SEARCH

SEARCH BY CITATION

Keywords:

  • Bacteria;
  • Gas chromatography-mass spectrometry;
  • Pan-bacterial PCR;
  • Rheumatoid arthritis

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Objective

To study the presence of bacterial components in the synovial tissue (ST) of patients with advanced rheumatoid arthritis (RA).

Methods

ST was collected during joint surgery from 41 RA patients. Tissue from 39 patients with osteoarthritis (OA), 4 patients with undifferentiated inflammatory arthritis (UA), and 3 cases of accidental deaths served as controls. The pan-bacterial polymerase chain reaction (PCR) with primers for the 23S ribosomal RNA (rRNA) and 16S rRNA genes was used to detect bacterial DNA. In addition, synovial fluid (SF) samples from patients with chlamydial reactive arthritis (ReA) were also examined by the same method. The positive controls, bacterial DNA or ST spiked with different living bacteria, were analyzed alongside clinical samples. Most of the ST samples were also analyzed by gas chromatography-mass spectrometry (GC-MS) for determining the presence of bacteria-derived muramic acid. Strict precautions were followed in the clinics and the laboratory to prevent contamination.

Results

In GC-MS analysis, muramic acid was observed in the ST from 4 of 35 RA patients and from 2 of 14 OA patients, but not in ST from 2 patients with UA and 3 cadavers. Bacterial DNA was not detected by either one of the PCR primers used in ST from 42 patients with RA and 39 patients with OA. However, 5 of 15 SF samples from ReA patients were PCR positive. The sensitivity of GC-MS to detect muramic acid was 2 pg/injected amount (227 pg muramic acid/mg ST), and that of the pan-bacterial PCR was 2–20 bacteria colony forming units/reaction.

Conclusion

These results indicate that a bacterial component, muramic acid, is detectable by GC-MS in ST from a few patients with advanced RA or OA. However, no bacterial DNA was detectable by PCR.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

It has long been hypothesized that infectious agents might play a role in the etiopathogenesis of rheumatoid arthritis (RA) (1). Among them, bacteria are the most important ones regarding diseases of the locomotor system (2–6). Several lines of observation have been used to support this hypothesis. 1) Certain forms of arthritides are triggered by bacteria. In addition to septic (bacterial) arthritis, they include arthritides connected to enterogenic and urogenic infections (reactive arthritis [ReA]), rheumatic fever, Lyme borreliosis, tuberculosis, leprosy, and Whipple's disease (4–9). 2) Humoral or cellular immune responses to certain bacteria and bacterial components have been observed in RA (10–12). 3) The intestinal microbiota of RA patients has been found to be different from that in controls (13). 4) Some antimicrobial agents have been reported to be effective in the treatment of RA, though it is not necessarily attributed to the antibacterial effect (14). 5) Certain bacteria or bacterial components are able to induce experimental chronic arthritis closely resembling RA (15–17). However, all this evidence is circumstantial, and no bacterial species has reproducibly been identified as a causative agent for RA. Nevertheless, it is possible that the causative agents or their degradation products are present at very low concentrations in synovial tissue (ST). Therefore, polymerase chain reaction (PCR), with its high sensitivity and specificity, appears a suitable technology to detect residual live or perhaps dead organisms within the ST. Both microbe-specific PCR (18, 19) and pan-bacterial PCR have been applied in RA studies (20–25). The latter technique, usually based on highly conserved bacterial 16S or 23S ribosomal RNA (rRNA) genes, is much more attractive because the potential trigger of RA is unknown. Using PCR for 16S rRNA, the presence of bacterial DNA in ST from people with a variety of inflammatory arthritides has been reported (23–25). To confirm this, we have applied the PCR with pan-bacterial 23S rRNA and 16S rRNA primers, both of which have successfully been used for bacterial identification in various clinical samples (26).

Gas chromatography-mass spectrometry (GC-MS) is a highly selective and sensitive method to determine trace amounts of bacterial components in complex clinical samples (27, 28). GC-MS analysis has also been used in arthritis studies (29–32). It is possible that only the bacterial components, not necessarily living bacteria, are present in rheumatoid ST. Such a possibility is supported by the experience gained from bacterial cell wall (BCW) arthritis (15–17). With this consideration, GC-MS analysis of ST to demonstrate a bacterial marker appears a suitable method in searching for the bacterial components other than DNA. Muramic acid, a compound unique to bacterial peptidoglycan and not present elsewhere in nature, is such a chemical marker for bacteria (33). For this reason, in addition to PCR, we have applied GC-MS to detect bacterial muramic acid in ST of patients with RA.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Patients and samples.

