Table of Contents:
***** Understanding the Western Blot *****
* The National Lyme Disease Network *
* LymeNet Newsletter *
IDX# Volume 4 - Number 13 - 9/23/96
IDX# ***** SPECIAL ISSUE *****
IDX# ***** Understanding the Western Blot *****
I. LYMENET: Understanding the Western Blot
Senders: The LymeNet Newsletter Editors
Revised: September, 1996
Inquiries about various issues relating to Western blot (WB) testing
are frequently posted to the Lyme disease discussion groups on the
Internet. Among the most commonly asked questions are: What laboratory
techniques are used to carry out the assay? What exactly is being
measured? What is a "band"? How are the results interpreted? What
are the CDC criteria for a "positive" test? Although some of the
medical jargon associated with immunology can be a little overwhelming,
the scientific principles behind these tests are not difficult to
grasp. The following article is offered as a primer in the techniques
and interpretation of Western blotting, and should help most patients
navigate their way through some of the medical and scientific
terminology associated with the assay.
First of all, it should be noted that the Western blot is usually
performed as a follow-up to an ELISA test, which is the most commonly
employed initial test for Lyme disease. "ELISA" is an acronym for
"enzyme-linked immunosorbent assay." There are ELISA tests and Western
blots for many infectious agents; for example, the usual testing
regime for HIV is also an initial ELISA followed by a confirmatory
Both the ELISA and the Western blot are "indirect" tests -- that is,
they measure the immune system's response to an infectious agent rather
than looking for components of the agent itself. In a Lyme disease
ELISA, antigens (proteins that evoke an immune response in humans) from
Borrelia burgdorferi (Bb) are fixed to a solid-phase medium and
incubated with diluted preparations of the patient's serum. If
antibodies to the organism are present in the patient's blood, they
will bind to the antigen. These bound antibodies can then be detected
when a second solution, which contains antibodies to human antibodies,
is added to the preparation. Linked to these second antibodies is an
enzyme which changes color when a certain chemical is added to the mix.
Although the methodology is somewhat complicated, the basic principle
is simple: the test looks for antibodies in the patient's serum that
react to the antigens present in Borrelia burgdorferi. If such
antibodies exist in the patient's blood, that is an indication that
the patient has been previously exposed to B. burgdorferi.
However, many different species of bacteria can share common proteins.
Most Lyme disease ELISA's use sonicated whole Borrelia burgdorferi --
that is, they take a bunch of B. burgdorferi cells and break them down
with high frequency sound waves, then use the resulting smear as the
antigen in the test. It is possible that a given patient's serum can
react with the B. burgdorferi preparation even if the patient hasn't
been exposed to Bb, perhaps because Bb shares proteins with another
infectious agent that the patient's immune system *has* encountered.
For example, some patients with periodontal disease, which is sometimes
associated with an oral spirochete, might test positive on a Lyme
ELISA, because their sera will react to components of Bb (like the
flagellar protein, which is shared by many spirochetes) even though
they themselves have never been infected with Bb. Therefore, some
positive Lyme disease ELISA results can be "false" positives.
To distinguish the false positives from the true positives, a more
specific laboratory technique, known as immunoblotting, is used.
(The Western blot, which identifies specific antibody proteins, is but
one kind of immunoblot; there is also a Northern blot, which separates
and identifies RNA fragments, and a Southern blot, which does the same
for DNA sequences.) In a Western blot, the testing laboratory looks
for antibodies directed against a wide range of Bb proteins. This is
done by first disrupting Bb cells with an electrical current and then
"blotting" the separated proteins onto a paper or nylon sheet. The
current causes the proteins to separate according to their particle
weights, measured in kilodaltons (kDa). From here on, the procedure
is similar to the ELISA -- the various Bb antigens are exposed to the
patient's serum, and reactivity is measured the same way (by linking an
enzyme to a second antibody that reacts to the human antibodies).
If the patient has antibody to a specific Bb protein, a "band" will form
at a specific place on the immunoblot. For example, if a patient has
antibody directed against outer surface protein A (OspA) of Bb, there
will be a WB band at 31 kDa. By looking at the band pattern of
patient's WB results, the lab can determine if the patient's immune
response is specific for Bb.
Here's where all the problems come in. Until recently, there has
never been an agreed-upon standard for what constitutes a positive WB.
Different laboratories have used different antigen preparations (say,
different strains of Bb) to run the test and have also interpreted
results differently. Some required a certain number of bands to
constitute a positive result, others might require more or fewer.
Some felt that certain bands should be given more priority than others.
In late 1994, the Centers for Disease Control and Prevention (CDC)
convened a meeting in Dearborn, Michigan  in an attempt to get
everybody on the same page, so that there would be some consistency
from lab to lab in the methodology and reporting of Western blot
Before we get to the recommendations that resulted from this meeting,
we need to understand one more facet of the human immune response.
