| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Laboratoire Universitaire de Biophysique, Unité d’Analyses Endocriniennes, ULP/CNRS UMR 7004, Faculté de Médecine, 67085 Strasbourg Cedex, France
2 Immunotech Beckman Coulter, 13276 Marseille Cedex 9, France
aaddress correspondence to this author at: Institut de Physique Biologique, Faculté de Médecine, F-67085 Strasbourg Cedex, France; fax 33-3-90-24-40-57, e-mail sapin{at}ipb.u-strasbg.fr
Prolactin (PRL) circulates in serum in three major molecular sizes identifiable by gel-filtration chromatography: monomeric PRL (23 kDa), big PRL (45–60 kDa), and big big PRL or macroprolactin (150–170 kDa). Macroprolactin is mainly a complex of PRL with human immunoglobulins G (hIgG) (1)(2)(3)(4), but aggregates of PRL may also be present (4). Identification of macroprolactin, which has reduced bioactivity but can be the cause of high PRL values in patient samples, can help resolve diagnostic confusion and avoid expensive investigations and inappropriate treatment. It is generally admitted that all samples showing apparent hyperprolactinemia should be examined for macroprolactin.
The immunoassays used to determine prolactinemia react variously with macroprolactin (5). With high-reacting assays such as the Elecsys® PRL assay from Roche Diagnostics, polyethylene glycol (PEG) precipitation has been validated as a rapid technique to detect macroprolactin (6). However, the PRL PEG precipitation test suffers from high nonspecific precipitation [
14–16% (7)(8)], and results of this test may not be definitive in any case, making it necessary to define a gray zone (6)(9). Comparison of the results of a high-reacting method with those of low-reacting methods, such as the ADVIA:Centaur or ACS:180 PRL assays (Bayer Diagnostics), has also been proposed as a screening method (10). However, results obtained for macroprolactinemic sera have been shown to be sample dependent, particularly with the Bayer assays (7)(11), and the presence of macroprolactin can not be excluded by comparing the results obtained with a high- and a low-reacting method.
Recently, a screening method based on the recognition of the hIgG component of macroprolactin by goat anti-human IgG-agarose (binding capacity, 1.5 mg hIgG/mL of resin) has been validated with the Elecsys assay (12). This method incorporates a 2-h incubation time and a 20-fold dilution of the serum, which may prevent its use in samples with moderate hyperprolactinemia (<1000 mIU/L). In the present study we describe the application of a simple and rapid method based on precipitation of hIgG-PRL complexes by a protein A-Sepharose suspension with high hIgG binding capacity (16 mg/mL of resin). Protein A is a polypeptide that binds the Fc region of human immunoglobulin molecules, especially hIgG1, hIgG2, and hIgG4 but only a fraction of hIgG3 (13). This method involves a short incubation time and a threefold dilution factor only. Results obtained with this new method were compared with those obtained with the PEG precipitation test.
Prolactin was measured in 116 sera from hyperprolactinemic individuals (PRL >600 mIU/L): 23 men (age range, 12–79 years) and 93 women (age range, 8–90 years). These sera were selected on the basis of their PEG-precipitated PRL values (7–96%) to study the whole range of values. The PRL concentrations in these samples were between 636 and 24 400 mIU/L. The procedures were in accordance with the Helsinki Declaration of 1975 and the subsequent 1996 amendments.
