| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1
Institute of Radiology,
2
Department of Cardiology,
3
Department of Dermatology, and
4
Department of Cardiac Surgery, University of Luebeck Medical School, 23538 Luebeck, Germany
5
Institute of Clinical Chemistry, Municipal Hospital, 18435 Stralsund, Germany
a address correspondence to this author at: Institut fuer Radiologie, Medizinische Universitaet zu Luebeck, Ratzeburger Allee 160, 23538 Luebeck, Germany. Fax 49-451-500-6190, e-mail missler{at}medinf.mu-luebeck.de
S-100 protein concentrations in serum are considered a quantitative marker of the extent of damage to the central nervous system (CNS) (1)(2)(3), including possible cerebral injury following procedures such as coronary artery bypass surgery (4)(5). There are 19 S-100 proteins, of which S-100A1 and S-100B are the most prevalent (6). S-100A1 and S-100B form dimeric proteins that previously were labeled S-100a (S-100A1-B), S-100b (S-100B-B), and S-100a0 (S-100A1-A1) (7)(8)(9). Initial studies of S-100 protein reported that this protein is present only in the CNS (7), but later studies found various concentrations of S-100A1 and S-100B in tissues outside the CNS, including the heart and aorta (10). At present, there are questions about the subtype specificity of S-100 assays, about which subtype of S-100 is associated with different clinical entities, and how the results obtained using various assays compare. To answer these questions, we used the Sangtec 100® LIA and the immunofluorometric (IFMA) S-100 assay to analyze purified recombinant monomeric S-100 proteins, purified dimeric S-100 proteins isolated from bovine brain, and blood samples from patients with various CNS diseases, malignant melanoma, or post cardiac surgery and compared the results.
Serum (Sangtec 100 LIA) or heparinized plasma (IFMA S-100) samples were drawn from 218 patients (141 males, 77 females; ages, 16–89 years; mean, 55.4 ± 16.2 years) and 121 healthy blood donors (64 males, 57 females; ages, 18–65 years; mean, 37.8 ± 12.7 years). The study was approved by the local Research Ethics Committee, and all subjects gave informed consent to the procedure. One hundred four of the patients suffered from CNS disease: subarachnoid hemorrhage (n = 41), intracerebral hemorrhage (n = 17), head trauma (n = 19), ischemic cerebral infarction (n = 4), cerebral tumor (n = 12), hydrocephalus (n = 1), or lumbar disc herniation (n = 10). Twenty-nine of the patients had malignant melanoma, and 85 had undergone cardiac surgery involving the use of cardiopulmonary bypass. Samples were centrifuged within 6 h and stored at -70 °C until analysis. Measurements were performed in duplicate.
The Sangtec 100 LIA immunoluminometric assay uses tubes, coated with two monoclonal antibodies, as solid phase and a monoclonal antibody for detection. The assay measures concentrations of S-100 protein over the range 0–20 µg/L. Measurements were performed according to the instructions of the manufacturer. The IFMA S-100 is an immunofluorometric assay that uses microtiter plates coated with a monoclonal anti-S-100B antibody and a polyclonal detection antibody as described elsewhere (1). A 1:1 (by volume) mixture of S-100A1-B and S-100B-B (Sigma) was used for calibration. S-100 protein concentrations between 0.015 and 25 µg/L were measured.
To determine the assay specificities for different S-100 subtypes, solutions with known concentrations of purified recombinant monomeric S-100A1, S-100B, and the dimeric proteins S-100A1-A1, S-100B-B, and S-100A1-B (Sigma) were measured. S-100A1 and S-100B were produced as described previously (11)(12). Stock solutions contained 6.39 g/L S-100A1, 1.3 g/L S-100B (as determined using the Lowry method), and 1 g/L S-100A1-A1, S-100A1-B, and S-100B-B (as specified by the manufacturer). Stock solutions were diluted in assay diluent (Sangtec 100 LIA) or horse serum (IFMA S-100). Analytical recovery was determined by mixing six patient samples containing various amounts of S-100. The lower detection limit for S-100 protein in each assay was calculated by adding 2 SD to the mean of 20 (IFMA) or 15 (100 LIA) measurements of the assay diluents. We sought to identify correlations between S-100 protein concentrations and clinical conditions by linear regression analysis of S-100 protein concentrations for all patients and for either type of assay in patients with neurological disease, malignant melanoma, or post cardiac surgery. Differences in the slopes of the regression lines were evaluated for statistical significance at P <0.05 (two-tailed).
