| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
a Author for correspondence. Fax 403-492-8241; e-mail msuresh{at}gpu.srv.ualberta.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Key Words: indexing terms: bispecific antibodies • tumor markers • screening • hybrid hybridoma • quadroma
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4.0 µg/L) in male blood plasma
(3)(4). Most of the naturally occurring PSA in
serum remains inactive by association with
1-antichymotrypsin (ACT) and
2-macroglobulin (3)(5)(6)(7). PSA
is not a neoantigen or cancer-specific antigen, but rather is expressed
by normal, benign hyperplastic, and neoplastic prostatic cells as well
other nonprostatic tissues (4). However, the total (free
plus complexed PSA) can be increased in conjunction with prostate
cancer, benign prostatic hyperplasia, and after surgical trauma to the
prostate. More recently, assays have emerged that measure free, total,
and complexed PSA forms, but the clinical significance and utility of
the measurement of these different forms are still under investigation
(5)(6)(8). Currently, there are many different assays for the measurement of PSA (7)(9)(10)(11)(12)(13)(14)(15). All of them involve monoclonal or polyclonal antibodies labeled with an enzymatic, fluorometric, or radioactive marker. Here we report the development of a novel bispecific monoclonal antibody (bsMAb) that could be used for enzyme-based measurement of PSA, with some potential advantages over current methods of detection.
BsMAbs are uniquely engineered antibodies bearing two different binding sites (paratopes) in a single antibody molecule, in contrast with the monospecific antibodies that possess two congruous paratopes (16)(17). This bifuctional design allowed us to develop a bsMAb with one site capable of binding PSA and the other an enzymatic marker (e.g., peroxidase). This immunoprobe with intrinsic enzyme marker binding capability can be used directly as a tracer in immunoassays (17)(18)(19) approaching the theoretical limit of the specific activity, with every bsMAb molecule uniformly bound to the enzyme marker.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
hybrid hybridoma generation
The fusion protocol used to generate the quadroma or hybrid
hybridoma was similar to a previously described method with some
modifications (17)(18). Approximately
2.5 x 107 drYP4 cells were fused with 2.4 x
107 B80.1 cells by using a 500 mL/L polyethylene glycol
(PEG) solution (Sigma, St. Louis, MO) for 7 min. The cells were washed
to remove all PEG and the pellet resuspended in 100 mL of standard
media supplemented with 100 mL/L growth factor (IGEN, Rockville, MD),
oxaloacetic acid, sodium pyruvate, bovine insulin (Sigma), 0.75 mmol/L
ouabain, 0.1 mmol/L sodium hypoxanthine, 0.4 µmol/L aminopterin,
and 16 µmol/L thymidine (Gibco BRL). The cells were plated at a final
concentration of 1.5 x 105 cells per well and
incubated at 37 °C with 5% CO2. Screening of the hybrid
hybridoma fusion supernatants was performed after 10 to 16 days of
culture with a sandwich assay as described below. Fifteen of the
stronger positive clones were selected, expanded, and frozen. The best
clone was recloned twice by limiting-dilution method.
bsmab sandwich assay for psa
Purified B87.1 MAb, kindly provided by Biomira, was used to coat
ELISA plates (Nunc, Naperville, IL). This MAb is a mouse IgG1 anti-PSA
antibody that recognizes a different epitope on the PSA and can be used
in a heterosandwich assay along with monospecific
125I-labeled B80.3 for the measurement of PSA (M.
