| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
Department of Medicine, Johns Hopkins University, Baltimore, MD
aAddress correspondence to this author at: 602 Mason F. Lord Bldg., Center Tower, 5200 Eastern Ave., Baltimore, MD 21224. Fax 410-550-8512; e-mail jvaneyk1{at}jhmi.edu.
Determining the circulating form of a biomarker is critical for the development of a highly sensitive and specific immunoassay. The hormone brain natriuretic peptide (BNP) is produced in the atria as a preprohormone, and the cleavage of its signal sequence produces the circulating prohormone proBNP (amino acid residues 1–108, human sequence). The prohormone is subsequently cut to yield a 76-amino acid N-terminal (NT) fragment (NT-proBNP) and the 33-amino acid active hormone BNP (comprising residues 77–108 of proBNP). The 2 commercially available immunoassays detect either the circulating NT fragment or BNP and, potentially, could detect proBNP. However, whether these assays detect all possible circulating forms of BNP remains unclear.
The potential circulating forms of BNP have different intrinsic physical characteristics that can dictate antigenicity, extent of nonspecific binding during isolation, endogenous clearance kinetics, circulating half-life, and/or affinity for other proteins including cellular receptors and putative serum/plasma carrier proteins. Consequently, characterization of the physical status of each circulating form of BNP is key to assay development. It is critical to determine whether the circulating BNP forms have posttranslational modifications (PTM), such as phosphorylation or N- and O-linked glycosylation, and/or bind to other protein(s). In a recent issue of Clinical Chemistry, an article by Seferian et al. (1) described 2 approaches to shedding light on the circulating forms of BNP. The 2 strategies examined differences in (a) antigenicity, (b) antibody binding, and (c) the theoretical and observed MWs of the unmodified recombinant form and the circulating endogenous form(s). The data presented indicated that the chemistry of circulating BNP is complex. This complexity needs to be taken into account during immunoassay development.
Is there a potential PTM? Seferian et al. (1) produced a large number of peptides and fragments of BNP (pro, NT-pro, and BNP) and used them as immunogens to raise monoclonal antibodies; then these investigators elegantly screened the circulating forms. It is important to realize that all of the immunogens were unmodified. The antibodies were raised specifically against the primary amino acid sequence, and hence could display reduced affinity (or no binding) if the sequence were modified or blocked in vivo because of tertiary or ternary conformations that inhibit the binding of the antibody to the linear sequence. Seferian et al. (1) indeed found that for both NT-proBNP and proBNP, antibodies reacted differently with the recombinant form than with the endogenous analyte. For example, antibodies raised against synthetic peptides to regions within the NT-proBNP sequence 46–60 (or proBNP residues 28–60) did not bind (or had minimal binding) to the endogenous antigen. The only antibody pairs that enabled high-sensitivity detection of both the recombinant and endogenous antigen were raised to residues between 5–27 and 67–76. This finding suggests that the central region of endogenous NT-proBNP and proBNP is altered and has little or no availability for antibody binding.
Interestingly, Schellenberger et al. (2) recently showed that the proBNP expressed in the mammalian Chinese hamster ovary cell line is extensively glycosylated (O-linked) at several serine or threonine residues located between amino acid residues 22 and 58 of proBNP. This domain is the same location where the antibodies raised by Seferian et al. (1) were inefficient at binding the endogenous circulating forms. Thus, it is possible that the modification preventing antibody binding is O-linked glycosylation. Although direct evidence was not provided, Seferian et al. (1) showed by size-exclusion chromatography (SEC) that the MWs of the endogenous circulating proBNP and NT-proBNP were higher than the unmodified recombinant versions (eluting at observed MW of 
35–37 kDa vs 25 kDa for the endogenous and recombinant proBNP and 35–37 kDa vs 15–17 kDa for the endogenous and recombinant NT-proBNP, respectively), a finding consistent with the presence of glycosylation.