All studies involving human subjects were approved by the local ethical committee. The patients enrolled were from Satalinna Hospital, Turku University Central Hospital, and Turku City Hospital. ST was collected aseptically during joint surgery from 85 patients for PCR study. Patients with a recent infection within 1 month, or with a chronic infectious history, or receiving antibiotic therapy within 2 weeks before the time of sampling were excluded. Each ST sample was divided into 4 pieces (1 cm × 1 cm), and was blindly coded so that the diagnosis was unknown to the investigators. Samples were immediately frozen and stored at −135°C without treatment with heparin or hyaluronidase.

Forty-two patients with advanced RA (duration of the disease >5 years) and fulfilling the 1987 revised American College of Rheumatology criteria are included (34). Thirty-nine patients with osteoarthritis (OA) and 4 with undifferentiated inflammatory arthritis (UA) served as controls (Table 1). As PCR controls, synovial fluid (SF) samples were collected during therapeutic aspiration from knee joints of 15 patients with chlamydial ReA. The diagnoses of chlamydial ReA were based on clinical and/or laboratory evidence, including in most cases microbe-specific PCR for the urinary or cervical samples. Clinical features of the patients are given in Table 1. Among those 85 ST samples used for PCR study, the ST from 51 arthritis patients (35 RA, 14 OA, and 2 UA) were blindly and randomly selected for GC-MS analysis. The same samples were also used in the PCR study. In addition, ST and muscle tissues derived from 3 cadavers (accidental deaths) served as negative controls for GC-MS.

Table 1. Characterization of patients in the 23S rRNA and 16S rRNA PCR study*
GroupNumber of patientsAge, years mean (SD)Duration of disease, years mean (SD)
  • *

    rRNA = ribosomal RNA; PCR = polymerase chain reaction; SD = standard deviation; RA = rheumatoid arthritis; OA = osteoarthritis; UA = undifferentiated inflammatory arthritis; ReA = reactive arthritis; ND = not determined.

RA4260.7 (11.6)12.7 (10.8)
OA3974.6 (8.3)>5
UA450.5 (19.1)>5
ReA1555.5 (16.7)ND

GC-MS.

Sample preparation for GC-MS.

Briefly, ST was first homogenized with an Ultra Turrax T23 tissue homogenizer (Janke & Kunkel, IKA Labortechnik, Staufen, Germany) in 2 ml of water. Suspensions of ST (100 μl) were evaporated to dryness under a nitrogen stream at 40°C, then methanolyzed under a nitrogen atmosphere at 85°C for 24 hours in 2 ml of 4 M methanolic hydrochloric acid. Hexane (SupraSolv purity; Merck, Rahway, NJ) extraction (3 ml) was performed for the methonolysate. The samples were derivatized with 50 μl of acetonitrile and 50 μl of trifluoroacetic anhydride at 80°C for 5 minutes. After cooling, the derivatized samples were diluted with 400 μl of toluene, followed by 2 extractions with 1 ml of water. Fifty-microliter aliquots were taken from the toluene phase into the sample vials and diluted with an equal amount of derivatized internal standard, and 1 μl was injected into the gas chromatograph. The internal standard of N-methyl-D-glucamine (Sigma Chemical Co, St. Louis, MO) was processed in the same way as the samples, except no water extraction was performed.

Standard for GC-MS.

Because free muramic acid cannot be used as a standard for negative chemical ionization (NCI), cell walls of Eubacterium limosum were used as the muramic acid standard (17, 27). The standard curve was linear with muramic acid (Sigma) concentrations from 2 to 23 pg (the final injected amount), with the final concentration of the internal standard at 15 pg.

Determination of muramic acid by GC-MS.