Many patients have noticed that their Western blot report is actually
comprised of two separate parts, IgM and IgG. These are
immunoglobulins (antibody proteins) produced by the immune system to
fight infection. IgM is produced fairly early in the course of an
infection, while IgG response comes later. Some patients might already
have an IgM response at the time of the EM rash; IgG response,
according to the traditional model, tends to start several weeks after
infection and peak months or even years later. In some patients, the
IgM response can remain elevated; in others it might decline,
regardless of whether or not treatment is successful. Similarly, IgG
response can remain strong or decline with time, again regardless of
treatment. Most WB results report separate IgM and IgG band patterns
and the criteria for a positive result are different for the two
Finally, in setting up a nationwide standard for a positive WB, one
makes several assumptions -- that all strains of Bb will provoke
similar immune responses in all patients, that all patients will mount
a measurable immune response when exposed to Bb, and that the IgG
immune response will persist in an infected patient. Unfortunately,
none of these is always true. Therefore, a judicious interpretation
of Western blot results in a clinical setting should take into account
both the vagaries of the human immune response and the possibility that
strain variations in Bb might produce unusual banding patterns.
The CDC criteria for a positive WB are as follows:
* For IgM, 2 of the following three bands: OspC (21-25), 39 and 41.
* For IgG, 5 of the following ten bands: 18, OspC (21-25), 28, 30,
39, 41, 45, 58, 66 and 93.
How were these recommendations arrived at? The IgG criteria were taken
pretty much unchanged from a 1993 paper by Dressler, Whalen, Reinhardt
and Steere . In this study, the authors performed immunoblots on
several dozen patients with well characterized Lyme disease and a
strong antibody response and looked at the resulting blot patterns.
By doing some fairly involved statistical analysis, they could
determine which bands showed up most often and which best distinguished
LD patients from control subjects who did not have LD. They found that
by requiring 5 of the 10 bands listed, they could make the results the
most specific, in their view, without sacrificing too much sensitivity.
("Sensitivity" means the ability of the test to detect patients who
have the disease, "specificity" means the ability of the test to
exclude those who don't. Usually, an increase in one of these measures
means a decrease in the other.)
The IgM criteria were determined in much the same fashion (by different
authors in different papers). Fewer bands are required here because
the immune response is less mature at this point. Several studies have
shown that the first band to show up on a Lyme disease patient's IgM
blot is usually the one at 41 kDa, followed by the OspC band and/or
the one at 39. The OspC and 39 kDa band are highly specific for Bb,
while the 41 kDa band isn't. That's why the 41 by itself isn't
considered adequate. Here's the rub, though: the CDC doesn't want the
IgM criteria being used for any patient that has been sick for more
than a month or two. The thinking here is that by this time an IgG
response should have kicked in and the IgM criteria, because they
require fewer bands, are not appropriate for patients with later
A number of criticisms have been offered of the CDC criteria since
their adoption in 1994. The first is centered on the CDC's failure to
make any qualitative distinction among the various bands that can show
up on a patient's Western blot. A number of Lyme disease researchers
feel that different bands on a WB have different relative importance --
that "all bands are not created equal." For example, many patients
with Lyme disease will show reactive bands at, say, 60 and/or 66 kDa.
However, these correspond to common proteins in many bacteria, not
just Borrelia burgdorferi, and so are of limited diagnostic usefulness,
especially in the absence of other, more species-specific bands. The
band at 41 kDa corresponds to Bb's flagella (the whip like organelles
used for locomotion -- Bb has several) and is one of the earliest to
show up on the Western blots of Lyme disease patients. But for some
reason it is also the most commonly appearing band in control subjects.
This may be due to the fact that many people are exposed to spirochetes
at some time in their lives and so their sera might cross react with
On the other hand, certain other bands are considered highly specific for
Bb -- the aforementioned 31 kDa band, for example, or 34 (OspB) or 39 or
OspC (anywhere between 21 and 25). The 83 and 94 kDa bands are also
thought to be species-specific. Many Lyme disease scientists believe
that any patient whose IgG Western blot exhibits bands at, say, any
three (or even two) of these locations almost certainly has Lyme
disease, regardless of whether or not any other bands are present. They
feel that these bands on a Lyme Western blot are simply more meaningful
than other, less specific ones and that a rational interpretation of a
WB result should take this into account. Unfortunately, this does not
often happen, and will happen even less with the new CDC criteria.
A second criticism of the CDC Western blot criteria is that they fail
to include the 31 and 34 kDa bands. This does indeed seem like an odd
decision, since antibodies with these molecular weights correspond to
the OspA and OspB proteins of B. burgdorferi, which are considered to
be among the most species-specific proteins of the organism. So why
didn't Dressler et al. include them? Answer: These bands tend to
appear late if at all in Lyme disease patients, and did not show up
with great frequency in the patients that the Dressler et al. group
studied (though they did show up sometimes). As a result, they weren't
deemed to have much diagnostic value and didn't find their way onto the
CDC hot list. However, while the absence of either of these bands from
a patient's immunoblot result does not rule out Lyme disease, their
presence is hardly meaningless. Thus, many Lyme disease experts
believe it is a serious mistake to exclude these two antibody proteins
from the list of significant bands. The CDC's decision to do so seems
particularly strange in light of the fact that it is the OspA component
of Bb that is being used as the stimulating antigen in the ongoing
experimental Lyme disease vaccine trials. As one immunologist remarked
shortly after the 1994 CDC conference, "If OspA is so unimportant, then
why the heck are we vaccinating people with it?"
