INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A preparation from spleen or lymph nodes of sarcoidosis patients, commonly called Kveim antigen, is the only reagent eliciting sarcoidosis-specific in vivo responses (1, 2). In patients with sarcoidosis approximately 4 wk after intracutaneous injection of Kveim antigen, a noncaseating granuloma identical with sarcoid granulomas can be found at the injection site. This reaction, named after the authors Kveim and Siltzbach (KST = Kveim-Siltzbach test), can be used to establish the diagnosis of sarcoidosis (3, 4); however, the active component of the reagent is still elusive.

There have been many efforts to establish an in vitro KST; however, these have been without conclusive results (5). A partial characterization of the molecules eliciting the response suggests a protein or proteins (6) located in immune cells or in membrane fragments of those cells (7). In view of the fact that hypotheses of a bacterial (8), slow-growing bacterial (9), or mycobacterial (10) etiology of sarcoidosis are maintained, we examined the KST reagent for the presence of bacterial including mycobacterialDNA using a polymerase chain reaction (PCR), employing universal primers for conserved DNA sequences encoding for bacterial and mycobacterial 16S ribosomal ribonucleic acid (rRNA), as an approach to gain information about an eventual bacterial mechanism in the KST granulomata. Recently the presence of mycobacterial DNA sequences has been excluded in sarcoid lesions using this technical approach (11).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

KST Reagents and Control Reagents

KST reagent was purchased from the Public Health Laboratory Service, Centre for Applied Microbiology and Research, Wiltshire, United Kingdom. A second KST reagent and a corresponding control preparation from normal spleen (CSP) were prepared following the original protocol (2), but the addition of phenol for conservation was omitted. The KST reagent employed in our study was a Type 1 KST reagent at a concentration of 6 mg/ml of the ethanol-precipitated proteins dissolved in phosphate-buffered saline. Penicillin/streptomycin (100 µg/ml) and fungizone (250 ng/ml) were added as preservatives. The CSP was prepared following the same protocol. The preparations were stored at 70° C until required. It has been demonstrated that the utilized sarcoid spleen yields a diagnostic KST reagent (7, 12). Both KST preparations were analyzed by PCR without further purification.

DNA Extraction from Sarcoid Spleen

DNA was extracted from several small pieces of a native spleen of another sarcoidosis patient undergoing splenectomy because of thrombocytopenia. The samples were incubated in 50 mM Tris-HCl (pH 7.4) containing proteinase K (1 mg/ml; Boehringer Mannheim, Mannheim, Germany) and 1% sodium dodecyl sulfate (SDS) for 1 to 2 h at 37° C. After digestion samples were alternately incubated in a water bath at 100° C for 1 min and snap-frozen for 1 min in liquid nitrogen four times, DNA was purified by phenol/chloroform extraction, precipitated with isopropanol, redissolved in 50 µl mM Tris (pH 8.0), and stored at 70° C.

PCR

The sequences for the oligonucleotide primers are summarized in Table 1. Primer A targets a sequence in the 5' region of the DNA coding for 16S rRNA of a variety of gram-positive and gram-negative bacteria (13). Primer D hybridizes to a sequence of the 16S rDNA, which is well conserved within eubacteria, archaebacteria, and eukaryotes (14). A PCR for human -actin served as control for inhibition in the Kveim suspensions of the KST reagent and the control. The PCR was carried out in an automatic DNA thermal cycler (Perkin-Elmer Cetus, Norwalk, CT). For the amplification of 1 to 10 µl KST suspension or DNA, 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl₂, 50 mM KCl, 0.01% gelatin, 200 µM of each deoxyribonucleoside triphosphate (dNTP), 2.5 U Taq polymerase, and 1 µM of each of the primers, were used at a final volume of 100 µl. The step-cycle program was set to denature at 94° C for 1 min, to anneal at 57° C for 1 min, and to extend at 72° C for 1.5 min plus a 2-s extension in every cycle for a total of 35 cycles.

PCR products were run with molecular weight marker VI (pBR328 DNA × Bgl1 + pBR328 DNA × Hinf1; Boehringer Mannheim) on 1% agarose gel containing 0.01% (vol/vol) ethidium bromide.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The sensitivity of the amplification protocol was tested with serial dilutions of a suspension containing a known amount of Escherichia coli DNA. As few as 10 genomes gave positive amplification results (Figure 1). Additional experiments evaluating the sensitivity of the PCR revealed that in the presence of 1 to 2 µl KST reagent the sensitivity of the PCR was 100 to 200 organisms (E. coli) per tube; in the presence of 5 µl KST reagent we needed 1,000 organisms per tube to be detected (data not shown).

View larger version (54K): [in this window] [in a new window]   Figure 1.   Agarose gel electrophoresis of PCR fragments of a serial dilution of E. coli using primers complementary to the DNA coding for the 16S RNA. The numbers of genomes are indicated below the lanes (M = molecular weight markers). As few as 10 genomes can be detected.

Employing the KST and the CSP of the nonphenolized preparations in a PCR assay using primers for human -actin, a positive signal could be obtained (Figure 2), indicating the absence of inhibitors of the PCR. The same results held true for the phenolized KST as well as the native spleen preparation (data not shown).

View larger version (65K): [in this window] [in a new window]   Figure 2.   Agarose gel electrophoresis of PCR fragments of human -actin. Lane 1: KST reagent; lane 2: control spleen preparation; lane 3: negative (no DNA); lane 4: positive (human DNA) control reaction. M indicates the molecular weight markers. Both KST reagent and control spleen preparation are positive for human -actin indicating that there are no inhibitors of PCR.

