DOCTORAL DISSERTATION SERIES

The Operd ion And ApplicariM R

TITLE

Of The Tiseh'us Electrophoresis
Apparaiw
AUTHOR

Chester H. fordt

UNIVERSITY. M ich ig an
DEGREE

fU l

S ta le

DATE
College

PUBLICATION NO

1

mm

111111!

'8

M

UNIVERSITY MICROFILMS

-P"

ANN

ARBOR

-

MICHIGAN

/ 9¥3

THE OPERATION AND APPLICATION
OF THE TISELIES ELECTROPHORESIS
APPARATUS

By
Chester R a Hardt

A THESIS
Submitted to the Graduate School of Michigan
State College of Agriculture and Applied
Science in partial fulfilment of the
requirements for the degree of
DOCTOR OF PHILOSOPHY

Department of Bacteriology

19U-3

*4

AC I0K W L 3DGLI3MT
I v»'ish to thank Dr. I. F

Hu dd 1.3son

The Operation and Application of the Tiselius
Electrophoresis apparatus

Infcroduction

........

Historical
Apparatus

......

Optical Systems

10

for follcs-ring and r e c o r d i n g e l e c t r o p h o r e t i c

separat ions
Kethods
Buffers,

>1 1

for d e t e r m i n i n g t o t a l p r o t e i n o o n c 9 n t r a t i o n * *
b u f f e r effects and

the p r e p a r a t i o n of b

C o ndu cbanco m e a s u r e m e n t

u

f

f

e

......

r

*3h

s

,.,..,., >.^1

Tre atmen t of e l e c t r o d e s and a s s e m b l i n g and f i l l i n g oells
ancl e l e c t r o d e

>5^

B o u n d a r y Cornpensauxon
H e a t i n g effects and convecscion c u r r e n t s

.......

»^9

Phocogi apiixng ^11*^ ooundar aos
C a l c u l a t i o n of m o b i l i t i e s

,71

a nd r e l a t i v e coneerit r

a

t

i

o

n

s

»

7

Bound a i y anotimlxe o

,91

Steps' to foil 0X7 iri a coniplete e l e c t r o p h o r e t i c
A p p l i c a t i o n

si-uxrxbure c

, , , , ,

i

5

b b b b b b ^ b b ,

^

o

u

b b b b b ' b , , 0 , 9 9 4

,

«

»

»

a n a l y s i s ® V,,",,'. » » , * 8 7 "

* * t * > a ***%*•;*

»

»

»

»

a »* * 0 & * > * * * * * * *

>

»

>

*

,

X13

INTRODUCTION

■When an electrical current is passed through a solution or
suspension of charged partioles, the particles migrate to the anode
or cathode.

The phenomenon is known as electrophoretic migration.

Abramson (1) has used the following illustrations to show that electro­
phoretic migration depends only upon the nature of the surface and is
independent of the size, shape, and nature of the bulk of the particles.
Figure 1 indicates the direction of electrophoretic migration of a wide
variety of charged particles in a neutral salt solution.

Figure 2 shows

the effect of adding a small amount of neutral gelati/v 'solution.

It

will be noted that the ferric oxide particle has reversed its direction
of migration and the following conclusions are evident.

"The electro­

phoresis of gelatin coated particles is primarily concerned with and
determined by a property of gelatin surfaces.

In addition there is every

reason to believe that we are investigating something very closely con­
nected to the properties of gelatin as it exists in solution" (l)«
Two widely different methods have been developed to study
electrophoretic migration, the microscopic and the moving boundary methods.
The advantages of the microscopic method have been considered by
Abramson (2) and are as follows:
" (l)

During the period of observation the environment of tine
particle does not change perceptibly*

” (2)

The electrophoretic mobilities of particles may be
compared by watching them move simultaneously.
Detection of slight differences between particles
that are visually indistinguishable may be accomplished
by this method.

B e fo r e

G e /a tm

Colon Bacillus

✓—
F e rr/c O xid e
—

*
Glass Fiber

A i r B u b b le
O

—

*

M in e r a l O il

Figure 1« The arrows indicate the direction and
approximate Telocity of particles of different
chemical constitution, size, shape, and orientation
in an electric field* (Abramson, Moyer, and Gorin,1)

Afi er Gelai in

Colon B acillus

/
F e r r ic O xide
0

—

G la s s F i b e r

A i r B a b b le

.

Mineral Oil

_______________________________

Figure 2. After addition of gelatin the direction
and approximate velocity of particles of different
chemical constitution,, size, shape, and orientation
in an electric field, (Abramson, Moyer, and Gorin,l)

-

” (3)

2-

High magnification may be employed, resulting in great
sensitivity.

H (J+)

The size, shape, and orientation of the particles under
observations are directly observed.

" (5)

Measurements of electric mobility may be made not only
in fairly concentrated salt solutions (up to about 1^5)
but also in very dilute solutions.

n (6)

The method is the only way in which certain biological
systems may be accurately investigated.

Thus bacteria,

fungi, blood cells, and protein particles are most easily
examined in this way.
ft (7)

There is no difficulty in calculating from the curve,
relating electric mobility to depth in the cell, the
electroosmotic mobility of the liquid relative to the
wall of the cell itself.

” (8 )

By means of the vertical cell, the same particle may be
observed over a comparatively long period.

Changes in

the surface of the particle during this time may be
followed.
The microscopic method may be employed only with microscopically
visible particles thus limiting its range of adaptation.
object of this paper to discuss this method.

It is not the

For the interested reader

an excellent review of the literature on this method has been presented
by Abramson (l).
The moving boundary method as developed by Tiselius (3) is
applicable to dissolved materials.

According to Abramson (l) Longsworth

has claimed the following advantages for this method.

” (1)

It is applicable to a wide variety of high molecular
weight substances in solution in both their native and
denatured forms;

(2)

it is applicable to mixtures of these substances, and
when applied to such mixtures yields information as to,
(a) the number of electrically separable components “in
the mixture, (b) the degree of electrical homogeneity
of each component, (c) the concentration and (d) the
mobility of each component;

(3)

moreover, the moving boundary method may be used to
separate in a pure state the components of a mixture;

(1+)

the mobilities of a given material as a funotion of
pH may be studied over the entire aqueous pH scale;

(5)

since the method is applicable over a wide range of
concentrations of the substance in a given solvent
it may be used to study the interaction between the
solvent and the substance and between the particles
of the substance itself*"

It is evident from the advantages of the two methods that
they supplant each other*

The microscopic method is designed for the

study of particles, whereas the moving boundary method is designed
essentially for dissolved material.
Many excellent treatises have been presented on the moving
boundary teohnique of studying electrophoretic migration.

Many of

these papers are of a theoretical nature and although much has been
written on the practicle application it is widely scattered throughout

the literature®

For this reason it seemed highly desirable to survey the

literature and supplement this material with the praotiole experience
gained in this laboratory vdth the sole purpose in view of preparing a
working manual on the moving boundary technique of electrophoretic
analysis®

Much of the following material has been drawn from the

excellent works of Tiselius and co-workers abroadj Longsworth and the
Rockefeller Institute group in this country and from Abramson and coworkers®

To the seasoned worker in the electrophoresis field this treatise

will seem rather elemental but it is hoped that newcomers in the field may
find much fundamental and practice*/ help on the subject as it is treated
in this thesis®

HISTORICAL

The Russian physicist Reuss (Ij.) appears to have been the first
to observe the phenomenon of electrophoresis*

In the year 1807 he con­

structed an apparatus consisting of two glass tubes driven into a d u m p
of moist clay.

In the bottom of each tube was placed a thin layer of

sand and the tubes were then partially filled with water*

Upon passing

a ourrent from a voltaio pile through the apparatus he noted that the
water surrounding the positive pole became cloudy due to the electro­
phoretic migration of the clay particles, while the water surrounding
the negative pole remained clear, but increased in volume, due to
electroosmosis*
More complete eleotroosmosis experiments were reported by
Porret (5) in 1816.

He divided the inside of a glass jar into two

compartments by means of a moistened bladder*

One compartment (anode)

was filled with water and the other (cathode) was only partially filled*
Upon passing a current through the compartments he observed that the
liquid moved from the anode to the cathode*

After one half hour the

levels in the two compartments were equal*

He summarizes his results

as follows t

”1 think that by the above experiment I have demonstrated

the existence of a power not before noticed in the voltaic current, namely,
that of conveying fluids through minute pores not otherwise pervious to
them, and of overcoming the force of gravity*”
A similar experiment using muscle strips as membranes was
performed by Kuhne (6)*

He too reported an increase in the volume of

water at the cathode accompanied by a decrease in volume at the anode*
Stern reports (7 ) that the first quantitative electrophoresis
experiments were undertaken by Wiedemann in I856,

-

6-

Using the microscopic method Quinoke (8) studied the migration
of microscopic particles of many types*

All particles investigated bore

a negative charge in distilled water but many were positively oharged in
other liquids*

He also conceived the idea of the electrical double layer

and showed that the rate of migration of particles in an electric field
was a linear function of the potential gradient*
Jflrgensen (9) in I860 reported on some results of experiments
by Heidenhain*

It was found that the passage of a current through cells

of "Vallisneria" oaused a migration of chlorophyll partioles and other
111 defined bodies towards the anode.

This migration could be reversed*

Jorgensen investigated this phenomena ^in vitro and found that the particles
moved towards the anode*
Lodge (10) used the moving boundary technique to follow the
eleotrophoretio migration of colored ions*
Picton and Linder (11) (1892-1897) were the first to study
proteins by the electrophoretic method.

They utilized a V shaped cell,

the forerunner of the TJ tube to study hemoglobin*
In 1899 Hardy (12) carried out electrophoretic studies on
denatured egg albumin particles*

He observed the behaviour of the

particles under three different conditions with respect to the acidity
of the solution,

"When the fluid has an alkaline reaction the effect

of the passage of a constant current is the formation of an opaque white
coagulum about the anode,
"When the fluid has an acid reaction the movements of the
proteid partioles and of the water are the reverse of those described,
"Fluid is neutral,,,There is now so little movement of the
particles under the influenoe of a current that it is difficult to detect,

and what movement there is, is due to the fact that the material is not
absolutely neutral*"
Hardy in 1905 (13) used the "boundary" method for studying the
migration of globulin suspensions*
Pauli (li;) found that dissolved salt free serum albumin behaved
in the same manner as the particles studied by Hardy*
Michaelis and associates (15) made use of the variation in the
rate and direction of migration with changing hydrogen ion concentration
to determine the isoelectric points of various proteins®

They found that

for each amphoteric substance there was a definite pH value, or zone of
pH values, corresponding to the absence of migration*
The earlier work on the electrophoresis of proteins is reviewed
by Pauli and Valko (16), Tiselius (3), and Prausnitz and Reitstotter (17)*
Svedberg and Jette (18) were the first to report a direct
observation of the electrophoretic migration of a soluble, colorless,
protein boundary*

In their novel method the green fluorescence emitted

by the protein when irradiated with long wave ultraviolet light was uti­
lized,

The migration was followed by photographing the fluorescing material

Wo quantitative results were reported at this time*

One of their origional

observations is shown in Figure 3®
In a later paper Scott and Svedberg (19) described quantitative
results by the fluorescence method on the mobility of egg albumin in
acetate and phosphate buffers of different hydrogen ion concentrations*
Figure Ij. illustrates the results of one experiment.
Several disadvantages were inherent in the fluorescent method,
namely, (1) a long period of exposure was necessary in photographing the

Figure 3 * One of the earlier direct observations
of the migration of colorless protein boundaries.
Photograph made by irradiating the protein solution
and photographing the resulting green fluoresoanoe.
(Svedberg and Jette, 18)

Figure 1+. Photograph of the migrating
boundaries of egg albumin solutions by
the fluorescent method*
(Soott and Svedberg, 19)

-

8-

boundary, (2) the V tube could not be immersed in a water bath, (3)
contamination by other fluorescing substances was not unoommon,

However

this method was the basis of work which followed and revolutionized
electrophoresis studies,
Svedberg and Tiselius (20) attempted to improve the method of
optical observation of the moving boundary.

They made use of the fact

that protein solutions absorb rather strongly, radiation of wave lengths
below 300 tijj. This method has been used previously by Svedberg and others
to study the sedimentation of proteins in the ultracentrifuge*

The general

arrangement of the apparatus employed by them is shown in Figure 5®

6 shows the type of electrophoresis cell used.

Figure

The portion of the U hi be

in which the migration of the protein solution is recorded by ultra-violet
photography consists of fused bubble-free quartz.

Figure 7 shows the

results of an experiment by this method.
Apparatus for preparative as well as analytical purposes was
designed by Theorell (21),

A number of successful applications with this

apparatus have been made in the biochemical field.

Of notable importance

was the purification of the yellow oxidation enzyme of Warburg by Theorell,
Stern (7) has reviewed these successful applications.

An improved pre­

parative apparatus of Theorell (22) is shown in Figure 8,
A now improved apparatus employing the nSohlierenT' method of
optical analysis was introduced by Tiselius in 1937 (23) (2l+),

This new

method of optical analysis using visible light was based on a principle
used by Toepler (25) for detecting flaws in telescope lenses.

The application

of this method requires only that the refractive index of the solution differ
from that of the solvent*
Figure 9*

The new cell and electrode assembly are shewn in

4

K

L

Figure 5« Tiselius electrophoresis apparatus for
light absorption method. L, lamphouse, double
walled and water cooled, containing as a light
source a quartz mercury lamp run from storage
batteries. In front of the housing is a screen
of ground, fused quartz. T, water thermostat
equipped with plane parallel windows of bubblefree fused quartz. The tank contains the electro­
phoresis cell(Figure 6). K, camera equipped with
quartz lens and a combination of chlorine and
bromine filters transmitting only radiations below
2800A°* (Svedberg and Tiselius,20)

Figure 6® Electrophoresis cell according to
Tiselius. The portion of the U tube in dark
line is bubble free fused quartz. The two
cylindrical limbs are ground with precision
to yield a uniform cross section® The bottom
of the TJ tube is conneoted through a capillary
and stopcock with the storage vessel for the
protein solution. The stopcocks at either side
allow concentrated potassium chloride to be run
around the electrodes in the electrode vessels.
(Svedberg and Tiselius, 20)

Figure 7» Eleotrophoresis of C-phyoooyan, a
chromoproteid from algae, by light absorption
method, (Svedberg and Tiselius, 20 )

■odA

F J

b§d
Figure 8* Preparative apparatus of* Theorell*s«
A* Separation chamber; B and C electrode vessels
(10 liter); D* Intermediary Chamber (10 liters)}
E and E communication pieces; F and Gr glass
cylinders; J. K. and L. Membranes, two of which
J and L are impermeable for the colloid to be
purified, while K is permeable for itj 1, 2, 3*
and !+, tops for admitting and withdrawing fluids;
P storage vessels for saturated potassium chloride
solution; T* water bath equipped with glass windows*
(Theorell, 22)

Figure 9. Cells and electrode vessels for
electrophoretic analyses. E]_» Eg represent
the electrode tubes, T rubber tubingsj I, IX,
III, IV, the four U tube sections of which
II and III can be moved with the pneumatic
arrangement P^, Pg and P.,. (Tiselius 23)

The glass U tube portion is divided into three separate
compartments and is connected vrith rubber tubing to the electrode vessels*
The glass sections are connected together with vaseline and are held in
place under slight spring pressure*

The sections are moved by a pneumatic

arrangement (dotted lines) and can be used for separation studies*

Silver-

silver chloride electrodes were immersed in a buffer solution in the large
electrode vessels and surrounded by a layer of saturated potassium chloride
solution*

A typical pattern obtained by using this technique is shown in

Figure 10*
The advantages of this new apparatus are (26)? (l) the reduction
of the disturbing effects of heat convection in the electrophoresis tube
by working at a temperature in the neighborhood of

(2) conventional

optical observation of the migrating boundaries by the Toepler-Sohlieren
method or similiar methods depending upon the refraotive index? (3) an
improved method for forming the boundary and dividing the contents of the
electrophoresis tube into portions for analysis; (1+) a compensation arrange
ment for preventing the boundaries from migrating out of the tube in
prolonged experiments, particularly valuable for separation studies*
Since this improved apparatus and its modifications are to be
the subject of this paper no further details will be discussed at this

Figure 10* The migrating boundaries of
normal horse aerum, sohlieren method*
(Tiselius, 2lj.)

-

10-

APPARATUS

The purpose of this paper is to discuss apparatus as well as
technique.

In conjunction with the discussion of the apparatus as a

whole some detail will be given as to the individual parts, the general
arrangement and construction.
Support

The electrophoresis apparatus is supported by three concrete
bases, A, B and A shown in Figure 11.

The specifications of the B block

are shown in Figure 12 and those of the A blocks in Figure 13*

The bases

extend into the soil below the floor and the sides are insulated from
the floor by means of asphalt.

Two steel channel irons (5M by 19 l/2’)

spaced 8W apart by means of steel plates, Figure 11 serve as the optical
bench.

They are anchored to the A blocks by means of threaded studs and

rest in the channel of the B block.

Figure li+ illustrates the manner in

which the various optical elements to be described later are attached to
the optical bench.

This arrangement using standard structural steel shapes

provides for flexibility of adjustment.
The electrophoresis apparatus on the optical bench consists of
three main parts, (l) a light source, (2) thermostat, (3) and camera*
Light Source
The light source L, Figure II is a General Electric Mercury
vapor lamp type Blj.* Placed in a horizontal position in the lamp housing
behind the slit it serves as an adequate source of illumination.
curvature of the glass bulb eliminates a condensing lens (27)*
of the lamp housing and slit are shown in Figure 15*

The
The details

The position of the

whole fixture may be adjusted by means of the slots in the angle iron.

c

—

30"-

30

30

---

30

10.5''■

60

-* /5 ‘

Fi-ure 11

39"

rr

/

Or

k si

H

’6

jfJL
_±-

_ A

30

274
Figure 12

Z 5

*

^p»
Figure 13

I inches

mum,

Figure l2|.. Diagram showing method of attaching
optical elements to the optical bench. The
steel angle iron, A, is clamped to the upper
flanges of the channel, C, with the aid of the
plate, P*, and the bolts b* The lens or other
optical element is carried by a steel plate,
P, which is bolted to the angle through slots
as shown at a. The iron pipe, B, is used to
space the channel irons 8” apart. (Longsworth,
27)

v<

-

11-

(The Hlj. type lamp requires a General Electric auto transformer $59^22
with 110 volt AC current.)
adjustable by screws.

The vertical dimension of the slit is

A length of 25 mm. and a width of .5 mm have

been used satisfactorily.
gauge when the lamp is hot.

The width is measured by means of a feeler
This has been found necessary as the slit

width contracts as the lamp heats up.
In work with cloudy or colored solutions th6 light waves of
a Hij. mercury vapor lamp may be absorbed by the solution.

Under such

conditions a 300 watt tungsten filament lamp and condensor (Spencer
Lens Co. Model 3^7) have been found to furnish satisfactory lighting
for most samples of this type,- If solutions show a marked turbidity
light from a carbon arc lamp (Bausch and Lomb Mechanical feed arc
lamp) furnishing light in the red spectrum will often penetrate through
the solution and permit a scanning photograph to be made.
Improved patterns are obtained by filtering the light and
permitting only a single wave length to pass through the solution.

For

blood plasma' and serum samples a "Wratten #22” filter is satisfactory.
With the mercury vapor lamp this permits only the light of 577 and 579
millimicrons through.

In addition to such filters, smoked glass filters

are often necessary to reduce the intensity of the light reaching the
photographic plate.

Over exposure of the plate causes a blurring of

the edges of the patterns.
Thermostat
A cross section view of the water bath is shown in Figure l6.
A 2[j. gauge tinned copper tank, 20” deep, 21;” long, and 12” wide is supported
by a wooden box lined with celotex.

The horizontal flange on the top of

the tank is covered with strips of black linen bake1ite.

n
vx- a.

■w.
□ BiKelite
[23 Celotex

E3 Gl4as
E 3

Steel
) Wood

Figure 16, Cross seotion diagram of water bath
with cell, E, in place, D, schlieren lens, W t ,
Wg, Wj windows of optical glass, (Longsworth, 27)

The Schlieren lens, D, Figure l6, forms one of the windows•
It is a lj.M achromatic lens of 36" fooal length,
Moffit).

(Perkin, Elmer and

The chromatic correction is for green and blue*

The double

window is necessary to prevent moisture condensation in the schlieren
lens.
glass*
ring.

The windows Wj, W 2 and

are I4. l/2 x 1/1*” discs of optical plate

The schlieren lens and glass windows are clamped to the bakelite
Moisture in the space between the windows is removed by circulating

dry air drawn through calcium chloride*

(A portable General Electric

vacuum pump #5KHi;3AB125 is adequate for this purpose).

Figure 17 shows

the outward appearance of the water bath with the schlieren lens in view.
This same figure also shows the stirring motor and standard and the
synchronous motor and syringe used to displace the boundaries*

It may

be remembered from Figure 11 that the water bath rests on the concrete
pier B, and is not supported by the optical bench or motor support as
might be inferred from Figure 17•
The synchronous motor with syringe attached is shown in more
detail in Figure 18* "When-the motor is

in gear and the current on, the -■

plunger is pushed into the syringe through a set of gears and a threaded
shaft*

The motor is attached in such a manner that its position may be

readily inverted and the plunger caused to withdraw from the syringe.
The speed of the plunger may be varied by changing the gear ratio between
the motor and the driving shaft.

The auxiliary gears are shown on a

spindle in Figure 18*
The water bath is circulated and cooled by the power stirrer
and refrigeration coil shown in Figure 19*

A mercury thermostat Figure

20 controls the electron tube relay switch, Figure 21, which in turn
regulates the power to the refrigeration unit.

A one quarter horsepower

Figure 17»

External view of water bath*

(Courtesy Klett Mfg. Co*)

Figure 19* Water bath refrigeration coil
and stirring motor. (Courtesy Klett Mfg.
Co.)

Figure 20* Mercury thermostat for controlling water bath temperature*
(Courtesy KXett Mfg* Co*)

Figure 21# Electron tube relay switch for controlling refrigeration
ualfe* (Courtesy Klett Mfg. Go.)

unit has the necessary capacity (Frigidaire-Model AF025)*

Since the box

remains at a low temperature at all times a provision has been made to
protect the stirring motor in case the bath freezes up*
is placed in series with the stirring motor*

A h amp* fuse

A second thermostat

(American Instrument Co. "Quickset") controls a buzzer that rings when
the temperature rises above or below the desired setting.
Camera
The camera barrel is shown in Figure 22.

The end of the camera

bellows facing the water bath supports the camera objective and the
schlieren diaphragm Figure 23®
lens of 36" focal length.

The camera objective, 0, is a 2" achromat

The circular disc, M, with slots of varying

width cut along radii of the disc, is the camera diaphragm.

This diaphragm

serves to control the amount of light reaching the lens and thus improves
its resolving power*

It may be rotated manually and once the proper slot

has been placed in front of the camera lens no further adjustment is nec­
essary®

The width of the slot employed will depend upon the light used,

the, filter,

and t h e type o f p h o t o g r a p h i c plate®

The camera shutter, C, is shown in place behind the camera
diaphragm in Figure 23*

It is a circular disc with a 2" hole in one

quadrant and a sector shaped opening in the opposite quadrant*

It is

not used in making Schlieren scanning photographs, in which case the disc
is set so that the circular opening is directly in front of the lens and
is left in this position.

It is used, however, in photographing Schlierin

bands and opaque boundaries that are occasionally encountered with very
turbid suspensions*

In the latter case the disc is rotated by means of

the electric motor, E, with the plate and plate holder in position.

Figure 23» Objective end of camera barrel showing sohlieren diaphraga D*
caas&f^ mask
and camera shutter C* Description in text* (Courtesy H e t t

As soon as the circular opening has rotated from in front of the lens
the dark slide is withdrawn from the plate holder and left out until
the sector shaped opening in the disc has passed in front of the lens,
thus making the exposure, after which the dark slide is inserted before
the circular opening again moves across the lens.

The exposure can be

controlled by loosening the nut holding the two leaves of the disc to­
gether and moving the two with respect to each other, thereby adjusting
the sector opening.

The achlieren diaphragm, D, Figure 23, is moved

synchronously with the plate carriage, C, Figure 2!+, by means of a
Bodine gear reduction motor (110 V - 60 cycle - Cat. No. B22li6 - 72
R. H» horizontal position) through the shaft, A, Figure 23*

The shaft

is equipped with a micrometer, D, Figure 21;, graduated to 0.01 mm. which
records the distance traveled by the schlieren diaphragm.

The distance

traveled by the plate carriage is read from a millimeter scale, V, Figure
2l+, directly above it.

The ratio of distance traveled by the schlieren

diaphragm to that of the plate holder can be calculated from these readings.
The ratio n a n be changed by changing the ratio of the gears shown at G

in Figure 21;.

The plate carriage is locked to the shaft, T, by means

of the ana L.

The clutch, K, serves to engage the motor, M, with the

driveshaft, A.

Safety switches, S, are shown in Figures 23 and 21;.

The camera mask (E, Figure 21;) which controls the area of
light striking the photographic plate is shewn in more detail in Figure
25.

The slit width of the mask is variable and can be adjusted by using

a feeler gauge.

The mask is so constructed that one-half of the vertical

length of the ceil may be blocked off if necessary.

