|
[ART: Centrifuge.jpg in artwork folder]
@head:Forensic Analysis of Blood Protein and DNA: A Brief History
@byline:By Phillip Jones, PhD, JD
@body:According to the Locard Principle of Exchange, when two objects come in
contact, traces from one will be transferred to the other. A physical encounter
between two people can result in the transfers of hair and fibers. In violent
crimes, a criminal may leave the scene with traces of the victim's blood,
compelling biological evidence of guilt.
For more than 100 years, forensic investigators looked for the right combination
of blood factors to link a bloodstain to an individual-a blood fingerprint. This
ambition is approaching realization, but in a way that the original
investigators could not have contemplated. But before forensic scientists could
begin to consider analyzing a bloodstain to identify who had shed it, they had
to contend with more basic questions. Is a stain a bloodstain? And if so, then
is it human blood?
@sub:Is it Blood?
@body:How can an investigator tell if a stain is blood? Color is a poor
indicator. As it dries and ages, blood loses its typical appearance: red
transforms to brown and then to greenish-yellow colors. By the late 19th
century, scientists had devised tests for blood with chemical indicators, but
these methods had several drawbacks.
Sometime during the early 1880s, Dr. John Watson met Sherlock Holmes in a
chemical laboratory of a London teaching hospital. In Arthur Conan Doyle's "A
Study in Scarlet," the great detective lectured the doctor on the problem of
stains found on a criminal suspect. Are the stains mud, blood, rust, or fruit
stains? This is a question, he said, that puzzled many experts because there was
no reliable blood test. The detective dismissed the "old guaiacum test" as
uncertain and inferior to the new Sherlock Holmes blood detection method.
In this passage, Holmes criticized a colormetric test that depended upon the
ability of hemoglobin to oxidize a West Indian shrub extract to a blue color.
Like many simple blood tests, the guaiacum test is not specific for blood and
can produce a false positive result. For this reason, simple blood detection
methods are considered presumptive tests. A currently popular test, for example,
uses phenolphthalein as a color indicator. When a bloodstain, phenolphthalein
reagent, and hydrogen peroxide are mixed, the hemoglobin triggers the formation
of a pink color. However, a similar reaction is observed if constituents of
potatoes or horseradish are present. Even the luminol test, the blue
luminescence method popular in TV shows, produces false positives.
During an 1878 trial, Dr. William Hodgson Ellis, assistant professor of
chemistry at King's College (now, the University of Toronto), testified about
his analysis of stains found on the defendant's recently washed trousers. The
results of Dr. Ellis' presumptive blood test showed that the stains were blood,
but he warned that he could not say that the bloodstains were human blood.
This limitation presented a significant dilemma. Suspects found with fresh
bloodstains on their persons could always maintain that the stains came from
slaughtering an animal or handling meat. Police could not refute these claims if
there were no way to distinguish between animal and human blood. A solution to
this problem arose from early studies in the field of immunology.
@sub:Is it Human Blood?
@body:At the end of the 19th century, Emil von Behring discovered that animals
inoculated with diphtheria toxin formed defensive substances against diphtheria
in their serum. When other scientists tried to devise serums against a variety
of infectious diseases, they made a significant observation. If an animal was
injected with a foreign substance, the animal produced defensive substances
specific for the foreign material. These defensive substances, termed
"precipitins," could be used to distinguish different types of protein.
By the summer of 1900, Paul Uhlenhuth, assistant professor at the Institute of
Hygiene at the University of Greifswald, had performed experiments in which he
injected hen's blood into rabbits. He found that the serum of injected rabbits
precipitated the protein in hen's blood, but caused no reaction with blood from
cows, horses, sheep or pigs. Uhlenhuth concluded that the blood of different
animals had one or more characteristic proteins. Eventually, he proved that he
could use serums from various animals to distinguish human from animal blood.
Otto Beumer, the coroner of Greifswald and professor of forensic medicine at the
University of Greifswald, learned about Uhlenhuth's experiments. Working
together, the scientists showed that animal serums could detect human blood in
dried bloodstains that were months or years old. Their discovery was soon put to
a practical use.
In July 1901, two children were murdered on the Baltic Island of Rügen. A fruit
peddler told the police that a journeyman carpenter named Ludwig Tessnow had
been talking with the children on the day that they were reported missing. The
police arrested Tessnow.
