close

Вход

Забыли?

вход по аккаунту

?

Mycoplasma infection and rheumatoid arthritis. Analysis of their relationship using immunoblotting and an ultrasensitive polymerase chain reaction detection method

код для вставкиСкачать
ARTHRITIS & RHEUMATISM
Vol. 40, No. 7,July 1997, pp 3219-1228
0 1997, American College of Rheumatology
1219
MYCOPUSMA INFECTION AND RHEUMATOID ARTHRITIS
Analysis of Their Relationship Using Immunoblotting and an
Ultrasensitive Polymerase Chain Reaction Detection Method
ROBERT W. HOFFMAN, FRANK X. O’SULLIVAN, KIM R. SCHAFERMEYER, TERRY L. MOORE,
DEBORAH ROUSSELL, ROBYN WATSON-McKOWN, MARY F. KIM, and KIM S. WISE
Objective. To examine the relationship between
infection with Mycoplasma and the development of rheumatoid arthritis (RA) and juvenile rheumatoid arthritis
(JU).
Methods. Immunoblotting of patient synovial
fluid and sera on detergent-phase membrane protein
extracts of various Mycoplasma species was carried out
to learn whether patients exhibited serologic evidence of
previous exposure to mycoplasmas. Moreover, an ultrasensitive polymerase chain reaction (PCR) method was
developed for assessing whether Mycoplasma DNA could
be detected in synovial fluid from patients and controls.
Results. Immunoblotting provided serologic evidence of previous Mycoplasma exposure in patients and
controls. The genus-specific PCR detected known human Mycoplasma species and could reliably detect <5
copies of Mycoplasma hominis, Mycoplasma fermentans,
or a molecular mimic control in synovial fluid. Repeat
testing revealed no evidence of Mycoplasma DNA in
patient synovial samples.
Conclusion. This study provided serologic evidence suggesting that, while previous exposure to Mycoplasma was common, there was no detectable persistence
of Mycoplasma DNA in the synovial fluid or tissue of
patients with RA or JRA.
Supported by NIH grants AR-42537 and AI-32219, the Leda
J. Sears Trust, and the Department of Veterans Affairs.
Robert W. Hoffman, DO, Frank X. O’Sullivan, MD, Kim R.
Schafermeyer, BS: University of Missouri-Columbia, and Harry S
Truman Memorial Veterans Hospital, Columbia, Missouri; Terry L.
Moore, MD: St. Louis University, St. Louis, Missouri; Dcborah
Roussell, PhD, Robyn Watson-McKown, MS, Mary F. Kim, MS,
Kim S. Wise, PhD: University of Missouri-Columbia.
Address reprint requests to Robert W. Hoffman, DO, Director, Division of Immunology and Rheumatology, MA427 HSC, One
Hospital Drive, Columbia, MO 65212.
Submitted for publication December 18, 1996; accepted in
revised form March 17. 1997.
There has been longstanding interest in a potential infectious etiology of both rheumatoid arthritis (RA)
and juvenile rheumatoid arthritis (JRA). Interest in this
possibility of an infectious etiology for RA has been
renewed by the discovery that a spirochete, Borrelia
burgdovferi, can cause chronic arthritis that often resembles RA and JRA (1). In addition, other infectious
agents have recently been recognized to cause persistent
synovitis in humans; these include the hepatitis B virus,
human immunodeficiency virus, and parvovirus (2-4).
Among the various infectious agents considered
as potential etiologic agents in RA, mycoplasmas have
been considered as candidates for a variety of reasons.
Mycoplasma hominis, Mycoplasma salivarium, and Mycoplasma pneumoniae have been identified as causing
acute monarthritis in humans (5-7). Mycoplasma genitalium and Mpneumoniae have been isolated from the
joints of patients with polyarthritis (7-10). Polyarthralgia
and polyarthritis have been described as occurring in 9%
of patients with pneumonia that is caused by Mpneumoniae (9). In addition, mycoplasmas are wellestablished etiologic agents of acute and chronic arthritis
in various animal species, including porcine (Mycoplasma hyorhinis and Mycoplasma hyosynoviae), avian (Mycoplasma synoviae), bovine (Mycoplasma bovis), and
murine (Mycoplasma arthritidis) species (11).
Furthermore, mycoplasmas are ubiquitous, and
several species have recently been recognized to be
human pathogens. For example, A4 genitalium has recently been recognized to be a common pathogen of the
genitourinary tract (12), and both Mycoplasma fermentans and Mycoplasma penetrans may be contributing
pathogens in the acquired immunodeficiency syndrome
(13,14). In addition, evidence consistent with a role for
mycoplasmas in RA was suggested by 2 recent controlled
studies that demonstrated the clinical efficacy of mino-
HOFFMAN ET AL
1220
cycline in RA (15,16). Minocycline is an antimicrobial
agent known to be active against mycoplasmas.
These previous data suggest 2 alternative hypotheses that can be considered with respect to the role of
mycoplasmas in RA. The first hypothesis is that infection with Mycoplasma is a cofactor that can trigger the
onset of RA in a genetically susceptible host, but it need
not persist after the autoimmune process is initiated.
The second hypothesis is that RA is caused by chronic
local infection with Mycoplasma, similar to that associated with Mycoplasma in some animal hosts.
Previous studies on the relationship between
Mycoplasma infection and RA have been hindered by
limitations in the methods available to address these
possibilities. Mycoplasmas are difficult to culture, and
can rapidly undergo phenotypic changes in culture.
Recent advances in immunologic methods and polymerase chain reaction (PCR)-based molecular diagnostic
techniques, however, have provided the opportunity to
apply new approaches to the detection of organisms in
the host. In the present study, we have developed in vitro
amplification techniques that allow for the reproducible
detection in synovial fluid of very limited copies of genes
present in a broad range of Mycoplasma species. We
have used these tools to examine synovial fluid for the
local presence of mycoplasmas in patients with R A and
JRA. We report herein the results of a comprehensive
analysis of synovial fluid and tissue from patients with
either R A or JRA, in conjunction with a serologic
assessment of humoral responses to antigens of 2 Mycoplasma species, M hominis and M fermentans.
