CLIN.
CHEM 26/10,
1450-1453
(1980)
Continuous-FlowZone Electrophoresis:ExperimentalDemonstrationand
Discussionof PotentialClinicaland ResearchApp’ications
Egil Fosslien, Donna N. Anderson, and Alan Nastir
We report continuous-flow sample application and continuous electrophoretic separation on a moving cellulose
acetate tape. This very simple automated electrophoretic
system requires minimum electronic control circuitry.
Samples and calibrators are pumped by a peristaltic pump
from sample containers onto a continuously moving cellulose acetate tape. Applied samples are interspaced with
rinse solution (buffer) and small air segments for sample
integrity. Electrophoretic separation, staining, and destaining are all continuous functions performed on the
continuously moving tape. The system appears suitable
for separation of proteins. Continuous-flow application
could be combined with discrete sample application onto
the same tape for continuous counter-immunoelectrophoresis; moreover, counter-electrophoresis
of two
continuously applied proteins, with one applied in a gradient, could potentially produce a system for automated
gradient counter-immunoelectrophoresis.
It should also
be possible to design a simple system for automated immunofixation electrophoresis.
Addftlonal K.yphrases:
cellulose acetate electrophoresis
hemoglobin
continuous-flow analysis
mechanized
analysis
gradient counter-immunoelectrophoresis
proteins
economics of laboratory operation
.
.
Several
principles
suitable
for partial
mation
of electrophoretic
procedures
have
ported.
Electrical field flow fractionation
or complete
already
auto-
been
re-
(1) and isotacho(3-6), are mainly
phoresis (2), as well as isoelectric focusing
used in research, isoelectric focusing having the disadvantage
of high cost of the separating medium. Laser spectroscopy
electrophoresis
(7, 8), potentially
a powerful analytical principle, has nevertheless found little application, probably because of problems with sample positioning. In automated
elution electrophoresis
(9), protein fractions of various electrophoretic mobilities are eluted sequentially from a porous
packed bed after a small aliquot of sample is injected directly
into the column. The electrical current passes through the long
axis of the column, and electrophoretic
separation and elution
take place at the same time. Separate sweep-buffer flows are
used to separate electrode electrolysis products from the eluates. The great advantage of this system is that continuous-flow analytical procedures can be applied directly to the
column effluent for (e.g.) automated isoenzyme analysis. The
limitation of the system at present is that only one sample at
a time is separated in the column.
Automated,
sequential electrophoresis with discrete sample
application and continuous separation on a Mylar-backed
cellulose acetate tape was previously reported (10-12). Recent
progress in this laboratory prototype system has allowed an
Department of Pathology, Abraham Lincoln School of Medicine,
University of Illinois Medical Center, Chicago, IL 60680.
Received Apr. 24, 1980; accepted June 24, 1980.
1450
CLINICAL CHEMISTRY,
Vol. 26, No. 10, 1980
rate of 300 hemoglobin or serum protein separations
per hour at a throughput
time of 30 mm each (details submitted to Clinical Chemistry for separate publication). Such
a rapid system for automated
electrophoresis
is suitable
mainly for large-scale
hemoglobinopathy
screening programs.
Recently, two commercial systems for automated
zone
electrophoresis on cellulose acetate have been introduced. The
first of these, the Rite electrophoresis
system (Olympus
Corporation of America, Hyde Park, New York, NY 11041),
is fully automated.
Segments of non-supported
celluloseacetate membranes are automatically
cut from a supply roll,
moistened, and moved to a sample-application
station, where
small aliquots of 20 samples are transferred
onto the moistened membrane from 20-well sample trays with a multisample applicator. Each cellulose-acetate
segment is fed into
one of two electrophoresis chambers, each of which holds one
cellulose acetate strip. After a 40-mm electrophoretic
separation, the membrane strips are fed onto a rotating drum in
a chamber, where they are stained, destained,
and dried. The
dry cellulose-acetate
strips proceed to a densitometer,
where
they are scanned as they pass through a well containing a
clearing solution. The densitometric scans and the percentage
distributions of the protein fractions are recorded on a thermal
recorder/printer.
Three hundred samples can be processed
in 7 h. Although fully automated, the Hite system thus far
performs only serum protein electrophoresis,
and the cost of
the instrument is relatively high ($50,000).
The Gelman Automated Electrophoresis
System (Gelman
Sciences, Inc., Ann Arbor, MI 48106) resembles systems previously. described
(10, 11). It involves use of a slotted,
Mylar-supported,
continuous
cellulose-acetate
membrane
tape for electrophoretic
separations. The tape is moistened
in a membrane-wetting
tank, then passes through a longitudinal electrophoresis
chamber, a process tank, a wash tank,
and a scanner, and onto a tape-drive spool. Samples are applied eight at a time to the wetted membrane by transfer from
a 128-sample tray with a multisample applicator, the application area being at the input end of the separation chamber,
The system sells for $45,000 and is therefore suitable mainly
for high-volume applications. Analysis of 128 serum protein
samples takes 3.5 h.