We used the automated Elecsys PRL immunoassay on the Elecsys 2010 analyzer. The PRL standard was the third International Reference Preparation WHO 84/500. PEG precipitation was performed as described previously (8). This method involves a twofold dilution. Results were expressed as the percentage of PEG-precipitated PRL (10). A serum was considered to be negative, i.e., to contain a low proportion of macroprolactin, when the percentage was <50% and positive when the percentage was >60%. The range between 50% and 60% was defined as borderline and constituted a gray area (10)(14). Ready-to-use protein A-Sepharose suspension, obtained from Immunotech Beckman Coulter (Marseille, France), was used to precipitate the hIgG fraction in serum samples, according to the manufacturer’s protocol (15). To this end, 300 µL of protein A-Sepharose suspension (Protein A CL-4B in 20 mmol/L borate buffer) was added to 150 µL of serum in conical tubes. Thorough mixing of the protein A suspension is necessary before use. After incubation at room temperature for 15 min with rotation (20 rpm), samples were centrifuged at 2000g for 5 min at 20 °C. PRL concentrations in the serum (PRL before) and in the supernatant (PRL after) were then immediately assayed using the Elecsys method. To determine the real PRL dilution factor, we assayed seven serum samples without macroprolactin (as shown by gel-filtration chromatography) before and after treatment with protein A. The dilution factor (D) was the ratio PRL before/PRL after. As expected, the dilution factor was close to 3 (mean ± SD, 2.90 ± 0.09), showing the absence of matrix effects from the protein A-Sepharose treatment and from nonspecific binding of PRL to the Sepharose. The percentage of protein A-precipitated PRL was calculated as follows: 100 (PRL before - PRL after x D)/PRL before. Interassay reproducibility was assessed over a 3-month period through repeated analysis (n = 10) of two patient sera, S1 and S2. CVs were 12% for S1 (protein A-precipitated PRL, 31.2%; PRL before, 2050 mIU/L) and 4.1% for S2 (protein A-precipitated PRL, 74.0%; PRL before, 1580 mIU/L). It has been shown that the percentages of macroprolactin established by protein A precipitation and by gel-filtration chromatography, the reference method to identify macroprolactin, are well correlated (15).
Comparative PEG- and protein A-precipitated PRL results are shown in Fig. 1
. In our population of 116 hyperprolactinemic samples, 76 yielded PEG-precipitated PRL values <50% and were considered as negative for macroprolactin. In these 76 sera, the mean precipitated PRL was 31.6% (SD, 6.8%; range, 7–49%) with PEG and 0.4% (SD, 6.8%; range, -13.3% to 13.7%) with protein A. The percentage of PRL precipitated was lower for protein A-Sepharose than with PEG precipitation. In fact, the mean percentage of protein A-precipitated PRL did not differ from zero in these macroprolactin-negative samples.
|
Thirty-one sera yielded PEG-precipitated PRL values >60% and were considered as positive. In these 31 sera, precipitated PRL results were 76.8% (SD, 8.1%; range, 61–96%) with PEG and 55.1% (SD, 16.3%; range, 24–92%) with protein A. In the positive group, the percentage of protein A-precipitated PRL was clearly increased in all samples, indicating that hIgG-PRL complexes were present in these samples. The percentage of protein A-precipitated PRL was positively correlated with the percentage of PEG-precipitated PRL (r = 0.82). Among the nine samples in the gray zone defined with the PEG precipitation test, four were clearly negative with the protein A test (PEG, 52%, 54%, 58%, and 58%, respectively; protein A, 10%, -1.7%, -0.3%, and 10%, respectively), two samples (PEG-precipitated PRL, 56% and 59%) had markedly increased protein A results (32% and 25%, respectively), and the last three samples (PEG-precipitated PRL, 50%, 50%, and 53%) had protein A results (15%, 18%, and 22%) between the higher value in the negative group (14%) and the lower value in the positive group (24%).
All samples with PEG results >60% yielded a high percentage of protein A-precipitated PRL (>24%), and all samples with PEG results <50% yielded a low percentage of protein A-precipitated PRL (<14%). As expected, the results obtained in the nine samples with PEG-precipitated PRL between 50% and 60% were less clear-cut. Four of the nine samples (there was no serum left available for the other five samples) were subjected to gel-filtration chromatography (15). Three samples with markedly (25% and 32%) or moderately (22%) increased protein A-precipitated PRL contained substantial amounts of high-molecular mass PRL (38%, 37%, and 29%, respectively) as determined by gel-filtration chromatography. The last sample with a low protein A-precipitated PRL result (10%) contained 35% high-molecular mass PRL, and the chromatographic profile of this serum showed a marked heterogeneity of macroprolactin with the presence of molecular forms of mass higher than the 150- to 170-kDa macroprolactin. This sample might contain forms of macroprolactin other than hIgG-PRL complexes, high-molecular mass aggregates of PRL or hIgA, or hIgM-PRL complexes (hIgA and hIgM are only in part bound to protein A) (13). This finding underlines the heterogeneity of macroprolactin.