Thus if all of the

protein constituents are migrating as anions and there are no boundaries

Description in t e J / nf c ^ r t ^ rK i ^ r^ g^

o^ hli,srsn scanning attachment.

Figure.25* Camera mask, M, and Philpot STensson cylindrical lens, L.
(Courtesy Klett Mfg. Co.)

-

15-

in •the lower anode and upper cathode section, these sections can be masked.
The same figure shows the Philpot-Svensson cylindrical lens which is in­
stalled in the camera bellows (22” from the camera objective) and can be
turned to intercept the light rays striking the photographio plate.
Electrical control Unit
The instrument panel of the electrical control cabinet is Bhovm
in Figure 2 6

(Ropf Apparatus Co.), The wiring diagram of the instrument

panel is sh p. in Figure 27.

The source of current for electrophoresis

comes from five J4.5 volt ”B” batteries connected in series.
resistance is connected in series to control the voltage.

A 10,000 ohm
If 110 volt

A. C. current from a power line is the source of power, a current rectifier
and voltage control unit

is necessary (Model V-United Transformer Co).

Current passing through the

electrophoresis cell is measured

roughly by a milliammeter and accurately by means of a potentiometer.
In order to increase the

range of the potentiometer two step resistances

of 100 and 1000 ohms are

connected in series with it. A standard cell is

used to standardise the bridge.

The other essentials of the control equip­

ment are a voltmeter, a de'Arsnoval type galvonometer and an automatic
timing recorder, (reading in seconds).
In practice the potentiometer is first standardized against
the standard cell.

In carrying out this operation the three-way control

switch is thrown to the position marked standard cell.

The potentiometer

is set at the standard cell rating (1.0185V.), the galvonometer is turned
on and the galvonometer damp released.

Standardization is accomplished

by adjusting the coarse and fine resistances until no galvonometer defleoticn
is evident when the key is depressed.

Figure 26. Electrical Control Cabinet
Description in text*

■o
O

B
ot
P o t e n t io m e t e r

Ext. e.m .f

Get/v.

o-

6 voits

HO A C.
M a

o

P -1 0,000

2 7 0 volts or
more
volts

to cell.

Figure 27

An accurate reading of the current flowing through the cell
can be made by determining the potential drop across the resistance
Ra Ra (Figure 27) which is in series with the cell, and applying Ohm's
law*

This measurement is made on the potentiometer with the three

way control switch in the internal E. Mo Fo position.

The value of

the resistance to be used can be figured roughly from Ohm's law (E=IR)
where E is the E. M. F. that can b® measured on the potentiometer
(1.6V being the highest) and I is the current reading on the milliammeter.
Set shunt R3 and R 4 to the value figured*

For current values under l6

miiiamperes through the electrophoresis cell, which will cover most
operating conditions, a resistance of 100 ohms is convenient.

By

multiplying the potential reading by 10 the current is read directly
in milliamperes.
A jack marked Ext. E. M. F. has been added in case one wishes
to make external voltage readings up to 1.6 volts.
the shunts R;; and R 4 are out of the circuit.
than 1.6 volts cannot be measured*

If this jack is used

Therefore, values higher

"When making such external measurements

the control switch should be thrown to the position marked "Ext. E. M. F."

OPTICAL SYSTEMS FOR FOLLOWING AND RECORDING ELECTROPHORETIC SEPARATIONS

Since most protein solutions are colorless the development of
the moving "boundary technique of electrophoretic analysis followed the
development of adequate and simple methods for abserving the rate of
migration of the boundary*

Tiselius (23) (2lj.) used the schlieren or

streak method to great advantage (28),

Abramson (29) has very clearly

explained the principle of this method with the aid of Figure 28*

n •..

light from a distant source is brought to focus by the lens to be tested*
A diaphragm is placed so that it just covers the image formed at f, the
focal point of the lens.

An observer looking through a telescope behind

the diaphragm and focusing on the lens o’serves only the rays which fail
to converge to a single point at f *

These rays which do not strike the

diaphragm come from defective portions of the lens and their position
is imaged in the telescope®

This procedure (used by Foucault), was

reversed by Toepler who by intercepting the deviated rays made the
defective portion of the lens appear dark in the telescope®

The present

method used for detecting concentration gradients in the electrophoresis
cell is en adaptation by Tiselius of the Toepler method»w

If the lens

being examined in Figure 28 is replaced by a protein solution and a con­
verging lens (schlieren) placed immediately behind the solution, dark
bands will appear at the boundary that have a greater index of refraction
than the adjacent solution®

It should be recalled that a beam of light

is deviated from its origional direction If its velocity changes in passing
from one medium into another®

If its velocity in the second medium be

less than in the first a beam is bent toward the perpendicular, that is
the angle of refraction is less than the angle of incidence®

When the

DEFECTIVE
spot

OPAQUE

EXPLORING

DIAPHRAGM

LENS

AT

SOURCE

TELESCOPE
LENS

LENS

TESTED

Figure 28. Diagram illustrating method of
deteoting defective spots in lens "by means
of an opaque, exploring diaphragm. Expla­
nation in text. (Abramson and Moore, 29)

IMAGE

OF

L IG H T

-

13 -

velocity in the second medium is greater than in the first a beam is bent
away from the perpendicular and the angle of refraction is greater than
the angle of inciderce.
The principles and theory of the schlieren method of optical
analysis as first used by Tiselius (23) (2ip) in electrophoretic work
have been so clearly and adequately described by Longsworth (27) that
any attempt to alter his description would be foolhardy#

For this reason

the following discussion will be for the most part verbatims
» k diagram of the schlieren method is shown in Figure 29#

An

image of the horizontal slit, 3, illuminated by the lamp, L, and condenser,
C, is formed in the plane, P, by the schlieren lens D.

The schlieren

diaphragm, A, a screen with a sharp horizontal upper edge, is placed in
the plane P and may be displaced vertically with, a micrometer#

The

electrophoresis cell, E, is placed as near the lens D as the thermostat
construction permits.

The camera objective, 0, placed immediately behind

the schlieren diaphragm, is focused on the cell and forms a full size
image of this on' the ground glass or p h o t o g r a p h i c p late at 'G®
"In the absence of refraction gradients, i.e., boundaries, in
the electrophoresis cell all of the light traversing the cell is brought
to focus in the image of the illuminated slit at P and enters the camera
objective.

If, however, a boundary is present in the tube the refractive

index, n, decreases with increasing heighth, x, through the boundary, and
the pencils uf light through this region are deflected downward.

If these

deflected pencils are intercepted by the schlieren diaphragm they do not
enter the camera objective and the region at G conjugate to the boundary
in the cell appears as a dark band on a light background.

D

E

„8'

Figure 29* Diagram illustrating the sehlieren method of optical
analysis* Explanation in text* (LongSTrorth* 27)

-

19 -

"For boundaries that are not too "sharp” i« e»» do not have
too large refractive index gradients, the angular deviation of a
horizontal pencil of light in a thin horizontal layer of the boundary
is proportional (a) to the gradient •g— , in the layer and (b) to the
breadth, a, (Figure 29) of the layer*

The displacement,^! , of the

schlieren diaphragm, A, from the position of the undeviated slit image
that is necessary to intercept the defleoted pencil is also proportional
to the optical distance, b, from the center of the cell to the diaphragm*
dn

Therefore 4 ' ab

in which a and b are constants of the apparatus and

dn
~ var5.es vertically through the boundary but is assumed to hs constant
in any thin horizontal section*”
"As the schlieren diaphragm is raised the first pencils of
light to be intercepted are those which have passed through the steepest
gradients of refractive

index.

In Figure 29 the paths of three pencils

of light through a boundary have been traced.

The p l a n e s , , and ' t f ' in

the boundary have been selected such that
dn' :■■ ■
• ......
33Tjb"
Max,
and the lines conjugates to these in the focal plane of the camera are
. .

- -

.dn,. „ ■■ dn- •

■■■■..dn. ■_■

t Z r

shown at

ti. *, 6 ’

and

1,

"The pencil through^ Figure 29 suffers the maximum deflection
to the position
throughob. and
plane.

1 near the plane of the schlieren diaphragm while those
are defleoted equally to a common position

With the upper edge of the diaphragm at 1 only the pencil through

£ is intercepted and the resultant dark band, or line, at
minimum width.
o( and

2 near that

• has a

With the diaphragm raised to position 2 the penoils through

, together with the pencils through all planes between <

and y-

-

20 “

are also intercepted and the band at G has broadened correspondingly.
With the proper conditions the displacement of the diaphragm from the
position of the undeviated slit image is proportional to the refraction
gradient at positions in the cell E conjugate to the edges of the schlieren
bands.”
Thus if a single component is present in a solution which has
been separated electrophoretically, a single dark band will show up on
the photographic plate (excluding boundary anomalies).

If several components

are present in the solution used then a dark band will show up for each
^
^ V J.
U U J lljjU iil/ iiU

p

i

O U U U V )

<

«

A.LJ.

n
C*

«
V W W W W * V iU V 4 . w n

n

A O

VN«. V»
J y " - '*

AOV\

4’

v u * .w

W *

ft

«VII«

M. *. VJ±*4.

4* V > «
V 4 iw

distances traveled by these bands from the initial boundary the mobilities
of the components may be calculated.
The schlieren method as origionallv used by Tiselius yielded
no information as to the relative amounts of electrophoretically separable
proteins present.

However, a method was suggested by Tiselius and developed

by Lararn (30) that overcame this difficulty.

The Lamm ’^cale” method gives

estimates o f ,the ..quantities, of .the various .constituents.,., their, mobilities ...

and relative homogeneities.
with the aid of Figure 30.

Longsworth (27) has explained this method
In this method the cell is moved toward the

camera objective a few centimeters and a transparent ruled scale, having
several lines per millimeter is placed in the position formerly occupied
by the cell,

,rIn the absence of refractive gradients in the cell, E,

a scale line, S, is brought to focus at s ’ by the lens 0,

If however there

is a gradient on the cell between the planes pp* the rays from S which are
collected by the lens have been deflected downward as shown in the figure
and intersect at approximately SH.

Due to the faulty lens action of 1he

gradient the image of a at s" is imperfect.

With sufficient depth of

focus s" appears on the plate as a line displaced by an amount § from

Figure JO* Diagram illustrating the Lamm
scale method of optical analyses* Expla­
nation in text, (Longsworth, 27)

Its position s* in the absence of gradients inthe cell.

In order for

the displacement of each line to be determined by the gradients in a
thin horizontal section of the column the lens 0 should have a long focal
length and be used at a small aperture.

The latter condition is also

desirable in order to increasethe focal length,”
In using this method the filled cell is put in

place and the

scale is photographed through the cell with the schlieren diaphragm
removed so that schlieren bands will not obscure certain of the scale
lines.

After completion of the experiment the scale lines are again

photographed through the oell,

Bv using a comparator the positions of

corresponding lines on the two photograph in the neighborhood of the
boundary are computed*
If the refractive gradients are not too groat the displacement
of a scale line is proportional to the horizontal breath of the
boundary* a, the optical distance, Q p from the scale to the center of
dn ■
the cell and the gradient ^ « For unit magnification
c

b -

„ dn
a(3

the value of s” mist be reduced by a factor since the electrophoresis cell
undergoes an apparent enlargement when moved out of the focal plane toward
the camera.

The factor is
L « £
L

TOiere L is the optical distance from the scale to the camera.
By plotting scale line displacement on the y axisagainstdistance
from the initial botmdary on the x ajiis a pattern is obtained thearea

of

which under certain conditions is proportional to the concentration of the
component.

The plotted results obtained by KeKwick (31) using this procedure

are shown in Figure 31*
The schlieren method was modified by Longsvrorth (32)» so that
graphs of the gradient dn/dx, in a thin horizontal layer

of the column

as a function of the position, x» of the layer could be recorded auto­
matically*

Figure 32 illustrated this "schlieren scanning" method.

An

image at P of the illuminated slit s-s is formed by the schlieren lens D ,
The camera objective 0 is focused on the cell E and forms an image on the
screen at G~G®
buffer.

A boundary, B, is shown between a protein solution and a

This boundary does not consist of a single geometric plane but

a region in which the composition varies gradually from that of one solution
to that of the other as shown by the density of the shading in Figure 32®
The refractive index in this region changes with the heighth h, in the cell
and the refractive index gradient varies from zero to a maximum and back
to zero again as shown by the pencils of light passing through the boundary.
The horizontal plate movement of the carriage C, Figure 22+, past the narrow
slit H, Figure 2i+, has been synchronized with t he vertical displacement
of the schlieren diaphragm D (see also Figure 23),

If the plate carriage

is connected with the driving mechanism of the schlieren diaphragm when
the latter is displaced so that no deflected rays are intercepted the
slit at G is entirely illuminated,

'’
/Then the driving motor is turned on

the plate moves towards the reader (Figure 22+) as the schlieren diaphragm
is raised.

Figure 32a shows a band in the slit G-G resulting from the

interception of a few pencils of light that have suffered maximum refractions
The area e-f-g shows that portion of the photographic plate that would have
been exposed up to this point and the darkened area in the slit to the right
shows the band as it would appear on a ground glass plate.

Figure 32b shows

M ite r a t io n

d ir e c T io n .

O-0 J

0 45

t-O

Figure 31* Electrophoresis of normal human
serum. Photographed by Lamm scale method.
Ordinate t soale line displacement in mm.
Absoessa* distance in U tube in cm.
(Kekwiok, 31)

a

s

g
'e

(a)

a'

g'

5

8e'
<t>

i-

s
e

(C)
Figure 32

the schlieren diaphragm at the position q ’ and more pencils of light have
been intercepted causing a widening of the band at G-G*

Since the plate

has moved over further the exposed area of the plate g ’-f’-g* is larger#
Figure 32c shows the schlieren diaphragm at the position 411 just below
the focal point o f the lens 0.

The penoils of light suffering the

minimum deflection have been intercepted and the area

el,“f n-g”has reached

its maximum area*

position q” the

As the diaphragm moves up from the

entire vertical length of the slit at G-G becomes dark.

By continuing

the operation the entire field has been covered all of the boundaries in
one

c tn x

xiuttgt?

turt?

p u o w u g i'tip i& e u *

ia o

ouuvvw i

wx

w uv

in c ic a v S S

both the position of the boundary and the magnitude of the refractive index
gradients existing in it.

If the change in refractive index is proportional

to the change in protein concentration in a boundary then the area e-°f«g
is a measure of the change in protein concentration from one side of a
boundary to another#

It has been experimentally shown that the schlieren

scanning method gives pattern that are coxaparable to the Lamb scale method
(2?). '

'

'

With a gear ratio

'
of 1-1 from the motor drive shaft to the

plate carriage shaft the ratio of the plate movement to schlieren diaphragm
displacement is 6*1.

The distance traveled by the schlieren diaphragm is

U .20 mm per minute making 25*20 mm per minute the rate of travel of the
photographic plate in the carriage past the slit.

Thus three minutes is

the average time necessary to take the complete scanning photograph of the
boundaries*
Another simple and adequate method for recording electrophoresis
patterns was suggested by Philpot (33) and modified by Svensson (3U)®

The

Philpot-Svensson cylindrical lens method has been explained by Longsworth

(35) "with the aid of Figure

33*

"An illuminatedhorizontal slit, present

on the left of S but not shown in the figure, is focused by means of the
schlieren lens, S, in the plane of the schlieren diaphragm, D«
contains a diagonal slit, kk, as shown in the front view, D*®

The latter
The camera

objective, C, is focuBed on the electrophoresis cell, E, and forms, in the
absence of the lens H, a normal image of the cell on the ground-glass or
photographic plate at Go

The cylindrical lens, H, with its axis vertical,

is focused on the schlierendiaphragm and also on the plate
from the side (Figure 33a),

at G© Viewed

the cylindrical lenshas no effect on the

pencils of light forming the cell image©

Thus the vertical coordinate of

each point in the image is conjugate to the corresponding level in the
channel E and, owing to the focusing action of the camera lens, C, this
also remains true for pencils that may be deflected by gradients in the
channel®

Viewed from above however (Figure 33^) the cylindrical lens,

in conjunction with the diagnoal slit, causes, as will be shown below, a
lateral deviation of a pencil of light that is proportional to the vertical
deflection the pencil has suffered

in a boundary®

The curve to the ri ght

of G in Figure 33®- represents the pattern of the boundary, B, as it would
appear on the screen if the later were hinged at the side and turned toward
the reader, where as if it were hinged at the top and turned, the pattern
would appear as in Figure 33b©
"If the fluid in the electrophoresis cell is homogenous, all of
the light through the channel is concentrated in an image of the illuminated
slit at the upper or normal level on the diaphragm, i.e., f ’ of D* or D*

As

can be seen from the figure, only the extreme left hand portion of the light
in this i®age passes through the diaphragm to form a straight verticle line,
i.e., the base line, on the screen at the position c-f»

The width of this

line varies with the width of the diagonal slit and, sine© a wide line is undersirable, Svensson has made the practical suggestion of tapering the end of

xl

CQ)

(b)

t

V

Figure 33. Diagram illustrating cylindrical
lens method for the observation of electro­
phoretic patterns. (Longsworth, 35)

the slit kk to a point*

If, on the other hand, a boundary B is present

in the cell, a pencil through the layer of solution in the boundary having
the maximum gradient, for sample, is deflected downward as indicated by
the line d ’d and forms an image of the slit at the lower level d* on the
diaphragm*

Owing to the angle the diaphragm slit makes with the vertical,

the portion of the light in the lower image d' that enters the slit is
shifted laterally from the position at which the normal pencil enters by
an amount proportional to the vertical, deflection in the boundary gradient*
The cylindrical lens consequently imparts to this pencil a corresponding
lateral shift xn the o p p o s ^ . w

kj

uii*

rx^ui'6

without affecting its vertical position i.e*, d of Figure 33a *
”The path of a pencil through another portion of the boundary
is indicated by the line e*-e and forms the corresponding element in the
pattern*

All other elements in the complete pattervare formed similarly*”
In a discussion on the cylindrical lens, Longsworth (35) has

summarized its use as follows®

nThe cylindrical lena method is convienient

for visual observation during an-"experiment'and for'the control of electro­

phoretic separations, since the pattern may be viewed directly on the screen
of the camera*

For visual observation a slit diaphragm is preferable to a

straight edge, since the greater quantity of light reaching the screen In
the latter case causes a decrease in the apparent contrast between the
pattern and the background*

For photographic work, however, the straight

edge Is to be preferred, owing, in part, to the superior resolving power
and simpler diffraction phenomena characteristic of the diaphragm*
Consequently the permanent photographic records of an experiment, on which
the analyses are based, are obtained with a straight edge as diaphragm
and with the scanning procedure.

With the latter method the optical errors

inherent in the unoorraotad cylindrical lens are eliminated”*

As desoribed in the section on apparatus our camera has been
modified so that either the scanning or cylindrical lens procedures can
be used interchangeably®

Figures 3U&, 3Ub and 3U« show complete electro­

phoretic patterns obtained from the electrophoresis of human blood serum
and photographed by the schlieren scanning method, the diagonal slit
method and the diagonal straight edge method respectively®

If a slit is

used with the schlieren scanning method patterns of the type shown, in
Figure 3U° are obtained (36)«

In either method a slit us-'d as a diaphragm

produces the pattern as an illuminated line on a dark background whereas
'
JLXX

t.T,-

n. n

-

— f
V X

r\
M

-4o WA

A

r-Vii*

rt /4
U S t g V

between a light and dark field®

4-

-4--4-

v »
■
■
»

4
x

h>

/-»!—,4-f» ' vs n >4
V /U W v S J .U 9 U

«s
UkO

A- is r<
V i i V

V V 4 iW U U A

F ig u r e

r ir u r o

3Ub

Figure Z'ho

Electrophoretic boundaries of nomal human plasma diluted to one per­
cent in phosphate buffer, pH 7«5> ionic strength 0.2. Electrophoresis
carried out for 10,800 seconds at a potential gradient of
volts
■er centimeter. Boundaries photographed by the schlieren scanning
r.ethod, Figure 3U&1 the cylindrical lens and diagonal straight edge,
"igure 3Ub; and the cylindrical lens and diagonal slit. Figure 3Uc.

METHODS FOR DETERMINING TOTAL PROTEIN CONCENTRATION

If the total protein concentration in a sample is known the
concentrations of the various components separating during an electro­
phoresis experiment can be oaluclated from certain data*

Since it is

desirable to obtain this information in most cases, several well known
methods for determining protein concentration will be discussed in
some detail*
Sinoe the schlieren method for photographing boundary positions
and areas depends upon differences in refractive index between buffer and
protein solution, it is not surprising that refractometric methods are
widely used for determining total protein concentration*

Siebenmann (37)

has modified Reiss’s (3$) refractometric method for determining serum
protein*

His so-called "graphic" method is applicable to native serum

and protein.

This method utilizes a dipping refractometer and the readings

are made on the solution in a constant temperature water bath at 20°C*
Three to four ml. of serum are required for a determination.

A standard

graph (Figure 35) is made by plotting refractometric readings against the
total protein concentration as determined by the k;jeldahl or some other
method*

The protein concentration of an unknown serum sample can then be

read directly off the graph if the refractometer reading is determined*
This method is reliable when the salt and non-protein concentration in
the sample are of the same order as the samples used for determining the
standard ourve*

The graphic method cannot be applied to serum dilutions

or other protein solutions containing unknown amounts of salt.
In our laboratory the "differential” method of Siebenmann’s (37)

9
Total

protein

10

°0

Figure 35* Standard ourve shewing percent protein
(microkjeldahl) against soale readings.(dipping
refractometer) (siebenmann, 37)

-23-

1b used*

This method is based upon the refractometric change taking place

when the protein solution is coagulated by heating*

The difference between

the refractometric reading of the protein in an acetate-acetic acid buffer
(pH U.6) and the same sample after heating at 100°C for 15 min. is taken
as the index of protein concentration*

For serum and plasma samples the

following technique is employed*
To 10 ml* of a sodium acetate-acetic acid buffer (11? gms
glacial acetic acid-236 gms* sodium acetate per liter) is added 2 ml* of
serum*

The refractive index is determined at 20°C by the dipping rafracto-

meter on ii ml. of this mixture.

The rest is used for heat coagulation,

employing a special double tube shown in Figure 36.

By use of the tube

arrangement aqueus solutions may be heated at lOO°C, without evaporation*
The size of the outer tube containing water is 20 x 150 mm.* that of the
inner tube 15 x 125 mm. (pyrex no. 2370)®

The double tube is placed in

boiling water for fifteen minutes.

It is then cooled and the inner tube

containing the sample centrifuged*

The refractive index on the protein-

free supernatant is determined as-before*

For the calculation of the percentage of protein from the scale
readings, the specific refractive increment of the protein must be known.
This represents the change in the refractive index of the protein solution
due to a change in protein concentration of one per cent®

The specific

refractive increments, a, have been carefully determined for horse serum
albumin and globulins.

The variations in the increments with changing

wavelengths of light have been reported by Pederson and Andersson (39)*

*9

-O uter tube
containing w ater

In n e r tube
containing b u iltrc d
p ro tein solution

Piece o f vaeuum
' ubing

Figure 3^» Diagram of "test: 1311)© arrangement
for heat coagulation of protein solutions.
VSiebenmann, 57)

Wavelength (*vO
366
436
5I4.6

(a) Albumin

(a) Globulin

0.00198

0,00202

589

0,00190
0.00185
0.00183
0.00183

656

0.00179

0.00195
0.00187
0.00186
0 .00I86
0.0C182

579

The scale readings

m ay

be converted into refractive indices

and the protein percentage calculated by means of the following equation
C “

where

a

* dilution

is the refraotive index before and ng the refractive index after

heat coagulation of the protein.
The conversion of scale readings to refractive indices may be
avoided by using the following equation
C = (R^ - Rg) »k.dilution
here R^ and Rg are the scale readings before and after protein coagulation
and k is the amount of protein in per cent which on heat coagulation leads
to a drop in the dipping refractometer reading of one scale unit.
constant k may be determined as follows*

The

Tables for converting dipping

refractometer readings to refractive indices show that in the range of

20 to 30 scale units the constant difference in refractive index for a
change of one scale unit is 0.000383.

Taking the refractive increment

value of horse serum as 0.00185 (value to be used with a tungsten
incandescent lamp) the constant k is calculated
.000363

« o.207?$

.00185
where 0.207 represents the percentage of horse serum protein equal to
one scale unit,

(0.205 in the range of 30 to U0 scale units).

Since

very little information is known concerning specific refraotive increments,

-30-

the values for horse serum are used for other protein solutions*
Examples

Serum 6l6, dilutions

buffer

2 ml serum plus 10 ml aoetate

“ 29*0, Bq *= 23*5
C ** (R^ - Rg) K • dilution

c = 5*5 * 0,207 x 6 *
The ’’Differential” method is the only refractometric method
available for determining protein concentrations in samples with an
unknown salt content.

The method estimates protein concentrations that

are not influenced by the amount of non-protein substances present.

For

determining protein in solutions other than serum and plasma slight
dilution modifications are neoessary depending upon the protein
concentration of the samples.