Many applications of electrophoresis do not require the high
analysis
rates of analysis
of the systems described
above, and most
users must amortize
the cost of the equipment
over fewer
samples. Simplifying
the instrumentation
would make automated electrophoresis
more cost-effective
for many clinical
applications
where at present
it is not. Additionally,
the
present commercial systems are not easily adapted for other
uses by individual researchers.
We report here our investigations
of continuous-flow
sample application and continuous electrophoretic separation
on a moving cellulose-acetate
tape. This novel combination
results in a somewhat slower, but greatly simplified, automated electrophoresis
system, simple enough for experimenters to assemble in their research laboratories.
In addition,
combining the principle of continuous-flow
application with
discrete sample application
onto the same tape should make
continuous
counter-immunoelectrophoresis
possible.
Materials and Methods
Hemolysates containing hemoglobins AS or AA were used
as samples. Blood was collected in 5-mL Vacutainer Tubes
(Becton-Dickinson,
Rutherford, NJ 07070) containing 7.5mg
of tetrasodium ethylenedmnitriotetraacetate
in water (Helena
Hemolysate Reagent; Helena Labs., Beaumont, TX 77704).
We prepared cellulose-acetate
tapes in our laboratory by
pumping a cellulose-acetate
casting solution (13; Formula 10)
onto a 0.1-mm thick, 12.5-cm wide polyester film (“Mylar”;
DuPont, Wilmington, DE 19898) at a rate of 3 mL/min while
the film moved at 25 cm/mm. After gelation under a partial
pressure of 1.67 to 4 kPa (12.5-30 mmHg) and at a temperature of 22 #{176}C,
a 0.1-mm thick membrane of cellulose acetate
was bonded to the Mylar base. The central 75mm of the tape
was cut out after the tape had dried completely, and was used
without further processing. Similar tapes from commerical
sources (Gelman; and Eyemark Corp., Houston, TX 77036)
are available in limited quantities on special order, but were
not used in these experiments.
Borate buffer (pH 8.4, 25 mmol/L; “super-Heme
Buffer,”
cat. no. 5802; Helena Labs.) was used to moisten the cellulose-acetate
membrane
and to serve as electrolyte.
The
sample application system was washed with buffer between
samples. Separated hemoglobmns were stained with Ponceau
S stain (Corning Ad, Palo Alto, CA 94306), then rinsed with
dilute aqueous acetic acid (50 mL/L). The experimental
separations
shown in the Figures were not cleared.
Apparatus
An experimental,
automated
electrophoresis
instrument
that has been developed for sequential, discrete zone electrophoresis described in earlier reports from our laboratory
(10) was modified for the experiments
described in this paper.
The discrete sample-application
system was replaced by a
simpler, newly designed
continuous-flow
sample-application
system; the methods for moistening
and staining the cellulose-acetate tape were also simplified.
The principal components of the modified electrophoresis
system are shown in Figure 1. Tape T moves from tape supply
roll TSR over alignment rollers Ri and R2 to tape uptake reel
TUR. The tape is driven by friction drive FD at a rate of 3
cm/mm. All wet-chemical processing stations are located along
the horizontal path of the tape (shaded area) where the cel-
lulose-acetate
membrane of the tape faces downward. The
distance
between rollers Ri and R2 is 80 cm. The first processing station consists of prewet tube PT made of glass (6mm
o.d., 2 mm i.d.) and containing a centrally located 1-mm diameter hole. This hole does not face the tape directly, but
rather is slightly downstream as shown in Figure 1, insert B.
PT is located 5 cm from roller Ri. Buffer is supplied to one
end of tube PT from a reservoir with a peristaltic pump; the
other end of tube PT is closed. A 22-gauge needle N made of
stainless steel hypodermic
tubing touches the membrane,
forming a sample application point SA located 12 cm from
roller Ri (Figure 1, insert A). The lower end of needle N is
connected
with a thin, flexible Tygon tube through peristaltic
pump P to a probe resting in sample-container
S. The electrophoresis chamber has been described previously (12) and
is not shown. The wicks contact the tape for 50 cm, starting
at a point 7 cm from roller Ri. A stain tube ST, constructed
just like tube PT, is located 62 cm from roller Ri. The open
end of tube ST is connected with a tube through a peristaltic
pump to a reservoir of stain. Rinse plate PL is next to the stain
tube, with the leading end of the platen touching the tape 9
cm from the stain tube.