Krantz and M.R. Suresh, manuscript in preparation). After coating (1
µg/well of B87.1 overnight at 4 °C) and blocking with 30 g/L
bovine serum albumin (BSA) in PBS for 2 h at 37 °C, the plates
were washed and 100 µL of the PSA containing LNCap supernatant
incubated for 2 h at room temperature. The plates were again
washed and 75 µL of supernatant from each well containing quadroma
clones were added. At the same time, 25 µL (100 mg/L) of peroxidase
in PBS were also added to each well. The mixture was incubated for
1 h at 37 °C. After washing,
2,2'-azino-di(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS; Kirkegaard
& Perry Labs., Gaithersburg, MD) substrate was added and the absorbance
measured at 405 nm. Affinity chromatography with a mixture of B80 and
B87 MAbs bound to a CNBr-Sepharose preactivated matrix (Sigma) was used
to obtain purified free PSA as described previously (20).
purification of bsmabs
Anion exchange
. Bulk cultures (1–2 L) of the selected
hybrid hybridoma (P57.3R2.21) cultures were prepared (17).
Approximately 1.5 L of supernatant was centrifuged to clarify and
remove cells, and solid ammonium sulfate was gradually added with
stirring to achieve 50% salt saturation. The stirring was continued
overnight at 4 °C, and centrifuged for 30 min at 3600g to
collect the pellet. The precipitated immunoglobulins were dissolved in
20 mL of 10 mmol/L sodium phosphate (Pi) buffer and
dialyzed exhaustively with two changes of 100 volumes of 10 mmol/L
Pi buffer. The crude immunoglobulin sample was loaded onto
a 50-mL bed volume DE52 column equilibrated with 100 bed volumes of 10
mmol/L Pi buffer. The column was washed with 10 mmol/L
Pi buffer. The absorbance at 280 nm was monitored
continuously. The washing was continued until all the unbound material
had been removed and the absorbance reading was again at baseline. To
elute the immunoglobulins, a linear ionic gradient was set up with 200
mL of 10 mmol/L Pi and 200 mL of 100 mmol/L Pi.
The column flow was 1.5 mL/min, and 50 fractions (7 mL each) were
collected. The fractions were tested with the bsMAb assay described
previously for screening of the quadroma supernatants. The fractions
with highest bispecific activity were pooled and the purity was
determined by a reducing sodium dodecyl sulfate–polyacrylamide gel
electrophoresis (SDS-PAGE) method with the Phast® gel
system (Pharmacia, Uppsala, Sweden).
Affinity purification
. HRPO (RZ 3.0, Sigma) was
covalently linked to a CNBr preactivated Sepharose by using published
protocols (21), at 10 mg of peroxidase per mL of gel.
Unbound peroxidase was removed by washing with PBS. Supernatant from
the hybrid hybridoma (500 mL) was repeatedly loaded onto this affinity
column by using a closed-loop system overnight at 4 °C, with a speed
of ~1 mL/min. The column was then washed with 10 mmol/L phosphate
buffer, pH 6.8. The antibodies bound to the column were eluted with 100
mmol/L glycine, pH 2.8. The fractions (2 mL) were neutralized with 50
µL of 1 mol/L Pi, pH 8.0. Each fraction was assayed for
bsMAb activity and pooled together on the basis of its activity.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
quadroma development
The protocol involving PEG and double drug selection was very
effective in generating anti-PSA x anti-HRPO hybrid hybridomas.
From the original four microtiter plates, >90% of the wells contained
clones. The initial screening to detect bsMAb was performed with a
two-step bsMAb sandwich assay described in Materials and
Methods. The results showed that almost 100% of these wells
contained bsMAb-secreting clones. Some of the positive wells gave an
absorbance of >3.0 in 15 min after addition of substrate. From these
wells, 25 primary clones were selected on the basis of their growth
characteristics and bispecific activity. Three of these were recloned
by limiting dilution, and the best of these reclones was chosen and
recloned again to ensure monoclonality of the quadromas and to develop
one subclone (P57.3R2.21).
psa purification
The affinity column was loaded with ~1.2 L of the LNCap cell
line with 0.990 mg/L PSA. The final yield after elution was 85% (1.012
mg). The purity by SDS-PAGE was ~95%, and only free PSA was present.
We calibrated this standard with a pure PSA sample from Scripps Labs.