Applying universal primers (Table 1) targeting the gene of the bacterial 16S rRNA for PCR, no bacterial DNA could be amplified in these preparations (Figure 3). Analysis of the phenolized KST and the native spleen preparation in the same PCR protocol yielded identical results (data not shown).

View larger version (70K): [in this window] [in a new window]   Figure 3.   Agarose gel electrophoresis of PCR fragments of DNA coding for 16S RNA. Lane 1: KST reagent; lane 2: control spleen preparation; lane 3: negative (no DNA); lane 4: positive (bacterial DNA) control reaction. M indicates the molecular weight markers. Both KST reagent and control spleen preparation produced negative results after amplification for bacterial 16S rRNA.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

PCR was performed to amplify a fragment of the DNA coding for the ribosomal 16S rRNA, which is an essential part of the bacterial ribosome. In the clinically used KST reagent, the KST reagent specially prepared without phenol, the control preparation, and the sarcoid spleen, no bacterial DNA could be amplified, whereas DNA coding for human -actin could be detected. The sensitivity of this method enables us to detect low numbers of bacteria (about 100 genomes even in the presence of 1 to 2 µl KST) and is comparable to those used in studies in search of bacterial DNA in sarcoid granuloma and bronchoalveolar lavage fluid (11, 15, 16). However, the KST reagent is a concentrated spleen tissue extract and the yield of the active principle varies from lot to lot. From this, a direct deduction to the original mass of spleen tissue reflected by our approach is difficult to estimate. A rough calculation reveals that 4 µg KST reagent can be extracted from 10 µg of spleen tissue (concentration factor: 2.5). This means that 2 µl of KST represents 30 µg of spleen tissue (2.5 × 2 µl × 6 µg/µl), which could contain less than 100 genomes, i.e., approximately < 3 × 10⁵ bacteria per 100 mg tissue. In experimental infections such a bacterial burden can be achieved by infections with avirulent mycobacteria without causing systemic disease (17). Nevertheless, the negative results obtained do not support the hypothesis that the cutaneous granulomas at the site of a positive KST are caused by bacteria. However, the possibility that bacterial products such as superantigens or membrane fragments (9) induce these reactions cannot be excluded. In the case that a positive result would have been obtained, the bacterial species could have been identified by the employed approach (11, 15). Because bacterial 16S rDNA contains several fragments that are specific for different phylogenetic levels, even at the species level, bacterial species can be identified by the amplification of this gene, followed by subsequent sequencing and comparison with sequences stored in gene databases (11). Otherwise, new bacterial species can be identified by this method (15).

If we take into account both the present study and the fact that recently the KST reagent has been used to analyze the granulomatous reaction of sarcoidosis (18), the specificity of the KST deserves consideration. The reagents have been tested extensively and only few false-positive KST results have been obtained (3, 4, 19). However, in Crohn's disease (20), lepromatous leprosy (21), granulomatous cheilitis (22), eosinophil pneumonia (23), Melkersson-Rosenthal syndrome (24), and in animal models of granulomatous diseases (25), positive KSTs have been observed. Furthermore, some control preparations from normal spleens induced positive reactions in patients with sarcoidosis (26). Other groups could not reproduce these "unspecific" positive results (27). To increase the specificity, lots of the reagent have been extensively characterized and only highly specific lots used for diagnostic purposes (2).

The Public Health Laboratory Service of the United Kingdom prepares a clinically validated KST reagent used for diagnostic purposes that contains phenol as a preservative. The PCR experiment on this preparation gave negative results, whereas the control reaction showed amplification of the human -actin gene demonstrating the presence of human DNA. Although the control reaction demonstrated a normal function of the PCR, a false-negative result for the reaction with the universal bacterial primers due to the phenol could not be excluded. However, nonphenolized lots of these preparations are not available from the Public Health Laboratory Service of the United Kingdom. For that reason, a phenol-free KST reagent was prepared at the East Carolina University School of Medicine, Greenville, NC and tested using identical PCR techniques and it was found to be also negative. A batch of this KST reagent is known to yield highly specific results for diagnostic and research purposes (7, 12). An identical preparation from a normal spleen was also negative in the PCR assay. The same result was obtained from an additional sarcoid splenic preparation, demonstrating consistent negative results, backtracking over the process of KST production from the phenolized diagnostic reagent to the final preparation (phenol-free KST) to the substrate (spleen).

The studies of Williams and Nickerson (28), and Kveim (1) in which the "Kveim" reaction was observed were pursued under the hypothesis that sarcoidosis is caused by an infectious agent and that a specific immune reaction against this agent, resembling the tuberculin-like reaction could be induced with the preparations used. While this hypothesis is still under debate (10), an infectious agent has not yet been identified in sarcoidosis. It should also be noted that KST reagents induce granulomatous inflammation in 3 to 6 wk but without any accompanying tuberculin-like reaction.

Numerous immunobiological studies, especially those dealing with lymphocytes, have shown that most of the immune characteristics of the sarcoid reaction are shared with the normal immune response, supporting the hypothesis of an infectious etiology of sarcoidosis (18, 29). The fact that bacterial DNA cannot be found in KST reagent or in sarcoid lesions suggests that if an infectious etiology of sarcoidosis exists at all bacteria or viruses may trigger the sarcoid response, which may be maintained by undegradable products of these species or by autoaggressive or cross-reacting host mechanisms after the causing agent has been cleared, as has been demonstrated for infectious arthritis (9). This notion is further supported by studies indicating that the active principle of the KST reagent is of proteinous nature (6, 30).