cord, which bore strands of McCourt’s hair. But, McCourt’s body was never found.

The problem the police faced was proving that the blood in Simms’s apartment was McCourt’s. With no body from which to obtain a blood sample, they turned to the girl’s parents. Samples taken from her mother and father were matched against the blood found at Simms’s home. At trial, Dr. Alec Jeffreys, the father of DNA fingerprinting, testified that the odds against the blood from Simms’ apartment being that of anyone other than that of the daughter of the two parents was 14,500 to 1. Simms was convicted in 1989 and received a life sentence.

So far, we have discussed the uses of nuclear DNA—the DNA that resides within the nuclei of our cells—but the body has other, non-nuclear forms of DNA that are extremely useful to the forensic scientist. Two of these are mitochondrial DNA (mtDNA) and Y-chromosomal DNA.

The discovery of mitochondrial DNA added an extremely useful tool to the forensic toolbox. It is useful for identification of perpetrators and human remains as well as for determination of ancestry. The first admittance of mtDNA evidence into a U.S. court occurred in Tennessee v. Wade in 1996.

Non-nuclear DNA is found within the mitochondria, small organelles that reside within the cytoplasm of the cell and serve as the cell’s energy production center. A small amount of DNA is found within the mitochondria, but each cell has many mitochondria.

Mitochondrial DNA has several characteristics that make it unique. It is passed from generation to generation by the maternal linage, mutates rarely, is found in places where nuclear DNA doesn’t exist, and is exceptionally hardy.

Your mtDNA is inherited unchanged from your mother and only from your mother. And she received hers from her mother, and her mother from her mother, and so on. Why is this? At fertilization, the egg supplies the cell and half the DNA while the sperm supplies only half the DNA (see Figure 10-6). The sperm cell itself breaks down and disappears after passing its genetic material into the nucleus of the egg cell. This means that the actual cell and all the cell components (including the mitochondria) of the developing zygote come from the mother. As the cell divides and multiplies, these mitochondria are copied and passed on, generation after generation. This means that all the cells of the body contain identical mtDNA.

Figure 10-6: The fertilization process. At fertilization, the egg supplies half of the DNA, the cell, and all the cellular structures to the fertilized zygote, while the sperm supplies only half the DNA.

Since mtDNA only undergoes a significant mutation approximately once every 6,500 years, it is unchanged over many generations. This means that your mtDNA is virtually identical to your mother’s, your great-great-grandmother’s, and your maternal ancestors’ from one thousand years ago. Thus, anyone’s maternal lineage can be accurately traced over many generations and this fact can be used to prove if two people share the same maternal linage. This became important in the famous Boston Strangler case.

FORENSIC CASE FILES: THE BOSTON STRANGLER

Albert DeSalvo was convicted of the series of rapes and murders attributed to a killer known as the Boston Strangler. He confessed to many of the murders, but frequently got some of the details wrong, raising suspicion that perhaps he did not commit all the crimes he took credit for. Recently, forensic science attempted to solve this mystery.

The Strangler’s last victim was Mary Sullivan. In October 2000, thirty-six years after her death, the exhumation of her body took place in Hyannis, Massachusetts. Investigators found a semen stain on her body. Tests for the presence of spermatozoa and for PSA were not possible due to degradation of the sample. However, the material did reveal the presence of mtDNA.

With DeSalvo dead and buried, blood was obtained from his brother Richard. Since the brothers should have identical mitochondrial DNA, matching the semen stain’s mtDNA to Richard’s would prove that Albert was indeed the killer. The result? No match. This means that even though Albert DeSalvo confessed to killing Mary Sullivan, he did not. Did he also confess to other murders that he did not commit? Was he truly the Boston Strangler? We may never know for sure, but at least in Mary Sullivan’s case, he was innocent.

This case also underscores two other advantages of mtDNA. First, it is very hardy and can often be extracted from older tissues and the bones and teeth of very old skeletons. Secondly, it is found in some tissues where nuclear DNA is not.

I mentioned earlier that hair is predominantly composed of dead cellular debris, with the only living part of hair being the follicle. The cells of the follicles contain nuclear DNA, which can be used for DNA profiling. The dead cells of the hair have no nuclei, and thus no DNA. This means that hair that has been yanked out or shed with a follicular bulb attached can provide nuclear DNA, while hair that has been cut or has no bulb attached will not. But, all is not lost.