An examination of the suspect's wardrobe revealed that some items contained
barely-dried spots. Tessnow claimed that the spots were from cattle blood or
from a wood stain that he worked with. The mention of a wood stain reminded the
Greifswald magistrate of another double murder that had taken place three years
earlier. In that case, the police had identified a suspect, a journeyman
carpenter who had numerous stains on his clothing. The police had accepted the
suspect's explanation that the spatters were wood stain, and he was released.
The suspect had been Tessnow.
The police sent Uhlenhuth two packages of Tessnow's clothing. After examining
nearly 100 stains with presumptive blood tests, Uhlenhuth used his precipitin
test and identified human bloodstains on Tessnow's jacket, pants, vest, hat and
shirt. By the time that Tessnow was executed in Greifswald prison in 1904, the
precipitin test had become a prime tool of forensic science.
After a stain is identified as human blood, is there a way to link that blood to
a particular individual? At the time of the Tessnow case, another major medical
discovery sparked the hope that a bloodstain could be used to identify the
person who had shed the blood.
@sub:Whose Blood?
@body:The therapeutic method of blood transfusion was often successful; but
sometimes, the patient died. Several hundred years of studies revealed the cause
of failure and provided a significant forensic technique.
During the late 1660s, Jean-Baptiste Denys, court physician to Louis XIV,
received a patient, a feverish and weak boy who had been bled by other doctors.
Deciding that blood loss was responsible for the boy's weakened condition, Denys
transfused about nine ounces of a lamb's blood into the vein of the patient. The
boy survived. Denys continued to perform animal-to-human transfusions until he
experienced a fatality. Transfusions involving humans were soon outlawed in
large parts of Europe.
About 150 years later, Dr. James Blundell, a London obstetrician, began
experiments with blood transfusion. He found that he could drain almost all of
the blood from a dog and then revive it by transfusing the blood of another dog.
On the other hand, the dog would die if Blundell attempted to revive it with
sheep's blood. When Blundell began transfusing humans with human blood, he found
that his patients occasionally died. Why should transfusion fail with human
blood?
The answer started to become clear in 1875 when German physiologist Leonard
Landois noticed that an animal's red blood cells would aggregate and burst if
mixed with the serum of an animal from another species. Significantly, the same
reaction could occur when blood samples from two people were mixed. This
suggested that there were different types of human blood.
In 1900, Karl Landsteiner, assistant professor at the Institute of Pathology and
Anatomy in Vienna, took blood samples from six people, centrifuged the samples
to separate serum from red blood cells, and then mixed red blood cells and serum
samples from different people. In certain mixtures, the serum seemed to attract
the red blood cells, causing the cells to clump together, whereas this reaction
did not occur in other mixtures. He labeled the two blood types A and B. Soon,
he found a third blood type that showed characteristics of both A and B, and he
called this C (later known as type O). A year later, an assistant discovered yet
another type of serum that did not cause aggregation of either type A or type B
blood. This one was called AB, the fourth major blood group.
We know now that, in the ABO system, blood type is determined by the presence or
absence of A antigens and B antigens on red blood cells. If an individual is
type A, then that person's red blood cells have A antigens located on the
surface. Similarly, a type B person has red blood cells with B antigens, whereas
a type AB person has red blood cells with both A and B antigens. Type O persons
have neither A nor B antigens on the red blood cells.
These antigens can stimulate an immune response, including production of
antibodies that bind the antigens. For example, an anti-A antibody binds with
red blood cells that carry A antigens. Since an antibody can bind at least two
antigens, an anti-A antibody can bind with A antigens on two red blood cells. A
collection of anti-A antibodies can create a network of cross-linked red blood
cells. The result: red blood cells aggregate.
The fatal consequence of transfusing incompatible blood is based on this
interaction between a person's antibodies and antigens in the transfused blood.
As an illustration, the transfusion of type A blood into a type O patient can
result in the binding of the patient's anti-A antibodies to the blood cells,
causing the cells to stick together. Transfusion reaction symptoms include red
blood cell lysis, kidney damage, shock, and blood coagulation.
Blood typing takes advantage of these antibody-antigen reactions. In one
approach, blood is tested with anti-A antibodies and anti-B antibodies. Anti-A
antibodies cause clumping, or "agglutination," of type A red blood cells, anti-B
antibodies agglutinate type B blood, both types of antibodies agglutinate AB
blood, and neither type of antibody will agglutinate type O blood.
Other scientists built upon Lansteiner's ABO discovery. In the mid-1920s, the
Japanese scientist, Saburo Sirai, found that blood was not required to determine
blood type: many people secrete blood group-specific antigens into other bodily
fluids. About 80 percent of the human population are secretors, meaning that the
specific types of antigens, antibodies, and enzymes characteristic of their
blood can be found in other bodily fluids and tissues. Investigators can
determine a secretor's blood type by examining saliva, teardrops, skin tissue,
urine or semen.
A blood sample can also reveal gender. Humans have 46 chromosomes in their body
cells; 23 are derived from the mother's egg and 23 from the father's sperm. In a
male, the body cells contain 44 autosomes, an X chromosome, and a Y chromosome.
A female's body cells have 44 autosomes and two X chromosomes. But in early
embryonic development, one of the X chromosome pair is inactivated, and becomes
shortened and condensed. In 1949, two British scientists observed these
condensed chromosomes in cell nuclei of female tissue. This chromosomal
structure, termed the Barr body, is most noticeable in a female's white blood
cells, and is rare in males.
By the end of the 1950s and 1960s, scientists had identified many types of blood
antigens. Yet blood characterization was still more useful to show who had not
shed the blood, rather than point out who had shed the blood. Forensic
serologists lacked sufficient tools to create the elusive blood fingerprint.
There was a glimmer of hope in 1967 when Brian Culliford of the British
Metropolitan Police Laboratory found that he could detect the enzyme
phosphoglucomutase (PGM) in dried bloodstains. This enzyme is polymorphic-that
is, the enzyme is found in multiple, distinct forms. The PGM polymorphic forms
are inherited and occur in the population in frequencies useful to the
investigator.
Here's how PGM characterization aids an investigation. Consider a bloodstain
that contains the PGM form PGM2-1 and that is type A blood. In the United
States, 42 percent of the population have type A blood. Regardless of blood
type, PGM2-1 is found in 36 percent of the population. This means that PGM2-1 is
present in type A blood in 15 percent (0.36 x 0.42) of the population. Although
blood samples from different people might have the same blood type or blood
enzyme, it is less probable that two unrelated people have the same blood type
and same form of blood enzyme.
A 1983 investigation illustrates the value of bloodstain analysis. On Oct. 24 of
that year, workmen arrived at the Laitner home located in a fashionable suburb
of Sheffield, England. They found the Laitner's 18-year old daughter in a
bloodstained nightgown. She told them that she had been raped and that her
parents and brother had been murdered. Alfred Faragher, who was in charge of a
local serology lab, had a problem: too much blood. But on the daughter's bed, he
did find a bloodstain that appeared about knee level of a person lying in bed.
Since the daughter had no cuts on her knees, the investigator decided that the
blood had come from the killer.
Faragher's team analyzed the bloodstain and found a combination of factors that
should occur in only one person in 50,000. Fortuitously, Faragher recognized the
combination; he had seen it in the blood of a rape suspect that had been sent to
him one month earlier. While in custody, the police had taken a blood sample
from the suspect, Arthur Hutchinson. Faragher's comparison of Hutchinson's blood
with the bloodstain found in the bed strongly indicated that Hutchinson had
killed the Laitners.
Unfortunately, Hutchinson had escaped. He had jumped out of a second story
window of the police station and had climbed a 12-foot wall topped with barbed
wire that had torn open his leg. To flush out Hutchinson, the police spread
reports that the barbed wire had been specially treated and was likely to turn
the suspect's leg gangrenous. The police captured Hutchinson as he made his way
to a hospital. Serological detection, coupled with clever police work, had led
to the identity of an unknown killer.
By the 1980s, about 100 protein polymorphisms had been identified; blood
analysis had come a long way in the century since the famous meeting of Holmes
and Watson. And in the early 1980s, the quest for a technique that could match a
person to a bloodstain took a large step closer to reality.
@sub:The Value of Junk DNA
@body:In 1984, Dr. (now Sir) Alec Jeffreys, a research fellow at Leicester
University's Lister Institute, discovered highly repetitive sequences in the
genetic code that could be used to identify an individual. Forensic science has
not been the same since. Before looking at Jeffreys' contribution, let's take a
quick hitchhiker's guide to genetics.
A DNA molecule is a polymer made of linked nucleotides. Nucleotides are composed
of a sugar molecule, a phosphorus-containing group and a nitrogen-containing
molecule called a base. The four types of bases, adenine, cytosine, guanine, and
thymine, are usually designated by the first letter of their names. If you
remember the science fiction film "Gattaca," you know the building blocks of the
genetic code.
Two DNA molecules coil into a double helix with paired bases from each molecule
providing the steps of this spiral staircase. The average human chromosome has a
double strand of DNA containing 100 million base pairs, and all of the human
chromosomes taken together contain about three billion base pairs. This genetic
information encodes every protein made by the body. However, not all nucleotide
sequences code for the production of proteins -- at least 90 percent of DNA is
considered "junk DNA" because it lacks a known function. This junk DNA contains
nucleotide sequences that are repeated numerous times, and although the
significance of the repeats is unclear, forensic investigators find them useful.
In 1980, scientists discovered hypervariable regions of DNA. These regions
consist of short tandem sequences repeated over and over again, and show extreme
variation between individuals. Jeffreys found that the repeats contain core
nucleotide sequences of 10 to 15 bases. He isolated several core sequences and
used them as probes to detect hypervariable regions in DNA samples that he
obtained from members of a family.
First, Jeffreys treated the DNA samples with restriction enzymes, which
recognize certain nucleotide sequences and cut DNA to produce DNA fragments of
various lengths. Jeffreys then placed cleaved DNA on gels and applied a high
voltage electric current to sort DNA fragments by size. Jeffreys transferred the
DNA fragments from gels to nylon membranes and treated the membranes with a
radioactive marker that binds with target nucleotide sequences. After the nylon
sheets were placed against X-ray sensitive film and the X-ray film was
developed, the DNA fragments carrying the radioactive markers appeared in a
series of bars that look like bar codes. When he compared X-ray film from
parents and children, he found that these bar codes varied between individuals,
and that the patterns were inherited. Jeffreys soon applied his technique in the
forensic science arena.
In the early morning of Nov. 22, 1983, a hospital porter walked along a footpath
on his way to work in the village of Narborough, near Leicester. He discovered
the body of a 15-year-old girl who had been raped and killed. Semen analysis
indicated that the killer was a Type A secretor with the enzyme marker, PGM1.
The two factors occurred together in ten percent of the adult male population,
which helped the police to narrow their suspects. But the investigation led
nowhere.
Three years later, the body of another 15-year-old girl was found within a mile
of the first victim. Semen tests suggested that this was the same killer. A
teenage boy fell under suspicion even though a blood test showed that he was not
a Type A-PGM1 secretor. The authorities contacted Jeffreys, who extracted DNA
from the killer's semen and compared it with DNA obtained from the suspect's
blood. The samples did not match. On the other hand, Jeffreys did confirm that
one man was responsible for both murders.
The police decided that if Jeffrey's DNA typing technique could clear a suspect,
then it might also identify the killer if blood samples were analyzed from the
local male population. Hoping to flush out the killer, they drew blood from
every local male between the ages of 16 and 34 for DNA testing. It worked.
In a Leicester pub, an eavesdropper heard a group of bakery workers discussing
the sexual adventures of an acquaintance named Colin Pitchfork. One of the group
mentioned that Pitchfork had forced him to take the blood test on his behalf.
The eavesdropper passed this information to the police who arrested Pitchfork.
Jeffreys tested Pitchfork's blood and showed that the suspect's DNA was
identical to the DNA of the killer-rapist. Pitchfork was the 4,583rd male to be
tested.
The Pitchfork case garnered international attention and DNA typing was hailed as
the most significant development in forensic science since fingerprinting. A
modification of Jeffreys' method became the first accepted protocol in the
United States for forensic characterization of DNA. The technique captured the
public's imagination after an FBI laboratory used it to analyze stains on Monica
Lewinsky's infamous blue dress.
The latest DNA typing technique relies on short tandem repeat (STR) analysis.
STR refers to locations on chromosomes that contain short repeating sequences of
three to seven nucleotides. Human chromosomes have hundreds of different types
of STRs. On each chromosome, the number of repeats varies greatly from person to
person. The probability of two random DNA samples having the same repeat pattern
at a single location in one chromosome is small. The probabilities become minute
after combining results of STR analysis from multiple chromosomal locations.
DNA typing is often called "DNA fingerprinting," but this term is misleading.
Fingerprints are considered unique to an individual, whereas the results of DNA
typing are expressed as probabilities. A DNA profile probability indicates the
probability that a person chosen at random from a certain population has the DNA
profile of the sample obtained from the crime scene.
The FBI has selected 13 STRs to serve as a standard battery of repeated
sequences for DNA typing. Suppose a DNA sample is obtained from a suspect and
analyzed for all 13 STRs to obtain a DNA profile. The probability that DNA from
an unrelated person will provide the same profile as the suspect's DNA is less
than one in 10 billion. Considering that the world's population runs just over
six billion, these results are fairly close to an identification.
@sub:The Future: Digital Detection
@body:Where might things go from here? The U.S. Department of Justice sees a
trend toward enhanced nucleotide sequence automation and miniaturization,
leading to transportable devices for DNA analysis. Miniaturized, handheld chips
may be in use by 2010, the Justice Department predicts. These improvements,
coupled with advances in communications technology, would allow investigators to
test DNA samples at the crime scene with remote links to databases.
Speaking of databases, the U.S. forensic science community has standardized the
13 STR series for entry into the FBI's national database, the Combined Index
System (CODIS). This system allows forensic labs to store and match DNA records
from convicted offenders and crime scene evidence. By June 2003, CODIS had
produced over 7,800 hits for more than 8,600 investigations.
The Forensic Science Service, an executive agency of the United Kingdom Home
Office, runs the National DNA Database. On July 15, 2003, Home Office Minister
Hazel Blears loaded the two millionth DNA profile into the database. The British
system uses eight of the 10 STRs stored in CODIS, enabling comparisons between
the U.K. and U.S. DNA databases. The capability for international comparisons of
DNA profiles will undoubtedly increase in the future.
Today, forensic scientists can copy a DNA fragment one million times or more in
just a few hours. This means that as little as one-billionth of a gram of DNA
can be used for analysis. Investigators can prepare DNA samples from hairs,
saliva, coffee cups, postage stamps and cigarette butts. The ability to create a
DNA profile from minute biological traces combined with advances in computer
analysis should allow law enforcement to identify a criminal faster and with
greater assurance.
From the earliest days, forensic scientists have wanted to find an attribute
unique to each individual that would be difficult or impossible to disguise. A
DNA profile fulfills that need. After all, a perpetrator's genetic identity
cannot be shed like bloodstained clothes. As Buckaroo Banzai wisely said, "No
matter where you go, there you are."
@bio:Phillip Jones, PhD, JD, is a biotech law specialist and has written more
than 76 articles for magazines, journals and newsletters on a variety of
subjects that blend law, history and science.
@sub:References:
@body:Benecke M. DNA Typing in Forensic Medicine and in Criminal Investigations:
A Current Survey. <I>Naturwissenschaften<$>. 1997; 84:181-188.
Campbell MF. <I>A Century of Crime<$>. Toronto, Canada: McClelland and Stewart,
Ltd.; 1970.
Evans C. <I>The Casebook of Forensic Detection<$>. New York, NY: John Wiley &
Sons; 1996.
Owen D. <I>Hidden Evidence<$>. Buffalo, NY: Firefly Books, Ltd.; 2000.
Rudin N, Inman K. <I>An Introduction to DNA Analysis<$>, Second Edition. Boca
Raton, Fla: CRC Press; 2002.
Saferstein R. <I>Criminalistics<$> Seventh Edition. Upper Saddle River, NJ:
Prentice Hall, Inc.; 2001.
Thorwald J. <I>The Century of the Detective<$>. New York, NY: Harcourt, Brace &
World, Inc.; 1964.
U.S. Department of Justice. <I>The Future of Forensic DNA Testing<$>. Available
at: http://www.ojp.usdoj.gov/nij/pubs-sum/183697.htm. Accessed August 29, 2003.
Wambaugh J. <I>The Blooding<$>. New York, NY: Bantam Books; 1989.
Wilson C. <I>Clues! A History of Forensic Detection<$>. New York, NY: Warner
Books; 1989.
More News
|