(La Jolla, CA). All work on the optimization of the bsMAb immunoassay
was performed with this affinity-purified PSA.
initial immunoassay kinetics
With the supernatant from one of the most positive clones (P57.3),
a preliminary study was done to develop a bsMAb PSA sandwich assay,
schematically shown in Fig. 1
. The antigen, two antibodies, and the enzyme form a tetrametric
complex to generate the ELISA signal. In this assay the second step was
done varying the incubation time from 30 s to 3 min, and the
absorbance was measured 2 and 5 min after the addition of the ABTS
substrate (Fig. 2
a). The results show that even with a crude supernatant that
contains competing monospecific antibodies that could potentially
decrease assay sensitivity, the second step was almost complete at
60 s of incubation. The high absorbance at 405 nm was not
significantly different when compared with the 3-min incubation period.
This assay was performed with nonpurified LNCap supernatant containing
close to 1 mg/L PSA. The high colorimetric yield and fast second-step
kinetics achieved with a crude cell supernatant from the bsMAb
(anti-PSA/anti-peroxidase) in a two-step forward sandwich immunoassay
was also seen in a single-step assay (Fig. 2B
). Here all the components
were incubated together for 1 to 10 min and the reaction kinetics
monitored. These preliminary results showed promise for pursuing
further work on the purification of bsMAb and optimization of the
assay.
|
|
bsmab purification
Ammonium sulfate precipitation
. We first prepared
purified bsMAb by using an ammonium sulfate precipitation method. With
this procedure, we could remove some of the protein contaminants of the
culture supernatant and also concentrate the bsMAb sample.
Approximately 1.5 L of tissue culture supernatant was precipitated with
ammonium sulfate and the immunoglobulins resuspended in PBS. The final
concentration factor after dialysis was 1:20. This concentrated sample
of antibody was used in a PSA immunoassay; the dilution curve is shown
in Fig. 3
.
|
A bell-shaped curve was obtained as a function of the crude bsMAb concentration. This unique dilution could be explained by the fact that the presence of anti-PSA monospecific MAbs secreted by the quadroma could compete with the bsMAb, decreasing the signal in the more concentrated samples. At low dilutions of the concentrated sample (1:10 and 1:50), a strong competition between the monospecific anti-PSA and the bsMAb could ensue, in favor of the former. This competition decreases upon further dilution and the signal increases significantly. Higher dilution (1:1000) results in decreased signal because of low bsMAb mass.
Anion exchange
. The ammonium sulfate-precipitated
antibody was loaded on a DE52 ion-exchange column and the antibodies
were eluted with a phosphate gradient from 10 to 100 mmol/L. The
fractions representing the highest bispecific activity were pooled
(data omitted). SDS-PAGE showed that the pooled fraction A was 65%
pure immunoglobulins.
Affinity chromatography
. Affinity-purified bsMAbs were
obtained as a single eluted peak by using a 3-mL HRPO-CNBr-Sepharose
column (data not shown). The affinity eluate is expected to contain
both the bsMAb and the monospecific anti-peroxidase. Monospecific
anti-PSA MAb is washed off as the unbound fraction. The electrophoretic
analysis of this purification showed that some HRPO was leaching from
the column during the purification step.
development of a single-step two-site immunoassay
The essential requirements of a sandwich enzyme assay are a
matched pair of antibodies with high combined affinity and specificity,
a solid-phase support for one antibody, an analyte preparation to be
used as a calibrator, and an enzyme marker that remains highly
detectable when linked to the antibody and does not reduce antibody
affinity. The big advantage of bsMAb as tracer (Fig. 1
) is the fact
that the antibody bears an intrinsic binding site for the enzyme. This
avoids the need for chemical conjugation, which could potentially (in
some less-optimized conjugation methods) lead to loss of activity of
the antibody or the enzyme, aggregation, and formation of undesirable
complexes. Here we explored the optimization of the bsMAb performance
as tracer in a single-step immunoassay. We chose a single-step format
because of its convenience, even though there was a possibility of a
hook effect at extreme high concentrations of PSA.
