| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Pathology and Immunology, Division of Laboratory Medicine, Washington University School of Medicine, 660 South Euclid Ave., Box 8118, St. Louis, MO 63110; fax 314-454-5208
ae-mail crimmins{at}pathology.wustl.edu
The cardiac hormone B-type natriuretic peptide (BNP) is synthesized in myocytes as a prepro 134-amino acid residue molecule. The 108-residue proBNP mature form of the hormone is proteolytically cleaved to a biologically active form of 32 amino acids (residues 77–108) and an N-terminal fragment (residues 1–76; NT-proBNP) with an as yet undefined biological function (1). Clinically, both BNP and NT-proBNP have shown great promise as secreted, bloodborne diagnostic markers of left ventricle dysfunction. Measurement of each is based on immunoassays; it therefore is likely that changes in the molecular form, e.g., posttranslational modifications, further proteolytic processing, and an oligomeric state for either analyte, could affect their measurements (2)(3)(4)(5). These types of confounding molecular issues are likely for many analytes, with the myocyte damage marker cardiac troponin I one of the better studied. In this case, the commercial assays use antibodies directed to different epitopes, making "universal" calibration and determination of absolute analyte concentration difficult (6).
A previous report has indicated that NT-proBNP exists as a coiled-coil trimer, based on size-exclusion HPLC (SE-HPLC) of human, plasma-extracted material and a computer algorithm that predicts coiled-coils (7). I reinvestigated this claim on synthetic NT-proBNP, using the physicochemical techniques of analytical sedimentation, equilibrium ultracentrifugation, and circular dichroism (CD), and demonstrate that NT-proBNP is a monomer and not a trimer.
NT-proBNP was produced by solid-phase peptide synthesis (AnaSpec, Inc.) and obtained as a gift from Dade-Behring (Newark, DE). I used N-terminal sequencing and mass spectrometry as quality assurance procedures. Edman sequencing (8) was performed by Midwest Analytical, Inc. on 2 Coomassie-stained Immobilon-PSQ (Sigma) NT-proBNP–containing membrane sections. The first 52 residues were positively identified, with no preview, before the signal was not discernable from background (data not shown). Matrix-assisted laser desorption/ionization mass spectrometry (8) gave an experimental mass of 8457.0 compared with a calculated value of 8457.6 (data not shown). Lyophilized NT-proBNP was dissolved in and exhaustively dialyzed vs phosphate-buffered saline (pH 7.2) at 4 °C. Analyte concentration was estimated gravimetrically and based on a molar absorptivity at 280 nm (
280) of 0.82 L · g–1 · cm–1; the 2 different techniques yielded better than 95% agreement.
The synthetic peptide in a neutral pH physiologic salt solution was run at 0.5 mL/min on SE-HPLC, and the resulting elution profile is plotted in Fig. 1A
as A214nm vs time in minutes. The chromatographic process is monitored at 214 nm, which measures "peptide bond" absorbance; there thus is no bias in analyte detection under these analysis conditions. This is in contrast to the antibody-based, postcolumn analysis of plasma-extracted peptide (7), where detection is strictly a function of antibody reactivity. Furthermore, it is unclear what effect, if any, the C18 solid-phase plasma extraction procedure used in that study has on the molecular state of NT-proBNP before chromatography. Fig. 1A
shows NT-proBNP eluting well before cytochrome C (12.4 kDa) and just after myoglobin (17 kDa). One might be tempted to interpret this result to imply that NT-proBNP is a dimer, i.e., 8.5 kDa x 2 = 17 kDa; this is incorrect, however, as discussed below. The analyte profile observed here is not identical to that of the previous data (7), and is likely a result of combined use of a different SE-HPLC column packing material, different detection procedures, different protein markers, and different sample preparation methods. Nonetheless, one common attribute is that the elution of NT-proBNP is earlier than what would be expected of a typical globular protein of 
8.5 kDa. It is invalid here, and in general, on SE-HPLC performed under benign conditions to use the elution position as a surrogate for analyte molecular mass. The elution position on SE-HPLC is dictated by hydrodynamic volume, which is a function of the degree of hydration, molecular asymmetry, and the polar/nonpolar nature of the analyte and not on molecular mass (9). It becomes possible to estimate molecular mass for single-chain species only when the protein calibrators and analytes possess the same tertiary structure, as occurs when denaturing/disulfide-reducing solvents are used, for example (10). In the neutral-pH phosphate-buffered saline solution used here, all one can conclude is that NT-proBNP elutes unexpectedly with a larger molecular volume than the corresponding globular protein of 8.5 kDa.