The concentration of muramic acid in the ST was analyzed as trifluoroacetylated methyl glycoside derivatives by GC-MS using NCI as previously described, with some modifications (17, 27). Briefly, the GC-MS analyses were performed with a gas chromatograph (model 6890; Hewlett-Packard, Little Falls, DE) coupled to a mass selective detector (model 5973; Hewlett-Packard, Palo Alto, CA). The gas chromatograph was equipped with HP-5MS capillary column (30 m × 0.25 mm × 0.25 μm film thickness; Agilent Technologies, Palo Alto, CA). Helium served as the carrier gas, and isobutane as the reagent gas.

Starting temperature of the oven (80°C) was held for 2.5 minutes and programmed at the rate of 8°C/minute to 210°C, and held for 1 min. As a postrun step, the oven was heated to 300°C for 7 minutes. Pulsed splitless injection mode with a slow plunger speed was used with 2.5 minutes as the pulse time. For muramic acid quantification, molecular ion m/z = 567 (m/z = mass to charge ratio) was used as the target ion, and m/z = 480 and m/z = 453 were used as qualifier ions. A computerized quantitative internal standard method for the determination of muramic acid was performed as described in the Hewlett-Packard G1701AA MS Chemstation workbook (DOS Series). The detection limit was 2 pg defined as the amount of muramic acid giving a signal to noise ratio of about 40:60 (peak to peak noise). Negative sample control (pure water) and reagent control were always included alongside the clinical samples. The cleaning run with toluene was performed after each sample run to prevent contamination. Analysis of all the samples by GC-MS was repeated at least 3 times.

Glassware used was first treated with Decon 90 (Decon Laboratories, Sussex, UK), and 10% Deconex 11 Universal (Borer Chemie AB, Zuchwil, Switzerland), heated at 170°C for 2 hours before autoclaving for 20 minutes. Water used for GC-MS was double distilled, reverse osmosis purified, and autoclaved.

PCR.

DNA extraction and PCR procedure.

DNA extraction and PCR were performed as previously described, with some modification (26). Briefly, DNA was extracted from fresh ST samples by proteinase K (0.1 mg/ml) digestion (56°C 2–14 hours) with 2 phenol-chloroform-isoamyl alcohol extractions followed by 1 ether wash. SF samples (500 μl) were first concentrated by centrifugation and were then treated similarly as ST (35).

The primers for 23S and 16S rDNA PCR have been described before. Both primer pairs, MS-37 and MS-38 for the 23S PCR and fD1 and rP2 for 16S PCR, have been used for diagnosis of bacterial infection for a variety of clinical samples (26, 36, 37).

A 5-μl aliquot of DNA was used as a template for DNA amplification. For the 23S rRNA gene PCR, the PCR steps (94°C for 45 seconds, 60°C for 1 minute, 72°C for 2 minutes) were repeated for 34 cycles using DNA Thermal Cycler 480 (Perkin-Elmer, Emeryville, CA). For the 16S rRNA gene PCR, the PCR steps (94°C for 1 minute, 55°C for 30 seconds, 72°C for 1 minute) were repeated for 35 cycles using GeneAmp PCR System 2400 thermocycler (Perkin-Elmer, Norwalk, CT).

After amplification, PCR products were separated by agarose gel electrophoresis and visualized as ultraviolet (UV) fluorescence after staining with ethidium bromide.

Prevention of contamination.

Broad range PCR is particularly sensitive to contamination. Strict precautions were followed in the clinics and the laboratory to prevent contamination. All steps of the PCR procedure (sample preparation, DNA extraction, PCR mixture preparation, addition of sample DNA, PCR amplification, and analysis of PCR products) were performed in the 6 physically separated areas. To prevent carryover contamination, 2 investigators were responsible for pre-PCR and post-PCR experiments, respectively. DNA purification was performed in the “PCR room” equipped with positive air pressure. In the “PCR laminar flow room,” PCR mixtures were done in a sterile laminar flow, used only for this purpose. Sample DNA was added in another laminar flow, using positive-displacement pipettes (Microman, Gilson, France). When not in use, the laminar flows and rooms were illuminated with UV lamps. Protective clothes and cap were changed on entering the “PCR room” and only selected investigators were allowed entrance. To destroy possible contaminated bacterial DNA present in PCR reagents, PCR reagents were irradiated with UV light (Hanau Fluotest, Chicago, IL) together with 8-methoxypsoralen. Moreover, a negative DNA isolation control, i.e., pure water prepared like the samples, was analyzed alongside each clinical sample. In addition, a negative PCR control consisting of all reagents necessary for PCR, with water as a template, was always included. Aseptic techniques were carried out through the experimental procedure. Laboratory coats and caps were worn at all times. Disposable gloves were changed frequently. Positive-displacement pipettes and laminar flow were used. Water used for PCR was double distilled, reverse osmosis purified, and autoclaved, and further sterilized by UV light for 2 hours prior to use. Only freshly prepared or properly stored unused reagents and buffer were used. Tested reagents were aliquoted for single-usage volume for storage.