One of the factors in the optimization of immunoassays was the concentration of the bsMAb. In this evaluation, HRPO was always used in excess (30 mg/L). A standard dilution curve of the purified antibody was used to determined the optimal dilution to be used in the single-step assay. The direct PSA binding assay showed (data omitted) that the pooled bsMAb in a 1:10 dilution had the highest binding. None of the purified samples presented the characteristic bell-shaped curve that may indicate the successful removal of the monospecific anti-PSA. On the basis of these results, we proceeded with the experiments by using the pooled bsMAb from the DE52 purification diluted 1:10 in all subsequent single-step assay optimizations.
The above pool, diluted 1:10, was preincubated with excess HRPO (30 mg/L) and used in a set of experiments involving four points in the low range of PSA (0, 4, 10, and 25 µg/L). The incubation time was varied from 10 min up to 2 h. In all these assays the solid-phase MAb B87.1 was coated at 1 µg/well in 100 µL and the additional binding sites on polystyrene blocked with 30 g/L BSA in PBS. There was no significant increase in the signal after 15 min of incubation time (data not shown). Hence, all subsequent experiments were performed with a 15-min incubation time.
detection limit
The lower limit of detection, defined as the unit value of antigen
above the zero value signal plus 2 SD, was determined with 12
replicates. Note that we used the less sensitive ABTS substrate; the
plate was read after 5 min and various times up to 1 h. The
detection limit was expressed in relation to the substrate incubation
time (data not shown). The detection limit of most commercial assays,
with incubation times from 1 to 3 h, are between 0.3 and 0.05
µg/L (Table 1
). Those limits could be reached with our 15-min assay even when
using a less sensitive peroxidase such as ABTS, which required longer
color development time. When we compared the detection limit with
tetramethylbenzidine (TMB; Kirkegaard & Perry Labs.) as a peroxidase
substrate under the same conditions as before, with 1 mol/L phosphoric
acid added to stop the colorimetric reaction after 5 min, the detection
limit of the assay was highly improved at 28 ng/L (Table 1
).
|
psa/act complexes (equimolar assay)
PSA has intrinsic proteolytic activity, and most of the naturally
occurring PSA in serum remains inactive by association with ACT. The
clinical utility of assay depends on the capacity of detection of the
total PSA, which represents free PSA plus ACT-complexed antigen
(3)(5). The assay is defined as equimolar if
it is equally reactive to ACT-bound PSA and free forms of PSA. We
evaluated the effect of added ACT in the assay and the capacity of the
assay to detect total PSA. Purified PSA was incubated with increasing
molar amounts of ACT for 24 h (6). The complex
formation was also confirmed by SDS-PAGE. ACT-PSA was used as
calibrator to compare with free PSA. ACT complexation had no
significant influence on the recovery of PSA (data not shown). This
indicates that this assay is also an equimolar assay, being equally
reactive to free PSA as well PSA-ACT. The above results were confirmed
also by using a commercially available (Scripps Labs.) pure PSA-ACT
complex. In a directed comparison, the Hybritech Tandem®
PSA calibrators were compared with the pure PSA-ACT calibrators for
relative recoveries. The two recovery results were superimposable (data
omitted).
methods comparison
To further evaluate our new bsMAb-based PSA assay, we tested 138
samples obtained from the Tumor Marker Laboratory, W.W. Cross Cancer
Institute, Edmonton, AB (courtesy of S. Popemma and Sheila
Stelmaschuck) against the results obtained with the samples previously
analyzed in an automated Hybritech enzyme immunoassay for PSA.