PATIENTS AND METHODS
Patients and controls. All studies involving human
subjects were approved by the University of Missouri Institutional Review Board. Patients were selected from the University of Missouri Hospital and Clinics, the Harry S Truman
Memorial Veterans Hospital, and the St. Louis University
Hospital and Cardinal Glennon Children’s Hospital. The 35
adult patients studied were classified as having RA using the
revised American College of Rheumatology (formerly, the
American Rheumatism Association) disease classification criteria (17). The mean age of the RA patients studied was 59.2
years (range 26-77 years). The mean duration of disease was
10.2 years (range 1-31 years). The 40 patients with juvenileonset arthritis were classified as having either systemic-onset
JRA (n = 5; mean age 14 years [range 9-16 years], mean
duration of disease <1 year [range 0-1 year]), pauciarticular
JRA (n = 21; mean age 8.6 years [range 3-16 years], mean
duration of disease 1.9 years [range 0-14 years]), or polyarticular JRA (n = 14; mean age 12.2 years [range 5-17 years],
mean duration of disease 3.8 years [range 0-8 years]) (18).
Matched samples of sera and synovial fluid from patients with
other types of arthritis served as a matched control group for
the studies. The control patients were classified as having
osteoarthritis (n = lo), gout (n = 5), bursitis (n = l), psoriatic
arthropathy (n = l), reactive arthropathy (n = l), polyarthritis
of unknown etiology (n = 1), or bacterial-mediated infectious
arthritis (n = 5) based upon standard clinical criteria (19).
Normal controls were healthy blood donors.
Serologic analysis. Sera were collected and processed
under aseptic conditions. Samples were aliquoted into sterile
tubes in a laminar flow hood and stored at -70°C. Sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) and immunoblotting of Triton X-114 (TX-114)
detergent-phase extracts of M hominis and M fermentans were
performed as previously described (20,21). Synovial fluid was
centrifuged at 300g for 10 minutes and diluted 1:200 prior to
incubation with blot strips. Blots were developed with horseradish peroxidase-conjugated secondary rabbit antibody specific for human IgG-Fc (Accurate Chemical and Scientific
Corp., Westbury, NY) (20). Monoclonal antibodies (MAb) to
various TX-114 proteins were used to identify some specific
Mycoplasma components bound by human IgG antibodies.
Immunoblotting was performed using MAb to M hominis proteins, as previously described (20). These MAb included
H3 (MAb to the size variable Vaa antigen) and G5 (MAb to
P70), both provided by Michael Barile and Lyn D. Olson
(Center for Biologic Evaluation and Research, Food and Drug
Administration, Rockville, MD), and 26.7 (MAb to P120),
provided by Gunna Christiansen (University of Aarhus, Aarhus, Denmark). MAb and other monospecific antibodies to the
M fermentans membrane proteins P150, P95, P78, P76 (Wise
KS: unpublirhed observations), P61, P41, P38, and P29 have
also been previously described (21,22).
Acquisition and physical isolation of clinical samples.
All samples were initially received, processed, aliquoted, and
stored at -70°C at the Harry S Truman Memorial Veterans
Ho\pital. These facilities are physically separate from laboratories at the University of Missouri. Aliquots were transferred
to the University of Missouri, where PCR reaction mixtures
were prepared and sealed in an isolation hood, and thermocycling was performed in a second hood in a limited-access
laboratory designed specifically for this purpose. Tubes containing PCR products were transferred to a third laboratory at
the University of Missouri, where amplified products were
analyzed to reduce the possibility of contamination with Mycoplasma or amplicon DNA. Samples always moved in a
forward direction through these 3 laboratories.
Extraction and isolation of DNA. DNA was isolated
from axenically grown M hominis (strain 1620) and M f e m e n tans (strain PG18), which were cultured by an investigator
(KSW) in 1 of our laboratories, or from broth cultures of
Ureaplasma urealyticum, Acholeplasma laidlawii, M genitalium,
M pneurnoniae, Mycoplasma orale, M salivarium, Mycoplasma
buccale, Mycoplasma faucium, Mycoplasma lipophilum, Mycoplasma primaturn, Mycoplasma spematophilum, and Mycoplasma pirum, which were kindly provided by Joseph G. Tully
(Frederick Cancer Center, Frederick, MD). Synovial fluid
2121 was obtained from the knee of a patient infected with
septic arthritis that was caused by M hominis (20,23), and
samples were kindly provided by Michael Barile and Lyn D.
Olson.
MYCOPUSMA INFECTION IN RA
1221
Table 1. Mycoplasma primers*
Primer
MGSO
GPO-1
MF-I
MF-2
MH-1
MH-2
AD-I
AD-2
P-1
P-2
* rRNA
Sequence
Annealing
temperature
Designed to amplify
5‘-TGCACCATCTGTCACTCTGTTAACCTC-3‘
(forward) Genus-specific Mycoplasmu 16s rRNA
S’-ACTCCTACGCGAGGCAGCAGTA-3’(reverse)
Genus-specific Mycoplasmu 16s rRNA
5’-GAAGCCTTTCTTCGCTGGAG-3‘(forward)
Species-specific Mfermentuns 16s rRNA
5’-ACAAAATCATITCCTATCTGCT-3‘ (reverse)
Species-specific Mfermentuns 16s rRNA
S‘-TGAAAGGCGCTGTAAGGCGC-3’
(forward)
Species-specific M horninis 16s rRNA
5‘-GTCTGCAATCATTTCCTATGCAAA-3’ (reverse)
Species-specific M horninis 16s rRNA
5‘-CGTGACCCATTTGCATCAG-3‘
(forward)
Arginine deiminase-specific
S’-GTGCATAAGlTTG’ITCAnr(TIA)GG-3’ (reverse) Arginine deiminase-specific
S’-GTGCCAGCAGCCGCGGTAATAC-3’
(forward)
Prokaryote-specific
5’-TCTACGCATTTCACCGCTACAC-3‘
(reverse)
Prokaryote-specific
=
64°C
64°C
55°C
55°C
55°C
55°C
60°C
60°C
60°C
60°C
Product
size Reference
(bp)
no.
718
718
272
272
28 1
281
330
330
190
190
25,26
25,26
25
25
25
25
27
27
28,29
28,29
ribosomal RNA.