|
Sedimentation equilibrium ultracentrifugation was used to assess the oligomeric state of synthetic NT-proBNP. This analysis was performed at Iowa State University Protein Facility (ISUPF) on a Beckman Optima XL-A rotor at 4 °C. A concentration of 0.26 g/L was used at rotor speeds of 20 000, 30 000, and 40 000 rpm for various run times with optical scanning at 280 nm, which monitors aromatic residues in proteins/peptides. Data were plotted as A280nm vs radius in centimeters. This plot is shown in Fig. 1B
, with the corresponding residuals for the 40 000 rpm rotor speed. For a single sedimenting ideal species, the instrument software (Ver. 2.01) transforms the absorbance vs radius data into a molecular weight of 8351, i.e., a monomer. The residuals, which are a measure of the goodness-of-fit of the curved line through the data points, are randomly yet narrowly distributed around 0. This attests to the high quality and lack of bias of the data. I attempted to fit the 40 000 rotor speed data to a 2-ideal-species monomer–dimer and a monomer–trimer equilibrium. After 11–13 computer program iterations, the results, given as concentrations of each species, were as follows: 0.262 g/L monomer with 1.4 x 10–4 g/L dimer, and 0.262 g/L monomer with 5.7 x 10–7 g/L trimer. Clearly, the only species present during ultracentrifugation was a monomer. It would not be possible to analyze the serum-generated sample (7) by this physicochemical technique because of the extremely low (ng/L) sample concentration.
NT-proBNP was reported to be a trimer containing a coiled-coil motif of repeating heptad units (7). Specifically, residues 17–38 were predicted to form a trimeric coiled-coil in a pattern represented as a-b-c-d-e-f-g. Positions a and d in this 7-residue repeat are almost invariantly hydrophobic residues, e and g are usually charged residues of opposite sign, and the remaining 3 residues are usually hydrophilic. The molecular forces, including sequence position and specific amino acids along the 7-residue motif, that determine coiled-coil formation have been studied extensively (11). Typically a 4- or 5-heptad repeat or greater is necessary to produce stable coiled-coils in benign neutral-pH buffer depending on the exact amino acid sequence. Thus, the predicted 3-heptad coiled-coil would have to be extraordinarily stable to exist as a trimer in benign medium, a point also discussed by Seilder et al. (7). This 22-residue stretch represents 30% of the sequence; it therefore is reasonable to expect that the helix content of this putative trimer would be at least 30%. CD provides an excellent physicochemical measurement of protein helix content because the helix spectrum has a large diagnostically distinct negative doublet minima pair at 222/208 nm (12). The CD run performed at ISUPF on a Jasco J-710 instrument (Fig. 1C
) showed no such minima pair for 3 different NT-proBNP concentrations. However, the spectra did show a minimum at <200 nm, which is indicative of a random coil (12), i.e., an unordered, possibly extended-like tertiary structure. The CD results showing no helix implies the absence of coiled-coils because such a quaternary structure requires association of slightly left-hand–twisted helices of 3.5 residues per turn.
The solution structure of NT-proBNP inferred from the respective results of the 3 experimental techniques is summarized in Table 1
. The experimental techniques so chosen allow for solution structural assignment of the peptide. Collectively, these data convincingly indicate that NT-proBNP is not a coiled-coil trimer and in fact is a monomer. This is a consequence of essentially no helix as assessed by CD, which implies no coiled-coil and therefore no quaternary association, i.e., oligomerization, of individual molecules. Finally, the ultracentrifuge data conclusively show that at moderate concentrations and in a benign medium, synthetic human NT-proBNP is monomeric.
|
It is not intuitively obvious how to reconcile the results from this study and previous work (1)(7)(13) regarding the oligomeric nature of human NT-proBNP. In the earlier work, the sample was prepared by hydrophobic solid-phase extraction and elution with organic solvent. It is unclear how this procedure could affect either association or disassociation of the analyte. The extractant was then chromatographed by SE-HPLC in benign buffer, and the column eluate was measured by immunoreactivity. The identified fraction was of "high molecular weight", the inference being oligomeric NT-proBNP. Another possibility involves a putative non-NT-proBNP binding component partner in serum, stable to SE-HPLC, that would produce an immunoreactive high–molecular-weight complex that collapses to "normal-eluting" (1)(7)(12) NT-proBNP after SE-HPLC run under denaturing conditions. This would be expected because the putative non-NT-proBNP–binding component partner is silent, i.e., unobservable, by immunodetection.
The actual solution structure of NT-proBNP must await high-resolution nuclear magnetic resonance or x-ray studies. One can speculate from the data presented here, however, that synthetic NT-proBNP is likely an unordered random coil with an extended-like structure. Whatever the case, human synthetic NT-proBNP is a monomer, and the potential confounding issue of analyte oligomerization is not a problem for this analyte.
Acknowledgments
I thank Vonnie Landt, Jitka Olander, and Jack Ladenson, in whose laboratory this work was performed, for suggestions and critical reading of the manuscript. The Mass Spectrometry Facility kindly provided instrument time and is supported by NIH Grants P41-RR00954, P60-DK20579, and P30-DK56341.
References
upregulation, processing, and antiapoptotic activity. J Invest Med 2002;50:421-434.[ISI][Medline]
[Order article via Infotrieve]
-helical coiled-coils. J Biol Chem 1984;259:13253-13261.The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  A. Hammerer-Lercher, B. Halfinger, B. Sarg, J. Mair, B. Puschendorf, A. Griesmacher, N. A. Guzman, and H. H. Lindner Analysis of Circulating Forms of proBNP and NT-proBNP in Patients with Severe Heart Failure Clin. Chem., May 1, 2008; 54(5): 858 - 865. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