The results (Fig. 4
A and B
) show an excellent overall correlation:
r = 0.98, slope = 0.433 [95% confidence interval
(CI) = 0.415–0.451], intercept = 0.88 (CI = 0.45–1.31),
and Sy|x = 1.83 µg/L. In the
lower range from 0 to 10 µg/L, r was 0.95, slope 0.584
(CI = 0.539–0.629), intercept -0.33 (CI = -0.57–-0.09),
and Sy|x = 0.45 µg/L. The
calibrators prepared by us were ~50% lower in assay values than the
calibrators provided in the Hybritech kit; hence the decreased slope.
No hook effect was detected up to 100 µg/L PSA.
|
The intraassay CV was 27% with PSA values of 0.05–1.0 µg/L, 11% at 1 and 4 µg/L, 6% at 4–10 µg/L, 4% at 10–30 µg/L, and 3% at 30–100 µg/L.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although there are new approaches for the development of bifuctional molecules (22)(23)(24)(25), we opted for the generation of hybrid hybridomas as a convenient source of bsMAbs. Previously we had developed a double-resistant anti-peroxidase hybridoma (drYP4) that facilitated the development of the new anti-CA125 x anti-HRPO hybrid hybridoma (18). In this example, CA125 was representative of a huge tumor antigen with a molecular mass in excess of 200 kDa. Despite this enormous size, we developed an ELISA based on bsMAbs (18). Here, we chose PSA as an example of a small-molecular-mass antigen and two well-established anti-PSA clones (B80.3 and B87.1) to develop a rapid ELISA. The standard PEG fusion protocol generated bsMAb clones with high frequency. The PSA used in all our experiments was produced by the LNCap cell line, which produced close to 1 mg of PSA per liter of supernatant. These supernatants were used directly for all initial screening assays and also as the source of antigen by affinity purification.
Our initial evaluation of the bsMAb probe showed very high colorimetric yield. With 1 mg/L PSA, the absorbance was ~2.000 after only 180 s of incubation. These extremely fast second-step sandwich assay kinetics were our initial starting point for further development and optimization of a new PSA assay. Ammonium sulfate fractionation was used to remove some of the protein contaminants and also concentrate our antibody preparation. The serial dilution of this preparation demonstrated an interesting observation, namely a bell-shaped curve. At high concentrations (low dilution), the signal was low, likely due to a strong competition between the monospecific anti-PSA and bsMAb. As the dilution factor increased, the signal increased up to a point where the mass of bsMAb became limiting and the signal diminished. This enhanced signal due to dilution was also seen in primary quadroma cultures, and this phenomenon could result in an apparent decrease in the number of positive clones identified in hybrid hybridoma fusions if screening is not performed at various dilutions of culture supernatants.
Two methods for purification of the bsMAb were attempted: anion exchange and affinity purification. The affinity purification with HRPO coupled to preactivated CNBr-Sepharose was designed to produce an eluate containing the bsMAb and the monospecific anti-peroxidase but not free monospecific anti-PSA. The presence of monospecific anti-HRPO would not interfere in the optimization of the assay because saturating amounts of HRPO could be added, preventing any competition for signal quenching. The limitation of this method was the lower yield and also the possible loss of antibody activity due to the harsh conditions of elution. The HRPO column also showed slow leakage of coupled HRPO, which could contribute to the low yields. The anion-exchange column was effective and produced good yields even though the purity was only 60% by SDS-PAGE. The different fractions of purified material, when used in a serial dilution assay, did not present the bell-shaped curve, which indicated that bsMAb was relatively free of the monospecific B80 (anti-PSA) species.