DNA was purified from the Mycoplasma species or
from synovial fluid using the IsoQuick DNA isolation method
(Orca Research, Bothell, WA), as described by the manufacturer. Synovial tissue was obtained aseptically at arthrotomy
and placed into sterile phosphate buffered saline. Tissue
was frozen by immersion in liquid nitrogen and stored at
-70°C. The frozen synovial tissue was crushed and DNA
was isolated by phenol-chloroform extraction and ethanol
precipitation (24).
Construction of molecular mimic. A nonhomologous
internal standard (molecular mimic) was constructed as a
control for these experiments. The mimic was constructed as
follows: the linearized Bum HIiEco RI fragment of v-erbB
(Clontech Laboratories, Palo Alto, CA) was used as a DNA
template. The primer pair 5‘-TGCACCATCTGTCAC
TCTGTTAACCTCCGCAAGTGAAATCTCCTCCG-3‘(forward) and 5’-ACTCCTACGGGAGGCAGCAGTATT
TGATTCTGGACCATGGC-3’ (reverse) was used to
generate a 528-basepair segment of the Bum HIIEco RI
fragment of v-erhB. A second PCR reaction was then performed using a Mycoplasma genus-specific primer pair,
GPO-1, S‘-ACTCCTACGGGAGGCAGCAGTA-3’ (forward), and MGSO, 5’-TGCACCATCTGTCACTCTGTTA
ACCTC-3’ (reverse) (25), to generate a 555-bp PCR product
(mimic). This was purified away from any residual primers by
centrifugation through a CHROMA-SPIN+TE-100 column
(Clontech Laboratories), and the molar concentrations were
determined by comparison with ethidium bromide-stained
DNA standards of known concentrations. Aliquots were stored
with glycogen at -70°C. The mimic was diluted with water to
the appropriate concentration before use.
DNA PCR primers. After a series of preliminary
experiments on several prototypic Mycoplasrna species, several
primers were chosen for use in analyzing patient and control
samples. The primers that were ultimately chosen are shown in
Table 1 (25-29).
PCR assay and gel electrophoresis. All components of
the PCR reaction mixture were initially tested over a wide
range of concentrations to optimize the sensitivity of the assay,
and DNA polymerases from a variety of sources were assessed.
Patient samples were then tested under the optimized conditions in a 50-pl reaction volume containing the following
components, at the final concentrations or amounts indicated:
MgSO,, 5.7 mM; P-mercaptoethanol, 20 pM; nucleoside
triphosphates, dATP, dCTP, dGTP, and dTTP, 400 pJ4 each;
primers, 0.3 pA4 each; Tuq Gold DNA polymerase (PerkinElmer, Branchburg, NJ), 3 units; and purified sample DNA, 1
pl. Samples were amplified for 40 cycles on a Perkin-Elmer
model 480 thermocycler with a cycle profile of 1 minute at
94”C, followed by 1 minute at the appropriate annealing
temperature for the primer pair, and then 1minute at 72°C. At
the end of the 40 cycles, a 5-minute extension cycle at 72°C was
included. Following PCR, samples were subjected to agarose
gel electrophoresis, stained in ethidium bromide, and visualized by ultraviolet transillumination. The image was then
documented by photography and a digital imaging system
(Alpha Innotech, San Leandro, CA).
RESULTS
Serologic findings. Amphiphilic membrane proteins isolated from wall-less mycoplasmas by detergentphase fractionation with TX-114 (21) are known to
contain major immunogenic surface proteins, several of
which display wide variations in size and expression
among the various Mycoplasma species (21,30). Due to
the relatively low level of nonspecific binding compared
with that in whole organism preparations, immunoblotting of SDS-PAGE-separated TX-114 proteins has been
used for serologic screening of animal and human
humoral responses during Mycoplasma infections
(20,31-34). We have used this approach to define various antibodies to M fennentans (21,22) and M horninis
(20), both of which are human Mycoplasma species
known or suspected to be associated with septic arthritis
(5,12,35-37).
We therefore examined TX-114 immunoblots of
these 2 species using synovial fluids from control sub-
HOFFMAN ET AL
1222
A
**
150+
95 -+
-+
78
76 -+
61 -+
41 -+
38 -+
29 -+
B
***
4- 120
+ 72
4- 50
Figure 1. Immunoblotting patterns of patient synovial fluids on Triton X-114 extracts of Mycoplasma fernenfans (A) and Mycoplasma
horninis (B). Arrows mark the size (in kd) of known membrane
proteins (indicated by single asterisks), which were stained with a
combination of corresponding monoclonal antibodies. A, Immunoblot
of patient sera on extracts of M fernentans, showing several distinctive
patterns of serologic reactivity among individuals. B, Immunoblotting
results of patient synovial fluids on extracts of M horninis, showing that
reactivity with Mycoplasma-specificantigens is common.
jects or patients with RA or JRA (Figures 1A and B).
Patterns of IgG antibody binding to the various mycoplasma1 proteins are summarized in Table 2 for all
patient and matched control synovial fluids studied. As
shown in Figure 1B, distinctive patterns of synovial fluid
and serum antibodies reactive with Mycoplasma antigen
can be recognized among individuals. The immunoblot
shows that the MAb that define specific Mycoplasma
antigens can identify some of the specific Mycoplasma
components bound by patient or control IgG antibodies.
These results reveal a widespread response
among patient and control groups, indicating the presence of serum and synovial fluid IgG that binds mem-
brane proteins of 2 Mycoplasma species, M fernentans
(Figure 1A) and M horninis (Figure 1B). For each of
these organisms, IgG binding to multiple protein bands
was seen in the majority of individuals, whereas some
failed to show a reaction. This suggests that antibodies
were being specifically targeted to the membrane
components, rather than binding through nonspecific
mechanisms.
Two additional features of the overall serologic
responses were noteworthy. First, multiple distinct components of the membrane preparations were typically
recognized by an individual responding to an organism.