Similar to precipitation with anti-hIgG-agarose, protein A precipitation is more expensive than the PEG test. Its use therefore could be restricted to the evaluation of borderline samples. Compared with immunoprecipitation with anti-hIgG-agarose resin, protein A-Sepharose precipitation has two advantages: a shorter incubation time and, above all, a much lower dilution factor, allowing macroprolactin to be detected in samples with moderate hyperprolactinemia (600–1000 mIU/L PRL) and in samples with PRL concentrations within reference values. In our 3-year experience with the PEG precipitation test and the Elecsys PRL assay, we have found 154 macroprolactin-positive sera. Among these sera, 51% had a PRL concentration <1000 mIU/L (macroprolactinemia is most frequently associated with moderate hyperprolactinemia), and 15% had a concentration <600 mIU/L. We compared PEG- and protein A-precipitated PRL values in seven sera with PRL concentrations <600 mIU/L and PEG-precipitated PRL 
50% (Table 1
). Four samples were positive with both the PEG and with the protein A precipitation tests, and, as observed previously in hyperprolactinemic sera, for the other three samples in the gray zone of the PEG test, the results of the protein A test were different: one was negative, one was positive, and the third sample was in the borderline zone of the protein A test (14–24%). In our opinion, it can also be useful to detect macroprolactin in a sample with a PRL concentration within reference values, e.g., to discontinue inappropriate anti-PRL therapy.
|
Contrary to the PEG test, the protein A precipitation test showed negligible nonspecific PRL fixation, as evidenced by the fact that dilution did not alter the result from the expected value. In the PEG-negative group, the mean protein A-precipitated PRL value was near zero. This study confirms that the PEG precipitation method can be used routinely with benefit with the Elecsys PRL assay. High amounts of PRL-hIgG complexes were detected by the protein A-Sepharose test in all samples with a PEG test result >60%. That might not be always the case if macroprolactin is composed of PRL aggregates or of hIgG3-, hIgA-, or hIgM-PRL complexes. The protein A test is also reliable and may help resolve cases in the gray zone of the PEG test (3.6% of 1091 sera screened for macroprolactin in our experience). In this case, however, should the protein A test result be negative, as clearly shown by the results obtained for one of our serum samples (PEG-precipitated PRL, 52%; protein A-precipitated PRL, 10%; high-molecular mass PRL by gel-filtration chromatography, 35%), the presence of high-molecular mass forms of PRL should be assessed by gel-filtration chromatography.
Acknowledgments
We thank E. Fischbach for expert technical assistance and N. Heider for reviewing the English in this manuscript. We thank Immunotech Beckman Coulter for providing the protein A suspension free of charge. This study was supported in part by a grant from the Hôpitaux Universitaires de Strasbourg.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  L. Kavanagh, T. J. McKenna, M. N. Fahie-Wilson, J. Gibney, and T. P. Smith Specificity and Clinical Utility of Methods for the Detection of Macroprolactin Clin. Chem., July 1, 2006; 52(7): 1366 - 1372. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Schiettecatte, A. Van Opdenbosch, E. Anckaert, J. De Schepper, K. Poppe, B. Velkeniers, and J. Smitz Immunoprecipitation for Rapid Detection of Macroprolactin in the Form of Prolactin-Immunoglobulin Complexes Clin. Chem., September 1, 2005; 51(9): 1746 - 1748. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
