| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Abstracts of Oak Ridge Posters |
1
Roche Diagnostics GmbH, Nonnenwaldstrasse 2, 82372 Penzberg, Germany
2
Technical University Munich, 85748 Garching, Germany
a author for correspondence: fax 49-8856-603341, e-mail wolfgang.hoesel{at}roche.com
Preparative purification of macromolecules, e.g., proteins, has reached a high standard because of efficient separation materials that are available for this task. However, this is limited to samples containing substantial amounts of the desired protein. When proteins that are present only in minute amounts (e.g., in the µg/L range) must be isolated from a limited supply of complex biological material (e.g., human serum), a one-step method is recommended to obtain the protein in sufficient yield. When suitable antibodies are available, immunopurification often is the method of choice (1), but nonspecific binding of proteins to affinity matrices is a substantial problem because elution with acid or chaotropic agents very often washes off the impurities as well as the analyte (2). Thus, there is a need for methods that allow the specific elution of the desired proteins without eluting the impurities. Specific elution by competing with a low-molecular weight analog is appealing, but suitable analogs may not be available.
We wanted to isolate prostate-specific antigen (PSA) from serum and
analyze its structure by mass spectrometry. The concentration of PSA in
serum from patients with prostate cancer (PCa) ranges from 
3 to
>3000 µg/L, 105–107
times less than that of other serum proteins, e.g., albumin. We have
developed a general indirect immunosorption method that follows to a
certain extent an immunoassay principle developed by Hashida et al.
(3) and Ishikawa et al. (4) and makes use
of a digoxigenylated anti-analyte antibody. Initially, the PSA from the
biological sample is bound to magnetic beads by an array of antibodies.
The key step is the competitive release of the PSA-antibody pair by
digoxigenin-lysine under neutral conditions. These conditions leave the
impurities almost entirely bound to the matrix and yield the
antibody-PSA complex in a pure form (5).
When we began isolating PSA from serum, we first tried a standard
immunosorption method using magnetic beads combined with acidic
elution. To this end, streptavidin-coated magnetic beads were loaded
with a biotinylated anti-PSA monoclonal antibody (Mab) and incubated
with serum from a PCa patient. After magnetic collection of the beads
and several washing steps, the bound protein was released by a small
amount of a mixture of either formic acid-water-acetonitrile (1:3:2, by
volume) or 1 mol/L propionic acid, and the eluted material was analyzed
by nonreducing sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE) and matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS).
It was found that mainly nonspecifically bound serum proteins were
liberated from the beads (see lanes A1–A4 in Fig. 1
). The high degree of impurities did not allow the unambiguous
detection of free PSA and the
PSA/
1-antichymotrypsin (ACT) complex. This
applied especially to the elution with formic acid-water-acetonitrile,
a solvent used for preparing the matrix solution for MALDI-TOF MS and
which would therefore be very suitable for that type of analysis. But
when the elution mixture with this solvent was applied to the MALDI-TOF
MS, almost all of the proteins detected in the mass spectrum were
impurities. Free PSA was barely detectable, and the PSA/ACT
complex could not be seen at all. Because the direct immunosorption
method using acidic elution was not suitable for the analysis of
proteins either by SDS-PAGE or directly by MALDI-TOF MS, we reasoned
that we should use a specific elution step at neutral pH to avoid
coelution of nonspecifically bound proteins.
|
Again using streptavidin-coated magnetic beads, we constructed a
complex consisting of biotinylated anti-digoxigenin IgG/digoxigenylated
anti-PSA IgG/PSA on the beads. Specific elution of the complex
consisting of digoxigenylated anti-PSA IgG/PSA was accomplished using a
solution of a digoxigenin-lysine conjugate (2.5 g/L) at pH 7.3.
The eluted material was analyzed in comparison with that eluted by acid
(see lanes B1 and B2 in Fig. 1
). Below a molecular mass of 100
kDa, only the target analytes PSA and PSA/ACT were present in the
indirect immunosorption eluates, except for a protein band at 65 kDa,
which very likely represents albumin. Evidently, even these very mild
conditions elute small amounts of this abundant serum protein from the
beads. The protein bands above 100 kDa consisted mostly of the
digoxigenylated anti-PSA antibody and did not interfere with separation
of the analytes because of their high molecular masses. Analysis of
free PSA and PSA/ACT complex in the eluates by MALDI-TOF MS revealed
that except for serum albumin, only the target analytes could be
detected in the spectra (free PSA, 28.3 kDa; PSA/ACT, 83 kDa).
In the first experiments, an analyte recovery of 20% was obtained.
It became obvious that the affinity of the anti-digoxigenin antibody
used was a key factor because most of the losses occurred when the
beads were washed. By switching to a different anti-digoxigenin
antibody with higher affinity and lower
Koff rate, the overall yield could be
increased to 40%. A second set of antibodies was also evaluated,
which consisted of biotinylated anti-ruthenium-bispyridyl Mab
and anti-PSA Mab labeled with ruthenium-bispyridyl complex.
With this system, a recovery of 60% could be obtained when the free
ruthenium-bispyridyl complex was used for elution. The improved yield
in comparison to the digoxigenin/anti-digoxigenin system was very
likely attributable to the better binding characteristics of the
anti-rubispyridyl Mab and a higher concentration of the more
water-soluble hapten in the competitive elution step. The comparatively
hydrophobic digoxigenin-lysine molecule could also be interfering with
the crystallization step in the MALDI matrix preparation, so it should
probably be removed (e.g., by dialysis) before the crystallization.
This is probably not necessary when the rubispyridyl complex is used.
After removal of the free PSA from a PCa serum, the PSA/ACT complex was
cleaved by treatment with ethanolamine under alkaline conditions. The
released PSA was then isolated by immunosorption as described above,
and the intact PSA as well as the peptides obtained by an
endoproteinase Lys C digest were analyzed by MALDI-TOF MS and compared
with PSA from seminal fluid. The intact PSA as well as the peptides
obtained after digestion (80% coverage of the sequence) did not
reveal any structural difference between the PSA released from the
PSA/ACT complex of human serum and PSA from seminal fluid. These data
were published recently in this Journal (6).
The indirect immunosorption method described above allows the isolation of small amounts of analytes from complex biological material (e.g., serum) in a form that is almost free of impurities. The method is generally suitable if a ligand of high affinity for an analyte is available (e.g., antibody, receptor, lectin, or aptamer) that can be haptenylated. It is especially suitable for the isolation of proteins for analysis by SDS-PAGE and mass spectrometry (e.g., MALDI-TOF MS of the proteins or their peptides).
Footnotes
1 present address: National
Institute of Environmental Health Sciences (NIH/NIEHS), Bldg. 101, Room
F011, Research Triangle Park, NC 27709 
2 present address: Lehrstuhl
für Bioorganische Chemie, Universität Bayreuth,
Gebäude NW 1, 95440 Bayreuth, Germany
References
1-antichymotrypsin (PSA-ACT). Clin Chem 2000;46:474-482.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