Some of these components could be unambiguously
identified by aligning gels (not shown in Figure lB), to
determine their recognition by MAb that are known to
recognize surface lipoproteins on these organisms. Second, not all membrane proteins of the mycoplasmas
were recognized by responsive individuals. The strain
(clonal isolates) of the mycoplasmas used to prepare
detergent extracts for immunoblotting had been selected
to represent the known repertoire of lipoprotein antigens. However, these components are known to undergo
high-frequency mutational-phase variation in vitro
(22,38,39). Our findings are therefore interesting in that
individuals appeared to vary in their precise antibody
response profile to the antigens displayed, consistent
with either an individualized response to antigenically
variable organisms, or a selective response in individuals
to the repertoire of antigens of the organisms. In any
case, and irrespective of the source and nature of this
response, no significant difference was noted between
any of the RA patient or control groups.
PCR assay and purification of DNA. Initial experiments focused on the development of an appropriately sensitive PCR-based assay that could detect a
range of Mycoplasma species in synovial fluid. A series of
PCR primers with different specificities was synthesized
and tested on purified DNA extracted from brothTable 2. Mycoplasma antibodies in patient and matched control
synovial fluids*
Juvenile
Other
Rheumatoid rheumatoid
inflammatory
Mycoplasma
arthritis
arthritis Osteoarthritis arthropathies
(n = 40)
(n = 10)
(n = 14)t
species
(n = 20)
11 (55)
3 (15)
M lzominis
M fernenfuns
35 (88)
36 (90)
9 (90)
5 (50)
12 (86)
6 (43)
* Values are the number (%) positive.
patients with gout (n = 5), bacterial-mediated infectious
arthritis (n = 5), psoriatic arthropathy (n = 1), reactive arthropathy (n =
t Includes
1),bursitis (n
=
1), and polyarthritis of unknown etiology (n
=
1).
MYCOPLASMA INFECTION IN RA
1223
1I 21 3I 4I 5I 6I 7l 8l 9I 1I 0 I1 1 1 ~ 1 ~ 1 ~ 1 ~ 1 ~ To address the potential problem of the presence
of inhibitors of the PCR in synovial fluid, several methA
ods of DNA template preparation were tested. For these
experiments, synovial fluid 2121 from a patient infected
with
M hominis was used. After testing several ap190-b
proaches, including boiling of synovial fluid, phenolchloroform extractions, and DNA precipitation methods
of extraction, the IsoQuick extraction method (Orca
Research) was selected.
I I
I
I
I
I
I
1
I
I
1
I
I
I
I
I
To determine the sensitivity of the PCR assay,
B
decreasing concentrations of purified Mycoplasma species DNA were added to synovial fluid that was negative
under optimized PCR conditions, and all the samples
718+
were again tested on at least 2 separate occasions. As
illustrated in Figure 3, it was determined that the
PCR-based assay could reliably detect <5 copies of M
hominis in the synovial fluid. Similar sensitivity was
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
observed in samples receiving M fermentans DNA. BeC
cause this genus-specific primer set was able to amplify
330
+
Figure 2. Polymerase chain reaction (PCR) products obtained using
various primer sets to amplify Mycoplasma DNA from the following
sources: lane 1, no DNA control; lane 2, Eschenchia coli; lane 3,
Acholeplasma laidlawii; lane 4, Ureaplasma urealyticum; lane 5 , Mycoplasma genitalium; lane 6, Mycoplasma pneumoniae; lane 7, Mycoplasma hominis; lane 8, Mycoplasma fermentans; lane 9, Mycoplasma
orale; lane 10, Mycoplasma salivanum; lane 11, Mycoplasma buccale;
lane 12, Mycoplasma faucium; lane 13, Mycoplasma lipophilum; lane
14, Mycoplasma primatum; lane 15, Mycoplasma spermatophilum; lane
16, Mycoplasma pirum. A, The prokaryote-specific primer pair P1 and
P2 was used. Arrow indicates the expected location for the 190basepair PCR product. B, The genus-specific primer pair GPO-1 and
MGSO was used. Arrow indicates the expected location for the 718-bp
PCR product. C, The deiminase-specific primer pair AD1 and AD2
was used. Arrow indicates the expected location for the 330-bp PCR
product. Note that lane 6 in B is negative due to omission of the
template in the series of reactions shown.
cultured M hominis, M fermentans, M genitalium, U
urealyticum, A laidlawii, M pneumoniae, M orale, M
salivarium, M buccale, M faucium, M lipophilum, M
primatum, M spermatophilum, and M pirum (see Figure
2). Primers were selected based on homology analyses
against known Mycoplasma DNA and RNA sequences,
and on pilot studies in which a larger series of primers
was tested to define optimal specificity and sensitivity.
Various DNA polymerases were also initially tested,
and several additional parameters of the PCR were
optimized.
1
2
3
1
1
1
4
1
5
6
1
7
1
8
1
9
1
1
1
0
1
718 -b
Figure 3. Titration of purified Mycoplasma hominis DNA in human
synovial fluid using the genus-specific primer pair GPO-1 and MGSO
under the optimized assay conditions. Lane 1 contains 100-basepair
DNA size markers. Lanes 2-4 contain polymerase chain reaction
(PCR) products generated from reactions containing (500 (lane 2),
<SO (lane 3), or <5 (lane 4) copies of target DNA template. Lane 5
shows a negative control in the same series, which contains the
identical components except the M hominis DNA template. Arrow
indicates the expected location for the 718-bp PCR product.
HOFFMAN ET AL
1224
I2345678910
I
I
I
I
I
I
I
I
I
I
718+
555-b
Figure 4. Polymerase chain reaction (PCR) with the Mycoplusmu
genus-specific primers, GPO-] and MSGO, on the mimic control and
on serial dilutions of M hominis DNA from synovial fluid. Lane 1
contains 100-basepair DNA size markers. Lanes 2-8 contain products
amplified from reactions involving a constant concentration of mimic
(3 X 10W2’ molar) mixed with serial dilutions of DNA extracted from
Mycoplusma horninis-infected synovial fluid 2121. Each lane represents the PCR products amplified from <5 copies of mimic. Lane 9
represents a negative control reaction lacking the DNA template. The
upper arrow indicates the PCR product amplified from Mycoplasma at
718 bp. The lower arrow indicates the expected location for the 555-bp
PCR product amplified from the mimic. As shown in lane 4, when the
mimic and the Mycoplusma DNA templates are present at near
equimolar concentrations, competitive inhibition of the amplification
from both DNA templates occurs.
16s ribosomal RNA sequences from all mycoplasmas
tested, this was the effective sensitivity for detection of
any Mycoplasma species in synovidl fluid samples.