Prevention of false-negative results.

It has been demonstrated that some clinical samples may contain inhibitors for the PCR, e.g., for DNA polymerase. As a sample positive control, a part of human β-globin gene was always amplified from all purified sample DNA preparations, as described before (26, 35). Furthermore, the PCR positive controls, including purified bacterial DNA from Neisseria meningitidis or Bacillus subtilis, were amplified alongside each clinical sample. In addition, in some cases, ST samples spiked with 40–4,000 colony forming units (CFU)/ml bacteria (Staphylococcus aureus, Streptococcus pyogenes, Escherichia coli, Salmonella enteritidis) were used as positive sample controls.

Statistical analysis.

Analysis of variance was used to test for differences between the groups. Samples with equal distribution were compared using the Student's t-test. Differences were considered statistically significant when P < 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

GC-MS.

The sensitivity of GC-MS analysis for muramic acid was 2 pg/μl (injected amount) at a signal to noise ratio of 40:60, giving a sensitivity of 227 pg muramic acid/mg ST (dry weight). With this detection limit, the negative sample and reagent controls were always negative. The results of GC-MS are summarized in Table 2. Muramic acid was detected in the ST from 4 of 35 RA patients, and in the ST from 2 of 14 OA patients. It was not observed in the ST from 2 cases with UA and in the ST or muscle samples of 3 cadavers. No significant difference is apparent between RA and OA regarding concentration of muramic acid or frequency of its occurrence. The authenticity of the muramic acid detected in GC-MS was confirmed by running all samples with added BCW (E limosum) containing muramic acid (8 pg). Representative results of GC-MS analyses are shown in Figure 1. All GC-MS analyses were repeated at least 3 times.

Table 2. Muramic acid concentration in ST determined by GC-MS*
GroupNo. of patients positive for muramic acid/ no. of patients studiedMuramic acid concentrations in ST, range (ng/mg)
  • *

    ST = synovial tissue; GC-MS = gas chromatography–mass spectrometry; RA = rheumatoid arthritis; OA = osteoarthritis; UA = undifferentiated inflammatory arthritis.

  • Detection limit: 227 pg muramic acid per mg ST (dry weight).

RA4/351.39–2.20
OA2/140.18–2.66
UA0/2
Cadavers0/3
thumbnail image

Figure 1. Detection, by gas chromatography-mass spectrometry (GC-MS), of bacterial muramic acid in the synovial tissue (ST) of rheumatoid arthritis (RA) patient. The x axis represents the running time (minutes), and the y axis represents the abundance of molecular ion of muramic acid (mass to charge ratio = 567). The samples were derivatized by trifluoroacetic anhydride and analyzed by negative chemical ionization method. A, Detection of muramic acid (5.2 pg) in the ST from an RA patient. B, The positive result was further confirmed by adding authentic Eubacterium limosum bacterial cell wall (BCW) of a known amount (8 pg) of muramic acid into the ST analyzed in A; both curves are shown. C, Failure of detection of muramic acid (<2 pg) in the ST from a cadaver. D, Standard muramic acid (2 pg and 8 pg) from BCW of E limosum.

Download figure to PowerPoint

PCR analyses.