The results of the S-100 subtype cross-reactivity evaluation are
summarized in Table 1
. Recovery rates of S-100B protein from analyzed samples were
94.8–113.2% for the Sangtec 100 LIA and 89.7–100.6% for the IFMA
S-100. The calculated lower limit for detection for S-100B was <0.015
µg/L in the IFMA S-100 and <0.02 µg/L in the Sangtec 100 LIA. The
precision was calculated for concentration ranges of <0.1, 0.1–0.49,
0.5–1.0, and >1.0 µg/L for 222 samples. The median CVs were
15%, 5.7%, 3.7%, and 2.5% for the Sangtec 100 LIA and 7.9%,
3.6%, 3.2%, and 2.2% for the IFMA S-100. The median concentration of
S-100 protein in specimens from the blood donors assayed using the
Sangtec 100 LIA was 0.014 µg/L (10th percentile, 0.0 µg/L; 90th
percentile, 0.089 µg/L). In a previous study using the S-100 IFMA, we
found a median plasma concentration of 0.052 µg/L S-100 (10th
percentile, 0.023 µg/L; 90th percentile, 0.097 µg/L) in 200 healthy
blood donors (ages, 18–65 years) and no dependence of S-100
concentration on age or sex (13). The scatter plot of
patient samples and linear regression analyses for each of the three
groups and all groups together are shown in Fig. 1
. S-100 measurements obtained using the two assays correlated
well (R2= 0.95). Comparison of the
slopes of the regression lines, however, showed significant differences
between S-100 concentrations in the blood of patients with a CNS
disorder and patients who had undergone cardiac surgery
(t = 6.92; df = 185; P
<0.05, two-tailed) and also between patients with a CNS disorder and
patients suffering from malignant melanoma (t = 5.64;
df = 129; P <0.05, two-tailed). The
patients who had undergone cardiac surgery did not differ from the
melanoma patients (t = 0.60; df = 110;
P >0.05, two-tailed).
|
|
To the best of our knowledge, this study is the first to report validation data of the Sangtec 100 LIA assay. Our results showed that both assays are reliable methods with comparable lower limits of detection, measurement ranges, precision profiles, and reference ranges. Both assays are specific for monomeric S-100B. S-100A1 was not detected by both methods. However, the extent of cross-reactivity between the different dimeric proteins (S-100A1-A1, S-100A1-B, and S-100B-B) differed between both methods. Only small amounts of the A1-A1 dimer were detected by both assays, and differences between the methods were small. Although both assays were comparable regarding the measurement of S-100B-B, which has been claimed to be relevant in neurological disorders, there was a clear difference in the measurement of S-100A1-B. The Sangtec 100 LIA returned ~10-fold higher S-100A1-B values than the IFMA S-100, which may be attributable to highly different affinities of the antibodies to the dimeric S-100A1-B molecule. It may be speculated that the conformation of the S-100B molecule is altered during formation of the S-100A1-B dimer. This may increase the affinity of the antibodies used in the Sangtec 100 LIA to their binding sites.
Linear regression analyses of the clinical samples demonstrated a good correlation between both assays. The absolute values of the Sangtec 100 LIA were ~twofold higher than the results of the IFMA S-100, which may be explained by two mechanisms: Because no international reference preparation for S-100 protein is available, this may reflect differences in calibration. In addition, the different subtype cross-reactivities of both methods may contribute to these results.
In immunohistochemical studies using antibodies specific for S-100A1, S-100B, or other S-100 proteins, the distribution of these proteins in human tissues has been reported (10). However, no study has been able to define the extent to which the different S-100 subtypes are released into cerebrospinal fluid and blood in distinct clinical entities. Moreover, for methodological reasons, we do not know whether the S-100 proteins are present in monomeric or dimeric forms in the human body. There are numerous reports in the literature about the measurement of S-100 in different diseases. Several authors have claimed, for example, that CNS damage should increase the concentrations of S-100B-B (1)(2)(3)(4)(5)(8)(9)(14)(15), whereas S-100A1-A1 is released from cardiac tissue (16). Controversial results have been reported in malignant melanoma, in which either only S-100B or S-100A1 and S-100B were found to be increased (17)(18). The conclusions of all authors, however, were based only on their usage of polyclonal or monoclonal antibodies against the subunits S-100A1 and S-100B. The subtype specificities of these antibodies for the dimeric S-100 proteins have not yet been studied. To be specific for a dimeric protein (e.g., S-100B-B), an antibody should be able to detect the connection region of the two subunits. There are no reports on such antibodies against S-100 proteins. Thus, it is not clear which forms of S-100 proteins have been found by the investigators cited above.
Using purified recombinant S-100A1 and S-100B proteins, we were able to define the subtype cross-reactivities of two widely used S-100 assays. If the assumption holds true that different S-100 subtypes are present in neurological disease, malignant melanoma, and heart disease, the results of two S-100 assays with different subtype cross-reactivities should differ between these groups. In our study, we were able to show statistically significant differences between the neurologically diseased patients compared with the patients who underwent cardiac surgery or suffered from malignant melanoma. In most clinical studies cited above, the authors have considered the monomeric S-100 subunits to be increased. However, there is evidence that in vivo specific functional properties of S-100B depend on its dimeric status (19). It has been questioned whether the monomeric forms can even exist in solution (6). Therefore, we conclude that although this study utilized assay systems based on monoclonal antibodies directed against S-100B subunits, our data may support the hypothesis that different patterns of dimeric S-100 proteins are present in various clinical entities and thus should be taken into account when interpreting "S-100B" values.
References
, CAPL, CACY: molecular cloning and expression analysis or three calcium-binding proteins from human heart. Biochemistry 1992;31:10258-10264.
[Medline]
[Order article via Infotrieve]
Read all eLetters
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