PCR assay and molecular mimic. As an additional control, a nonhomologous internal standard, or
molecular mimic, was constructed and used in these
experiments. PCR results using the mimic are illustrated
in Figure 4. The mimic was constructed to generate a
555-bp segment of DNA using the same primers, GPO-1
and MGSO, that generated a 718-bp sequence from
Mycoplasma DNA. This mimic was added to synovial
fluid samples before IsoQuick extraction, at a copy
number that yielded no more than 5 gene copies per lane
on the final agarose gel. The inclusion of this control
confirmed that, if the mimic was detected in the final gel,
all conditions were optimal for amplification of Mycoplasma DNA. This control was critical for interpretation
of negative results, in that it excluded trivial explanations
for the inability to detect Mycoplasma DNA amplification, such as omission of the DNA sample or other
reagents, or the presence of an inhibitor in the template
preparation. Finally, because competition between Mycoplasma and mimic DNA could occur and obscure
detection of Mycoplasma when both were present at a
near equimolar ratio (see Figure 4, lane 4), all samples
were tested both with and without addition of the mimic
(Figure 5). These results also confirmed the sensitivity of
the primers for the Mycoplasma DNA template. The
presence of a defined quantity of mimic in the samples
suggests that the sensitivity of the assay approached 1
genomic copy of Mycoplasma DNA.
Analysis of synovial fluid and tissue. Multiple
independent assays of all synovial fluid samples using 4
different primer pairs revealed no Mycoplasma DNA in
1
2 3
1
1
1
4 5
I
1
6
7 8 9 10 11 12 13 14 15 16 17 16 1920
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Figure 5. Polymerase chain reaction (PCR) products from the extracted synovial fluid DNA template of rheumatoid arthritis (RA)
patients using the Mycoplusma genus-specific primer pair GPO-1 and
MSGO. Pairs of synovial fluid samples are shown, with or without
addition of mimic, prior to DNA extraction. Lane 1 contains 100basepair DNA size markers. Lanes 2 and 3 contain PCR products from
purified Mycoplusma fermentans DNA, used as positive controls. Lane
2 also contains the mimic control. The upper arrow indicates the
Mycoplastnu-specificproduct at 718 bp. The lower arrow indicates the
mimic product at 555 bp. Lanes 4 and 5 contain PCR products from
purified Mycoplusmu hominis DNA as positive controls, with or
without mimic, respectively. Lanes 6 and 7 contain PCR products using
the DNA template purified from synovial fluid 2121, known to be
infected with Mycoplusmu hominis. This is shown with or without
mimic controls. Lane 6 is with, and lane 7 is without, mimic. Lane 8 is
a negative control from the same PCR, but lacking added DNA
template. Lanes 9-20 are paired RA patient synovial fluids, with or
without mimic added to them prior to DNA extraction. Note that there
are no Mycoplusmu-specificproducts (718 bp) in any of the RA patient
samples (lanes 9-20) or in the negative control (lane 8), as indicated by
the open arrow.
MYCOPLASMA INFECTION IN RA
A
9 10 11 12 13 14 15
1225
2
3
4
5
6
7
8
1
1
1
1
1
1
1
1
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1,6 1,7118
lp 2p
1
1
1
1
1
1
1
1
DNA could be reproducibly detected in synovial fluid to
which purified DNA from M hori1ini.s or M fernentans
had bccn added at the detection level of <5 gene copies.
DNA could also be reproducibly detected in synovial
fluid 2121 obtained from a patient infected with M
horninis.
To test the possibility that Mycoplasma DNA
might be present in synovial tissue and not detectable in
ynovial fluid, we examined both synovial fluid and
synovial tissue from 8 patients with RA, in the absence
or presence of the mimic. We failed to detect any
evidence of Mycoplasma DNA i n these tissue samples
using our PCR-based assays.
1
DISCUSSION
272
+
Figure 6. Polymerase chain reaction (PCR) products from purified
patient synovial fluid using species-specific primers for Mycoplasma
horninis (A) and for Mycoplusma ferrnentans (B). A, Lane 1 contains
100-basepair DNA size markers. Lane 2 contains PCR products from
a positive control using purified M hominis DNA template. Lane 3 is
a positive control using synovial fluid 2121 infected with M horninis.
Lane 4 represents a negative control from the same PCR, performed
without added DNA template. Lanes 5-20 contain PCR products using
purified DNA from rheumatoid arthritis (RA) synovial fluid samples.
B, Lane 1 contains 100-bp DNA size markers. Lane 2 contains
products from a positive control using purified M femientans DNA
template. Lane 3 represents a negative control from the same PCR,
performed without added DNA template. Lanes 4-20 contain PCR
products using purified DNA from RA synovial fluid samples. In A,
there was no M homznis species-specific product in the PCR from RA
patient samples, although there was species-specific product in the
positive control (lane 2 at 281 bp, as indicated by the arrow). In B,
there was no M fernentans species-specific product in the PCR from
RA patient samples, although there was species-specific product in the
positive control (lane 2 at 272 bp, as indicated by the arrow).
the synovial fluid from either the patient or control
groups. Representative results are shown in Figures 5
and 6. As shown in Figure 5, the molecular mimic was
detected in all samples, thus confirming that neither
inadvertent omissions from, nor inhibition of, the PCR
could account for the findings. Appropriate positive and
negative controls were also included in each experiment.
There has been longstanding interest in the potential role of mycoplasmas as etiologic agents of RA.
To address this question, we used immunoblotting of
detergent-phase extractions of M hominis and M jermentans in the present study, and we developed and
tested an ultrasensitive PCR-based molecular diagnostic
method for detection of Mycoplasma DNA.
As illustrated in Figure 1 and summarized in
Table 2, serologic reactivity against Mycoplusma, using
immunoblotting with membrane protein extracts of M
hominis or M fernentans, was common. We found that
55% of RA and 88% of JRA patient synovial fluid
samples exhibited IgG antibody binding to 2 1 M hominis antigen. Similar results were found using matched
samples of sera and synovial fluid from representative patients as controls. Targets of this serologic reactivity were identified in some cases, using a series of
Mycoplasma-specific MAb (Figure 1). However, serologic reactivity with Mycoplasma antigens was not restricted to RA or JRA synovial fluid or sera. A similar,
high frequency of serologic (IgG) reactivity with Mycoplasma antigens was also seen in the synovial fluid from
controls with either inflammatory or noninflammatory
arthritis (Table 2), and in the sera of normal donors.