Different assay conditions were optimized, resulting in the development
of a 15-min single-step assay. This demonstrated that the bsMAb could
be used in the generation of fast assays. Although the detection limit
was comparable with most commercially available assays with longer
incubation times, this 1-h substrate incubation would remove some of
the advantages of fast kinetics of the assay. For this reason we used
TMB as the substrate to reach a sensitivity of 0.028 µg/L in only 5
min of substrate development. The Hybritech Tandem-E PSA assay is a
single-step EIA with 2 h of incubation and 30 min of substrate
development, with a detection limit of 0.1 µg/L. It appears from
Table 1
that our bsMAb assay has good sensitivity, even with a short
incubation time. More recently, an ultrasensitive time-resolved
fluorometric assay for PSA was reported, involving alkaline phosphatase
plus diflusinal phosphate as the signal-generating system
(26). This assay has a detection limit of 0.3 to 1 ng/L in
a 90-min assay. It would be interesting to compare if (a)
our current peroxidase-based bsMAb assay would exhibit similar
ultrasensitivity with luminescent substrate, or (b) our new
anti-PSA bsMAb with alkaline phosphatase binding arm (unpublished
results) can achieve those remarkable sensitivities.
Bispecific MAbs have intrinsic binding to any two predetermined antigens, and, in fact, these molecules can be considered macromolecular cross-linkers (16)(17)(18). This uniformly reproducible 1:1 binding with the signal-generating arm has several advantages in immunohistochemistry and immunoassays (27). Although conjugation chemistry is better refined now than earlier methods, cross-linking of two macromolecules cannot be controlled to form 1:1 heteroconjugates. Some MAb molecules are either not labeled at all or have more than one signal molecule that could present size and steric problems due to random linking. In the bsMAbs, with excess HRPO, we expect that every molecule is uniformly bound without sterically interfering with the anti-PSA arm (Kreutz and Suresh, unpublished data). This could explain the high specific activity and fast kinetics. In addition, bsMAbs provide unique advantages in product development, with simplified production and greater assay consistency from lot to lot.
A portion of the naturally occurring PSA in serum remains enzymatically inactive by association with ACT. This fact should be considered in the development of assays capable of measuring the total PSA [free PSA + PSA-ACT (equimolar assay)]. Externally added ACT in increasing concentrations did not interfere in the recovery of PSA, which implied that the B87/P57 sandwich assay is an equimolar assay. Further, the recoveries were essentially similar when either pure PSA-ACT (Scripps Labs.) or Hybritech Tandem PSA calibrators were used.
Finally, we compared the bsMAb assay with the Hybritech Tandem EIA, also an equimolar assay. The Sy|x from 0 to 100 ng was excellent (1.83 µg/L), and even in low range (0 to 10 ng), Sy|x was 0.45 µg/L. No hook effect was observed up to 100 µg/L PSA, and with our exceptionally clean backgrounds, the tracer mass could be increased if necessary, to decrease potential hook effects. Also, the values obtained in our assay were approximately half those obtained by the Hybritech assay, which indicated a need to improve the preparation or conservation of our PSA calibrators. We have also recently used this bsMAb to map the potential epitope on the PSA molecule by synthetic peptide scanning techniques (20).
BsMAbs are powerful immunoprobes that could be used as ultimate enzymoimmunotracers, presenting a high uniform specific activity, wherein every bsMAb molecule is labeled uniformly with an HRPO molecule. We developed a bsMAb anti-PSA/anti-HRPO-secreting quadroma and optimized a manual sandwich single-step assay that can be performed in 20 min (incubation plus substrate development), preserving a high analytical sensitivity. This probe could also be used in the next generation of automated immunoassays as well as in rapid screening formats for physicians' offices or home diagnostics.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 Nonstandard abbreviations: PSA, prostate-specific
antigen; ACT, 1-antichymotrypsin; bsMAb, bispecific
monoclonal antibody; HRPO, horseradish peroxidase; PEG, polyethylene
glycol; BSA, bovine serum albumin; ABTS,
2,2'-azino-di(3-ethylbenzthiazoline-6-sulfonic acid); Pi,
sodium phosphate buffer; SDS-PAGE, sodium dodecyl
sulfate–polyacrylamide gel electrophoresis; TMB, tetramethylbenzidine;
EIA, enzyme immunoassay; and CI, confidence interval. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin. Clin Chem 1993;39:2098-2103.
[Abstract]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