In all cases the ratio of protein

solution to buffer should be made so that upon heat coagulation a suit­
able drop in scale readings is observed (l to 10 units),

Siebenmann

(37) gives the following dilutions to be used covering a wide range of
protein concentrations.
Protein

percentage
range

Acetate - buffer (ml)
pH 4*6

4-12

1-4
0.2-1

10
5
1

--■■■■■■Protein
Solution (ml)

-r

2

+

5
10

+

Dilution

6
2
1.1

For protein solutions containing less than 1 per cent protein
a more concentrated buffer is used (56.6 gms, glacial acetic acid-118 gms.
sodium acetate per liter).
In our laboratory a Bausch and Lomb dipping refractometer with
constant temperature control equipment (No, 33-45-86. Bausch and Lomb)
is employed for making refractometric readings.

The refractometer and

constant temperature water bath are shown in Figures 37 and 38.

MQ*

-B^®6 fc o ^ ® SV

9

Figure 38* Bausch and Lomb constant temperature
oontrol equipment showing attachment to water
bath* (Courtesy Bausch and Lomb)

-31-

Siebenmann (37) also describes two gravimetric methods for
determining protein concentrations.

In the heat coagulation method

2 to 5 ml. of serum or protein solution are diluted in a porcelain
dish (acetic aeid-sodium acetate buffer pH I+.6) so that the dilution
contains not more than 1 per oerrt protein (2 ml serum + 18 ml buffer).
This solution is covered and heated for 20 min. on a steam bath®
protein precipitate is then filtered and washed with hot water*

The
The

filter paper is then dried in an aluminium dish for 16 hours at 100®C*
and weighed*
The acetone gravimetric method of Eierrv and Vivarrc (JLj.Q)
modified by Gfuilaumin etal. (ijl) has been further modified by
Siebenmann (37)*

Two to 3 ml» samples of plasma or serum are

pipetted into 10 ml. of acetone in a centrifuge tube.

The acetone

precipitates the proteins and dissolves some of the lipides, after
stirring and standing centrifugation is employed and the supernatent
discarded*

The precipitate is washed twice with acetone and again

centrifuged.

2 0 ml. of 0 * 6 per cent sodium chloride solution is

added to the precipitate causing partial solution of the precipitated
protein*

To the suspension is added acetic aoid-sodium acetate buffer

pH Iu6 and the mixture covered and heated on the steam bath for 20
min.

The coagulated precipitate is then filtered, washed and weighed

as described under the heat coagulation method*
According to the author, all of these methods give sharply
reprodi

ible re stilts and agreement between the various methods, where

applicr' le, is good®
A semi-microkjeldahl nitrogen method for determining protein

concentration may be required for oertain types of samples*

A modification

of two methods (1*2) (1+3) has been used with great satisfaction
laboratory.
attached.

in this

An all glass distillation apparatus with digestioi tube
(M-3065 Scientific Glass Apparatus Co.) is shown in Figure 39.

In making a determination the digestion flask is removed and
2-5 ml of the protein solution (depending upon N content) are added.

To

this is added 10 ml. of a digestion mixture of the following composition
per liter*
2 gms. copper sulfate
100 gms. potassium sulfate
200 ml, concentrated sulfuric acid
After the addition of a few glass beads to prevent bumping, the flask is
heated in a hood over a micro-burner.

Heating may be rather rapid at

first if frothing does not occur, but when the white fumes of sulfur
trioxide begin to evolve the heating must be cautiously observed to
prevent bumping and subsequent loss of material.
digested until it is clear.

The mixture must be

The time required for this may be shortened

by using small quantities of perhydrol near the end of the digestion.
In adding the perhydrol, remove the flame and allow to cool slightly;
add the perhydrol slowly down the side of the flask and then continue
heating*

Several treatments of this type may be necessary.

When the

digestion is complete, the flask is allowed to cool and is then attached
to the distillation apparatus.

The receiving flask contains 10 ml. of

saturated boric acid solution (14+) and 10 drops of a mixed indicator (1+5)
(methyl red-methylene blue).

The indicator mixture as described by the

above authors has been found to be unsatisfactory.

Ball (1+6) has found

that by adding one drop of aqueus .057° methylene blue to 10 drops of aqueus
.02^ methyl red at the time of using, the indicator changes from violet

Figure 39. Diagram of an all glass distillation
apparatus for micro-W
dahl determinates!
(Courtesy Scientific Glass Apparatus Co.)

-33 -

(acid) to green (alkaline) with, a definite intermediate grey oolor as
the end point*

Fifteen ml. of sodium hydroxide solution (500 gms per

liter) is then added through the funnel and stopcock above the digestion
flask.

Steam is then passed through the apparatus until the volume in

the receiving flask totals about 250 ml.

The contents of the receiving

flask are then titrated with standard acid to the grey end point.

The

amount of protein origionally present is calculated by multiplying the
amount of nitrogen found by 6 .25*
Another rapid and reliable method for determining total protein
in serum was published by Kingsley (1+7)*

This method is based up^n the

blue oolor developed in the biuret test for proteins (i+8).

The procedure

calls for the addition of copper sulfate and sodium hydroxide to the
serum sample in such a concentration that precipitation is avoided.

A

photoelectric colorimeter is used to estimate the amount of color developed
and standardization is accomplished by using diluted (0.85 per cent sodium
chloride) pooled blood serum and determining the actual protein concentration
by the Kjeldahl method.

To determine total protein 0.1 ml. of fresh serum

(free from cells) is pipetted, from a Folin micropipette, into exactly 1;
ml. of 10 per cent sodium hydroxide solution.
0*5 ml. of 1 per cent copper sulfate is added.
vigorously 5 to 6 times.

After mixing by rotation,
The solution is shaken

If the serum is jaundiced or lipemio, 2 ml. of

ethyl ether is added and the sample is shaken vigorously for about 20
seconds.

After standing 25 minutes the solution is read in a photelectric

colorimeter.
about 1 hour.

The readings remain stable until opalescence appears, usually
The authors experimental results show that excellent checks

can be obtained between this method and the Kjeldahl method.

In addition to

determining total protein, techniques are reported for determining albumin
and globulin in serum.

BUFFERS, BUFFER EFFECTS AND THE PREPARATION OF BUFFERS

Emil Fischer (1889-1918) showed that amino acids are the
primary decomposition products of proteins®

Adams (I4.
9) first suggested

that amphoteric compounds such as amino acetic acid exist largely in a
special state, called an inner salt
+

m z - CHa-COa

Bjerrum (50) coined the term "zwitterion" for this type of
compound*

He pointed out that all of the amino acids existed in the
«

salt-like double ion state*
+
n h 3-r -c o s
The zwitter ion theory changes the viewpoint on the reactions
of amino acids with acids and bases (51)•

"Acids react with the acidic

groups of amino acids and proteins, and bases react with the basic groups
of amino acids and proteins*'1 The reaotions between glycine and hydroohloric
acid and sodium hydroxide are represented as*
+

+

NH3-CH2-COO + H C l -------- >
Hh
«»
NH3-CH2-COO + NaOH ------- >

-

NHa-CHa-COOH - Cl
.
*f
NHa-CHa-COO + Ha + HOH

(l)
(2)

Proteins are multivalent ampholytes and exist largely as
dipolar ions in the neighborhood of their isoelectric point*
present as salts and possess a net charge at other reactions*

They are
The valence

of a protein might be expected to-influence mobility in two ways (52)*
(1)

"the greater the valence the greater the net charge and therefore

the greater the mobility

(2) **,the greater the valence the more the

activity coefficients deviate from unity at any given concentration and
the more the mobility will diminish from that which would obtain in
infinitely dilute solution."

Since the sign and net charge of a protein moleoule depend
upon th8 surrounding reaction (53) ^

is necessary that eleotrophoresis

experiments be carried out in suitable buffers in order to make
comparable results.
From reactions (l) and (2) it should be evident that there must
be some pH value at which the sum and magnitude of the charges will be
equal resulting in an electrically neutral molecule*
in this state it is said to be isoelectric*

"When a molecule is

The isoelectric point has

been defined in terms of electrical transport as*

"that hydrogen ion

concentration at which there will be a tendency for as many cations to
migrate towards the cathode as there are anions migrating toward the
anode®M

(51)
The isoeleotric point may be calculated for ampholytes from

the apparent acidic, and basic, dissociation constants (5U)»

Brohsted

(55) writes all ionization constants in terms of acid constants.

By

considering an amino acid as a dibasic acid this method can be applied
to the ionization of an amino acid.

The following equations are essentially

those of Schmidt (5U)«
+

+

-

HHaRCOO
+

v=i
+

NH3RCOO

1

=

a

=

activity

y

=

activity coefficient

+

+

R~ + H

+ H or R~+=? R

aR
aH+aR-

I<2 ’
where

+

HH3RCOO + H or R

KHgRCOOH

k

+

"

+ H

(3)
(I4)

(H+) U 1 ) (r/) (rR-) (5)
(R ) (rR )+
<H ) (R“) b * ) fc-lT)
^ n ^ )

(6)

Multiplying equations (ff) and (6 ) together
_ _
. a
Kl %

a R^ T ~

„

.
U>

C* ) » *

■J*
H
At the isoeleotrio point (R ) and (R ) are equal*

Hence, if I is

the value of (H ) at the isoelectric point
I2 - K lK2

— XaL,

(8)

If pi* is used to represent the value of pH at the isoeleotrio point,
’.ire have
pi* - 1/2 (pK

+
+ p K p - log * 2 L )

(9)
1
d
yR
where pK^ and pKg are the negative logarithms of the acidic ionization
constants of the ampholyte.

If the ampholyte has more than two ionizable

groups all of the ionization constants will he effective in determining
the location of the isoelectric point.

A complete review of the theory

of the isoelectric point has recently been presented by Hill (5&).
Several factors must be considered in choosing a buffer to
be used as the solvent in an electrophoresis experiment*
ares

(1)

These factors

H-ion concentration, (2) total salt concentration, (3) ionic

strength, (1+) buffer capacity, (5) conductance, (6) solvent power, and
(7) resolving power.

These factors will be discussed in some detail,
H-ion Concentration

It has been determined by many workers that there is a certain
hydrogen ion concentration or range of hydrogen ion concentrations at
which an ampholyte in solution has an equal number of positive and negative
charges and is electrically neutral (isoelectric point).

If an ampholyte

is put into solution at a hydrogen ion concentration above its isoelectric
point the ampholyte has an excess of negative charges and therefore behaves
as an anion in an electric field.
it behaves as a cation.

Conversely, below its isoelectric point

Equation (9) shows that the net charge on an ampholyte results
in part from the sum of its acidic and basic dissociation constants.
Therefore any effeot of the surrounding medium on the dissociation constants
will be reflected by the sign and magnitude of the net charge carried by
the molecule.
This effect can be illustrated by means of a general formula
for an amino acid.
-ffla (+)
I

NHa
f

^3
I

R

R

nnn

1

R

nnn (

COQW

(a)

(b)

(c)

t

(+)

I

An amino acid in its isoelectric condition is shown by formula
(a).

Now let us suppose that this molecule is placed in solution at a

pH value above its isoelectric point.

Free hydroxyl ions in the solution
+

oombina with the coordinated hydrogen ion of the acidic NH3

group to

form slightly dissociated water and as a result the amino acid becomes
electrically negative

(b). ^Placing t h i s

same mol e c u l e

in s o l u t i o n at &

pH value below its isoeleotric point results in free hydrogen ions
combining with the basic COO" group leaving the molecule eleotrieally
positive (c).

The greater the difference in hydrogen ion concentration

of the ampholyte in solution from its isoelectric point, the larger the
magnitude of the charge.

This point is well illustrated in Figure I4.O in

which the changes in mobilities with varying pH at constant ionic strength
and potential gradient for the components of normal bovine serum are shown
(57).
Whether a mobility determination should be carried out at a pH
above or below the isoelectric point of the sample depends on several

Tsoe/e ctr/c

€

4
3

-/

O
+/

3

4

6

4

Figure UO®
cow serum*

6

Mobility-pH relationship of
(San Clemente and Huddleson, 57)

factors:

(l)

the pH of the isoelectric point (2) the stability of the

sample in acidic and basic solutions, and (3) the solvent power of the
buffer at a particular pH value.
In addition to this charging process which results from the
effect of the solvent on the dissociation constants of the protein, are
to be considered the following:

(l) polarisation effects resulting

from a charge approaching a surface and inducing a separation of charges
on the surface of the molecule (58), this induced moment in turn may bind
the charge to the surface, (2) ion pair formation due to the fact that
<-»n o

coulombic

rn«nr >»4r><4 o

n Via

nnrte 4 +.A o 4 epri jltig +».q

forces (59), (3 ) & specific chemical reaction can determine

the charge on protein surfaces (60).

Hydrogen bond formation would be

of the latter type.
Salt concentration
The effects of salt on the mobilities of proteins are of
two

kinds (6l).

The first effect is concerned with the ion atmosphere

surrounding the-protein.

It

is non-specific and depends upon the

contribution of the salt to the ionic strength.
discussed in more detail later.

The second effect results in an actual

change in the charge of the protein.
and can be the result of (l)

This effect will be

This interaction is highly specific

chemical processes, (2) ion pairformation,

or (3 ) adsorption of the ion on the protein.
Figure I4.I shows the effect of adding salt at a constant pH to
an ideal spherical protein molecule.
decreases as salt is added.

At each pH value the mobility

Abramson points out that with a real protein

the net charge and isoelectric point would usually be affected by the
addition of salt.

In addition to these two effects the solubility of

the protein in salt solutions should be considered.

50
-05

0

OJ

02

05

05

05

Figure Ip.* The electric mobility of an ideal
spherical protein in various coneentrations of
salt* (Abramson, Moyer, and Gorin, 1)

c

-39-

Ionio Strength
X-ray studies have shown that a orystal of sodium chloride
does not contain molecules of the salt, but in contrast the crystals
are composed of sodium and chloride ions arranged
called the ion lattice.

in a definite order

If a crystal is placed in water the ions are

separated from the definite lattice form and are distributed throughout
the water*

According to this evidence, sodium chloride is 100 per cent

dissociated in any solution*

However, the apparent effects of various

concentrations as measured by colligative properties, vary widely*
T

<■* f^ AO \ V»o*• <-3
W
fc
—
y

S
A
VVUXVjV'k'W
i

V\■>« c

**A

W
.VVA

tr?4*tipV

«A

variations of strong electrolytes*

A
4
-W

4* jo aa
W
4

W
W<*4*V

AW
«

The activity theory applied to

strong electrolytes considers these electrolytes to be 100 per cent
dissociated in all cases, but that each ion possesses an activity or
capacity to function as an ion according to the concentration of ions
surrounding it*

From known behaviours of particles possessing like

and unlike charges it would appear that an ion, surrounded by other
ions of the same and- opposite charges, would possess various activities

according to variations in the concentration of the surrounding ions*
The apparent or effective concentration of a substance in solution as
shown by its properties, is termed its "activity."

If a 0*1 molal

solution of a strong electrolyte appears to have an activity in lowering
the freezing point corresponding to only 0.08 molal as based on theoretical
activity (concentration =■ 0), the value of the activity a is 0*08.
general the activity of an ion decreases as the ionic concentration
about it increases*
Lewis defines the activity coefficient as the ratio of
activity a to molal concentration mt

In

If 0.1 molal electrolyte shows an activity of 0.08 the aotivity
coefficient at this concentration is,
(11)
The aotivity coefficient equals unity -when the concentration
equals 0 (theoretical).
Equation (10) may be rewritten*

(12)

a

by the concentration.
The aotivity coefficient depends upon the total ion concentration
regardless of source.

This effect is an electrostatic one and the charges

carried by the ions are a determining factor.

Lewis has used the term

”ionic strength”, u,to include all of these ionic effects.

Hi3 equation

for determining the ionie strength of a solution is*

u » l/2

51CY

where

51 ** sum, C * concentration and

Y = valence of the ion.

(13)

In dilute solutions the activity coefficient of a given strong electrolyte
is the same in all solutions of the same ionic strength.
Gorin (63) states that variations in the ionic strength at
constant pH can effect the electric mobility of proteins in two ways,
(l) a change in ion atmosphere will occur, and (2) a change in the charge
of the protein might occur due to specific interactions with the ions of
the buffer.

However the influence of ionic strength on mobility is determined

by the type of buffer employed.

Cannan ( 6 k ) has shown that with uni-univalent

-u-

electrolytes the influence of ionic strength on the net charge pH ourve
is not very groat*

Buffers made of salts of other higher valence shew

very marked differences in mobilities with change in ionic strength at
constant pH (52).

"At an ionic strength of 0.02 and a pH of 5*65 &
q
2
mobility of 10.5 * 10“^ cm / sec/volt was observed for carboxyhemoglobin,
in phosphate buffers at 25°C.

At the same pH this mobility was reduced

to 2.7 x 10“5 at an icr-ic strength of 0.15® At the lower ionio strength
this mobility was observed at pH 6.7®

A change from 0.15 to 0.02 in

ionic strength thus produces a change in mobility in these systems
equivalent to a change of over one pH unit."
Tiselius and Svensson (26) have investigated the effect of
phosphate buffers of constant pH and varying ionic strength on the
mobility of egg albumin.

The results of their experiments are shewn

graphically in Figure ij2.
Hot only the observed mobility but the apparent isoelectric
point is a function of the ionic strength.

Smith (65) has summarized

the results (see Figure I4.3) of the effect of various acetate buffers oh
the isoelectric point of egg albumin adsorbed on collodion particles.
Longsworth (35) has found that although ascending ana descending
pattern asymmetries are reduced by increased ionic strengths, improvement
is generally more marked if the increase is due to a buffer salt, which
simultaneously raises the buffer capacity, than to a neutral salt.
Substituting one monovalent salt for another or neutral sodium chloride
at constant pH and ionic strength, does not result in appreciable changes
of mobility.

However, changing from a univalent buffer salt mixture to

one containing the divalent HP04 = ion has a marked effect on the mobilities

(66).

26
24

22

20
18
16
14

12
10

8
6
4
2
0

0.OO 0.05 0.10 0.15 0.20 0.25 030 035 OAO 0*S

V~\

Figure 42. Mobility of egg albumin at pH 7 * 1 .0 ,
T m 0.5°, and varying ionic strengths. Upper
curve* ideal mobility, calculated on the
assumption of free ionic migration. Lower
curve* calculated mobilities on basis of the
Uebye-Hflokel-Henry theory. The crosses repie sent
the observed values. (Tiselius and Svensson, 26)

3

o
3r

CL
4.30
NH
Na

0.02

0.04
onic

0.06
S

O.OS

0.10

t r e n g t h

Figure i+3* The effect of various acetate "buffers
of varying ionic strength on the isoelectric point
of egg albumin adsorbed on collodion particles,
(Smith, 65)

Buffer Capacity and Conductance
The buffer ehoaen as a solvent should have a high buffer
capacity itself so that the protein buffer capacity is relatively
reduced.

This will result in fewer boundary anomal fes

(67 ),

A buffer

with a low specific conductance is desirable in order to reduce the
disturbances due to the heating effect of the current.

The generation

of heat results from the friction of the ions passing through the sol*
ution and is related to the speed of migration of the ions.

The speed

of migration of ions (mobilites) can be caluclated by combining conduct­
ivity and transference number measurements •

"The anion transference of

an eleotrolyte expresses the fraction of current that is carried through
the solution by the anions, so that multiplying the equivalent conduct­
ivity of an electrolyte by this quantity gives the mobility of the
anion"

(68 ),

v/ater at 18°C,

Table (l) shows the mobilities of a number of ions in
Since the lithium ion has a lower mobility than either

the sodium or potassium ion its salts are preferred in making up buffers.
This allows r e l a t i v e l y m o r e current to b e 'carried by the protein ions

and is desirable since there is an upper limit to the total current which
can be

used to avoid convection artifacts due to electrical heat (69 ).

Both buffer capacity and conductance increase with the concentration of
buffer salts, and it has been pointed out that because of this incompatibility
a compromise must be made (67),

Since

ionic mobilities, buffer salts and ions

-iffor capacity does not depend upon
vhich have low mobilities should be

selected wherever possible.
Solvent Power and Resolution
The sample to be used in an electrophoresis experiment should
remain clear when adjusted with the buffer solvent.

If a precipitate

-

mu­

table

l

Mobilities in Water at 18°C
(cm./see. x 10 ■5)

Ion.
H

&

Mobility

Ion.

Mobility

316.55

ho3

61.71+

CHS

56.58

Li

35.28

Piorate

25.1+

Ha

1+5.59

C10 4

56

K

61+.liU

N(C2H s)4

28.1

Rb

67.5

HCOO

1+7.0

Cs

67.65

HCO3

1+0.1

Ag

55.77

Ca

51.5

?1

65.7

Sr

55.5

OH

176.6

Ba

55.1

.MS,.....

1+5.52

F

I+6.65

Cl

65.I1I

Zn

1+5.5

Br

67.1+

Cd

1+5.9

I

67.1+

Cu

1+5.9

C103

51+.99

Pb

60.5

BrOa

1+7.9

S04

68.0

IO»

51+.02

cao4

61.5

,•

develops and represents a quantity that can be detected in the electro­
phoretic patterns, then the pattern is not a true one.

Buffer solvents

for use in the electrophoretic analysis of human plasma and serum have
been studied at length by Longsworth (35)«

Table 2 summarizes his results.

The results of these experiments show that in resolving power
none of the buffers are superior to the diethylbarbiturate solution at pH

8,6, Column 8 gives the ratio of the maximum refractive index gradient in
the descending albumin boundary to that in the rising albumin boundary©

It

furnishes an index of the symmetry of the patterns, the value approaching
unity in the ideal case.
Figures I4I;, i+5 and Aj.6 show the differences in resolving power of
three different buffers at the same ionic strength and potential gradient
on normal bovine plasma.

The patterns shown in Figure I4J4. were obtained in

a barbital buffer of pH 8,6 and u 1,

The various components are well sepa­

rated from eaoh other and the peaks are sharp and well defined.

In addition

the descending and ascending patterns approaoh eaoh other symmetrically.
Figure 1+5 obtained from an experiment carried out in a phosphate buffer of
..

pH 7 «7 &nd u 1, shows a marked asymmetry between ascending and, descending
patterns and the separation of the alpha globulins is incomplete.

Figure l\.6

shows tire same sample separated in a carbonate buffer of pH 9,9 and u 1,

The

descending pattern shows a poor separation of the various components and the "
globulin has not separated from the albumin.
defined.

The peaks are not sharp and well

In contrast the ascending albumin peak is so steep that it could not

be completely recorded on the plate,
Preparation of buffers of known pH and ionic strength
srxderson (?0) and Washburn (71) simultaneously characterized the
equilibrium of a buffer solution in terms of the law of mass action, asF.rming
that the concentration of the negative Ion approached the total concentration
of salt and that the concentration of the undissociated acid is equal to the

H u f f e r s o l u t i o n s ust'tl as s o l r v n l s f o r the r l r c t r o p h o r t i i r . a n a l y s i s <>f h u m a n p l a s m a s a n d ser a
(1)

U)

!

(0

(5)

(4)

(6)

f7)

•

•

F

cz <

(8)

(9)

*

SERUM OR
PLASMA

>
H

7 SEPARA­
TI ON

.

(5.8

0.10

P

i

2

X

0 .65

Yes

0.2

7.7

-1.81

P

i

2

6§

0.39

Yes

0 . 0 0 1 .1/ N a l l d M ) r 0 . 0 : 5 2 .1/ N n , H P < > 4 ................... 0 . 1

7.7

2.62

s

: i

4

5

i 0.52

YesT

irt
C't
n c Ft ant

f-

X

06

2 *

A

2

S
!/3

i
i

<

'A

0

1

V

X a C a c t-O

02

.V

IlC 'u c

.1/ X n I I - P 0 4 - 0 . 0 6 1

0.00S

0 . 0 2 5 .V L i Y ,
0 0 2 5 .V L i V

N a . H P < ) 4 ......................

. 1/

0 . 0 2 5 .V I ! Y

...........................................

7.9

1 .96

p

i

4

5

i 0.75

No

7.9

3.97

s

i

4

4

0.75

No

7.9

4 .(X)

s

i .4

4

i 0.70

No

0.1

7.9

5.13

p

i

2

5

0.54

No

A' N a C l .................. 0 . 1

8.2

4.58

p

i :2

7

0.56

Yesf

0 . 0 2 5 ,Y L i C ! .......... 0 . 0 5

0 . 0 2 5 .V I l Y - 0 . 0 7 5

. V L i < ' 1 ............. 0 . 1

0 . 0 2 5 Y L i Y - 0 . 0 2 5 .Y I I Y - 0 . 0 ( 5 7 5 . V L i C l
1/ (

1.

,Y X a Y

0 025

0.0025

........................... 0 . 1

0 . 0 2 A' N a Y - 0 . 0 2
0.01

0.1

A'

H Y - 0 . 0 S A* N a C l .................

0 . 0 2 A' 11 \

V N a llC O rO

0.1

A' X a Y

0 . 0 2 .Y

0.1

,Y L i Y

0.1

A' N a O U

0 02
0.(5

Y

0.0(5

8.2

6.64

s

i :4

4§

0.54

No

11Y ........................................................ 0 . 1

8.6

3.03

p

i :2

7

0.81

Yes

..................................................... 0 . 1

8.6

2.41

p

i :2

7

0.81

Yes*;

9.0

3.48

p

i :2

G§

0.44

Yes"

1 A' N a C l

IIY

A’ g lycine

........................... 0 . 1 2 5

................................... JO. I

* W i l l i a d i l u t i o n o f 1 :2 ilu* n e w t a l i c e n t e r s e c t i o n w a s u s e d , o t h e r w i s e t h e o l d s h o r t
section.
t ( 'lie =
J Partial

cacodylatc.
p r e c i p i t a t i o n o f the p r o te i n s o c c u rr e d in

this hulTer s o lv e n t.