Taken together, these data suggest that a high cumulative background frequency of Mycoplasma exposure
existed in our study population. These results are consistent with those obtained by Hooton et al, who conducted a cross-sectional analysis of the infectious agent
Mgenitalium. Using PCR, Hooton et al could detect M
genitalium in the urine of 7% of asymptomatic heterosexual men, by PCR (40).
While this apparent broad range of serologic
recognition of M hominis and M fernentans antigens
occurred among all groups analyzed, a striking feature of
1226
the overall pattern was the recognition by individuals of
a distinctive set of membrane antigens. Moreover, some
of those recognized components (e.g., P78, P61, and P29
of M fermentans, or P120 and the size variable Vaa
antigen of M hominis) were subject to high-frequency
phase variation even in clonal populations (see refs. 22
and 38, and Zhang and Wise: unpublished observations)
Although speculative at this time, it is an intriguing
possibility that individuals might be exposed to variants
of these agents which differ in their detailed antigenic
makeup. However, neither the status of Mycoplasma
infection, nor the actual stimulus for the IgG response to
these 2 species, could be formally determined.
To examine the possibility that mycoplasmas
might act as agents that colonize the major site of
inflammation in RA patients, it was necessary to develop
an assay that could detect very small numbers of a target
gene present in a wide spectrum of Mycoplasma species,
even in the presence of synovial fluid. Based on previous
studies (25) and independent computer analysis, the
genus-specific primers selected had a high degree of
homology to all known Mycoplasma species found in
humans, as well as to those found in other hosts. Figure
2 illustrates that the genus-specific primers selected for
these experiments could amplify DNA from many Mycoplasma species known to be human pathogens. Extensive studies were then done to optimize the conditions of
the PCR assay, such that <5 gene copies could be
reproducibly detected. Figure 3 illustrates that, using the
genus-specific primers, the assays developed had this
level of sensitivity. Similar levels of detection were
obtained using M hominis and M fermentans speciesspecific primers.
Based on the results of our pilot studies and on
the published work of others, it was known that synovial
fluid is a difficult fluid to use for PCR because of poorly
characterized inhibitors (41). To address this problem,
synovial fluid from a patient with proven septic arthritis
caused by M hominis was used (20,39). Several alternative methods of DNA purification were tested using this
synovial fluid. Subsequently, the fluid was used to optimize the PCR assay and to test the sensitivity of the final
PCR-based assay reported herein.
Particular attention was given to the handling of
samples and to the inclusion of appropriate positive and
negative controls in each experiment. As discussed in
Patients and Methods, rigorous care was taken in the
handling, processing, and flow of all samples in these
experiments. Furthermore, because there is the potential for sample loss during DNA isolation, pipetting, or
other stages in the process, we constructed a nonho-
HOFFMAN ET AL
mologous internal standard, or molecular mimic, to
include in these experiments. The presence of the amplified mimic on ethidium-stained agarose gels served to
exclude the possibilities that a negative result could be
due to sample loss, or to an inhibitor in the PCR (i.e., a
false negative). Finally, because the molecular mimic
could compete with the Mycoplasma gene for amplification when both were present at near equimolar amounts
(shown in Figure 4, lane 4), separate PCR analyses were
done on all samples with and without the mimic.
The results of the PCR analyses on patient and
control samples are illustrated in Figures 5 and 6 .
Mycoplasma DNA could not be detected in any of the
samples, either in the intial experiments, or after repeat
testing utilizing the ultrasensitive PCR-based detection
method. All of the positive controls and the molecular
mimic were detectable at a detection level of <5 gene
copies in these experiments. Thus, there was no evidence
for Mycoplasma DNA at this level in any of the RA or
JRA synovial samples.
There are at least 2 interpretations of the data
reported herein regarding the relationship between Mycoplasma infection and RA. First, serologic results suggest that Mycoplasma exposure is common in this study
population (Table 2). For example, antibodies to M
hominis were present in 77% of the study patients. These
findings are consistent with a model in which Mycoplasma is a cofactor for the development of RA in a
genetically susceptible host, but in which, once this
autoimmune process is initiated, the organism is cleared
from synovial fluid and tissue. Other studies examining
correlations between reactive arthritis and potential
microbial etiologic agents are noteworthy in this regard
(for review, see refs. 35,42, and 43). While organisms or
material from Chlamydia or Yersinia species have, in
some cases, been reported in the synovial tissue or fluid
of patients with reactive arthritis, other evidence from
animal models indicates that the Yersinia agent may be
absent from the inflamed joint, yet may be revealed at
other anatomic sites during periods without joint inflammation (43). Thus, the possibility that slow Mycoplasma
infection at various sites may contribute to RA cannot
be formally ruled out, but the persistence of its presence
at the site of joint inflammation is strongly refuted by
our present study findings.
An alternative interpretation consistent with our
findings is that Mycoplasma does not play an etiologic
role in RA or JRA. Our results show that the methods
used are extremely sensitive, reliably detecting <5 gene
copies of Mycoplasma DNA from synovial fluid. We
would expect that, if there was persistent Mycoplasma
MYCOPLASM INFECTION IN RA
infection in the RA joint, these experiments would have
shown its presence. Moreover, in support of this interpretation, recent studies have shown that antibiotics that
have efficacy in clinical trials on RA (15,16) are potent
inhibitors of metalloproteinases and have antiinflammatory properties in addition to their antimicrobial properties (44). This tends to weaken the indirect argument
that microbial agents are involved in RA.