Finally, it is important to keep in mind that no matter how carefully
the Western blot test is carried out and interpreted, its usefulness,
like that of all tests that measure B. burgdorferi antibodies, is
ultimately contingent on the reliability of the human immune response as
an indicator of exposure to B. burgdorferi. There are several
scenarios in which the lack of a detectable antibody response may
falsely point to a lack of B. burgdorferi infection. First, it is well
established that early subcurative treatment of Lyme disease can
abrogate the human immune response to B. burgdorferi . Although
this is not thought to be a common phenomenon, a recent comparative
trial for the treatment of erythema migrans found that a majority of
patients who failed early treatment and suffered clinical relapse were
seronegative at the time of relapse . Even treatment for
disseminated Lyme disease, in which the patient's IgG immune response
was previously well-established, can render a patient seronegative
after treatment despite post-treatment culture-positivity for
B. burgdorferi [5,6].
In addition, patients with Lyme disease may not test positive for
exposure to B. burgdorferi because their antibodies to the organism are
bound up in immune complexes . Once steps are taken to dissociate
these immune complexes, free antibody can be detected; however, this
is not routinely done when performing serologic tests for Lyme disease.
Finally, an indeterminate number of patients with late Lyme disease are
simply seronegative for unknown reasons . The actual percentage of
such cases as a proportion of all Lyme disease cases is impossible to
estimate, since most studies of late Lyme disease enroll only
seropositive patients, which tends to reinforce the circular and
erroneous notion that virtually all patients with late Lyme disease
It should also be noted that a positive Western blot is not necessarily
an indication of active Lyme disease. A patient's immune response to
B. burgdorferi can remain intact long after curative treatment for a
Lyme infection; therefore, the results of a Western blot assay should
always be interpreted in the context of the total clinical picture.
 Proceedings of the Second National Conference on Serologic
Diagnosis of Lyme Disease, October 27-29, 1994.
 Dressler F, Whalen JA, Reinhardt BN, Steere AC. Western blotting
in the serodiagnosis of Lyme disease. J Infect Dis 1993;167:392-400.
 Dattwyler RJ, Volkman DJ, Luft BJ et al. Seronegative Lyme
disease: dissociation of specific T- and B-lymphocyte responses to
Borrelia burgdorferi . N Engl J Med 1988;319:1441-6.
 Luft BJ, Dattwyler RJ, Johnson RC et al. Azithromycin compared
with amoxicillin in the treatment of erythema migrans. Ann Intern Med
 Häupl T, Hahn G, Rittig M, et al. Persistence of Borrelia
burgdorferi in ligamentous tissue from a patient with chronic Lyme
borreliosis. Arth Rheum 1993;36:1621-6.
 Preac-Mursic V, Marget W, Busch U, Pleterski Rigler D, Hagl S.
Kill kinetics of Borrelia burgdorferi and bacterial findings in
relation to the treatment of Lyme borreliosis. Infection 1996;24:9-18.
 Schutzer SE, Coyle PK, Belman AL, et al. Sequestration of antibody
to Borrelia burgdorferi in immune complexes in seronegative Lyme
disease. Lancet 1990;335:312-5.
 Liegner KB. Lyme disease and the clinical spectrum of antibiotic
responsive chronic meningoencephalomyelitides. Clin Infect Dis, in
II. ABOUT THE LYMENET NEWSLETTER
For the most current information on LymeNet subscriptions,
contributions, and other sources of information on Lyme disease,
please refer to the LymeNet Home Page at:
To unsubscribe from the LymeNet newsletter, send a message to:
On the first line of the message, write: unsub lymenet-l
LymeNet - The Internet Lyme Disease Information Source
Editor-in-Chief: Marc C. Gabriel <a229@Lehigh.EDU>
FAX (for contributions ONLY): 908-789-0028
Contributing Editors: Carl Brenner <firstname.lastname@example.org>
John Setel O'Donnell <jod@Equator.com>
Frank Demarest <76116.2065@CompuServe.com>
Advisors: Carol-Jane Stolow, Director <email@example.com>
William S. Stolow, President <firstname.lastname@example.org>
The Lyme Disease Network of New Jersey
WHEN COMMENTS ARE PRESENTED WITH AN ATTRIBUTION, THEY DO NOT
NECESSARILY REPRESENT THE OPINIONS/ANALYSES OF THE EDITORS.
THIS NEWSLETTER MAY BE REPRODUCED AND/OR POSTED ON BULLETIN BOARDS
FREELY AS LONG AS IT IS NOT MODIFIED OR ABRIDGED IN ANY WAY.
SEND ALL BUG REPORTS TO a229@Lehigh.EDU