§ I n c o m p l e t e s e p a r a t i o n of a , front a l b u m i n ,
f

S e p a r a t i o n w a s less c o m p l e t e t h a n i n 0. 1 A

|; \ ’ =

\a V

a t p H 8.(5.

d ieth ylb arh itu rate.

Table 2. A summary of the electrophoretic
analysis of human serum and plasma in a wide
range of different buffers. (Longsworth, 35)

Figure l+i*. Electrophoretic boundaries of normal bovine plasxna diluted
to 1.5 per cent in barbiturate buffer of pH 8.6 and ionic strength 0.1®
Electrophoresis carried out for 10,000 seconds at a potential gradient
of 6,10 volts per centimeter. Boundaries photographed by the schlieren
scanning method.

a
Figure i;5* Electrophoretic boundaries of normal bovine plasma diluted
to 1.5 per cent in phosphate buffer pH 7*7 a n d ionic strength 0.1.
Electrophoresis carried out for 10,000 seconds at a potential gradient
of 6.1tj. volts per centimeter. Boundaries photographed by the schlieren
scanning method.

A

On l*J

Figure 1|6. Electrophoretic boundaries
to 1.5 P©r cent in carbonate buffer pH
lectrophoresis carried out for 10,000
.08 volts per centimeter. Boundaries
scanning method.

of normal bovine plasma diluted
9*9 and ionic strength 0.1*
seconds at a potential gradient
photographed by the schlieren

total concentration of the acid*

This is expressed as:

- K (!8 t }

( 1U

where salt and acid refer to concentrations and K is the apparent
dissociation constant*
Hasselbach used SBrensens term pH and used the above
relationship in its logocri T h m ic form*
pH K pK - !c E

(15)

Michaelis and KrOger (72) found that diluting phosphate
buffers with water caused the reaction to become more alkaline which
indicates that pK does not remain constant*

Cohn (73) made an extensive

study of this phenomena and found that pK varies with the ionic strength
and is constant only at infinite dilution*

2

1

coefficients of the salt C y O and acid ( y )

He introduced the activity
into the Henderson Hasselbach

equation, and accounted for these deviations*

In place of log

-2C-^ '

a new value pK* can be introducted for pK and this new value is such
that it corrects for the activity coefficients* ( 7 k )

That neutral salts may have a marked effect on the pH of
buffer mixtures has been demonstrated by Robinson (75)•

Tables have

been prepared by Cohn (73) (76) and extended by Green (77) from which
the molecular ratios of salt to acid can be obtained in preparing
phosphate and acetate buffers of constant ionic strength and varying
pH or constant pH and varying ionic strength*
For routine laboratory purposes the small correction due to
the change in activity coefficients is neglected and the Henderson
Hasselbach equation is used with the pK value to determine buffer

-U6-

compositions.

A list of the pK values of the more common aoids used

for buffer mixtures in electrophoresis work is as follows s
Acetic
Barbituric
Cacodylic
Glycine
Phosphoric

- 7 *90

- 6.20
- 2.J5, 9*77
- 6.77 (second dissociation)

Following are several examples of the type of buffer problems
that arise in practice.
Exaiaple
pH b * 3 h

u 0.3

•

Prepare a sodium acetate-acetic acid buffer of

(pK acetic acid ij.*^),
(salt)
pH = pK + log t t v t n
“
~ \aciuj
I+.JU

» U.64

■* log

h*3b -

= -.JO

logarithA\

-.JO

*

ratio of

^acidj

1.70
= .501

Since weak acids do not contribute appreciably to the ionic
strength (69) the salt concentration has to be 0.3 M in order to give
an ionic strength of *1
zL

x

- .501

X * .3997-molar concentration of acetic acid.

Therefore a mixture

being .1 molar with respect to sodium acetate and .1997 molar tc
acetic acid will have a pH of i+.Jlj and ionic strength 0.1.
The ionio strength of this solution may be checked using
equation (1J).
u - l/2 £(.1) x (l)2 + (.1) x(l)2
Example 2,

«= o.l

Prepare a sodium phosphate buffer of pH 5*59

-It*7—

and u Q.2.

(pK second dissociation constant phosphoric acid 6.77)
5.59 - 6.77 + log
5.59 - 6.77 - log
5.59 - 6.77 - -1.18
logarithm -1*18 = 2.82
ratio °r

f e w

‘ ,o66

By substituting the value x (acid concentration) and *066 x

(salt concentration) in equation 13 the molecular concentrations of acid
and salt to be used may be calculated.

Equation 13 must be used in

making this calculation since the acid in this oase contributes to the
ionic strength.

It will be recalled that the ions contributing to the

ionio strength are »
Na 2HP04 -^r=t Na+

+

NaHP04“ “^

MaHoP04 "--

+ H 2P0 4

Na+

+

HPO^

and
Na+

Thens

0.2 = l/2 (x) x (1) + (x) x (l)

+ (,066x) x (1)

+ («066x) x (1)

2
+ (.006x) x (2 )
0«ii » x + x + 0.c66x + 0.066x + O.26I4X
o.L. = 2 .396X
x = 1.669 = molar concentration of acid

0.066 x 1*669 * 0.0110 molar concentration of salt
Example 3.

Prepare a diethylbarbituric acid-sodium

diethylbarbiturate buffer of pH 8*6 and ionic strength 0.2*

8*6 * 7.9 + log

(salt)
taoTcTJ

_ ,
(salt)
8.6 - 7.9 - log T ^ S T d y

-|.3-

8.6 -7.9 - 0.7
logarithM of ratio “ 0*7
** 5*01
(acid)
Since the acid does not contribute to the ionic strength
**• ratio of

the salt concentration has to be 0.2 M.

x

= 5.01

x = *0599 * molar concentration of acid.
However the salt is not soluble to the extent of 0*2 M.

In

a oase of this type the buffer salt and acid concentration may be
decreased in the same ratio and the ionic strength may be increased
by using a neutral salt such as sodium chloride.
we can use 0.3

In the above example

M salt and .0199 M acid (the same ratio) and make the

solution 0.1M with respect to sodium chloride.
buffer composition.

Thisgives

thedesired

The salt may be prepared by adding an equivalent

of sodium hydroxide to an equivalent of the acid*
The calculation of ionic strengths is facilitated by the
following generalisations.

The Ionic strength of any uni-univalent

compound is equal to the concentration.
O J M A B 7Z I A +

+

u

“ I* ,1! * l l ) 2. . _+ (» ;) x

Likewise for

a uni-divalent compound,

u “

(if

2

.31/1 ASB 7-- * A+

2

(.3) x (1)

+ (.1) x (1) x (1)

2

B~

+

A+

- .1

+

B

2

+ (.1) x (2) x (2)

---g -

= .3

A di-divalent compound has an ionic strength equal to I4.times
the concentration
•1M
u

-

AB

(1) x (2)2

A++ +
+
2

b ’~

(.1) x (2)2
“ ~

- *!■

Table (3) shows the compositions of several different buffer
mixtures that have been used by Longsworth and his associates.

The

barbital buffer of pH 8*6 and ionic strength 0®1 has been highly recommended
as superior to other buffers for use with human serum*

It resolves the

various electrophoretic components appearing in the ascending and descending
patterns into more symmetrical patterns of human serum than do other buffers*
(78)*

Equvlly good results have been obtained

guinea pig n and fowl serum.

1 our laboratory on bovine,

The symmetry and asy vnetry of patterns

obtained wiuh the various buffers under similar conditions of electro­
phoresis are illustrated in Figures ill*., I4.5 and 1+6*

TABLE I
BUFFER SOLUTIONS
Gms/lOOO ml.

Gms/lOOOm.1,

.1

HCI-.729I4

NaCl-l4.676

3*05

.1

HCl-3.6>47

Glyoine-37»535

0.02 N NaAc-0.2N HAc - 0.03 N NaCl

3.62:

.1

Na^o-161408

HaC-12.010

NaCl 14.676

0.02 N NaAo-O.lN HAc - 0.03 N NaCl

3.91

.1

NaAo-1.#408

HaC-6.005

NaCl £4.676

.1

NaAc-3.20l4

HaC-12.010

Buffer Composition

pH

Ionic
Strength

0,02 N HC1 - 0,06 N NaCl

1.78

0,1 N HC1 - 0.5 N Glycine

0.1 N NaAo • 0*2 H HAc
0.1 N NaAc - 0.15 N HAc

i+Jtf

.1

NaAc-8,20l4

HaC-9.0075

0.1 N NaAc - 0.1 N HAc

14-.61}.

.1

NaAc-3.20l4

HaC-6.005

0.1 N NaAc - 0*02 N HAc

5.33

.1

NaAo-8.21+014

HaC-1.2010

.1

HCac-l6.5538

Na0H-.8002

0.02 N NaCac - 0.1 N HCac-0.08 N NaCl

Gms/lOOO ml

NaCl I4.676

0.1 N NaAo - 0.01 N HAc

5.65

a

NaAc-8.20.l4

HAC-.6005

0.02 N NaCac - 0.02 N Hcac-0.03 N NaCl

6.12

.1

HCac-5.5196

NaOH-,8002

NaCl U*676

0.02 N NaCac - O.OOI4 N HCac-0.03 N NaCl

6.79

a

HCac-3*3H7

NaO0-.8OO2

NaCl I4.676

0.02 N NaV - 0.02 N H V - 0.08 N Na Cl

7.83

.1

HV - 7.3676

NaOH-,8002

NaCl I4.676

0.1 N NaV - 0.02 N H V

8,6

HV - 22.1028

NaOH-l+.OOl

NaOH-l+.OOl

Glycine I.5OII4

NaOH-l4.001

Glycine 9*3837

NaOH-^.OOl

Glycine 7*507

0.1 N NaOH - 0.02 N Glyoine

10.28

0.1 N NaOH - 0.125 N Glycine

10.83

a
a
a

0,1 N NaOH - 0.1 N Glycine

11.31

.1

Ac
Cac
V

= Acetate
*»Cacodylate
= Diethyl Barbiturate

-51-

CONDUCTANCE MEASUREMENTS
In order to calculate the mobilities of the various components
in an electrophoretic pattern it is necessary to first calculate the
potential gradient maintained throughout the cell during the electrophoresis
experiment (see section on calculation of mobilities)*

In order to make

this calculation the specific conductance of the protein solution and buffer
must be known*

In this connection a brief summary on the theory and measure­

ment of conductance is presented,
",*t.7be passage of electricity in electrolytic conductors is
cht racterized by the movement of natter* that is to say, by particles
larger than electrons, in contrast to metallic conductance in which the
movement of electrons alone is involved*

The carriers of electricity

have both positive and negative charges.

In a solution of sodium chloride

in water, a portion of the current is maintained by the movement of sodium
ions in a positive direction, and another portion is due to chloride ions
traveling in a negative direction” (79)o
The specific resistance of an electrolyte may be defined as
the resistance in ohms of a column of solution 1 cm. long and 1 sq. cm*
in cross section.
Specific conductance L is the reciprocal of specific resistance

( 80) .

!

L=H

(17)

Since specific conductance is the reciprocal of specific
resistance and is expressed as reciprocal ohms, indirect methods may
be employed to measure the conductivity of a solution.

If the resistance

of a solution to the passage of an electric current between two electrodes
immersed in the solution is determined, the specific conductance can be

calculated.

A diagram of a Ylheatstcne bridge apparatus adaptable to

such measurements is shown in Figure I|7<»

The current passes from S

to A where part of it is allowed to pass through the unknown resistance
(solution in cell) and the known adjustable resistance Rg*

The rest

of the current passes through the slide wire resistances R j and
is

D

the current detector (earphones or some suitable device) and when it

shows that no current is passing between C and F the following relationship holds.

!i

=

R2

h.
%
(18)

R1 - R2 . Rj

Ri;
Since any current that passes from an electrode to an electro­
lyte caxises a chemical reaction, special precautions have to be taken
or the current used in making the resistance measurement will cause a
chemical change in the solution.

These changes cause variations in the

conductance and new potentials are developed at the electrodes.
term U3ed to describe these changes is polarization.

The

To overcome this

polarization effect a high frequency oscillating alternating current is
employed.

In this manner any change that occurs in the positive direction

tends to be immediately reversed in the negative direction*
In our laboratory the current source of the conductivity
apparatus is maintained from three dry cells connected in series.

The

current passes through an audio oscillator (Type 213-General Radio Co)
which gives a high frequency alternating current of a pure sinewave
type.
The conductance cell itself Figure I4.8 is of the Shedlovsky
type (l|l)» (82) (Hopf Apparatus Co).

nThe electrodes c-c are hollow

A~

Figure itf* Diagram of a VSheatstone bridge
apparatus used for conductance measurements*
Explanation in text*

Figure lj.8* Shedlovsky type conductance cell*
Description in text* (Longsworth, Shedlovsky
and Maclnnes* 82)

platinum cones and are sealed to the glass walls of the cell, and make
contact with mercury in the tubes t-t by means of platinum wire fused
into the glass*

Due to the relatively unimpeded flow when the cell is

filled by means of the tubes f-f the protein solutions do not foam*

The

glass rod r-r is used for strengthening and the projections are used for
supporting the cell®*1
A cell of this type is easily filled by means of a glass syringe
and stainless steel needle®
Since it is not practicle to prepare a conductivity cell whose
electrodes are exactly 1 sq. cm. in area anu exactly 1 cm. apart, equation
(18) cannot be used to calculate the specific conductance.

However the

cell can be standardized using a solution whose specific conductance is
knovm and the equation is then applicable*

Potassium chloride is well

suited for a standard since, (l) it is easily purified, (2 ) is non
hygroscopic, (j) sufficiently soluble, (I4.) stable both in solid form
and in solution, and (5 ) is non poisionous to the electrode.

Unfortunatel

the experimental results reported by various workers for the specific
conductances of potassium chloride standard solutions vary widely*

Jones

and Bradshaw (8J) have very carefully and painstakingly redetermined the
values of several concentrations of potassium chloride.

They point out

that since atomic weight values change from t ime to time, and are very
apt to change in the future, molar, normal etc® solutions should not be
used as standards.

They define their standards in terms of weight of

potassium chloride per kilogram of solution corrected to vacuum.
one ’’denial” solution contains
1000 grams of solution in vacuum*

Their

1352 grams of potassium chloride per
In preparing standards these authors

recommend that the use of ’’ultra-pure” conductance water be avoided and

that conductivity water prepared in contact with the laboratory air be
used.

They prepared suitable conductance water by sweeping out the

conductance water container with air feed from carbon dioxide and ammonia
and by bubbling the air through the water.

In practice it has been found

that if distilled water is redistilled in a seasoned glass vessel and
condenser with ground-glass joints water can be obtained with a specifio
conductance of about 1 x 10

-6 reciprocal ohms if a small amount of

potassium permanganate is added to the distilling flask (8ij.)«
dioxide can be removed as previously suggested*

Carbon

The constant values

for several cf their standard solutions are shown in table )jTable I*.
Specific Conductance of Standard Potassium Chloride
Solutions in Ohms
cm*
Concentration

(Demal)

ID
0.1D
0.01D

Grams KCl per
1000 g. of sol*.
in vac.
71.1352
7.1+1915
.71+5263

Specific Conductance
0®C

0.065176
.0071379
.00077361+

18°C

25*C

0.097838
0.1113i42
.0111667
.0128560
.00122052 .00ll|0877

The specific conductance of the water was determined beforehand
and subtracted from that of the standard solution.
The following equation may be used for correcting weights to
vaowim (85) (B6)

^

M
Where M
m
da
dm
dw

=
=
=
“
=

- m * mda ( ^ -

^

(xg)

wt. corrected to vacuum
apparent mass
density
ofthe air
density
ofthe mass(m)
density
ofthe weights
Having determined the resistance of a known standard solution,

at a definite temperature, the cell constant K can be calculated by means

of the equation*
K

-

LR

(20)

K

" E <- i m z a

or
5

L

<21>

Assuming the slide wire Figure i f f to be divided into 1000
division."? the specific conductance of an unknown solution can now be
determined from its observed resistance, by the following equation®

L - (

1000“a
a.R

)

K

(22)

The cell constant should be rechecked occasionally, using
a standard solution®
It has been found advisable to clean the platinum electrodes
thoroughly before restandardizing.

Concentrated sodium hydroxide or

concentrated sulfuric acid plus a few drops of perhydrol have been used
satisfactorily as cleansing agents®
Considerable difficulty is often encountered in duplicating
results with the earphone method of determining the null point®

This is

due to the fact that the null point appears to cover several scale units
on the slide wire.

An electronic nu*frpoint indicator using a J5 electron

tube (87) (88) has been modified and adapted (89) for determining the
point at which the bridge is balanced.
apparatus is shown in Figure bB *

The wiring diagram of this

The point of balance is indicated

visually by means of a fluorescent target*

This method of determining

the null point is very accurate and varying the resistance R2 as little
as 0®1 ohm causes a visible change on the target®

ALTERNATING

GJS

CURRENT

57

BRIDGE

DETECTOR
6E5

IN P U T

+ •
R ,

R*

4 5 0 VOLTS

IO O O O

/o ^

R j = I O 6^ n _

R%-50CJO_r~L
R s - /y O O O j- x

c , - o . o 5 m f Co n d e n s e r

Figure 1±9» Diagram of an electronic null point
indicator using a J5 electron tube. (Clark 89)

In order to improve this accuracy of the measurement the slide
wire may always be sot at 500 and the null point adjusted by means of the
resistance box.

With this ratio of 1 - 1 for the slidewire the specific

conductance is calculated from the equation*
L

-

|

(23)

A Kohlraush slide wire bridge (Leeds and Northrup ^Olj.669) in
conjunction with a 1000 ohm variable resistance box (Leeds and IJorthrup
# 167699) completes our conductance equipment®
A consideration of the nDonnan equilibrium” will show that the
conductance of the two solutions (buffer and protein) will always vary
slightly even after prolonged dialysis of the protein solution against
the buffer*

This is due to the fact that the protein solution contains

protein ions which cannot diffuse through the semipermeable membrane.
Two gases separated by a membrane permeable to both will diffuse

through the membrane in both directions, so that at equilibrium the mixture
will have the same composition on both sides of the membrane.

If two

solutions of different concentration are separated in the same manner and
if the membrane is permeable to both solute and solvent, equilibrium will
be attained wheh the liquid on both sides of the membrane is identical,
Donnan found that very different conditions prevail when the
membrane is impermeable to one of the ions (90)*

Consider a protein in

solution above its isoelectric point being dialyzed against a potassium
chloride solution.

The original state can be represented as
»
Na + 1 K +
P -

and the state of equilibrium as*

!I

■s Cl

-

s‘

Ba +

K +

K +

Ha +

P -

Cl -

Cl The osmotic pressure of the two solutions are equal but the
system containing the protein ions will have a lower conductance.

Small

differences in pH between the two systems can also be explained by in
this manner.
Inasmuch as the electrophoresis cell is maintained

near 0®C

during a run the conductance is measured with the conductance cell in
a Dewar flask containing ice and water*

The conductance of both the

protein solution, Kp, and the buffer solvent, Kg are determined and the
two averaged together for purposes of calculation.

-58-

TREATMENT OF ELECTRODES AND ASSEMBLING AND FILLING CELLS AND
ELECTRODE VESSELS.

Current is svipplied to the electrode vessels from the silversilver chloride electrodes at E and E', Figure
are shown in detail in Figure 50.

These electrodes

They are composed of a flat strip

of corrugated silver sheet wound in a tight spiral*

The ends of the

spiral are anchored into the core with silver screws*

A hollow silver

tube insulated by a glass tube is also threaded into this core*
When an electric current is passed through the electrodes an
oxidation takes place at the anode and a reduction at the cathode*

In

the presence of potassium chloride, chlorine ions lose an electron and
are deposited on the silver anode.

While at the cathode, chlorine ions

are given off to the surrounding solution as the chlorine atoms in the
silver chloride electrode gain an electron and become ions*

Because one

side of the assembled cell is closed off and the other remains open, the
fittings on the electrodes are different and cannot be used interchangably.
It is then necessary to occasionally reverse the current through the
electrodes to remove chloride from the anode and deposit it on the cathode.
In carrying out this operation the electrodes are immersed in N HCl in
separate leakers connected with a liquid bridge and the former cathode
is now made the anode,
passed*

A direct current of 10 milliamperes is then

Hydrogen will be liberated at the cathode after the electrode

becomes dechloridised and an adherent brownish purple deposit of silver
ohloride forms at the anode*
at the anode also*

After a time gas begins to be liberated

When this occurs the connections are reversed and

the silver chloride deposited on what was origionally the cathode until
it begins to gas.

The connections are then reversed and another cycle begun.

A

B

Figure 50. Detailed sketch showing electrode
construction. A* side view and B, top view.
(Longsworth and Maclnnes, 67)

This is repeated until the desired capacity is obtained*

It will bo

observed that with each cycle more silver chloride can be formed before
gas evolution begins at the anode than during the previous cycle.

More­

over, as the deposit of available silver on the one electrode and avail­
able silver chloride on the other are thus built up, heavier currents
can be carried by the electrodes without blistering or peeling of the
deposit.

Longsworth (91) using this method has built up eleotrodes with

a capacity of $ 0 0 milliampere hours that are capable of carrying currents
of 100 milliamperes without blistering.
In our laboratory electrodes with a capacity as high as this
have not been obtained with the above method.

However, by using the

same arrangement and starting with a low current, say 5 to 10 milliamperes,
and reversing the current when gassing occurs at the anode and gradually
increasing the current up to 50 to 60 mill amperes, capacities of the above
order have been built up®
If one is chloridizing new silver electrodes they should be
momentarily immersed in concentrated nitric acid followed by thorough
rinsing.

This cleans the electrodes and etches the silver surface

slightly and favors the formation of an adherent deposit of silver chloride
(69)*

Old eleotrodes can be cleaned, but with loss of silver by dissolving

the silver chloride in a concentrated solution of potassium cyanide
followed by the nitric acid treatment (69)®

Grease is easily eliminated

from electrode surfaces by dipping in an organic solvent such as ether
or acetone followed by distilled water.

The electrodes should be immersed

in distilled water when not in use.
There are two widely used types of electrophoresis cells.

The

four piece separation cell designed by Tiselius and the three piece cell
designed by Longsworth.

These cells are diagramatioally shown in Figure 51®

oeuH

5b ,

Mfe. Co.)

bo«I™ct1ofT^C°ft:pe0trt
°fh0”SiS
p

™

”tudle8-

(Courtesy Klett

-60 —

The top and bottom sections are the same for either cell, but with the
separation cell the single middle section is replaced with two smaller
sections*

Both types of middle sections are shown in the above figure*
Figure 52 shows a diagramatic drawing of the assembled four

section cell that was first employed for separation studies and illustra­
tes the method of assembling the cells*

Each section is connected to

the other by means of flat horizontal ground glass plates that slide
from side to side at the planes a-as, b-bT» and c-c_’«

A uniform U

shaped channel d-d* of rectangular cross section runs through the cells*
A top view of the center section is shown in the lower part of the

figure*

The other type of cell Figure 53 has the center two sections

incorporated into a single section with the elimination of the b-bf
plane*

The rack an pinion device designed by Longsworth (92) and used

to slide the cells from

side to side in orderto separate the channels

is shown in Figure 5U® rtThe rod a, of Figure

5U **&» carries a pinion,

jd, which engages the rack, r, carried by the collar, o.

The rod, a,

slides vertically through holes in the supporting frame while a tongue,
t, Figure 5J4. -b, of the collar moves in a groove,

in the bar, b, and

prevents the collar from rotating when the rod is turned.

The rod a ’,

Figure 5^4- -a, oarries a similiar rack and pinion arrangement for displace­
ment of the cell sections to the left.

The racks, r, and r ’» are provided

with pins to limit their movement,”
In assembling the cells some provision has to be made for sealing
the plates of the separate cells together and at the same time permit the
cells to slide from one side to the other.

Tiselius recommends the use of

vaseline thinned with paraffin oil (2 parts to l) for this purpose*

How­

ever, this proportion has to be varied depending upon the viscosity of

A

IV
0 1 2 cm.

3

Figure 52* Diagramatic view of I4. piece separation cell, I,
top section showing arms used to connect cell with electrode
vessels* II, and III oenter seotions of cells attached at
planes a-a*, b-b*, and c-cf* IV, bottom section of cell*
B, top view of bottom section* (Longsworth and Ma 0I1m.es, 67)

Figure 53® Electrode vessels and cell in supporting carriage,
(Courtesy Klett Mfg, Co,

Figure 5U* Section of cell support illustrating
the rack and pinion device used to displace cell,
(Longsworth, Caiman, and Maclnnes, 92)

- 61-

the particular lot of vaseline used*

If 2 parts of medicinal vaseline

are melted and one part of paraffin oil added, and the mixture allowed
to cool a suitable lubricant may result.