Only one previously published study, to our
knowledge, has examined RA synovial fluid for Mycoplasma using PCR. Schaeverbeke et a1 examined synovial fluid or tissues for evidence of DNA from M
fermentans using PCR (45). They found evidence of
Mycoplasma using M fermentans-specific primers on
synovial fluid or tissue from 15 of 154 specimens,
including 8 of 38 with RA, 2 of 10 with spondylarthropathy, 1 of 5 with psoriatic arthritis, and 4 of 31 with
unclassified inflammatory arthritis (45). The presence of
M fermentans in patients with a variety of rheumatic
diseases of presumed varying pathogenesis appears consistent with passive carriage of M fermentans in the study
populations, rather than with an etiologic role in these
diseases. In this regard, it is of interest that, using a
PCR-based assay, Chingbingyong and Hughes recently
found that M fermentuns could be detected in the saliva
of 49 of 110 healthy adults (44.5%), thus suggesting that
M fermentans commonly colonizes human mucosal surfaces (46). In the present study, using an ultrasensitive
PCR-based assay, we failed to detect evidence of M
fermentans DNA in either the synovial fluid or tissue.
This difference compared with the results obtained by
Schaeverbeke et a1 could reflect a higher frequency of
passive carriage of M fermentans in their study population, or a persistence of the organism in patients earlier
in their disease course.
The data presented herein are useful in addressing the 2 alternative hypotheses regarding the pathogenesis of Mycoplasma in RA. Our results provide strong
evidence against chronic local Mycoplasma infection as
the cause of RA or JRA. Additional studies, such as
those on early synovitis, will be required to exclude the
alternative hypothesis that RA is caused by an immune
response to a previous Mycoplasma infection.
In summary, we have described a systematic
analysis of RA and JRA synovial fluid and tissue using
immunoblotting and an ultrasensitive PCR-based assay
for the detection of Mycoplasma DNA. We report
serologic evidence that suggests that previous exposure
to Mycoplasma may be common, but, using an ultrasensitive PCR-based assay, there is no detectable persis-
1227
tence of Mycoplasma DNA in the synovial fluid or tissue
of patients with established RA or JRA.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the excellent technical assistance of Don L. Hill, as well as Herman T. Miller,
PhD, for his assistance in numerous aspects of this project, the
patients and their physicians for generously providing sample
material for the study, and, especially, Nancy Brown, DO,
Geetha Komatireddy, MD, Barry J. Gainor, MD, William C.
Allen, MD, and Walter M. Greene, MD, for their assistance in
providing clinical material for this study.
REFERENCES
1. Burgdorfer W, Barbour AG, Hayes SF, Benach JL, Grunwald TE,
Davies JP: Lyme disease-a tick borne spirochetosis? Science
216:1317-1319, 1982
2. Alpert E, Isselbacher KJ, Schur PH: The pathogenesis of arthritis
associated with viral hepatitis. N Engl J Med 285:185-189, 1971
3. Winchester R, Bernstein DH, Fischer HD, Enlow R, Solomon G:
The co-occurrence of Reiter’s syndrome and acquired immunodeficiency. Ann Intern Med 106:19-26, 1987
4. Smith CA, Woolf AD, Lenci M: Parvoviruses: infection and
arthropathies. Rheum Dis Clin North Am 13:249-263, 1987
5. Andrews BE: Mycoplasma hominis. Commun Dis Rep Public
Health Lab Serv (UK) 31, 1974
6. So AKL, Furr PM, Taylor-Robinson D, Webster ABD: Arthritis
caused by Mycoplasma salivarium in hypogammaglobulinemia.
BMJ 286:762-763, 1983
7. Weinstein MP, Hall CB: Mycoplasma pneumonia infection associated with migratory polyarthritis. A m J Dis Child 127:125-126,
1974
8. Taylor-Robinson D, Gilroy CB, Horowitz S, Horowitz J: Mycoplasma genitalium in the joints of two patients with arthritis. Eur
J Clin Microbiol Infect Dis 13:1066-1069, 1994
9. Davis CP, Cochran S, Lisse J, Buck G, DiNuzzo AR, Weber T,
Reinarz J A Isolation of Mycoplasma pneumoniae from synovial
fluid samples in a patients with pneumonia and polyarthritis. Arch
Intern Med 148:969-970, 1988
10. Furr PM, Taylor-Robinson D, Webster ADB: Mycoplasma and
ureaplasmas in patients with hypogammaglobulinemia and their
role in arthritis: microbiologic observations over twenty years. Ann
Rheum Dis 53:183-187, 1994
11. Cole BC, Washburn LR, Taylor-Robinson D: Mycoplasmainduced arthritis. In, The Mycoplasmas. Edited by S Razin, MF
Barile. New York, Academic Press, 1985
12. Taylor-Robinson D: Genital mycoplasmas. Curr Opin Infect Dis
8:16-21, 1995
13. Ainsworth JG, Katseni V, Hourshid S, Waldron S, Ball S, Cattell
V, Taylor-Robinson D: Mycoplasma fermentans and HIVassociated nephropathy. J Infect 29:323-326, 1994
14. Lo C, Hayes MM, Wang H, Pierce PF, Kotani H, Shih K Newly
discovered mycoplasma isolated from patients infected with HIV.
Lancet 338:1415-1418, 1991
15. Kloppenburg M, Breedveld FC, Tenviel JPH, Mallee C, Dijkmans
BAC: Minocycline in active rheumatoid arthritis. Arthritis Rheum
37529-636, 1994
16. Tilley BC, Alarcon GS, Heyse SP, Trentham DE, Neuner R,
Kaplan DA, Clegg DO, Leisen JCC, Buckley L, Cooper SM,
Duncan H, Pillemer SR, Tuttleman M, Fowler SE: Minocycline in
rheumatoid arthritis: a 48-week, double-blind, placebo-controlled
trial, Ann Intern Med 122:81-89, 1995
1228
17. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF,
Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS,
Medsger TA Jr, Mitchell DM, Neustadt DH, Pinals RS, Schaller
JG, Sharp JT, Wilder RL, Hunder GG: The American Rheumatism Association 1987 revised criteria for the classification of
rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988
18. Cassidy JT, Levinson JE, Bass JC, Brewer EJ Jr :The development
of classification criteria for children with juvenile rheumatoid
arthritis. Bull Rheum Dis 38:1-7, 1998
19. McCarty DJ, Koopman WJ, editors: Arthritis and Allied Conditions. Twelfth edition. Philadelphia, Lea & Febiger, 1993
20. Zhang Q, Wise KS: Molecular basis of size and antigenic variation
of a Mycoplasma hominis adhesin encoded by divergent vaa genes.