In the growth of hair, the cells of the bulb multiply, undergo change, and become incorporated into the growing hair. Part of this transformation is the loss of the nucleus from each cell. Thus, hair has no nuclear DNA. But the dead cellular debris that is incorporated into the hair shaft might contain mtDNA. If so, it can be extracted and used to identify the person who shed the hair.

The Y-chromosome is what makes males males. It is found only in males; a father passes his Y-chromosome on to his male offspring. So, Y-chromosomal DNA is passed down the paternal linage. By testing for STR repeats on the Y-chromosome (Y-STR), it is possible to show that two or more men share a common paternal ancestry. As with mtDNA, this can be useful in genealogy and in identifying corpses and suspects. It has been shown that this technique can connect two males through their paternal genealogy across as many as thirty generations.

CODIS stands for Combined DNA Index System. It is a database of DNA finger-prints taken from felons and from biological fluids obtained at crime scenes such as assaults, homicides, and rapes. It began as a pilot project in 1990, and then in 1994 the DNA Identification Act authorized the FBI to set up the National DNA Index System (NDIS), which became operational in 1998. This allows any CODIS-participating lab to compare DNA samples nationally.

For example, the DNA fingerprint of a crime scene or suspect sample can be plugged into the CODIS database and a computer compares it to all the profiles in the system. If a match is made to a particular individual, fresh suspect samples are taken and repeat testing in the crime lab is done to confirm or refute the presumptive match. Alternatively, the DNA sample could match DNA obtained from another crime scene; in this case a match would serve to link the crimes. This way evidence from two or more scenes can be considered together. This alone may lead to the identity of the perpetrator.

In 1997, in an attempt to universalize the DNA testing system, the FBI selected thirteen STR loci as the core of their database. Laboratories that use the CODIS system examine these thirteen core loci as part of their DNA analysis, making comparing samples in the CODIS database uniform.

CODIS has already enjoyed many successes. As of September 2003, the system had registered nearly nine thousand hits. One such case was that of Norman Jimmerman.

FORENSIC CASE FILES: THE NORMAN JIMMERMAN CASE

In March 1989, Debbie Smith was forced from her home in Williamsburg, Virginia, and into a nearby wooded area where she was raped. Her attacker warned her that he knew where she lived and that she should not tell anyone or he would return and kill her. Debbie went to the police. Blood from the prime suspect was tested against seminal DNA obtained from the victim’s rape examination. No match. In 1994, Smith’s neighborhood suffered a series of sexual assaults. Another suspect attracted the attention of police, and again DNA analysis was undertaken. Again, no match.

Meanwhile, the state of Virginia began the process of developing a databank of DNA profiles taken from convicted felons. As new profiles were obtained, the Virginia Department of Forensic Services periodically matched these against unsolved crimes. One of these matches identified Norman Jimmerman as Smith’s attacker. He was already serving time for robbery and abduction. His current sentence is 161 years.

FORENSIC CASE FILES: SNOWBALL THE CAT

In 1994, Shirley Duguay of Prince Edward Island disappeared. A few days later her corpse was discovered in a shallow grave along with a leather jacket, which was soaked with her blood and dotted with white cat hairs. Her estranged husband, Douglas Beamish, owned a white cat named Snowball. DNA in blood taken from Snowball matched that of the cat hairs found at the burial site, proving that those hairs came from Snowball and no other white cat. Beamish was convicted, marking this case the first time that animal DNA was used to gain a conviction.

FORENSIC CASE FILES: THE SCHMIDT HIV CASE

In 1994, Dr. Richard Schmidt injected his girlfriend with blood taken from one of his AIDS-infected patients. Six months later, she was diagnosed with HIV and went to the police. Sifting through Schmidt’s records, investigators discovered that he had drawn blood from one of his AIDS patients on the same night that he had injected the victim. This was critical since the virus can only survive a few hours outside the human body.

The problem facing the crime lab was that HIV mutates often, so making a match between the virus taken from the unsuspecting source patient with that found in the victim could be problematic. Either could have mutated enough so that no conclusive match could be made. To get around this, samples were taken from thirty-two other HIV-positive individuals in the area. Testing revealed that the samples taken from the patient source and the victim matched almost exactly, while the others did not. Schmidt was convicted of second-degree attempted murder and sentenced to fifty years. This case marked the first time viral DNA analysis was used to convict a felon.