On the other hand the preparation

may not be viscous enough to seal the plates of the cell together or the
mixture may be so viscous at zero degrees that it does not permit the
various parts of the ©ell to be slid over each other*

In either case a

suitable viscosity may be obtained by remelting and adding either vaseline
or paraffin oil*

Stern (7) recommends the use of noellosealH for this

purpose.
In assembling the cell, the upper surface of the lower section
is covered with a thin film of the lubricant.

This is worked around the

surfaoe with a finger, being careful to keep it out of the channel.

The

lower surface of the center section is then treated in the same manner.
These two surface are then superimposed and gently slid over each other
until all air bubbles are removed and the two sections are sealed together.
The top section of the cell is then attached in the same manner.

The two

metal clips used in sliding the cell sections over are then inserted in
the position shown in Figure 5

These clips serve to evenly distribute

the pressure from the rack and pinion device.

The assembled cell is

then placed in the supporting frame, Figure 55 e^d secured in position
by spring clips*

The electrode vessels are then fastened in their

respective positions in the frame.

Figure 56 shows the cells and electrode

vessels in place in the supporting carriage.

The top section outlets of

the cell are connected to the arm of the electrode vessels by means of
neoprene rubber sleeves (Pioneer Rubber Co., Willard, 0.).

To facilitate

slipping the sleeves on, and to insure water tight insulating connections
lubricant is smeared on these sections before the sleeves are attached.

II[IIII

Figure 56* Drawing of carriage support with cells,
eleotrodes, and electrode vessels in position*
Neoprene rubber sieves, N, are used to connect the
top arms of the cell to the side arms of the electrode
vessels* (Longsworth, 35)

- 6 s -

The complete assembly is shown in place in the water bath in
Figure 57*

Milled brass strips placed across the top of the water-bath

suspend the supporting frame*
Figure 58 illustrates the method of filling the assembled cells*
As shown in Figure 58A, the lower section of the cell is first filled with
the sample.

A twenty ml. glass syringe attaohed to a long 18 gauge stain­

less steel needle (MacGregor Inst. Co.) is very convenient for this purpose*
The carriage is then placed in the water bath and left until the sample has
reached the temperature of the bath as evidenced by no further contraction
of the liquid in the cell*

Before removing from the bath, the lower section

is shoved to the left until it is isolated from the center section.
is accomplished by the rack and pinion device.

This

A slight positive pressure

is applied to the center section in the opposite direction of the pressure
being applied to the lower section.
all operations of the above type.

This precaution should be observed in
The carriage is then removed from the

bath and the left channel is washed three times with buffer solution to
remove traces of the sample.

The right channel is filled with the sample

to a point above the middle and top section joining plane.
introduced to the same extent in the left channel.
are now filled up to the arms with buffer.

Buffer is

The electrode vessels

A one liter erylenmeyer flask

equipped with a rubber bulb for forcing the buffer out helps to prevent
formation of bubbles in the electrode vessels during the filling operation.
The carriage is again placed in the water bath until the temperature of
the sample and buffer have reached that of the water-bath (0®C)a

The center

seotion of the cell is now shoved to the right until it is isolated from
the top section*

The bottom section is then shoved back to its original

position as in Figure 588.

The remainder of the sample in the upper section

Figure 57* Top view of water bath with thermostat
and assembled cells in place. The supporting carriage
is suspended from the milled brass strips, (Courtesy
Klett Mfg. Co,}

Buffer

■Buffer

a

D
El

Protein

III

A

Figure 5®* Diagram of assembled eleotrophoresis cells
illustrating the method of filling the cells. Discussion
in text. (Longsworth, 35)

of the cell is now washed out with buffer.

The electrode vessels are

then completely filled with buffer solution.

As the vessels fill up

buffer runs over into the cell and fills the top sections.

The

electrodes are then rinsed with buffer and inserted in place after
lubricating the ground glass joints.

After submerging, the electrode

in the filled electrode vessels, they are shaken to remove any air
bubbles.

The three way stopcock and insert are then lubricated and

placed in the electrode arms as shown in Figure 53*
now placed in the water bath.

The carriage is

Any air bubbles in front or behind the

cells are removed using a wire with cotton on the end.
is shoved as close to the schlieren lens as possible.

The carriage
The syringe

shown in Figure 18 is removed and filled with buffer, and a rubber
tubing attached to it*

It is then put back into position and the end

of the rubber tubing attached to the arm of the three way stopcock*
Buffer can then be forced into the electrode from the syringe as the
buffer contracts due to cooling.

A rubber tubing is attached to the

top portion of the three way stopcock to serve as ah overflow*
At this stage it is necessary to surround the electrodes
with N potassium chloride solution.

This is necessary if the buffer

is to maintain a constant composition during the run*

Any changes

resulting from electrolysis will then occur in the potassium chloride
solution.
purpose.

Saturated potassium chloride is not desirable for this
In the first place, the electrolysis affects a net transfer

of potassium chloride from one eleotrode to the other so that super­
saturation and crystallization may occur.
:I

In the second place,

saturated potassium chloride has a solvent action on silver chloride.

It will be recalled from the disoussion on electrode construction
that the silver tube is hollow.

A 25 ml. pipette with a stop-oook attached

and supported by an arm which can be screwed to the carriage is used to run
the cold potassium chloride solution slowly into the electrode vessel through
the hollow silver tube.
electrode.

Twenty ml. is sufficient to completely surround each

A rubber stopper is placed over the silver tube on the closed

side after adding the potassium chloride to completely seal this side of
the apparatus during electrophoresis.
As soon as the sample and buffer have been equilibrated to the
a

jfl

U^UiQ VI

that the oell

J U T * -a

— rn.-U

Ut, „

W
A
O

m

m

a

a

a

a

J

„ j

4 *

V
1
X
O O
VU
pU
U
'U
.IU W
A
X U
J
A
t3 VIUOOU «
s
>
J.U
Q
I 10

on this side is completely closed.

- r L-, ^

vuruo u ou

The center seotion of

the eleetrophoresis cell is then pushed over so that the three sections
are in alignment with each other (Figure 58C).

A mask with two vertical

slits, the width and length of the rectangular channels of the center
section, is then placed over the schlieren lens and is lined up with ihe
center seotion of the cell.

This mask is so constructed that either of

the slits can be covered independently of the other and is used to mask
out scattering rays of light from one side of the cell while the other
side is being photographed*
The light source is turned on and the schlieren diaphragm is
lowered to such a position that no rays of light passing through the
camera objective are intercepted.

A mask with two vertical slits with

dimensions slightly larger than the channels in the center section of
the cell is placed in the viewing end of the camera barrel.

The oell

channels are then viewed through the slits on a ground glass plate inserted
in the plate holder carriage.
be centered within these slits.

The light coming through the channels should
T^is alignment is adjusted by sliding the

center section of the cell to the right or left*

After the proper alignment

has been made the boundaries are shoved out from behind the plated holding
the sections of the cell together.

This is accomplished by turning the

three way stopcock so as to connect it to the compensation syringe.

At

the same time the motor driving the plunger is turned on and this slowly
pushes the boundaries into view.

The position of the boundaries can be

followed by viewing the ground glass plate.
out from 1.5 to 2,0 centimeters®

The boundaries are pushed

The motor is then turned off and the

same time the stop-cook is turned so that the closed side is completely
sealed.
where,

The initial boundaries are then photographed as described else­
One may now apply the proper potential to the electrodes for

electrophoresis®

- 6 6 -

boundary

COMPENSATION

One of the outstanding features of the new apparatus described
by Tiselius was the provision for the so-called Compensation movement,”
the theory of which was described by him as follows (23;b
”When both substanoes migrate in the same direction the
possibility of sufficient separation is limited by the fact that long
before the desired separation has been reached, both substances have
migrated, out of the electrophoresis tube.

This is a somewhat serious

limitation, since the absolute differences in mobilities are very often
much larger at pH regions where both components have mobilities of the
same sign (e.g. the serum proteins).

Moreover, lack of solubility often

prevents a choice of pH between the isoelectric points.

For this reason

we arranged for a slow and uniform movement of the solution in the
electrophoresis tube at an exactly known rate and in a direction opposite
to the migration, by slowly lifting a cylindrical glass tube by clock­
work out of the liquid in one electrode tube during the electrophoresis.
If the rate at which the tube is lifted is 1 cm. per hour, its cross2
section area p cm , the free surface of the liquid in each electrode

2

2

tube Q cm , and the cross-section area of the electrophoresis tube q cm '
than a movement of a given level in this tube will take place, at a rate
of
..IpQ
f _ cm. per *
hour
q[2Q-p)

W

By suitably choosing 1 and p any desired rate can be obtained} even in
the narrow tubes used in our apparatus a rate of several centimeters
per hour did not markedly blur the boundaries in the electrophoresis tube.

For fractionation purposes, this "compensation movement" is adjusted so
that the observed boundary separating two fractions obtains a suitable
"apparent mobility" and consequently, at the end of the run, the column
of the solution can be cut off exactly at the right place,.."
As an example of the use of this "compensation movement" consider
a protein solution containing two tsqpes of impurities.

One type of

impurity has a faster mobility than the protein and the other a slower
mobility.

If the shift in the level of the boundary of the protein is

offset by the change in hydrostatic pressure caused by the compensation
drive the boundary will remain stationary in the tube.

The impurity

with the faster mobility continues migrating in the same direction but
at a slower rate and the slower impurity assumes an apparent reversal
of sign and migrates in the opposite direction.
cell the pure protein component can be isolated,
attention to another useful application.

By using a separation
Tiselius called

Thus if two proteins of

different mobilities are to be separated the arithmetic mean is calculated
and the compensation rate adjusted to this mean.
in opposite directions.

The proteins then migrate

The "compensation movement" extends the separation

capacity of the apparatus to its fullest extent.
Figure 59 from Longsworth (67) illustrates the use of "compensate
movement" in electrophoretic separations.

Three components A, B, and C

moving with the (positive) relative mobilities A B C

are present.

Movement

of the lower section of the cell from position a, to position b, brings the
protein solution in contact with the buffer solution at the plane, o(. »
Passage of an electric current then causes the components to separate as
shown in Figure 59b,

However, as illustrated in Figure 59*> before any

Solution

S * c
A+B+C

B u ff e t * + P r o t e in s A ,B ,C

&

u A > u B> u c> °

b

c

Figure 59* Diagram illustrating the ideal electrophoretic
separation of a protein mixture* Explanation in text,
(Longsworth and MacInnas, 67)

»
-63-

large proportions of A and C have teen separated, the boundaries 15111
have migrated out of the cell in one case and into the bottom section
in the other.

By applying a compensation movement, as indicated in

Figure 59o» the boundaries due to mixtures of A + B and B + C (Figure
59b) can be given an apparent velocity of zero.

This leads to the

separation of pure A in one side of the cell (cathode).

The "compensation

movement” gives C an apparent negative mobility and it can be recovered
in a pure form from the upper anode section.
Several different devices have been used to bring about this
"compensation movement”.

As previously mentioned Tiselius used a clock—

work motor with variable speeds to lift cylindrical rods of glass out
of one electrode vessel.

Stenhagen (93) slowly injected buffer into

one of the electrode vessels to push the boundaries out before the
i

current is turned on.
to accomplish the same.

Smith ( 9 b ) withdrew mercury in such a manner as
Longsworth has closed one electrode vessel and

moves the boundaries out or carries out "compensation movement” by
forcing buffer into the closed side of the cell by means of the synchronous
motor and syringe (Figure 18).

By virtue of the exchangeable transmission

gears driving the syringe the rate of compensation can be varied.

j
j

i
I

HEATING EFFECTS AHD CONVECTION CURRENTS
The resolving power of an electrophoresis apparatus is limited
by the magnitude of the potential gradient which may be applied without
causing appreciable thermal convection currents due to heat developed
within the solution in the cell*
through the walls of the cell*

The heat flows to the thermostat
The temperature of the solution along

the axis of the tube is thus warmer than at the cell wall*

Normally

therefore, the solution e.t the cell wall will be heavier, and in falling
will give rise to convection currents*

Longsworth and Maolnnes (67) have

investigated these effects in the following ways
,fWe shall consider, as a typical example, a current of 0*006
ampere passing through a solution the specific conductance of which is
*0038 mhos (0»1N sodium acetate buffer at 0°C) in a cylindrical tube of
5 mm* internal and 10 mm* external diameter*

The formula which describes

the temperature of the solution, ts, in the steady state as a function
of the distance, r, from the axis is

2

r

(25)

The corresponding formula for the temperature of the glass, t

(26)
In these equation a and b are the inside and outside radii, respectively
of the tube, Ks and

are the thermal conductivities of the solution and

glass, I is the current density, £ is the electrical equivalent of heat
and tQ is the thermostat temperature *

Using equation 25, the solution

along the axis of the tube is 0*65-’ notter than at the ’wall and from

-70-

•

equation 26 the drop in the wall ia Q*67®C.

The computed temperature

distribution is given in Figure 60, in which the temperature increase,

4 t over that of the thermostat is plotted against the distance from
the axis of the tube.

If the thermostat is regulating at 25°C., this

temperature gradient in the buffer solution is accompanied by the
density variations shown in Figure 60 and, as has been stated, it is
these differences which cause mixing by convection currents, of the
solution in the tube#

If on the other hand the thermostat is regulating

at 0°C», the density differences in the solution are much less and the
variation is in the opposite direction, as shown in Figure o0»

The

contrast between the curves 60b and 60c arises from the fact that this
buffer solution has a maximum density at 2.85#C*

If, in this particular

example, the thermostat temperature were regulated at
temperature

the average

in the tube would be 2«85"C., and the density gradient would

be a minimum, as indicated by the horizontal line in Figure <£0<& Some
preliminary measurements of the temperature variations in a rectangular
channel indicate that they are of the same order of magnitude as those
indicated in the example just given for a cylindrical tube.'1

1.4

Wall

Solution

Wall

1.2
1.0
0.6

0.6
04

0.2
0

0 .9 9 6 0 0
0 .9 9 6 8 4
0.99680
0 99676

1 .00900
0.99996

1
1
-5 -4 -3 -2 -1 0 1 2 3 4
5
Distance in m m from axis of tube

Figure 60.
Diagram showing the distribution of
temperature and density in a salt solution, in a
cylindrical tube during passage of eleetrie current.
(Longsworth and Maclnnes, 67)

-71-

PHOTOGRAPHING THE BOUNDARIES

After the "boundaries have been formed but before they are shoved
into view it is necessary to record the base line on the photographic plate.
The base line is the position on the pattern of undeviated light, including
any slight refractions due to irregularaties in the cell walls, the water
bath windows or the schlieren lens.

In photographing the ba3e line a mask

with a 0,2 mm, slit is placed in front of the photographic plate and one
side of the cell is blacked out.

The schlieren diaphragm is then raised

until it nearly intercepts the normal slit image at the camera objective*
The plate carriage is then locked to its driving shaft and the spur gears
synchronizing the plate carriage and the schlieren diaphragm are engaged.
The synchronous motor is then oonnected to the drive shaft by means of the
clutch,

A loaded plate holder is placed in the plate carriage and the

driving motor is turned on at the same time the plate shutter is pulled
over exposing the plate.

The motor is left on until the normal slit image

has been intercepted and the field has changed from light to dark.
then turned off and at the same time the plate is covered.

It is

In practice the

plate is exposed at one end and the diaphragm is plaoed in such a position
that it only has to travel one to two millimeters to intercept all of the
undeviated light.
Since the base line recording is due to the manner in which the
entire field becomes dark as the schlieren diaphragm intercepts the normal
slit Image, it will also vary with the focal point of the light source*
It will be recalled that the slit image is really a composite of all the
images made by each element of the schlieren lens.

If the cell image in

the focal plane of the camera becomes dark from top to bottom a3 the

diaphragm is raised it follows that a normal slit image formed by an
element at the bottom of the schlieren lens is intercepted before the
image formed by an element at the top of the lens is intercepted*

This

situation arises when the schlieren diaphragm is not situated at the
proper fooal point of the slit image and oan usually be remedied by refoeusing the slit*

This may be accomplished by moving the light housing

backward or forward*

If the field becomes dark from bottom to top, the

diaphragm is probably behind the focal point of the slit image*

The

position of the base line can be located with reference to the micrometer
scale and since the distance traveled by the plate per unit distance
traveled by the diaphragm ia known, it is possible to locate it on the
enlarged pattern*
In order to measure the distances traveled by the various
components in an electrophoretic pattern the initial boundary must be
known*

For this reason as soon as the boundaries have been pushed out

a sufficient distance they are photographed*

A mask with a 1 / V hori­

zontal slit superimposed over one side of the oell image is placed in
the viewing end of the camera barrel*

The schlieren diaphragm is lowered

until the boundaries are narrow and sharp*

A plate is then planed in the

carriage and exposed for a few seconds close to the area previously exposed
during the scanning of the cell wall,

A separate plate is used for the

ascending and descending boundary*
At the completion of an electrophoresis experiment the boundaries
in each of the cells are photographed using the Longsworth schlieren scanning
method as described previously.

The schlieren diaphragm is lowered manually

to the point where no refractive gradients are intercepted.
contains no boundary images.

The cell field

The micrometer reading (Figure 21+) is reoorded.

The schlierin diaphram is now raised manually until the cell field becomes
to tally dark*

The micrometer reading is again recorded.

The difference

between the two micrometer readings is the distance in millimeters that
the sohlieren diaphragm has to travel to intercept all of the deflected
gradients of light rays.

Since the plate carriage travels six millimeters

for each millimeter traveled by the diaphragm, (on our apparatus), the
distance that the plate has to travel to record the complete photograph
of the boundary images can be calculated by multiplying the above
difference by six.

The plate holder is plaoed in the carriage, the

carriage and schlieren diaphragm are placed in their proper positions
and the driving mechanism thrown in gear (see Figure 2l+),
the plate shutter is removed the motor is turned on.

At the time

When the plate

has traveled the necessary distance as indicated by the vernier scale
above the plate carriage, the motor is shut off and the shutter plaoed
over the plate.

During the scanning of the boundary images, the narrow

slit mask (Figure 25) employed in scanning the base line is used.

The

same operation is repeated for the other side of the bell using a separate
photographic plate.
Figure 6l illustrates the photographic patterns obtained using
the previously described technique.
cell base lines.

The top bands on either side are the

The initial boundaries are shown below the base lines,

and the schlieren scanning photographs of the boundaries below these.
The electrophoretic patterns shown in Figures 62, 63, 64, and

65 were obtained with the same sample photographed at different time
intervals by the Longsworth schlieren scanning method.

These figures

illustrate the manner in which the course of the separation can be
followed on a ground glass plate using the Philpot-Svensson cylindrical

a
b

*+*

/9

o

Figure 6l• Electrophoretic boundaries of normal bovine plasma diluted
to 1.86 per cent in barbiturate buffer of pH 8.6 and ionic strength
0.1. Electrophoresis carried out for 12,000 seconds at a potential
rradient of 5*56 volts per centimeter. Boundaries photographed by
the schlieren scanning method. The scanning photographs of the base­
lines are shown at a, and the initial boundaries at b.

Fi~ure 62

Figure 63

a\

hd
Figure 6i+

J- d

------ —

- ------Figure 65

Figures 62, 63, 6/4., and 65. Electrophoretic boundaries of normal bovine plasma diluted to 2 per
oent in barbiturate buffer of pH 8,6 and ionic strength 0,1, Electrophoresis carried out for
10.000 seconds at a potential gradient of 6 J4I volts per centimeter. Boundaries photographed
at intervals by the schlieren scanning method. Figure 62 photographed after electrophoresis for
2,500 seconds,' Figure 63 after 5»000 seconds, Figure 6^4 after 7#5>00 seconds and Figure 65 after
10.000 seconds.

*\

lens and diagonal straight edge*

Figure 62, taken after electrophoresis

for 2,500 seconds shows a separation of the albumin and globulin components.
The slower moving globulins cannot be identified due to incomplete separation.
Figure 63 photographed after electrophoresis for 5.000 seoonds permits
identification of the various globulin components.

More complete separations

are noted in Figures 64 and 65 photographed after electrophoresis for 7.500
and 10,000 seconds respectively*
The selection of the type of photographic plate to be employed
depends upon the character of the sample being analyzed.

For colored

solutions Eastman C-T-C. or Wratten Process Panchromatic 9 x 12 cm* plates
are satisfactory.
A Wratten $22 mercury monochromat filter should be placed in
the optical system when a panchromatic type of photographic plate is used.
This type of filter isolates the yellow line and improves the resolving
power of the lens system.

In the case of colorless samples 9 * 12 cm

contrast lantern slides are adequate.

In developing the plates the

manufactures recommendations should be followed.
In our laboratory a "Solar” enlarger (Model I4.5CL) fitted with
a 6 l/2 inch F:6,p Wollensak enlarging velostigmat lens is employed to
enlarge the negative 2.5 times.
on graph paper.

The enlarged tracing is made in pencil

All measurements are made from the enlarged pattern.

It is often necessary to touch up the negative for printing.
Photographic masking opaque (Eastman Kodak Co.) is used to darken light
areas of the negatives.

Elurred or ragged edges can be improved by using

masking paste and a reducer (Fanner's reducer, Eastman Kodak Co.).

The

reducer has given excellent results when one tube of the powder is dissolved
in 100 ml. of water and the solution flooded on the wet emulsion.
Outline bromide semi-matte paper is satisfactory for making prints.

CALCULATION OF MOBILITIES AND RELATIVE CONCENTRATIONS

From electrophoretic patterns it is possible to calculate
mobilities of the various constituents if the following data are avail­
able!

(1)

The distance each element has traveled from the starting

position of the boundary between the protein solution and buffer, (2)
the specific conductance, Kp, of the protein solution, (j) the current,
i, in amperes, (J4.) the time, t, during which a given boundary has moved
the distance, h, from the starting position, and (6) the cross sectional
area of the cell*

The latter value is a constant and is determined

independently for each side of the center section of the cell in the
following manner*

The length of each channel is accurately measured using

a precision caliper square.

One connecting plane of the center section

is then greased and superimposed on a flat piece of glass*
section and glass plate are then weighed accurately*
cell is then completely filled with clean mercury.

The center

One side of the
The weight of the

canter section and m e r c u r y is then determined as before#

The difference

between these two weighings is the weight of mercury necessary to fill
one side of the cell*

If the temperature of the mercury is known the

volume of mercury can be calculated from temperature density tables.
Having obtained the volume of mercury and the heighth of the cell the
average cross seotional area of the cell can be calculated.
The derivation of the equation for calculating mobilities has
bean carefully reviewed by Longsworth and Maclnnes (95)*

The following

derivation applied to protein solutions is essentially theirs.

The

movement of an average protein particle in the body of the protein solution

-7 6 -

corresponds to the transport of u A (P) grams of protein through a reference
plane.

P being the protein concentration, A the average cross sectional

area of the cell, and u_ the mobility.

If the transport continues for t,

seconds in an electric field differing from unity the total quantity of
protein, p, transported through the reference plane is

P

-

u A (P) Ft

u

=

p

(27)

and

FAi'T?;
(28)
The field strength or potential gradient in the body of
the protein solution is

V

,(29)X

in which _i is the current in amperes and K£ the specific conductance
of the protein solution.

Equation 28 then becomes

u

=

oKp

,

,

.

.

....

(3®)

The quantity p is determined by following the movement of
the boundary in the moving boundary method.

By reference to Figure

66 the amount of protein passing the reference plane can be calculated
(9 5 ) «

If a boundary descends from a to d , It sweeps through a volume

A(a-d) « Yd.

liulti— plying by the protein concentration this becomes
A(ft-d)

(P)

»

Yd (P)

(31)

Equation 3” then becomes
ud

-

A(a-d)(P) Kb

tfrrr-

(32)

which is the formula used to calculate mobilites from measurements on

C athode

A node

C athode

Anode

B

a

a

a
Figure 66. Diagrams of electrophoresis cell
illustrating the formation and relative position
of the boundaries. (Longsworth and Maclnnes, 95)

_77»
i

the descending side.

Figure 67 shows the descending pattern obtained

from an electrophoresis experiment on normal bovine plasma enlarged 2.5
times.

The center of each peak has been located with the aid of a planimeter

and the distance between each of these peaks and the center of the initial
boundary has been indicated in om«
In locating the center of each peak it should be recalled that
the center

of the peak point of a schlieren band photograph does not

locate the position of the boundary correctly if the gradients in the
latter are not symmetrical about the ordinate passing through the maximum
value of the gradient. When the boundaries are first formed they are quite
sharp and no difficulty is encountered in determining the center.

However,

after migrating in an electric field the boundaries may become rather
diffuse.

Such a boundary is shown in Figure 68.

The ordinate d* has

been located so that itbisects the area of the gradient curve.
of

The value

should be such that when the distance migrated by the plane d ’ from

the initial boundary is multiplied by the cross sectional area of the cell,
the same number of grams of protein are obtained as have simultaneously
migrated through some reference plane in the body of the solution.
integrating the gradient curve, d may be located with precision.

By
On the

curve shown in Figure t8 , d*has been looated with the aid of a planimeter
and it is that ordinate which cancels the shaded areas on either side*
It has been found that the determination of d f with the aid of a planimeter
gives just as satisfactory results as the laborious determination of _d
in Figure 69.
From the following data the mobilities on the descending side,
Figure S j s may be calculated -with the aid of equation 32.