Is there more to the story? Interestingly, results of nondenaturing SEC revealed that when added to plasma, the observed MWs for circulating BNP and unmodified recombinant forms of BNP were also much higher than their predicted MWs (1) (theoretical pI/MW are 5.42/8.184 kDa, 11.46/3.719 kDa, and 9.8/11.948 kDa for NT-proBNP, BNP, and proBNP, respectively). Although one explanation for such a huge discrepancy may be glycosylation, it cannot account for the differences in the MW of the unmodified recombinant forms of BNP (once mixed with plasma). In fact, the MW difference revealed by SEC is even greater than that reported by Schellenberger et al. (2), who used SDS-PAGE (denaturing and reducing conditions) to analyze recombinant O-linked glycosylated proBNP and endogenous forms from patient serum. Schellenberger et al. (2) showed a reduction in MW from 25 to 12 kDa in response to treatment with deglycosylating enzymes [See Figs. 1 and 5 in Ref. (2)]. These MW data can be reconciled by the presence of another noncovalent interaction, one that remains when SEC is carried out under nondenaturing conditions. Thus, it is possible that the circulating forms of NT-proBNP, proBNP, and BNP do not circulate in the blood as monomers, but rather are bound to as-yet-unidentified proteins.
The idea that circulating proteins and peptides bind other serum proteins is not new. Recently, we published results from the investigation of albumin and its binding partners in serum. In this example, albumin was found to bind to intact proteins as well as peptides (3). In fact, of the 35 albumin-binding proteins we found in serum, 24 were intact. Other studies of high-abundance serum proteins as carrier proteins have also revealed that small peptides bind [e.g., (4)(5)]. Additionally, analytes may remain bound to their tissue-originating binding partners, as is the case for circulating cardiac troponin I, which is preferentially bound to troponin C, another subunit of the regulatory troponin complex [e.g., (6)]. The implication for BNP is that complex forms may be present that alter antibody binding, again highlighting the importance of having knowledge about PTMs and binding partners during the development of an immunoassay.
Differentiating between an endogenous binding partner and a PTM should be straightforward, because the interaction with a binding partner is noncovalent (normally) and the binding partners will dissociate under denaturing conditions, whereas PTMs are covalent and will not be disrupted so easily. In addition to traditional protein biochemical methods, mass spectrometry (MS)-based approaches are available for identifying and characterizing protein isoforms, single-nucleotide polymorphisms, and PTMs, and for mapping protein complexes. MS can be used to measure the MW of (i) protein complexes, (ii) intact proteins, (iii) protein fragments, and (iv) peptides (in increasing order of accuracy). As a general rule, the high-accuracy instruments have mass accuracy of <1 ppm (under optimal conditions for small peptides <2 kDa, for example), allowing for absolute identification of a protein or protein modification based on the measurement of a single unique peptide. Therefore, MS can be used to characterize PTMs and binding partners.
In addition to the current utility of MS in characterization of PTMs and binding partners MS has the potential to be useful for clinical assays. An example of a relevant MS technique is multiple reaction monitoring, in which a previously defined peptide representative of the biomarker/analyte is quantitatively measured [reviewed in Ref. (7)]. Multiple reaction monitoring provides a high-throughput and accurate method for detecting and characterizing analytes of interest alone or in a multiplexed fashion (e.g., 30–100 analytes per analysis). The utility of multiple reaction monitoring for the detection and quantification of analytes from serum in a multiplex assay (8) or singly (9)(10)(11)(12) shows great promise. An alternative MS-based approach involves capturing a specific protein from a biological sample using immunoaffinity, subsequently eluting the protein from the antibody, and finally analyzing it by MS [e.g., reviewed in Ref. (13)]. In this way, the intact (predigested) form of the protein is measured to determine the presence of large PTMs such as glycosylation or the presence of multiple protein isoforms that bind to a single antibody. Furthermore, the captured proteins may also be enzymatically digested to allow for tandem MS analysis leading to the identification and localization of PTMs.
In conclusion, the potential complexity of the circulating forms of proBNP and NT-proBNP is reminiscent of the disease-induced modifications present on cardiac troponin. In these cases, disease-induced degradation and/or phosphorylation can occur within the tissue or after release of the protein from the cell or tissue [reviewed in Ref. (14), (15)]. Regardless of the analyte, efforts must be taken to determine the circulating forms and how they change among patient groups. The use of protein biochemistry and MS to characterize PTMs and protein binding partner(s) will not only shorten the time required to produce an effective assay, it could also allow for the exploitation of disease-induced changes of the analyte in a clinical assay for risk stratification.
Acknowledgments
Grant/funding support: R.L.G. and J.E.V.E. are supported by the National Heart Lung Blood Institute Proteomic Initiative (contract NO-HV-28120) and the Donald P. Amos Family Foundation.
Financial disclosures: None declared.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