PT-
Fig. 1. ExperImental
electrophoresis
instrument
for continuous-flow zone
Tape T moves from left to rlt; samples are applied atSA. See text for abbreviations. Separation time: 17 mm
During operation,
tape T moves continuously.
Buffer for
moistening
is supplied continuously to tube PT and stain is
supplied
continuously
to stain tube ST. A continuous
flow of
fluid is pumped through needle N onto the cellulose-acetate
membrane.
In our experiments,
we sequentially
applied
sample, air, and buffer solution onto the tape. We did not use
an automated sampler because we did not initially know what
sample rates could be achieved with acceptable carryover.
Results
Continuous
Sample Application
Results of continuous
application
and separation
of a he-
AS hemolysate
is shown in Figure 2 (top). The extra
band between the line of application and the hemoglobin S
fraction is probably carbonic anhydrase (14), which is clearly
visible here because of the heavy application and the use of
Ponceau S stain. The use of hemolysates permitted us to follow the sample application as it occurred and to observe the
separation
pattern through the electrophoresis
chamber.
Applying the sample continuously, we adjusted the application rate by varying the sample-application
pump speed; the
rate was set so that the line of sample application was approximately
1-mm wide. The sample application line is the
lower line in Figure 2 (top). The width of the line varies
slightly, indicating a slight variation in the rate of sample
application, possibly due to variation in the speed of the tape;
however, it is more probably caused by slight variation in the
pump rate of the peristaltic pump, which was controlled by
a variable-speed
motor. If you need to know only the percentage distribution
of various protein fractions after separation, a slight variation in the applied amount of sample is
permissible; however, for precise separations a miniaturized
peristaltic pump especially designed for uniform pump speed
at slow rates would be desirable.
As Figure 2 (top) shows, there is some variation in the migration distance for both hemoglobin A and hemoglobin 5,
indicating either variation in the properties of the tape itself
or inconsistency in the electrophoretic
separation. We found
that the homogeneity of the cellulose-acetate
membrane on
moglobin
CLINICAL CHEMISTRY, Vol. 26, No. 10, 1980
1451
buffer rinse solution, and air, and then studied
the carryover. Figure 3 shows a timing diagram for the sepaair segments,
rations shown in Figure 2 (middle).
Figures
2 and 3 show separations
of hemoglobmns
AA and
AS after application
with the continuous-flow
application
system. The sequence of samples wasasfollows: hemoglobin
AA, AS, AA, AS, AS, and AA. The approximate
rate of analysis was 30 samples
per hour. There is a small amount
of
visible carryover between samples and the buffer rinse, as
indicated by the thin lines in Figure 3. However, no carryover
is visible from hemoglobin AS samples to the next hemoglobin
AA samples. Migration
distances
apparently
remain fairly
constant from sample to sample. The first sample (at the left
in Figure 3) was applied for 25 5; however, as shown in the
fourth sample (AS), the application time could be somewhat
reduced, probably to about 15s, without affecting separation.
Bromphenol
blue could be added to every tenth sample for
sequential
sample identification.
Other hemoglobin
variants and proteins have not yet been
applied with this application
system. However, such separations have been performed on a similar type of separation
system,
involving rapid discrete applications at rates up to 300
samples/h
(data submitted
for separate publication). Scanning
could be performed with an automated, built-in scanner as
previouslydescribed
(i4), or the tape could be removed for
off-line scanning with an ordinary
densitometer.
For hemoglobin screening, trait conditions can be differentiated
from
clinically
significant
hemoglobinopathies simply by visual
Fig. 2. Top, continuous application and separation
of hemoglobin AS; bottom line represents application line; mkkile and
bottom, sequential application of hemolysates,
air, buffer, and
air
One hemotysate
contained
hemoglobin
application line Is cerbonic anhy&ase.
AA. the other AS. The line above the
Separation tIme: 17 mm
comparison
New Concepts
the tape was the most important
variable; consequently,
the
continuous application of sample can be used to indicate the
quality control of the tape itself.
In addition to separation of
5
Sam#{216}.
I
__
__
-
__
-
-
-
-
I
I
I
I
I
1
2
3
4
5
6
I
I
I
I
I
I
I
#{149}
.
f1
-I
Vol. 26, No. 10, 1980
mm
I___
it
I
CLINICAL CHEMISTRY,
.-
,
I
of sequential applications shown in Fig. 2, middle
I
r-1__,---1
I
ri
r
r-
1452
-
0
Air
Fig. 3. Time diagram
the present system of
anti-alpha-fetoprotein.
The analysis could easily be reversed,
with one serum applied continuously against a series of dif-
I
__
proteins,
continuous
sample application
could be used for automated
immunoelectrophoresis
in at least two ways. One application
would be that of counter-immunoelectrophoresis
of samples
and antisera, either applied both with continuous-flow
application or with one applied continuously and the other discretely (Figure 4, top), with counter-immunoelectrophoresis
of samples from several patients against one antibody,
e.g.,
Sequential
Sample Application
To make several separations
at once, the samples must be
applied sequentially, as shown in Figure 2 (middle and bottom). A major concern
in continuous-flow
systems
is the
amount of carryover that occurs between samples during their
transport
through the tubing system,
particularly
when very
small volumes are pumped. We utilized
a sequence
of sample,
Ce’
(i5).
I
This work was supported in part by USPHS
grant
R01HL20947-
01.
14 13 12
11 10
9
References
8765432
Fbtisnts Sem’
1. Giddings,
J. C., Yang,
F. J., and Myers,
M. D., Flow field-flow:
1320-1322 (1979).
Mohan, R., Steiner, R., and Kaufmann, R., Laser doppler
7.
Fig. 4. Concept of automated
continuous immunoelectrophoresis
Top, counter-lmmunoelectrophoresls
of several patient’s sera against monovalent antibody; bottom, gradient counter-immunoelectrophoresis: gradient of
multivalent antibody electrophoresed
against constant concentration ofpatient
serum. Horizontal locatIons ofprecipitmnlines should indicate concentration of
antigen Inserum
applied discretely
or against the same
antibody applied discretely but in a series of prearranged
dilutions for semiquantitative
counter-immunoelectrophoresis. Another interesting variant of counter-electrophoresis
isshown in Figure 4 (bottom),
where the antibody
isapplied
in a gradient by programming the rate of antibody application;
in this case precipitin lines should form, the horizontal location of which should be related to the amount of antigen and
the vertical location to the type of antigen in the patient’s
serum. Cost savings from automation might be offset by increased use of antibody; however, hybridization techniques
ferent
antibodies
for antibody
in the future.
production
should reduce the cost of antibodies
A
versatile new separation method. Science 193, 1244 (1976).
2. Haglund, H., Isotachophoresis-a
principle for analytical and
preparative separation ofsubstancessuch as proteins, peptides, nucleotides, weak acids, metals. Sci. Tools 17, 2 (1970).
3. Svenson, H., Isoelectric fractionation, analysis and characterization
of ampholytes in natural pH gradients. I. The differential equation
of solute concentrations ata steadystate and its solution in simple
cases. Acta Chem. Scand. 15, 325 (1961).
4. Svenson, H., Isoelectric fractionation, analysis and characterization
of ampholytes in natural pH gradients. II. Description of apparatus
for electrolysis in columns stabilized by density gradients and direct
determination of isoelectric points. Arch. Biochem. Biophys., SuppL
1, 132 (1962).
5. Harada, S., Isoelectrofocusing inacetatemembrane: The method
and application. Clin. Chim. Acta 63, 275-283 (1975).
6. Ambler, J., and Walker, G., Isoelectric focusing of serum proteins
on modified thin cellulose-acetate membranes. Clin. Chem. 25,
spec-
troscopy as applied to electrophoresis in protein solutions. Anal.
Biochem. 70,506-525 (1976).
8. Haas, D. D., and Ware, B. R., Design and construction of a new
electrophoretic light-scattering chamber and applications to solutions
of hemoglobin. Anal. Biochem. 74,175-188(1976).
9. Scott, C. D., Lee, N. E., and Perkins, A. W., Automated elution
electrophoresis: A potential clinical tool. Clin. Chem. 21, 1217
(1975).
10. Fosslien, E., Chen, B., Kavechi, A. E., and Preckshot, G. W.,
Automated
hemoglobin
electrophoresis. Am. J. Clin. Pathol. 62, 134
(1974). Abstract.
11. Foaslien, E., Automated
electrophoresis
instrument. U.S. patent
no. 3,896,021. July 22, 1975.
12. Fosslien, E., Automated
zone electrophoresis-expeiments
and
new concepts. Clin. C/tern. 23, 1436-1443 (1977).
13. Chen, Z. N., Development and production of cellulose acetate
membranes for an automated hemoglobin
electrophoresis
system.
Dissertation in chemical engineering. University of Missouri, Columbia, MO, 1975.
14. Nerenberg, S. T., Elect rophoretic Screening Procedures. Lea &
Febiger, Philadelphia, PA, 1973,p 133.
15. Schneider,R. G., Identification of hemoglobinsby electrophoresia.
In The Detection of Hemoglobinopathies,
R. M. Schmidt, T. H. J.
Huisman, and H. Lehmann, Eds., CRC Press, Cleveland, OH, 1974
pp 11-14.
CLINICAL CHEMISTRY, Vol. 26, No. 10, 1980
1453