Component

A

oc

OC

/3

Z+0

Distance traveled cm. //. oe 9.63 8.28
Mobility x 10~s 6.91
A r e a --------------- 157
Ad
A.d
A
Percentage

-------------

4186

s

5.8! 4.16
600 5.16 3.60 2.59
26
63
65 64 34
.165 .40! .414 .407
6.93 16.80 1733 1706

ksw nu v t t l r u r n s i f

Initial boundary

pigure 67

a

d'

Figure 68. Diagram showing the variation of
the refractive index gradient, assumed proportional
to the concentration gradient, through a typical
boundary. (Longsworth, 66)

H -

Figure 69* Concentration distance curve obtained by
integrating the gradient curve (Figure 66), The
ordinate, d, has been located so that it bisects the
level in the boundary at which the protein concentration
is one half of its origional value* (Longsworth 66)

-73-

Average cross sectional area of
cell

•7555

Average amps, passed during run

0llji|B8

Tims of run

10,000 see.

Specific conductances of protein
solution

.00299 mohs”^

As an example, the mobility of the albumin component would
be,

.7555 x 11.08 x .00299
.Olitl+BB x 10,000

.
T

2.5

■ 6.91 X 10"5 cm2/sec/volt

The computation of mobilities from data on the ascending
boundary is, however, complicated by the change of protein concentration
at the 6 boundary (see discussion on boundary anomalies).

Longsworth

(95) has shown that if correctly interpreted, the data on rising and
descending boundaries will yield the same value u. of mobility.

Equation

33 has been derived to correct ascending mobilities to descending.
va

X

Ar + A W
Ad + A

(33)

in which
Ar = area in arbitrary units of ascending pattern
Ad « area in arbitrary units of descending pattern.
A

* area in arbitrary units of boundary

In the above treatment of data no correction has been made
for volume changes which occur at the electrodes on the passage of a
current.

Since the observed boundary displacements refer to a plane

fixed with respect to the apparatus they should be corrected for
solvent displacements if the absolute mobilities are to be determined
(95) (96) (97)•
of mobilities.

However, this correction does not affect the order

In determining the relative concentrations of the various
components from schlieren scanning photographs, the area of each peak
must be determined in some arbitrary unit.

In order to determine the

size of the areas from the photograph it is necessary to make a more or
less arbitrary separation of the peaks since the gradients of the different
boundaries overlap.

Two methods have been

operation.

and Kabat (98) draw an ordinate from the lowestpoint

Tiselius

suggested for carrying out this

between two poles with the exception of the^3 globulin weak.
(99) resolves the pattern into a series of symmetrical curves.
in Figure C?/ have been separated by both methods.

Svedberg
The peaks

Longsworth (35) finds

an average deviation of 15 percent in the areas as determined by these
two mathods*

Because of

Tiselius and Kabat..

the simplicity he

recommends the method of

The area of eaoh peak

can be determined by the

integral

where ng and nq are the refractive indices of the two solutions which
meet at the boundary and a, b, and k, are constants of the apparatus.
The integration indicated in equation 3i+ may be carried out with a
planimeter.

The specific refractive increments of the electrophoretically

separable plasma proteins are not known and the concentrations are
determined as differences in refractive index and not in terms of protein
nitrogen or dry weight.

’’The available evidence indicates that the

refractive index is proportional to the protein concentration as determined
by other methods, although the proportionality factor doubtless varies
with the nature of the protein” (35)®

-80-

Th© measured areas, A, are the sum of an area, P, due to
gradients in the protein and an area S_ due to salt gradients.

These

salt gradient areas are small and certain assumptions are made concerning
them for purposes of calculation®

Thus in expressing the concentration

of a component relative to another, say albumin, on the descending side
it is assumed (l) that S is proportional to P and (2) that the proportion­
ality factor is the same for all components.

The plasma globulins and

fibrinogen concentrations relative to albumin were calculated from the
relationship, Pd/Pad « Ad/Aad and the results are indicated in Figure 67,
The calculation may be illustrated from the following aataj
albumin in planimeter units 157* area of

Area of

9 globulin 26 units,

= ,165
157
The first assumption is valid as indicated by experimental evidence (35)*
The second probably induces a small error.
By making another assumption, namely that the specifio refractive
increments of the electrophoretically separable components are the saoa
as the increment used in calculating the percentage protein from the
dipping refractometer res dings (,00185) the concentrations of the vari­
ous component.s can be expressed as percentages (8‘L),

Thus the total area

of Figure 6 / , excluding the € boundary, is 375 planimeter unite.

Since

the area oi the albumin pe&K. ie 157 units
■i~L x 100 =* 14.1,86 per cent albumin.
In malcing the same calculations on the ascending side a third
assumption is necessary, namely, that all of the protein components are
held in the same proportion through the J* boundary.

Values calculated

from the ascending side are in fair agreement v.dth values calculated on
the descending side even though the third assumption is not strictly valid (35)*

- 8 1 -

■J

BOUNDARY ANOMALIES
Several of the more common boundary anomalies are shown in
Figure 6l»

The <5 boundary is more pronounced than the €r boundary*

(2)

The rising albumin boundary is much sharper than the descending one*
(3)

The concentration distributions in the boundaries are not symmetrical

about the maximum as shown by the shapes of the curves*

(I}.) The total

area of the rising and descending sides are equal but the partial area of
each of the components on the rising side are less than the corresponding
areas on the descending sides*
globulins is not complete*
anomalies in some detail*

(5)

Finally, the separation of the alpha

Longsworth and coworkars have explained these
(67), (95)* (35)» (100), and (92)*

As discussed elsewhere in this paper, a difference due to the
Donnan equilibrium of salt concentration exists between the protein so­
lution and the buffer after dialysis*

In addition, the protein solution

contains conducting constituents, the protein ions that are not present
in the buffer solution*

Referring back to Fi^ire 66 the passage of a

certain amount of current causes the boundary, a, to migrate to, d, and
in the intervening volume V a buffer solution of composition B* is formed*
”This composition has bean adjusted in general, to a value different from
B in such a way that its ’’regulating function” has the same value as that
of the protein solution it has replaced*

This regulating function defines

a property of the solution which, at any given point, retains a constant
value independent of changes of concentration caused by electrolytic
migration*

If, as a result of such migration species of ions different

from those initially present appear at a point, their concentrations will
be adjusted to values compatible with the constant determined by the initial
composition of the solution” (67) (101)*

The boundary £ Figure 66, then

- 8 2 -

forms between two solutions of the same salt, but at different concen­
trations B and B*.

Simple cases of this type have been studied and

are understood (102) (103)*

At the rising boundary r, in the same

figure there is a similiar but more complicated adjustment of the
composition of the protein solution which replaces the buffer as the
boundary rises.

The theory for complicated systems of this type has

not been developed (95)*

The § is more pronounced than the fc

boundary

since the former involves a gradient of protein concentration while the
latter does not.
The specific conductance of the.protein solution, Kolias been
found experimentally to be less than, Kp*, of the adjusted buffer
solution (95)*

Therefore variations in the potential gradient occur at

the boundary d, of Figure 66,

The protein ions in the dilute uppermost

layers of the boundary are- therefore in weaker fields then are those in
the concentrated layers and thus tend to lag behind causing the descending
boundary to become diffuse as shown in Figure Ota* However, in the case of
the rising boundary the potential gradient is greater in the solution ]>*
than in B - so that the dilute, slowly moving protein ions in the upper­
most layer of the r boundary tend to be overtaken by the faster moving
ions in the loxver part of the boundary. s As a result this boundary remains
sharp as shown in r, Figure 69<t. The variations in potential gradient at
the boundaries, as well as pH gradients which may result from Donnan
equilibrium, may cause the boundary to be unsymmetrioal about the ordinate
passing through the maximum of the refractive index gradient curve.
Diffuse boundaries may result from protein boundaries in which
the protein molecules forming the boundary have a range of mobilities
instead of a single value.

Tiselius and Horsfall (lCllj.) have found that

••• »**••*•

O H }

C

a-

L

,

• *%
••

_ m <1•

d- i*.:2*Vv-Vv
4I
I
I

i

T
0
Figure 6<?i. Diagramatic v i ew of section of electro­
phoresis cell sho^/ring initial boundaries at a-a’ and
boundaries after electrophoresis at r and de
(Longsworth, 66)

certain proteins show "reversible boundary spreading”.
both the rising and descending boundaries.

Longsworth (92) has studied

the reversible boundary spreading of ovomuccid.
his results.

This occurs at

Figure 70 illustrates

The initial boundary r, becomes diffuse as ourrent is

passed through the solution as evidenced by r^ and r^

Reversal of the

current caused the boundaries to retain their origional shape.

A

certain amount of spreading due to diffusion is shown by the fact that
the boundaries do not regain their origional sharpness.

Spreading due

to diffusion occurs independently of the electrophoresis*
The fact that the areas on the two sides of Figure 6l do not
exactly compare has been discussed under calculations and is due to the
protein gradients in the boundary.
Svensson has stated (105) that a rapid spreading of the
descending albumin boundary is not desirable as the adjacent globulin
peak is apt to be masked.

Figure 6l shows a complete masking of the

c<_'globulin and incomplete separation of theoi/ and oC^globulins.
An anomalie of a different type called the nj£
disturbance is shown in Figures 71 end 72.

globulin

The sharp spikes, s, may

be due to convection, resulting from reaction in the neighborhood of
the boundary following electrophoretic separation of the constituents
(82) (107).

Moore and Lynn (107) have made a careful study of the anomaly

and explain it in the following manner*

"The descending pattern of the

globulin carries in many cases a spike which extends indefinitely out of
the pattern.

This corresponds to the presence of turbidity and is not

entirely dependent on a refraotion gradient.

Figure 73 from Moore

describes the physios of the production of the anomalous spike.

Part A

Figure 70. Eleetrophoretic patterns illustrating
the reversible boundary spreading of ovomucoid.
(Longerworth, Carman and Mac limes, 92)

h o --------------------------------— *---■1ectrorhore t i;
ounda:
E:
1 ,)r5 nor'dent in barbiturate 'buffer of
Elec '.ropharesl 5 carried out for 10,000
6.6? molts p*r centimeter,
Boundaries

?r>ion«
descendin' 3ide.

n

O *■v-i - r*'

ti

TV.,

<7-|

of normal v o-ine
r.a
luted to
_________
pH 3-6
_ and ionic
.6
-strength 0.1#
seconds at a potential gradient of
ohoto.-raphed by tlie schlierer.
ur ■>ance, s, is indicated on the

Electrophoretic boundaries of normal fowl plasma diluted
Figure 7 ° -,
to 0 per cent in barbiturate buffer of pi! 3.6 and ionic strength 0,1*
Ilectrophoresi s cars
;ential gradient
out for 11,200 seconds at
of 6.6l molts per centimeter.
Boundaries photographed by the schlieren
scanning method.
The beta boundary disturbance, s, is indicated on the
'o 7r‘n ■'i'in side.

a n cr o,
q

SAQ ET L //V B

Figure 73* Diagram illustrating a possible explanation
of thepboundary disturbance* Description in text.
(Moore and Lynn, 107)

is a small section of the electrophoresis cell in the region of the
descending beta boundary.
similiar to that in C«

JB represents a greatly magnified^pattern

Above a. Figure A there is no beta globulin.

An inorease in the concentration occurs at a and is shown by an
elevation in the pattern.
no light.

At b there is turbid material which transmits

The spike is, therefore, of indefinite length*

c the concentration is decreasing.

In the region

This refracts a horizontal light ray

upward and causes the projection below the base line.

A relatively low

concentration is found at d which increases again in the region e leading
to the third elevation in the pattern.
uniform.

Below e the concentration of J3 is

The high concentration in the turbid layer at b accounts for

the sparcity at d, so that unless one is able to determine the increment
in the dense layer and the decrement below it, it is impossible to
calculate the total change in concentration caused by the boundary*

In

not all cases, however, is the dense layer too opaque to transmit light*
In these samples the spike does not have indefinite length but simply
shows a steep gradient at that point.

The concentration in these oases

can be deduced if the area extending below the base line is subtracted
from the area above the base line.”

Some possible explanations for the

turbidity may be, (l) an interaction between

and

globulin (Abramson

p 186), (2) a stability change by removal of other components (82),
(55).

There might be two different globulins of different stabilities,

the slower one being more unstable in the absence of the faster moving
one (Abramson p 186). Longsworth and coworkers (82), (106), have found
the cold ether extraction of certain types of pathological serums
causes a marked decrease in the ^

globulin concentration.

However,

only unappreciable quantities of the lipids of normal serum can be

extracted with cold ether.

Sorensen (108) says "...the perfect olearness

of such liquids as serum and plasma, in spite of their contents of lipoids,
is explicable only by assuming linkage between lecithin and sterols on
one hand and the proteins on the other”.

Tiselius (109) has found that

most of the lipid material of normal serum travels with thejicomponent.
Tiselius (2ij.) and Longsworth (106) have noted that the opalescence of
normal serum, due presumably to suspended fat globules, migrates with
the

globulin.
In view of this evidence it may be possible that the Q

globulin

disturbance is due to the lipid content of the globulin*
In studying mixtures of yeast nucleic acid andovalbumin
electrophoretically, Longsworth and Maelnnes (100) found evidence of the
formation of a dissociable complex.

In buffer solutions of 0.1 ionic

strength and at a pH somewhat above the isoelectric point, each of the
components migrated independently of the other.

However, at a pH near

the isoelectric point a new component appeared at the expense of the
origional components.

High concentrations of the proteins and low ionic

strength favored the complex formation.

Future work of this type may

help in explaining certain boundary anomalies.
An additional type of boundary disturbance was noted by McFarlane
(110).

He found that suspensions of elementary bodies of vaccinia showed

distortions at the moving boundaries in the form of a stream of the particles
moving ahead from the center of the boundaries in the
electric current,

direction of the

A stream of suspended material arose in the center of

one side of the center of one side of the cell and a corresponding stream
of the solvent descended in the center of the other cell.
were reversible if the current was reversed.

These effects

McFarlane (110) attributed

-36-

the effect to endosmooses*

Shedlovsky and Smadel (111) found that this

type of boundary could be obtained also with particles of a similiar
size but of an entirely different nature*

The lack of density gradients

of sufficient magnitude at the boundary were shown to cause these effeots*
Figure 'jU & t was obtained with a dilute suspension of washed virus and
shows marked streaming 74h# the same virus in a concentrated form showing
little streaming 7^+c, the dilute virus suspension in the presence of
added soluble protein showing no streaming*

Figure 7 h * Photographs of eleotrophoretio boundaries
of elementary bodies of -vaccinia. (Shedlovsky and
Smadel. Ill)
(a) Boundaries obtained with dilute suspension of
washed virus (*2?Q showing marked streaming.
(b) Boundaries obtained with a concentrated suspension
of washed virus (.5^) showing little steaming.
(o) Boundaries obtained with a dilute suspension of
washed virus in the presence of added soluble protein
showing no streaming.

-37-

STEPS TO FOLLOW IN A COMPLETE ELECTROPHORETIC ANALYSIS

(1)

Determine total protein concentration of sample.

(2)

Adjust protein concentration of sample with buffer.

(3)

Dialyze sample (Visking casing) against three changes (2[(. hours
for each) of buffer to equilbrate conductance.

(1|)

Assemble electrophoresis cell.

(5)

Place electrophoresis cell in supporting frame*

(6)

Place electrode vessels in the supporting frame.

(7)

Connect electrode vessels to cell with rubber sleeves.

(0)

Fill bottom Bection with sample and isolate by sliding to right.

(9)

Wash out left middle section with buffer and fill with same*

(10)

Fill right middle section of cell with sample*

(11)

Partially fill electrode vessels with buffer used in 3rd change
in step 3*

(12)

Place in water bath to equilibrate temperature,

(13)

Isolate middle sections of cell by sliding to the left.

(12;)

Remove supporting frame from water bath and slide bottom section
of cell to original position*

(15)

Rinse out right side of top cell with buffer,

(16)

Fill top of cells and electrode vessels with buffer*

(17)

Rinse electrodes with buffer,

(18)

Insert electrodes in electrode holders and shake to remove air
bubbles*

(19)

Place three way stopcock in place in right eleotrode vessel,
upright a m *

(20)

Place hollow plug in place in left eleotrode vessel, upright a m .

(21)

Place supporting frame in water bath*

(22)

Remove air bubbles from in front and back of center section of cell*

(23)

Push supporting frame as near schlieren lens as possible.

-88-

Intpoduoe N KCL through Bilver tube into chamber containing silver
electrodes. Admit solution slowly#
place rubber cap on top of opening of silver tube on closed vessel.
Push up plunger in syringe to remove air bubbles in tubing.
Connect overflow tube to top arm of stopcock,
Cheok lenses for moisture and dry if necessary.
Turn on light source.
Adjust mask in front of sohlieren lens.
Close stopcock on closed side after temperature equilibrium has
been reached.
Shove center section of cell to right until aligned with mask.
Check light to see that it is centered on sohlieren lens and
focused on sohlieren diaphram in front of camera lens.
Cheok base line of Tiselius cells on ground glass plate at end
of camera barrel.
Insert proper filter in holder in front of light souroe.

(Wratten

#22)
Scann the base line.
Push out boundaries with compensating syringe attached to motor.
Shut off motor when boundaries are out sufficiently.
Photograph both initial boundaries.
Connect leads to electrodes.
Adjust potentiometer to zero volts against standard cell.
Cheok cells and eleotrode vessel assembly for current leaks by
connecting one pole of current source to frame and one to silver
electrode tube.
Determine speci f io conductance of buffer (0°c)»
Determine specific conductance of sample (0*C).
Turn on current and timer.
Check amperage at regular intervals.

(1*7

Follow migration of boundaries using Philpot-Svenfison lens.

(1+8

Turn off current at completion of run.

(1+9

Calculate distance plate has to travel to photograph patterns
of boundaries.

(50

Place narrow slit mask in position.

(51

Insert plate holder containing photographic plate and connect
apparatus for scanning photograph*

(52

Make separate scanning photograph of descending boundaries and
ascending boundaries.

(55

Develop plates.

(5U

Enlarge photograph 2.5 time in enlarger and trace on graph paper.

(55

Make measurements and calculations.

(56

Retouch negative if necessary for printing.

(57

Make prints from
If the length of
more than 12,000
6 volts, one may
out changing the
following steps:

(58)

Isolate center section of cell.

(59)

Remove assembly from water-bath*

(60)

Remove glass cocks from top of arms of electrode vessels.

(61)

Drain out buffer to point where arms connect with electrode vessels.

(62)

Remove rubber sleeves connecting top Tiselius cells*

negative on semi-matte bromide paper.
time of the first experiment has not required
seconds at a potential gradient of not more than
conduct a second experiment on a new sample with­
buffer in the eleotrode vessels by observing the

(65) Remove Tiselius cell assembly.

(61+) Replace with clean cells and connect to arms of eleotrode vessels
by means of rubber sleeves.
(65)

Repeat steps from 8 to 57 omitting steps, 11, 17, 18, 2i+, and 25*

-90-

APPLICATION
The improved electrophoresis apparatus of Tiselius has found
its widest application in the characterization and study of proteins.
A brief review of some cf the applications of studies of this type will
be presented*

No attempt will be made to review all of the numerous

experiments that have been reported, but it is intended to present
enough material so that the wide application of this technique can be
appreciated.
Human Serum and Plasma:
The four components of normal human serum were first identified
by Stenhagen ( 93) &s albumin, A, and ot 3 f t

, and y * globulins.

Human

serum albumin or> isolation migrated as a single boundary and was iso­
electric at pH if.7I4 (u D 0.10).

Fibrinogen the fifth component of normal

plasma was reported to have an isoelectric point between pH 5®2 and 5®6.
Jameson (112) confirmed the presence of four components in normal human
serum.

From data of Longsworth (106), Scudder (113)» and themselves,

Moore and Lynn (107) report the average ratios of the components of normal
human plasma as follows.
a/ g

V a

*/a

1.99

0.12

0.21

>/a
0.08

0.19

Figure 75 from Moore and Lynn (107) shows the range of
variations of twelve normal human plasmR patterns (shaded areas).
These same authors find the average mobilities of human plasma proteins
on the descending side to be.
(values x 10 p cm

/

Component

A

Mobility

6.6

sec / volt)
(3
5*0

5®5

r
2.1

"7T
.1+

Figure
in the
twelve
(Moore

75* Diagram illustrating the variations
electrophoretic patterns obtained from
normal plasma samples (shaded area).
and Lynn, 107)

Other data on normal human serum and plasma have been
reported by Svensson ( 3 b ) i Luetscher (11*^), Kekwick (115), Longsworth
(82), Blix (116), Cohn (117), and Tiselius (2L|.).
Longsworth and associates (82) have made an electrophoretic
study of a number of normal and pathological human blood serums and
plasmas &

76 and 77,

The patterns of several of these samples are shown in Figures
It will be noted from Figure 7ithat in all of the patho­

logical conditions there is an increase in 0^ globulin.

It is pointed

out by the authors that a common characteristic of these conditions is
that they are accompanied by fever.

Figure 77 shows a considerable

increase in /3 globulin, or, more exactly, of material moving with the
mobility ascribed to f t

globulin.

pattern is also abnormally high.
and an increase in the oL and /9

The fibrinogen peak of this same
Figure 77© shows a low albumin peak

globulin peaks.

Figures 77© and 77f

upon analysis showed an abnormal increase in cholesterol.

Kekwick

(118) found that the albumin globulin ratio of myelomatosis serum is
dccldidly different from the normal.

One group of serums of this type

contained an additional component, which appeared to be a globulin,
Blix (119) has found that pneumonia patients show an inorease in CC
globulin while the $ and r

globulins appear normal,

Luetscher (lll|.)

has found that certain pathological serums reveal a component similiar
to the Pg globulin reported by Green (120).
Shedlovsky and Scudder (121) in a study of normal and
pathological human serum and plasma found that an increase in the of
globulin level as well as an increase in cell sedimentation rates occur
when there is present in the body any considerable inflammation or tissue
destruction, irrespective of its cause.

lb)

(d)

Peritonsillar abscess

Pneumonia

(«) Rheumatic fever

(f) Peritonitis

(/) Lymphatic leukemia

Figure
A comparison of several pathological
human serums with a normal human serum sample*
(Longsworth, Shedlovsky and Maclnnes, 82)

(<i )

Normal plasma

‘ In

Mu'tipV myel->m:’ plasma

if) Xcphn sis scrum

r

A.
(ii) Aplastic anemia
perum

(<•) Obstructive jaundice
plasma

(/) Obstructive jaundice
plasma (ether extracted)

Figure 77* A comparison of several pathological
human plasma samples with a normal human plasma
sample. (Longsworth, Shedlovsky and Maclnnes, 82)

-92-

The presence of a large amount of viscous protein which sepa­
rated on prolonged dialysis, against distilled water from the "blood
plasma from a patient with multiple myeloma has been reported (122).
The protein contained 13«9 P©** cent nitrogen and following hydrolysis
yielded lj.li per cent reducing substance (as glucose).
at 6UeC.

It coagulated

Electrophoretic examination of the plasma revealed a very large

peak with a mobility of 1.9 x lo”^ (barbiturate buffer, pH = 7*8, ionic
strength = 0.05)•

This component was shown to be neither fibrinogen

nor Bence-Jones protein.
a,
AA
J
.V
AA V
-

t

A-.i

n AMyyo
yrn rl
wW
A
A
a
W V

— A

The authors summarize their results as follows.
+-/'■*
WW

A. A
*,

O

W

-f-Vto 4“
A* A
A
AA

4 4-A A

/ 4- 4^
^ A
,aaw

1 o
a. aaA g-^W

r- A A
AA
AJ

1 o
A. A
>

t »a t »w o T
A
A AAW
A -■
-A
AA-

serum globulin which reacts with another serum component to yield the
viscous protein which settles out on freeing the blood plasma of salts
by dialysis.”
An electrophoretic study of the serum proteins from patients
with diseases of the liver revealed the following facts (123)i
"Electrophoretic analyses of the serum proteins yield lower albumin
a nd h i g h e r g l o b u l i n d e t e r m i n a t i o n s ,

and consequently lower albumin

■

globulin ratios, than are obtained bjr fractional precipitation*
"The di53tribution of the serum globulin fractions may be
definitely abnormal electrophoretically in spite of a normal albuminglobulin ratio on chemical analysis.”
Horse serum and plasmaj
In describing his nev.r technique of electrophoretic analysis,
Tiselius (23) (2Lj.) reported on the electrophoretic analysis of normal
horse serum.

Using a phosphate buffer of pH 8.03 and ionic siren r-h

0.1 he identified four separate components.

The faster component

with

a mobility of 7»^1 x 10“^ cm^/sec/volt, was identified with serum albumin.

-9 3 -

The three slower moving components were labeled oC , (3
-5

and shoived mobilities of 5*79, h - 5 1 an<i 1*90 x 10
tively.

an<i

globulins

2/
/
om /sec/volt, respec­

The relative concentrations of the various components were

reported as A 35*£ percent, oC. globulin = 12.3 percent, /3 globulin =
18.7 percent, y

globulin

+ § = 3 3 d percent.

Tiselius was able to

separate the four components of horse serum eleotrophoretically.

Each

of the components was then studied separately and the following mobilities
and isoelectric points were observed.
Mobility
A1 u u i—t^u u

Isoelectric

7.15

1..61.

cA globulin

6.16

5*o6

( i globulin

U .20

5.12

y globulin

1.51

6.oo

—

The mobilities of the separated constituents agree well with the
mobilities observed with whole serum.
In comparative electrophoretic analyses of horse, swine and
monkey'serum Tiselius (121}.) found the mobilities of the normal components1
to be approximately the same.