The sensitivity of pan-bacterial PCR was 2–20 CFU/reaction, which was determined by mixing living bacteria with homogenized ST and using exactly the same experimental procedure as for the patient samples (Figure 2). Analyses of ST from 42 RA patients and 39 OA patients with pan 23S rRNA and 16S rRNA primers did not reveal the presence of bacterial DNA (Figure 3). Furthermore, using the 23S rRNA PCR, 5 (33%) of 15 SF samples from patients with chlamydial ReA were found PCR positive (Figure 4). Representative PCR results are shown in Figures 2–4.

thumbnail image

Figure 2. Sensitivity of polymerase chain reaction (PCR) for bacterial 16S and 23S ribosomal DNA (rDNA). Representative results of agarose-gel electrophoresis of PCR-amplified fragments are shown. The sensitivity was determined by mixing serial dilutions of living bacteria together with homogenized ST from the same osteoarthritis patient. A,DNA extracted from synovial tissue (ST) spiked with Staphylococcus aureus, amplified using primers for 16S rDNA. B and C, DNA extracted from ST spiked with S aureus (B) or Escherichia coli (C), amplified using primers for 23S rDNA. Lane MWM = molecular weight marker; lanes 1–6 = DNA extracted from ST spiked with 2000, 200, 20, 2, 0.2, 0.02 colony forming units (CFU) bacteria/reaction; lane ST = DNA extracted from ST, no spiking; lane + = DNA extracted from Neisseria menigitidis, as PCR positive control, lane – = DNA-free water, as PCR negative control.

Download figure to PowerPoint

thumbnail image

Figure 3. Failure to detect bacterial DNA in synovial tissue (ST) from patients with advanced rheumatoid arthritis (RA) or osteoarthritis (OA), by using PCR for 16S ribosomal DNA (rDNA) and 23S rDNA. Results of agarose gel electrophoresis of polymerase chain reaction (PCR)-amplified fragments are shown. A, DNA extracted from ST samples were amplified using primers for 16S rDNA. B,DNA extracted from ST samples were amplified using primers for 23S rDNA. C,DNA extracted from the same ST samples were amplified using primers for human β-globin gene, as a positive sample control. Lane MWM = molecular weight marker; lanes OA 1–3 and RA 1–4 = DNA extracted from ST from 3 patients with OA and 4 patients with RA, respectively; lane N = DNA-free water, as negative sample control; lane + = DNA extracted from Neisseria menigitidis, as positive PCR control; lane – = DNA-free water, as negative PCR control.

Download figure to PowerPoint

thumbnail image

Figure 4. Detection of bacterial DNA in synovial fluid (SF) from patients with chlamydial reactive arthritis (ReA). Representative results of agarose gel electrophoresis of polymerase chain reaction (PCR)-amplified fragments are shown. DNA extracted from SF was amplified using primers for 23S ribosomal RNA gene. Lane MWM = molecular weight marker; lanes 1–7 = DNA extracted from SF of 7 chlamydial ReA patients; lane + = DNA extracted from Neisseria menigitidis, as positive PCR control; lane – = DNA free water, as negative PCR control.

Download figure to PowerPoint

In the PCR analyses, the negative controls were always negative. Likewise, the positive controls, a part of human β-globin gene and purified bacterial DNA from N meningitidis or B subtilis were always positive. In addition, a few ST samples from RA or OA patients were spiked with 40–4,000 CFU/ml of bacteria (S aureus, S pyogenes, E coli, S enteritidis) and gave positive PCR results. All the analyses were repeated at least twice.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

In the present study using GC-MS, we observed the presence of bacterial muramic acid in ST of a few (11–14%) patients with RA or OA. GC-MS is a highly sensitive method, which has been applied to determine trace amounts of bacterial components in the SF; the arthritides studied include septic, reactive, and undifferentiated inflammatory arthritis (29–32). Muramic acid, a structural component of peptidoglycan, is a chemical marker for bacteria and not present elsewhere in the nature. Thus, the presence of muramic acid indicates that bacterial components exist in the inflamed joints of some RA or OA patients. Our results are consistent with several lines of previous observations. Bacterial peptidoglycan has been demonstrated in the joint tissues of patients with RA and other types of arthritis by using immunohistochemistry (23, 38). Proliferation responses of SF mononuclear cells to several BCW have been observed in RA (12).