Infect Immun 64:2737-2744, 1996
21. Wise KS, Kim MF, Theiss PM, Lo S-C: A family of strain-variant
surface lipoproteins of Mycoplasma fermentans. Infect Immun
6113327-3333, 1993
22. Theiss PM, Kim MF, Wise KS: Differential protein expression and
surface presentation generates high-frequency antigenic variation
in Mycoplasma fermentans. Infect Immun 61:5123-5128, 1993
23. Olson LD, Renshaw CA, Shane SW, Barile MF: Successive
synovial Mycoplasma hominis isolates exhibit apparent antigenic
variation. Infect Immun 59:3327-3329, 1991
24. Strauss WM: Preparation of genomic DNA from mammalian
tissue. In, Current Protocols in Molecular Biology. Edited by FM
Ausubel, R Brent, DD Moore, J G Seidman, JA Smith, K Struhl.
New York, Wiley, 1995
25. Van Kuppeveld FJM, van der Logt JTM, Angulo AF, van Zoest
MJ, Quint WGV, Niesters HGM, Galama JMD, Melchers WJG:
Genus- and species-specific identification of mycoplasmas by 16s
rRNA amplification. Appl Environ Microbiol .58:2606-261.5, 1992
26. Van Kuppeveld FJM, van der Logt JTM, Angulo AF, van Zoest
MJ, Quint WGV, Niesters HGM, Galama JMD, Melchers WJG:
Genus- and species-specific identification of mycoplasmas by 16s
rRNA amplification: authors’ correction. Appl Environ Microbiol
59:655, 1993
27. Harasawa R, Koshimizu K, Kitagawa M, Asada K, Kato I:
Nucleotide sequence of the arginine deiminase gene of Mycoplasma hominis. Microbiol lmmunol 36:661-665, 1992
28. Spaepen M, Angulo AF, Marynen P, Cassiman J-J: Detection of
bacterial and mycoplasma contamination in cell cultures by polymerase chain reaction. FEMS Microbiol Letters 90:89-94, 1992
29. Schmidt TM, Relman D A Phylogenetic identification of uncultured pathogens using ribosomal RNA sequences. Methods Enzymol235:205-222, 1995
30. Wise KS, Yogev D, Rosengarten R: Antigenic variation. In,
Mycoplasmas: Molecular Biology and Pathogenesis. Edited by J
Maniloff, RN McElhaney, LR Finch, JB Baseman. Washington,
DC, American Society for Microbiology, 1992
31. Wang RY-H, Shih JW-K, Hayes MM, Alter HJ, Wear DJ, Lo S-C:
Antibodies to Mycoplasma penetrans in HIV-infected patients:
reply (letter). Lancet 341:558, 1993
HOFFMAN ET AL
32. Wang RY-H, Shih JW-K, Weiss SH, Grandinetti T, Pierce PF,
Lange M, Alter HJ, Wear DJ, Davies CL, Mayur RK, Lo S-C:
Mycoplasma penetrans infection in male homosexuals with AIDS:
high seroprevalence and association with Kaposi’s sarcoma. Clin
Infect Dis 17:724-729, 1993
33. Rosengarten R, Wise KS; The Vlp system of Mycoplasma hyorhinis: combinatorial expression of distinct size variant lipoproteins
generating high-frequency surface antigenic variation. J Bacteriol
173:4782-4793, 1991
34. Wise KD, Kim MHF, Watson-McKown R: Variant membrane
proteins. In, Molecular and Diagnostic Procedures in Mycoplasmology. Edited by S Razin, JG Tully. New York, Academic Press,
1995
35. Cole BC: Mycoplasma interactions with the immune system:
implications for disease pathology. Am Soc Microbiol News 62:
471-475, 1996
36. Verinder DGR: Septic arthritis due to Mycoplasma hominis: a
case report and review of the literature. J Bone Joint Surg [Br]
60:224, 1978
37. Hakkarainen K, Turunen H, Miettinen A, Karppelin M, Kaitila K,
Jansson E: Mycoplasmas and arthritis. Ann Rheum Dis 51:11701172, 1992
38. Theiss PM, Karpas A, Wise KS: Antigenic topology of the P29
surface lipoprotein of Mycoplasma fermentans: differential display
of epitopes results in high-frequency phase variation. Infect Immun 64:1800-1809, 1996
39. Olson LD, Shane SW, Karpas AA, Cunningham TM, Probst PS,
Barile MF: Monoclonal antibodies to surface antigens of a pathogenic Mycoplasma hominis strain. Infect Immun 59:1683-1689,
1991
40. Hooton TM, Roberts MC, Roberts PL, Holmes KK, Stamm
WE, Kenny GE: Prevalence of Mycoplasma genitalium determined by DNA probe in men with urethritis. Lancet I:266-268,
1988
41. Malawista SE, Schoen RT, Moore TL, Dodge DE, White TJ,
Persing DH: Failure of multitarget detection of Borrelia
burgdorferi-associated DNA sequences in synovial fluids of patients with juvenile rheumatoid arthritis: a cautionary note. Arthritis Rheum 35246-247, 1992
42. Toivanen A: Infection and arthritis. Ann Med 26:245-248, 1994
43. Toivanen P, Toivanen A: Role of micro-organisms in the pathogenesis of arthritis: lessons from reactive and Lyme arthritis. %and
J Rheumatol 24:191-197, 1995
44. Nip LH, Uitto VJ, Golub LM: Inhibition of epithelial cell matrix
metalloproteinases by tetracyclines. J Periodontal Res 28:379-385,
1993
45. Schaeverbeke T, Gilroy CB, Bebear C, Dehais J, Taylor-Robinson
D: Mycoplasma fermentans in joints of patients with rheumatoid
arthritis and other joint diseases (letter). Lancet 347:1418, 1996
46. Chingbingyong MI, Hughes CV: Detection of Mycoplasma fermentans in human saliva with a polymerase chain reaction-based
assay. Arch Oral Biol41:311-314, 1996
Документ
Категория
Без категории
Просмотров
7
Размер файла
1 091 Кб
Теги
immunoblotting, using, reaction, chains, detection, method, relationships, arthritis, mycoplasma, analysis, infectious, polymerase, rheumatoid, ultrasensitive
1/--страниц
Пожаловаться на содержимое документа