Kekwick (125), Moore (126) and Fell (127)

have confirmed the findings with respect to the normal components of horse
serum.

Svensson (105) identified six components in normal horse serum.

1
2
In the order of descending mobilities these were albumin,dl , ^ >
e

1

*

2
e ”, and

r

globulin*

Other Sera and plasma*
San Clemente (57) has made a thorough electrophoretic study
of serums from normal and brucellosis infected cattle.

The mobilities

of the eleotrophoretically distinct proteins fell into four well defined
groups corresponding to albumin, A. , (£

, and y

globulins.

The average

mobilities of these components were 6.2; l+.U. 3*^* and 1*^ respectively

(barbiturate-sodium chloride buffer of pH 7*9 arul ionic strength O.l).
The average relative concentration, in per cent of each component with
respect to total protein was found to be 142.1+ for albumin, 18.3 for
1
2
ot ( ol + 0C_ ) 8*3 for @ and 31*1 for

globulin.

The isoelectric points

of the various components were determined eleotrophoretically as shown
in Figure i+O®

In this study it was found that serum from a new born calf

is extremely high in

globulin and

low in P globulin.

ingestion of colostrum there is a marked increase in y

Following the
globulin.

At

the end of two weeks all of the protoins components were in the relative
concentration usually found in young heifers*
Jameson (112) found that rat serum shows three eleotrophoretically
separable components, one albumin and two globulins corresponding to the
/3 and

*

By increasing the applied potential a partial disintegration

of the globulins occurred and two new components appeared.

Later studies

indicated that the o(_ component was present but in such low concentrations
that the band disappears on dilution (328).
The electrophoretic patterns obtained in this laboratory covering
a wide variety of supposedly normal serum and plasma samples are shown in
Figures 78 through 85®

The terms normal serum and plasma are used to denote

that the samples were taken from animals that were not afflicted with any
known disease or disorder.

Corresponding serum and plasma samples were

obtained from the same animal at the same time with the exception of the
guinea pig samples which were taken from two different animals.
Figure 78 shows a typical electrophoretic pattern of normal
human serum.

The various components have been identified as indicated

on the patterns.

The

$

boundary disturbance indicated as S, has nearly

d
Figure 78. Electrophoretie boundaries of normal human serum
diluted to 2.0 per eent in barbital buffer, pH 8,6. Eleotrophoresis carried out for 10,000 seoonds at a potential gradient
of 6*90 t o Its per centimeter. Boundaries photographed by the
sohlieren scanning method.

/3

<x

Figure 78a. Electrophoretic boundaries of normal human plasma
diluted to 2.0 per cent in barbital buffer, pH 8,6. Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 6.71 Tolts per centimeter. Boundaries photographed by the
schlieren scanning method.

-9 5 -

obscured the @

globulin peak on the descending side®

the corresponding plasma pattern.

Figure JQ& shows

One additional component, fibrinogen,

is present.
The normal components of horse serum and plasma aro shown in
Figures 79 and 79®-»

The descending serum pattern shows poor separation

of the globulin components.

The ascending pattern reveals a better

separation and two distinct peaks are shovm at the &

globulin position.

The descending plasma sample (Figure 79a ) shows a poor separation of the
globulins and fibrinogen.
_ • i. •
pus
x v-j.uij.«

mx.

Two components are again evident at the <3 boundary

i. . . . . .
... .
•
*>
«
«•*
pfy uounuary
u w uurL'tuice xs
noo* evxuenu
m» e m• ni ie r cne
serum

or plasma pattern®
Typical patterns of bovine serum and plasma are illustrated
in Figures 80 and 80a®

The patterns bear a close resemblance to human

serum and plasma patterns exc-pt for the f t boundary disturbance*

This

disturbance is not evident in the serum patterns and is not marked in
the descending plasma pattern.

It is of interest to note that this

disturbance migrates well'"ahead of the f t globulin component in bovine
serum and plasma while in the case of human serum and plasma it is
superimposed on the f t globulin peak.
The patterns resulting from the electrophoretic analysis of
fowl serum and plasma are described in Figures 31 and 81a.

The descending
. I

^

serum and plasma patterns show an incomplete separation of the «. and ck,
globulins while the ascending plasma pattern shows no evidence of the
alpha globulins.

The f t boundary disturbance so prominent in the descending

patterns is also discernable in the ascending patterns.

In contrast with

the human and bovine serum and plasma patterns the f t boundary disturbance
migrates slower than the B

globulin component.

Figure 79* Electrophoretic boundaries of normal horse serum
diluted to 2*0 per cent in barbital buffer, pH 8.6. Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 6.50 volts per centimeter* Boundaries photographed by the
sohlieren scanning method*

Figure 79®1* Eleotrophoretio boundaries of normal horse plasma
diluted to 2*0 per oent in barbital buffer, pH 8*6* Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 6*6l volt8 per centimeter* Boundaries photographed by the
sohlieren scanning method*

Figure 60* Electrophoretic boundaries of normal bovine serum
diluted to 2*0 per oent in barbital buffer, pH 8*6* Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 5*13 volts per centimeter* Boundaries photographed by the
sohlieren scanning method,

d

Figure 80a* Eleotrophoretio boundaries of normal bovine plasma
diluted to 2*0 per oent in barbital buffer, pH 8*6* Eleotrophoresis carried out for 10,000 seconds at a potential gradient
of 6'JA volts per oentimeter* Boundaries photographed by the
sohlieren scanning method*

Figure 81„ Electrophoretic boundaries of normal fowl serum
diluted to 2.0 per oent in barbital buffer, pH 8*6* Eleotrophoresis carried out for 12,000 seconds at a potential gradient
of 6.33 volts per oentimeter. Boundaries photographed by the
sohlieren scanning method.

). d ----------------- ►

o -\

Figure 81a. Electrophoretic boundaries of normal fowl
plasma diluted to 2,0 per oent in barbital buffer, pH 8.6*
Electrophoresis carried out for 12,000 seconds at a potential
gradient of 6.68 volts per centimeter. Boundaries photographed
by the schlieren scanning method*

The electrophoretic examination of guinea pig serum and plasma
has resulted in the patterns shown in Figures 82 and 82a*

The descending

serum pattern shows no evidence of the d j globulin component and the*^*"
globulin is composed of two peaks.
the pattern is relatively Ioy/.
component and both
more pronounced*

The concentration of & globulin in

The ascending pattern revaals the6*-

globulins.

In addition the # globulin peak is

The guinea pig plasma patterns (plasma from a different

animal) show an incomplete separation of fibrinogen and yglobulin.

There is no evidence in either the serum or plasma patterns of the/3
boundary disturbance.
Electrophoretic patterns of dog serum and plasma are shown in
Figures 33 and 83a.

The descending serum pattern reveals four components

in the**.' and cO"globulin regions.
three alpha globulins.

Two components can be seen at the &

position on the ascending side.
from the

globulin®

The ascending side apparently shows

The

globulin

(3 globulin is not completely separated

The descending plasma pattern shows two alpha

. globulins, while .in the ascending, pattern, there ...are .four® ■In, both, ascending ,
and descending patterns the
from the fibrinogen.

The

(3 globulin has failed to completely separate

(3 boundary disturbance resembles that of the

bovine in that it migrates ahead of t h e /3 component.
The sheep serum and plasma patterns, Figures
quite different from those of human serum and plasma*
serum pattern reveals two alpha globulins.
obscured by the # boundary disturbance.

The

8k and 8l|a, are
The descending

globulin peak is nearly

The ascending pattern shows no

dk. globulins but there appears to be two Q globulins.

Both patterns

reveal an additional component migrating just ahead of t h e y globulin.
The descending plasma pattern shows one alpha globulin, while in the
ascending pattern there are two.

The & peak is obscured by the# boundary

Figure 82, Electrophoretic boundaries of normal guinea pig
serum diluted to 2,0 per oent in barbital buffer, Ph 8,6,
Electrophoresis oarried out for 10,000 seconds at a potential
gradient of 6,76 volts per centimeter. Boundaries photographed
by the sohlieren scanning method.

(• d -------- —

<7-|

Figure 82a, Sleotrophoretio boundaries of normal guinea pig
plasma diluted to 1*5 por oent in barbital buffer, pH 8,6,
Electrophoresis carried out for 10,000 seconds at a potential
gradient of 5*79 volts per centimeter. Boundaries photographed
by the sohlieren scanning method.

Figure 8J- Electrophoretic boundaries of normal dog serum
diluted to 2*0 per cent in barbital buffer, pH 8*6* Eleotrophoresis carried out for 10,000 seconds at a potential gradient
of 7*06 volts per centimeter. Boundaries photographed by the
sohlieren scanning method.

a
Figure 8Ja, Electrophoretic boundaries of normal dog plasma
diluted to 2,0 per cent in barbital buffer, pH 8,6, Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 6,73 volts per centimeter. Boundaries photographed by the
sohlieren scanning method.

Figure SU* Eleetrophoretio boundaries of normal sheep serum
diluted to 2.0 per oent in barbital buffer, pH 8.6. Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 6Jjl volts per centimeter* Boundaries photographed by the
sohlieren scanning method.

Figure 8i+a. Eleotrophoretic boundaries of normal sheep plasma
diluted to 2.0 per cent in barbital buffer, pH 8.6. Electro­
phoresis oarried out for 10*000 seconds at a potential gradient
of 6.71 volts per centimeter. Boundaries photographed by the
sohlieren scanning method.

-97*

disturbance on the descending side and a double peak is revealed in the
ascending pattern.

Apparently the fibrinogen has not separated from the

^/globulin on either side.

The additional component is seen on the

ascending side.
Figures
serum and plasma.
serum pattern.

The descending serum pattern is similiar to the human

However, the $

of the{3 component.
rather high.

85 and 85a show the electrophoretic patterns of swine

boundary disturbance migrates well ahead

The concentration of the -ft* globulin appears to be

The ascending pattern shows only one alpha globulin and

the "^globulin appears to be divided into three closely combined peaks.
The (3 and "^globulins and fibrinogen components in the plasma patterns
failed to separate into well divided areas.
Serum and Plasma Fracti onation 1
In addition to offering a method for fractionating serum and
plasma protains the Ti.sel.ius apparatus Is a valuable tool for following
fractionations by other methods.
has been presented by Cohn (ll?).

An excellent review on this subject
In this paper Cohn states that "...

each of the fractions revealed represents not a single protein but a pop­
ulation of proteins varying in size, shape, solubility, physiological
and immunological function, and in many other respects."

In fact some

believe that blood is a continuous mixture of proteins and that all the
apparent fractions are artifacts.

This subject has been carefully

discussed by Hill (129;*
Svensson (lOSj fractionated serum of several species by adding
various percentages of ammonium sulfate.

After precipitation was complete

the serum was filtered and analyzed electrophoretically.
as high as

Up to concentrations

60 per cent saturation of ammonium sulfate, only the^globulin

Figure 85* Electrophoretic boundaries of normal swine serum
diluted to 2,0 per oent in barbital buffer, pH 8.6, Electro­
phoresis carried out for 10,000 seconds at a potential gradient
of 6J+5 volts per centimeter. Boundaries photographed by the
sohlieren scanning method.

dTf P + /3

p rtf'

Figure 85a, Electrophoretic boundaries of normal swine plasma
diluted to 2,0 per cent in barbital buffer, pH 8,6, Eleotrophoresis carried out for 10,000 seconds at a potential gradient
of 6,58 volts per centimeter. Boundaries photographed by the
sohlieren scanning method.

-98-

was completely precipitated.

The ek and /3 globulins were found to have

solubilities parallel to their electrophoretic mobilities#

Svensson

found that euglobulin (water insoluble globulin fraction) is a mixture
of

<*- j

^

and

globulins#

Cohn (130) fractionated normal horse serum

by equilibration across cellophane membranes with ammonium sulfate
solutions of known pH, concentration, -volume and temperature.

The

course of the fractionation was followed electrophoretically#

He too

found that the components showing higher mobilities were more soluble
than the components with lower mobilities#

After separating the globulin

fractions and showing that they were electrophoretically homogenous, they
were further separated by water dialysis into euglobulins and pseudo­
globulins (water soluble globulins)#

This study demonstrated the presence

of many more components than are revealed by analyses of unfractionated
serum#

In a separate study MoMeekin (131) was able to separate normal

horse serum albumin into two fractions, one high in carbohydrate and one
free from carbohydrate#

Both of these preparations were found to be

homogenous' by ultracentrifuge and electrophoretic analysis.

The mobilities

of the two albumins in buffer of pH 7*7 an^ ionic strength 0«2 were
5»3 and 1+.5 x

10

cmr/see/volt respectively for the carbohydrate free

and carbohydrate containing albumins#
Cohn and collaborators (132) have developed the method of
ethanol-water fractionation of serum and plasma proteins to a high
degree#

Electrophoretically they have shown that bovine plasma is

separated as follows.

Fifteen per cent alcohol at 0®C precipitates

largely fibrinogen and fibrin.

Fraction 2 is precipitated by 20 to 25

per cent ethanol at -5°C and is largely

globulin#

J>0 to I4O per cent

ethanol precipitates a mixture of the < ana ft globulins.

Fraction ij.

remaining in solution at ^0 per oent ethanol concentration is largely
albumin.

Figures 86,

87 and 88 show results obtained in this laboratory

in the ethanol-water fractionation of normal bovine serum.

All samples

wars dialyzed against the ethanol-water concentrations indicated at from
0° to -5°C until specific gravity measurements indicated the desired
concentration had been obtained.
with 10 per cent ethanol.

Figure 86 shows the precipitate obtained

It is largely fibrinogen but traces of (5 , and

^'"globulins and albumin are present.

The ascending pattern in particular

shows a component traveling behind the^globulin that has not been identified.
The presence of this component in normal unfractionated bovine plasma has
been indicated in unpublished work from this laboratory.
the precipitate obtained from 20 per cent ethanol.
has been identified as Y globulin;
are also present.

Figure

87 shows

The chief component

globulin and a trace of albumin

The precipitate formed with 30 to i+0 per cent alcohol

is shown in Figure 88.

It is largely Y

globulin with a fairly high

concentration of (3 globulin and traces of ot7" globulin and albumin.
The mate rial soluble in 1+0 per cent alcohol was found to be largely
albumin but traces of the alpha and beta globulins were present.

As

yet we have been unable to prepare pure fractions of any of the normal
bovine plasma or serum components by the ethanol-water method.
An interesting and timely application of electrophoretic
analysis is in the field of blood plasma and serum processing.

Scudder

(133) in studying the effects of refrigeration on human blood plasma
obtained the patterns shown in Figure 89®
limit reported) the plasma appears normal.

Up to 53 days (the upper
The only change indicated

is in the spike due to the {^boundary disturbance.

It is possible that

i~ure 36. Electrophoretic boundaries of material insoluble when bovine
la sms. is dialyzed against cold 10 per cent ethanol.

Figure 87. Electrophoretic boundaries of material insoluble when bovine
plasma is dialyzed ayainst cold 20 per oent ethanol.

Fi.311.re ° 8 . Electrophoretic boundaries of material insoluble when
bovine plasma is dialyzed arairst cold 30-J.iO per cent ethanol.

Different refrigerated
blood samples

F re s h p la s m a

18 days

24 d a y s

37 d a y s

22 days

53 d a y s

Figure 89* Patterns obtained from refrigerated human
blood plasma taken at various intervals after freezing*
(Scudder, 133)

~ 1 0 0 ~

aging causes a change in the lipid protein cor\pl9X to which some have
ascribed the basis of this anomaly*

Figure 90 shews a normal human

plasma sample and the same sample after drying.

It appears that no

noticeable changes occur in the pattern when plasma is dried either
b y the lyophile process or under partial vacuum*

Studies On Other Body Fluids:
In a study of nephrotic urine Longsworth (106) found that the
anomaly is not apparent, which may indicate that one of the components
of the constituents of the complex is missing or that conditions which
it s

io n

&***?

ls . d c 3 . n g *

p h t o 1 3.c

u r in s

p s -tte m s

- i h b l t h e s e

of normal serum*
Kabat (13i-i-) found that the proteins of cerebrospinal fluid are
similiar to those of the plasma proteins.
neither

However the normal fluid shows

globulin or fibrinogen®
Other investigations have been made on urine, pleural eff, sions,

cerebrospina}. fluid, hydrocele testes, ascites fluid, cow aqueus humor
■■ and p e ricardial fluid » {135)• ' In several cases the faster moving component

was isolated and identified as albumin®
, and

r

Fractions corresponding to o(. ,

globulins were found in some or all of the samples,

Blix

(I36) studied hyalmucoid from the vitreous body of the eye, synovial
mucin and submaxillary muoin.

Hesselvek (137) studied human synovial

fluid®
Immunological studies:
In a n early work on antipneumococcus serums from the horse,

swine, and monkey, Tiselius (121+) (138) demonstrated the presence of a
new component, all of which appeared to be antibody, since it was absent

Electrophoretic patterns of same plasma

Dried hy
lyophile
process
A

Normal
control

Dried at
37°C. under
partial
vacuum

B

Figure 90* A comparison of the eleotrophoretic
patterns obtained from normal human plasma,
dried by two different methods, (Seudder, 133:

......

1 0 1 -

on electrophoresis of the same serum after absorption of the antibody
with antigen.

The new component migrated between the 3

and ^globulins.

The new component was found in swine serum and in this case its mobility
very closely approached that of they'globulin.
anti-egg albumin serum was investigated.
identical with the

Y component

In the same study, rabbit

The antibody was found to be

of the serum.

Another investigation of

anti-pneumococcal horse serums was made by Vander Scheer and coworkers
(126) (139)*

They found that the antibodys always shewed the same

mobility as ^globulin and no new component was demonstrated.

These

authors explain the uesorepenoy between their results and those of
Tiselius (1214.) (138) as follows,

"It has been thought that the

different mobilities fo^md for pneumococcal antibodies arise from the
use of different pneumoccal antigens.

It has also been proposed that

the electrical mobilities of antibody molecules as well as their mole­
cular weights change with the length of time over which hyperimmunization
has been proceeding,"

In one case the new component as described by

Tiselius was demonstrated,' Fell (127) has found both types of antibodies
in antipneumococcus horse serum.
In electrophoretic studies w i t h horse tetanus antitoxic serums
the presence of a new component not present in normal horse serum was
demonstrated (ll+O) (1I4.I),

The new component was labeled "T" and was

found to have an average mobility of 1.98 x 10"5 cm2/ s ec /volt in a
phosphate-sodium chloride buffer of pH 7 *6 ,

The same component was

demonstrated in certain other antiserums (lli2 ) (see plate I Figures

3Jl-,5,6 ,7,8,9, 10,11, 12,1I4, and 15.)
authors state,

In a summary of these results the

"in some of these sera antibody is ex; ressed by an

increase in amount of the normally present ’^ g l o b u l i n

»

In others its

ifiM A i. uORSE

ANn.-'NEUMOCOCtX

I f . TANUS A N TI

D ip h t h e r ia a n t e

SEKUM 1 YP[
:and n

TOXIC SERUM

TOXICSERUM

scrum

DOTUUNUS

Antitoxic serum

Scarlet Fever
A ntitoxic serum

S tapwylocgccu
A ntitoxic serum

FI6URE 10

FIGURE II

A nti-pertussis
SERUM

a0 t r

FlGURF r

FI6URE 9

FIGURE -4

FIGURE 3

FIGURE 2

FT
TTUuXlL U! RUM
'-L YVElC'lll

I" A n ittoxic

5 fftU M |
T a\NTlTOXlC
*
i

SERUM

I

A o A lN ii I

i Cl

O F D E M A T lE N S |

1 anu

C l O O R D F ili

i

A 6AIN 3T

1

Cl HlSTCX YflCUM

I

A n t it o x ic SERu m
AGAINSI

|A

n TIMENIN60-

! -COCl .C SCRUM

C l .s f p t i u u E

UERUM AGAINST
B

s m i GAE

AND

5

e RUM AGAINST

F k y k h u s io p a th ia e

FIGURE 12

Pa s t e u r e l u a
antjserum

B F L t'X N F K l

i

a

FIGURE 5

FIGURE 6

FIGURE 7

Plate I

FIGURE S

FISURE IS

FIGURE 14

(Van der Scheer, Wyckoff and Clarke, li^Z)

FIGURE 15

FIGURE IS

- 1 0 2 -

appearance is accompanied by the development of a new ”T” component.”
It was stated further, that in spite of this obvious association of
antitoxic activity with ”T”, the areas under the "T”

peaks cannot be

taken as proportional to the measured antitoxic activity.
Kekwick (125"? followed the changes in horse serum during the
course of producing diptheria antitoxin.

The normal constituents were

found to remain but their relative concentrations changed.
serums showed a (3* and f3

globulin component.

The

All antitoxic

component may be

the same as the ”T” component but the mobility was so close to that of
(^ globulin that it did not GGToplottily separate.

ine buicioGxic serums

were fractionated and the{3 and Jglobulins were the only fractions that
showed antitoxic activity.

Rothen (li+3) found that a large increase in 3^

globulin occured during immunization of horses against diphtheria toxin®
It is known that certain antitoxins can be partially digested
with pepsin without losing their abilities to neutralize toxin (II4I;)*
Diphtheria antitoxic horse plasma when digested with pepsin under certain
conditions shows a

69

per cent loss of coagulabla protein which is"

accompanied by only a 20 per cent loss of antitoxic activity.

By this

procedure the antigenicity of the anti scrum is greatly reduced.

Figure

91 fro in Var. der Scheer (II4.3 ) shows the electrophoretic changes accompaning
pepsin digestion of defibrinated diphtheria-antitoxic horse plasma.
digestion was carried out with acid pepsin (pH I4..O) at 37°0e

The

Figure

91-1 shows the electrophoretic pattern of the antiserum before digestion.
In Figure $ 1 - 2 the ”1” component is indicated only after l/2 hours
digestion*

Figure 91-3 shows the results after i+8 hours digestion.

That the heat bv itself does not cause this effect is shown in Figure 91~U®

Diphtheria
Antitoxin

#4635

1

#4635
a
a fte p Vzhoup
d iq e s tio n

#4635
3
a fte p 4 8 hour
d iq estio n

#4635 | 4
at pH4 & B37*c
forfehoup* (no
p e p s in )H
/H

Figure 91• Electrophoretio patterns shewing the
effect of pepsin digestion of diphtheria-antitoxic
horse serum* Discussion in text* (Van der Soheer,
Wyokoff and Clarke, l?/5)

-1 0 3 -

It has been observed previously that many diseases cause a
marked change in the blood protein albumin-globulin ratios*

Frequently

the globulin concentrations, gamma in particular, show a marked increase*
A comparison of Figure 91&* & normal bovine plasma pattern with Figure

91b, obtained from a brucellosis infected cow shoves a marked increase in
the gamma globulin concentration*

By combining agglutinin absorption

techniques with electrophoretic analyses, San Clemente (57) was able
to show that Brucella agglutinins migrate with the gamma globulin component*
It was not possible to remove all of the gamma globulin by the absorption
method indicating that the gamma globulin fraction was a mixture.
Figures

91c and 91d show the electrophoretic patterns obtained

from serum and plasma of a brucellosis infected horse.

The globulin

concentrations are abnormally high as can be seen by comparing these
patterns with normal horse serum and plasma (see F-gures 79^- and 79b),
It is also of interest to note that the alpha two globulin concentration
approaches that of the albumin*
In a study of the proteins of tuberculin (li+6) and the blood
serum response in tuberculosis (li|7) Siebert has made the following
observations.

Two distinct proteins are present in tuberculin.

phosphate buffers of pE ?« 6-7»6 (ionic strength
of these two proteins are of the order of

In

0 .1; the mobilities

3 to U and 6 to 7 y- 10-5

cm.^/sec/volt the mobility varying slightly with various preparations*
It is shown that there is always a progressive deorease in the amount
and percentage of serum albumin with prog-ression of tuberculosis.
early tuberculosis the ^globulins increase in percentage.

In

At the same

time a new component, X, with a mobility greater than albumin occurs.

Figure 91a. Electrophoretic boundaries of normal bovine plasma
diluted to 2.0 per oent in barbital buffer, pH 8.6. Electro­
phoresis oarried out for 10,000 seoonda at a potential gradient
of 6.32 volts per centimeter. Boundaries photographed by the
sohlieren scanning method.

y d -------------------- »»

a -\

Figure 91b. Eleotrophoretio boundaries of bruoellosis infeoted
bovine plasma diluted to 2.0 per oent in barbital buffer, pH 8,6.
Electrophoresis oarried out for 10,000 seconds at a potential
gradient of 6 Jj 3 volts per centimeter. Boundaries photographed
by the sohlieren scanning method.

Figure 91c# Electrophoretic boundaries of bruoellosis infected
horse serum diluted to 2.0 per cent in barbital buffer, pH 8,6*
Electrophoresis carried out for 10,000 seoonds at a potential
gradient of 6,80 volts per centimeter# Boundaries photographed
by the schlieren scanning method#

j-

•

—

dH

Figure 91d# Electrophoretic boundaries of brucellosis infected
horse plasma diluted to 2#0 per cent in barbital buffer, pH 8,6#
Eleotrophoresis carried out for 10,000 seoonds at a potential
gradient of 6#9U volts per centimeter# Boundaries photographed
by the sohlieren scanning method#

It is believed that these changes represent sensitization to the tuberculin
proteino

In the terminal stages of tuberculosis the ft globulin increases.