Our results obtained by using PCR with pan-bacterial 23S rRNA and 16S rRNA primers did not reveal the presence of bacterial DNA in any of the ST samples analyzed. This observation is most likely not due to false negative results, because the sensitivity of the PCR was about 2–20 bacteria CFU/reaction. All the positive controls showed positive results. Moreover, the bacterial DNA was detected in SF of 33% of ReA patients (Figure 4), and was consistent with the previous observations (39, 40). The PCR primers used have also been tested in previous studies and shown to cover most bacteria (26, 36, 37). It should be noted that the pan-bacterial PCR is extremely sensitive to contamination. To avoid contamination, we chose a simple PCR method as described in the Patients and Methods, because less sample manipulation was required. Furthermore, to balance the risk of contamination and the sensitivity, a very sensitive PCR method, such as the nested-PCR was not applied. In our hands, the sensitivity could not be increased simply by increasing the cycle number. We also observed that by increasing the PCR cycles up to 40, reagents alone would give false positive results (data not shown) (26). It has been reported that even PCR reagents alone, such as DNA polymerase, may contain trace amounts of bacterial DNA (41–43).

Recently, the presence of bacterial DNA in inflammatory ST has been reported in several studies using pan-bacterial PCR (23–25). Our findings are in contrast to those. Several possible explanations for this exist. 1) To avoid contamination, we used a simple PCR method with sensitivity (2–20 CFU/reaction) probably lower than in the PCR used by many others (23–25). It is still possible that very low concentrations of bacteria or bacterial DNA may be present in the inflamed ST, even though the detection limit of our PCR method was low. 2) All of our patients had an advanced disease of more than 5 years' duration, which is longer than that in other studies (23, 25). It is possible that the permeability of inflamed ST from advanced disease is less than in the early phase of RA (44). This may decrease the chance of detecting bacterial DNA. Alternatively, there may be etiopathogenic organisms present in the early stages of the disease development, and they have later disappeared. 3) Finally, by using an extremely sensitive pan-bacterial PCR, which could detect even as little as 1 DNA copy of any bacteria, the possibility of contamination cannot be completely ignored (41, 45); bacteria are abundant everywhere in the environment, including the body of the investigators. Furthermore, it is known that trace amounts of bacterial nucleic acid may be present even in high quality reagents, such as PCR polymerase (41–43). When the results obtained by using GC-MS or PCR are compared, it is worth noting that bacterial muramic acid could be detected in affected joints at least 3 months later after a single intraperitoneal injection of BCW in a rat arthritis model (17, 46). The other possibility is that only microbial debris or degradation products are present in the inflamed joints, and live organisms are absent. Thus, bacterial muramic acid can be detected via GC-MS, but not bacterial DNA via PCR.

The presence of muramic acid in ST raises a question about the source of the bacterial components entering the joint. One possibility is that they are derived from indigenous intestinal flora carried by macrophages migrating from the gut into the joints (47–49). Muramic acid has been demonstrated in the peripheral mononuclear cells or spleen even in healthy individuals (50–52). It appears reasonable that local inflammation may upregulate expression of cell adhesion molecules and production of chemokines, increase endothelial permeability, and recruit new inflammatory cells transporting bacterial components to the inflamed joint (53). This suggestion is in agreement with the present observation that muramic acid was not detected in ST or muscle samples from healthy controls (Table 2). Nevertheless, the significance of muramic acid in RA joints remains unknown. It is well documented that bacterial peptidoglycan has several biologic activities, including stimulation of proinflammatory cytokines (51, 54) and capacity to induce experimental chronic arthritis (15–17). However, muramic acid was also detected in OA. Further investigation, particularly in early RA and OA, is required to answer whether the bacterial components present in ST are linked to the etiopathogenesis of chronic arthritis, or whether they simply reflect increased endothelial permeability in the inflamed joints. The possibility cannot be excluded that immunocompetent cells containing bacterial degradation products may contribute to synovial inflammation. Their exact significance and etiological role in the development of chronic arthritis remains unresolved.

In conclusion, our results demonstrate that a bacterial component, muramic acid, can be detected by GC-MS in ST from a few patients with RA or OA. However, bacterial DNA was not observed by using 2 methods of pan-bacterial PCR. This may indicate that the presence of bacterial DNA in the ST from patients with advanced RA is not as prevalent as previously suggested.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The authors thank Minna Suominen and Marja-Riitta Teräsjärvi for excellent technical assistance. We also thank Dr. Taru Savolainen for providing the cadaver samples. Dr. Kaisu Rantakokko-Jalava is acknowledged for valuable discussion and comments.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES