MAY 2000 


ELISA and Western blot testing

Appreciating Giraldo’s conclusions requires an understanding of the ELISA and Western blot techniques.

Both ELISAs and Western blots detect proteins, either antibodies or antigens. Antigens are foreign proteins, such as those belonging to viruses, which the immune system targets for destruction. One way that immune systems destroy antigens is by producing antibodies that lock onto, or "neutralize," them. A particular virus may contain about ten different proteins that get exposed to the immune system. Each one of these proteins triggers the production of different species of antibodies, just as a door with ten bolt locks requires ten different keys.

ELISA stands for Enzyme-Linked Immunosorbent Assay. Special enzymes linked to the test proteins luminate at intensities according to the amount of targeted protein that exists in the serum. Western blots work similarly. But, unlike ELISAs, which hold the different species of test proteins together, Western blots separate the different test protein species into different bands, according to their molecular weights. Thus, whereas a positive ELISA indicates that serum contains target proteins that react with at least one test protein, it cannot determine how many or which species of test proteins have reacted. Western blots can.

(The word "Western" honors the scientist who developed the technique, first used for DNA. The DNA procedure is called the "Southern blot," after the scientist’s last name, and when employed for RNA, the procedure is called the "Northern blot.")

These tests are inexpensive, easily performed substitutes for the only absolute method of determining if a person is actively infected with a microbe: isolation of the microbe from fresh patient tissue. In the case of a microbe that infects immune cells, which is what HIV is said to be, that would mean isolating HIV from fresh blood.

The accuracies of these tests are determined by successfully obtaining positive results in people from which isolations can be obtained (sensitivity), and successfully obtaining negative results in people from which isolations cannot be obtained (specificity).

Establishing test parameters

Though reactivity of a single target protein earns a positive ELISA designation, a positive Western blot may not require that every target protein react–not if microbial isolation data shows that a certain combination of positive reactions corresponds to a maximum accuracy in identifying people who are and who are not actively infected with that microbe.

Giraldo emphasizes a point that most HIV professionals overlook, and which plays a salient role in his investigation: ELISAs and Western blots are not only qualitative (they indicate if the target proteins are present in the serum), they are also quantitative (they indicate how much of the target proteins are present in the serum). Each can measure the amount of target proteins in the sera by the intensities of the test reactions, as determined by their observed luminosities.

ELISA and Western blot test instructions stipulate what luminosity level constitutes a positive reaction, and that the level varies according to the microbe being tested for. That opens the question: At what level of luminosity should a reaction be regarded as "positive"? The answer, as always, lies with isolation of the microbe. Only isolation data can logically determine the ELISA and Western blot luminosity levels that most accurately distinguish who has or doesn't have an active infection with a particular microbe.

ELISAs and Western blots can test for either antigens or antibodies, depending on what the test kit contains. Antigen tests contain antibodies and react if the serum contains antigens (the actual virus proteins, in the case of a viral test); antibody tests contain antigens and react if the serum contains antibodies. The HIV antibody tests, then, contain presumed HIV proteins, which are the viral antigens. They react with sera that possess antibodies that neutralized these antigens. What is called "the HIV test" consists of a battery of sequentially administered antibody tests, two ELISAs followed by at least one Western blot.

Both antibody and antigen tests can be reliable and valid indicators of viral infections. But only virus isolation can demonstrate if either accurately identifies who has and who doesn't have an active viral infection.

ELISA and Western blot tests exist for HIV antibodies and for HIV antigens. But the HIV antigen tests are not used for diagnosing HIV infections. Like the questions Giraldo has asked about the unusually high dilution levels for HIV tests, nobody has ever explained why HIV-antigen tests are not used to diagnose HIV infections. But the technical literature is very clear: while many members of the risk groups, including most who have "AIDS" conditions, test HIV-antibody positive, only those with "AIDS" conditions tend to test HIV-antigen positive as well (Piatak, Science 259, 1993). So whereas HIV antibody tests identify as positive lots of healthy people, antigen tests do not.

Giraldo says that even diluting of serum can be a valid practice and produce reliable results. But, again, only if the diluting has been established by isolation to improve accuracy.

– P. P.

--Paul Philpott