There was some evidence obtained which indicated that a rise in ^f'globulin
accompanied r -sistance to the disease.

In another study Siebert (li|.8)

separated the protein, nucleic acid, and polysaccharide fractions of
tuberculin electrophoretically#
types electrophoretically®

The polysaccharide fractions were of two

One fraction shewed no migration in the

electric field and the other a slew migration.
Electrophoretic studies on guinea pigs compliment fractions
have shown the following (iipy)*

The mid-piece has been separated as

a euglobulin, with an electrophoretic mobility of
buffer of ionic strength 0.2 at pH 7*7*

2.9 x 10-5 in phosphate

The end piecs and fourth component

were found together in a euglobulin fraction of serum which contained

10®? per cent carbohydrate and had an eleetrophoretic mobility of Ip.2
x 10~5 in phosphate buffer of ionic strength 0.2 et pH 7®7»
After certain of the viruses had been isolated in supposedly
pure form, electrophoretic studies were carried out in order to obtain
physical property measurements and to determine the homogeniety of the
preparations®
In a study on tobacco mosaic virus (lyO) solutions of the
pure virus protein were found to be isoelectric at pH

Electro-

phonetically the preparations were uniform but they were not monodisperse
in the ultracentrifuge®

It is of interest to note that Pfankuck (151)

has reported that it is possible to split the particle of tobacco mosaic
virus protein into
properties .

?0 to 100 pieces having identical electrophoretic

-105-

McFarlane (152) noted that, the tomato bu 3hy stunt protein vms
monodisperse electrophoretically over a pH range of 2.L to

8.6 and

isoelectric at pH JLp.11 *
An attempt to electrophoretically isolate and crystallize the
vaccaria virus was made by Douglas (155).
with rabbit tester proteins.

The virus was obtained mixed

The testes proteins were isoelectric at pH

6.8 and the virus proteins carried a negative charge from pH 5*5 to 8.1;,.
By carrying cut electrophoresis

tudi.es below pH

6.8 and above pH 5*5 the

testes proteins migrated as cations and the virus proteins as anions,
p, A r>0

t t

v r# a r'j»

r 'o r - ,

tH u a H

Itl

3TSMH1S2*

St

pU 2TC

C 2 7 y 2 * fc £ .iX i2 1 C

p ro d u c t

was net obtained.
An electrophoretic separation of the fowl leucosis virus vms
claimed by Lee (15U)* using a U tube apparatus.

The progress of the

separation vms followed by injecting material from the anode, cathode
and center sections of the cell into susceptible animals.

The isoelectric

point of the virus protein was found to lie somewhere between pH 6.01
■and 7 * 0 1 * ■

-

"-

■■

'■

■ '- "" ■" - ......■

Using electrophoretic techniques Smadel and Shedlovsky (155)
have shown the virus of vaccinia tc have a complex structure.

At

least five antibodies developed in animals following injection or
hyperimmunization with active elementary bodies, vie., a neutralizing
antibody, an agglutinin designated X, antibody against a nucleoprotein
constituent (HP) of the virus, and finally, antibodies against a heatlabile (L) and heat stable (s) soluble antigen.

L and S antigens,

although irnrunologically distinct, are not separate substances; they
are components of a single substance, LS,

-

-lo6-

Comphrensive studies have been carried out by Abramson (156)
on pollen extracts causing hay fever.
been analyzed electrophoretically.

A wide variety of pollens have

The patterns are all similar and

contain a pronounced peak which may be a protein, and several pigment
peaks.

Because of the low molecular weights of the proteins (about

5,000) it is believed that they m y be intermediate between polypeptides
and proteins*

Some of the pigments have been separated electrophoretically

and like the protein component, they are biologically active.
Other Studies of Biological Interest:

Ringer (157) m s

possibly the first tc attempt to purify

enzymes by electrophoresis*

He found that purified pepsin had no

apparent isoelectric point.

Tiselius (158) found on electrophoretic analysis

that unless special precautions are observed, m»st supposedly pure pepsin
preparations contain one main component and several contaminants.

Upon

further purification (electrophoretic) the contaminants were removed and
pepsin acted like an acid remaining negative at all values of pH investigated.

Herriot " ( 1 5 9 ) ' h a s - c o n f i r m e d t h e s e ' o b s e r v a t i o n s ' and reports the isoelectric”
point of pepsin to b e below pK 1.5*
Theorell

(l6o) succeeded in isolating and purifying the yellow

respiratory enzy m e of W a r b u r g and C h r i stian b y electrophoretic technique*
He was able t o

i d e n t i f y the purified enzyme as a protein.

Cytochrome- c

was isolated e l e c t r o p h o r e t i c a l l y b y this same w o r k e r (l6l).
Ot h e r e l e c t r o p h o r e t i c studies on enzymes have b e e n reported.

Choline esterase (162), ribonuclease (163), zymohoxase (l6iq), yeast
carboxylase (165)* and chynotrypsin (166).

In the case of zymohexase

it is of interest to note that the isoelectric point of the purified m t i v e

enzyme and of heat coagulated particles was pH 6.? in all cases.
Figure 93 shows the electrophoretic pattern obtained in this laboratorv
on a beef liver catalase preparation in barbital buffer, pH 8,6, u Q.X.
The preparation appears to be quite homogenous and hns a mobility of
3.12 om^/sec/volt calculated from the descending pattern.
Studies on the pitiutary-lactogenic hormone have been reported
by Shipley (167) and Li (l68).
to be homogenous.

Pure preparations of the hormone appear

The crude gland extracts are largely composed of

physiologically inert proteins.

The isoelectric point of the pure

hornuue appears to be somewhere in the neighborhooa of pH 3*6 to 3*7*
Shedlovsky and coworkers (169) (1J0) have succeeded in the
electrophoretic isolation of the interstitial cell-stimulating (luteinizing)
hormone from swine pituitary glands.

The initial extract after chemical

purification showed three electrophoretically distinct components.

The

main component was electrophoretically separated and found to contain all
of the biological activity®

The main component after separation was

homogenous in both electrophoretic and ultracentrifuge studies with an
isoelectric point of pH

7 *^-5«

Other electrophoretic studies on hormones have been reported
as follows:
(173)* ( y j k ) ,

gonadotrophic hormones (171) (172), posterior pituitary

(175) and thryoglobulin (176).

A comparison of the mobilities of various hemoglobins from
different species has been made by Landsteiner (17 '')*

7he mobilities

of soma of the hemoglobins from unrelated species were very similiar.
Muscle juice obtained by pressing muscle has been shown to
contain the following proteins: myohemoglobin, myogens A ana B, globulin
X, and myoalbumin.

Herbert (178) claims that at least 50 enzymes are

present in varying amour's in muscle juice.

LL

y

------------a - \

Figure 92.
Electrophoretic boundaries of
Bruce 11a protein nucleate. Electrophoresis
carried cut for 3»*i-00 seconds in phosphate
buffer of pH J
and ionic strength 0.1 at
a potential gradient of 5*33 volts per
centimeter.

H <f

Figure 93. Electro­
phoretic patterns of
beef liver catalase.
Description in text.

Jj
<M

Figure 9U. Electrophoretic boundaries of
Brucella protein nucleate. Electrophoresis
carried out for 2,900 seconds in phosphate
buffer of pH J .6 and ionic strength 0.02
at a potential gradient of 5.30 volts per
centimeter.

1

1
\d ~

1
~0*

Figure 95, Electro­
phoretic patterns of
Brucella gluco-lipid.
Description in text.

- 1 0 8 -

A chlorophyll-protein complex from leaves has been reported
by various workers.

One electrophoretic study has shown that suspensions

of these particles are isoelectric in acbtate buffers (u/50 at a pH of
lu7). (179)®

These particles are believed to have uniform surfaces.

Fibrous proteins have been investigated as followsi

1
Wool (180),

(181), silk (60), hair, skin and fingernails (182), (183), and steer collagen
(I8 I4).
Martin (185) and Kemp (186) have studied extracts of wheat proteins.
The isoelectric point of gliaden was found by Kerap to be near pH
was isoelectric at about pH 1+.9*

6*6 Gluten in

The isoelectric points of some other

prolam ines have been reported as follows i zein fractions, pH 6.29,
7*17 8r'^ 6*95, (187), and secalln pH

6.67 (in 63 per cent alcohol by

weight; (188).
Cow casein (189) has been shown to consist of three ©lectrophoretieally
separable components c a l l e d
mobilities*

0

and

asein in order of their decreasing

A boundary corresponding to lactaP'umin was also observed.

The electrophoretic behaviour of certain nueieoproteins has
been investigated (190).

The nueleohistone of the calf thymus, two

additional nueieoproteins of the calf thymus, and the nucleoprotein of
the hog thyroid were characterized as uniform substances from their electro­
phoretic behavior.

Figures 92 and 9-U show patterns obtained in this labora­

tory on the protein nucleate from Brucella cells (191)®
is the bacterial protein.

The major component

The faster moving component is nucleic acid.

The effect of ionic strength on electrophoretic separations is readily
seen by comparing Figures 92 and
Figure

9If®

95 shows the electrophoretic pattern obtained with the

endo-antigen from Brucella cells (192).

The electrophoretic homogeniety

-109-

of this aubstanoe is apparent.

This is interesting since the endo-

antigen has been found to contain a nitrogenous substance, glucose on
hydrolysis and lipid.
It has been s h o w that when blood serum is heated to about

65°C it becomes opalescent and prolonged heating causes the entire sample
to gel.

Electrophoretic examination of serum heated at 65°C. showed the

presence of a new component (193) called "C".

On further study on this

component it became evident that it v/as a colloidal aggregation product
resulting from denaturation of components of the serum.

Experiments in

this laboratory have shown conclusively that glucose ha'S’' an inhibitive
action on nCu component formation when samples of serum or plasma are
heated at 65°C in the presence of glucose (19U)«
pattern of a sample of normal bovine plasma.

Figure

Figure 97 shows the pattern

that results from heating the same sample at 65°C»
is very much in evidence.

96 shows the

The ,fC" component

As discussed elsewhere (1970 the peaks labeled

X and Y are believed to be lipid disturbances.

Figure

of the sample saturated with glucose and heated at

98 shows the pattern

65°C for one hour.

T^e ”C *’ component is not formed under this condition of heating.
The use of electrophoretic techniques in following chemical
separations has been clearly demonstrated by Longsworth (92) in an
electrophoretic study of the proteins of egg white*
the pattern of untreated egg white at pH 3*92.

Figure 99a shows

G , G , and G , are

1

d

p

globulin fractions, C the conalbumin fraction. A the ovalbumin and 0,
the ovomucoid fraction.

Part (b) shows the soluble fraction resulting

.from half saturation with ammonium sulfate.

This fraction contains all

of the components shown in (a) with the possible exception of globulin Gg.

d i

(a '

Figure 98
Figures 96, 97# and 98* Electrophoretic patterns of bovine plasma
in barbiturate buffer of pH 8,6 and ionic strength 0,1, Electrophoresis
of all samples carried out for 10,000 seconds at a potential gradient of
6,00 volts. Boundaries photographed by sohlieren scanning method. De­
scription in text.

Kgg white
(f) Ovalbumin -1st crystals

A

(b)

A

Albumin fraction
(g ) O v a l b u m in - 2 n d c r y s t a l s

{ r )

Fi U .rn t e f r o m 1 s t c r y s t a l s

Cel)

G lo b u lin f r a c t io n

----------------J

(.•;

(V .o b u lm

|----------------*

fraction

4,

*

(h) O v a lb u m in

(

''

*

3 rd c r y s t a ls

Figure 99* Electrophoretic patterns of egg white
proteins illustrating how chemical fractionations
can be followed electrophoretically* (Longsworth,
Cannon and Maolnnes, 92;

-n o

The insoluble fraction is shown in (d) and it too is

amixture of ail

of the original components.

effect of crystallization

ox the albumin fraction*
component is reduced.

Figure 99 (f) shows the

The G, globulin has dissappeared and the ovomucoid

Figures 99 (g) and (h) illustrate the effect of

further recrystallization of the ovalbumin fraction.
The chemical fractionation of the water soluble, crushed cell
fraction obtained from Brucella cells as described by Huddleson (195) ha 3
been followed eleetrophoretieally.
crushed cell fraction.
5,000

Figure 100 shows

atypical pattern of a

Electrophoresis of this sample was carried out for

seconds at a potential gradient of i+.lS volts per centimeter in a

barbiturate-eodium chloride buffer of pH

and ionic strength of 0.1.

7.8

The mobilities computed from the descendin' boundaries were as follows:
(a) U.ii.0 x 10~5 , (b) 9.L.3 x ID’5 , (c) 11.81; x 10"5, (d) II4 .7 6 x 10~ 5 cm2/
sec/volt.«

Figure lQOb shows the insoluble fraction obtained by dialyzing

the crushed cell fraction against J+5 per cent ammonium sulfate.

Electro­

phoresis of this fraction and the subsequent fractions was carried out at
a potential gradient of
of pH

8.6

6

volts for

and ionic strength C.l.

the d component has been removed.

3*000

seconds in a barbiturate buffer

The electrophoretic pattern shows that
In addition the concentration of the

a component has been reduced and at the same time the relative concentrations
of the b and c components have been increased.

Figure 100 c was obtained

by electrophoresis of the material remaining hi solution after the crushed
cell fraction was made
(l0°c).

30

per cent (by weight

Three components are now evident.

with respect to p-dioxane

Tho steepest peak does not

migrate under the conditions of this experiment and is believed to be an
electrophoretically inert polysaccharide.
believed to be the b and c fractions.

The other two components are

Figure 100 d was obtained upon

Figure 100. Electrophoretic patterns o.f water soluble Brucella crushed cell fractions
in barbiturate buffer. Description, in text.

-1 1 1 -

©loctrophoretic analysis of the material soluble in I4.0 percent d'ioxane.
The inert material is still evident and the other component has a mobility
indicating that it is the b fraction.

After dialysis of the crushed cell

fraction against 30 percent ethanol (0°C.) electrophoretic analyses showed
the b_ and & components to be present as shown in Figure 100 e*

Increasing

the ethanol concentration to liO per cent removed the d component as shovm
in Figure 100 f.

The electrophoretic pattern shovm in Figure 100 g was

obtained from the supernatant after the crushed cell fraction had been
dinlyzed against an acetate buffer of pH 3»91 and ionic strength of 0.1
(concentrated fear tames).

This material is elwctropuoretically inert

and does not give a positive biuret test.

The molisch test is positive

indicating that it may be a. polysaccharide.

Immunization studies have

been carried out using these various fractions and the results will be
published at a later date.
The standard Tiselius electrophoresis cell is applicable to
and satisfactory for diffusion studies (6 6 ).

Boundaries are formed as

in electrdph'ores is experiments and then shifted into "view." The center ‘
sections of the cell are then isolated by displacement to one side and
diffusion patterns (schlieren scanning photographs of the refractive
indgx gradients) are obtained at intervals until the diffusion process
has

cached the closed eiids of the sections.

may take several days.

At zero degrees this

Monopolization of the equipment oar, be avoided,

by supporting the diffusion cells in the water bath in such a manner
that they can be moved Info view 'without distrubing the ciii fusion.

In

the case of ideal diffusion the maximum value of the refractive index
gradient (dn) varies Inversely as the square root o . the time, t,
dx m
according to the relation

-112-

xn which D is the diffusion coefficient and d ^ is the refractive index
increment•

In practice the following procedure may be follo'wed (66).

"A series of ordinates are dravm on s. sheet of coordinate paper at values
of the abscissae equal to the reciprocals of the square roots of the times
at which patterns wero obtained.

Ah enlarged image of the pattern is then

traced on the paper after the position of the latter has been shifted to
bring the base line of the pattern into coincidence with the axis of
abscissae and the proper ordinate passes through the maximum value of
the gradient..

If diffusion has been normal these maxima fall as

.l1i.ustrad lix ra^ui'o luij Oil a Si/Jruign.u xixie passing onrougn une origin.
The diffusion coefficient is then computed from the slope m, of this
line and the areas under the curves, the latter being proportional to
p.

the refractive index

according to the formula

_ j

If —

f

\J~Z*T)

The areas A, under each of the curves are indicated in Figure JO | and
it will be noted that they aro essentially constant as required by the
simole theory.11

378

.0

3.6

3.2

373

2.8

378

2.4

2.0

372

1.6

1.2

0.8

100 v 'T

Figure 101.Diagram illustrating the adaptability
of the Tiselius apparatus to diffusion studies:.
{ Longsworth,66 )

0.4

0.0

-113-

LITERATURE CITED
1*

Abramson, H« A., L« S« Moyer, and M. H. Gorin*, Electrophoresis
of proteins, Reinhold Publishing Co. New York. (I9I42), p. 13*

2.

Abramson, H. A., The microscopic method of electrophoresis. J* Phys .
Chem. 36 II4.5U (1932)

See also reference (l) pages

3.

Tiselius, A., Thesis. Nova* Acta. Soc. Soi, Upsala IV, 7 1; (1930).

ha

Reuss, F . F., Notice sur un nouvel effet de 1* ectricite galvanique.
Mem. Soc. Imp. des Naturalistes de Moscou. 2 327-337 (1809).

5.

Porret, R«, Curious galvanic experiments.
r\ ’

Jf*

6.

•71

f*+~ I

T—

t —

<-■'•«-*-,/

(l

Annals of Philosophy

O T / J V

^ j.t_ul.v.»/®

Kuhn®, Yf., Ubsr das Porret’ache phfinomen am muskel. Arch. Anat*
Physiol. £ 51+2 (i860).

7.

Stern, K. G., The moving boundary method for studying electrophoresis.
Ann. N. Y. Acad. Soi. X X X D U lij.7-186 (1939)*

8® Ouincke, G«, Uber die fortfuhrung materieller theilchen duroh strBmende
elektricitSt.
■.

Pogg. Ann. 10?, 1 (1859); 113*

513-598 (186l).

"•jSrgensen, T®, 'Uber'die bewegung'" fester in fliissigkeiten suspendirter
kBrper unter dem einfluss des elektrischen stroms.

Arch. Anat*

Physiol. 2 673-897 (i860).
10®

Lodge, 0., Report of the British Association for the Advancement of
Science* 389 (1886).

11.

Picton, H., and Linder, S., Solution and pseudosolution.

-J. Chem.

Soc. 61, 1^48-172 (1892).
12.

Hardy, W« B.„ On the coagulation of proteid by electricity.

J.

Physiol, 21; 288-50U (1899).
13.

Hardy, W. B., Colloidal solution.
251-337 (1905).

The globulins.

-J. Physiol. 33

- 1 1 2j -

1)4. Pauli, W., Untersuohung fiber phys ilealische zustandsEnderungen der
Lolloide.
eiwei»3*

Pfinfte mitteilung.

Die elektrisohe ladung von

Beitr. Ohem. Physiol, Path, J , 531-5^7 (1906),

15,

Michaelis, L,, Die Wasserstoffionenkonzentration. Berlin (191I4), (1922).

l6 o

Pauli, W., and Yalko, L», Electrochemde'ider Kolloide,

Dresden and

Leipzig (1933)*
17«

Prausn.it z, P • H„, and Re its totter, .J,, Electrophorese, Electroosmoso,
Pro3den and Leipzig (1931)*

Electrodialyse in Flussigkeiten.
.13,

Svedberg, T®, and Jette, E» R,,
A—«

.* vucw»

190

C* - -

Ir

ij2

*

r
*
M
*
l

Scott, Y.D®, and Svedbei’g, To,

The cataphorosis of proteins,
f ^

J,

•
*\

Measurements of the mobility of egg

albumin at different acidities.

-J. Aon. Ohem. Soo, 146, 2700-2707

(192)4),
20*

Svedberg, T., and Tiselius, A., k new method for determination of
the mobility of proteins,

21*

-J, Am. Chem, Soc.

2272-2278 (1926)*

Theorell, H., Ein neuer kataphoreseapparat ffir untersuchungswecke.
Bioohem. Zeit. 275 1-10 (193U)«

22.

J48,

'

Theorell, H., Ein kataphoreseapparat ffir preparative zweoke.

Ibid

278 291-297 (1935),
23*

Tiselius, A,,
mixtures.

2)4. Tiselius, A.,

A new apparatus for electrophoretic analysis of colloidal
Trans,

li$ 4-1)477 (1937).

Topler, A., 0pti3che studien nach der methode der schlierenbeobachtung,
Pogg. Ann. 131

26.

52)1-531 (1937),

Electrophoretic analysis of normal and immune sera.

Bioohem.J. 31,
25.

Faraday Soo. 33

33“55 (1367).

Tiselius, A., and H. Svensson.,

The influence of electrolyte concentration

on the electrophoretic mobility of egg albumin.
Soc. 36 16-22 (19^0).

Trans® Farsday

-115-

27.

Longsworth, L. G®, The observation of electrophoretic boundaries.
Ann. N. Y* Acad. Soi. 3j+ 187-202 (1939).

28.

Schardin, H., Das toeplersche schlierenverfahren®

Grundlagen ffir

seine anwendung and quantitative anwertung VDI - Verlag. G®

b.H.
29®

Berlin (I93I).).

Abramson, H* A., and D« H* Moore®,

A discussion of electrophoresis

with special reference to serum and allergens.

J. of Lab.

and Clin. Med. 26, 176-181+ (191+0).
30. .Lamm, 0®, Nova Acta Regiae Soc* Scient. Upsaliensis. IV, 10, 6 (1937)®
31*

Kefericjc, R. A.,

Ob sorvat ions at the else bruphorecio moving boundary

with Lamm's scale method.
32.

Longsworth, L. G., A Modification of the schliereti method for use in
electrophoretic analysis®

33®

Nature ll+l, 282-283 (1938)•

Svsnsson, H. V®, Direkfce photographisohe aufnahme von elsktrophoresodiagr&mmeri.

35*

Roll. Zeit. 87 181-186 (1939)V 90 li+l-ll+9 (191+0).

Longsworth, L. G„, Recent advances in the study of proteins by electro­
phoresis.

36.

Chem. Rev. 30_ 323-339 (191+2).

Stern, K® G®, and D. DuBois., Aids for the optical analysis of electro­
phoretic diagrams.

37.

J* .Am® Chem® Soc, 6l, 529-530 (1939)*

Phiipot, J® St. L®, Direct photography of ultracontrifuge sedimentation
curves,

3U«

Trans. Faraday Soc. 36 1+7-52 (I9I+0).

Yale J. Biol. Med, 13_ 509-512 (191+1)*

Siebenmann, C®, Refractomatric methods for determining total protein.
Bioohem. J®, 31_ 205-211 (1937).

38.

Reiss, P«, Eine neue methods der quantitativen eiweissbastimmug
Arch® Exp. Path® Pharmack., 51* 16-29 (190l|-)*

39*

Quoted as a personal communication by A. S. MoFarlane.
centrifugal investigation of the serum proteins®

29, 1+07-1^9 (1935).

An ultraBioehem. J,

- 116-

hO»

Bierry, H., and R. Vivario., Determination of albumins in blood
plasma. Compt. Rend. Soc. Biol. 89, 13-15 (1923).

Jj.1*

Guillaumin, C», R. Wahl, and M* Laurencin., Determination of serum.
proteins.

Comparison of results obtained by the gravimetric

method and determination of nitrogen and refractive indices.
Bull. Soo. Chem. Biol. 11, 387-399 (1929).
1|2.

Mierokjeldahl nitrogen method.

J. Assoc. Official Agr. Chem. 2!+,

• 97-100 (19I4.I).

^3*

Niederl, J* P., and V* Riederl, Mioromethods of quantitative organic
axial'jx ja., John Wiley and Sons, Inc., Now York, (19142) p. 69-78.

Jjil* Wagner, E. C„, Titration of ammonia in presence of boric acid.

Ind. .

Eng. Chem. Anal. Ed. 12 771-772 (I9I4O).
J45*

Sobel, A. S., H* Yu ska, and J. Cohen., A convenient method of determining
small amounts of ammonia and other bases by the use of boric acid.
J. Biol. Chem. 118, 1 ( 1 9 3 7 ) ®

I46*

Ball, C. D», Associate Professor of Chemistry.

Mich. State College.,

East Lansing, Michigan. Personal Communication.
)47,

Kingsley, G. R., The determination of serum total protein, albumin,
and globulin by the biuret reaction.

J. Biol. Chem. 131, 197“

200 (19.39).
i(.8«

Riegler, E«, A colorimetric method for the determination of albumin.
2. Anal® Chem. 53, 2l|2-2l}_7 (19114)®

U9®

Adams, E« C®, Relations between the constants of dibasic acids and
of amphoteric electrolytes.

J. Am. Chem. Soc. 38, 1503-1510

(1916).
50®

Ejerrum, R., Die konstitution der ampholyte, besondere der aminoeluren
und ihre dissoziationkonstanten.
(1923).

Z. Physik, Chem. 1014, ll|7-173

-1J I."?
..
|
j

51.

Gortnar, R* A., Outlines of biochemistry.

John Wiley and Sons Inc.,*

Row York (1938) p. 359.
52®

Davis, B* D®, and E® J* Cohn®, The influence of ionic strength and
pH on electrophoretic mobility.

Ann. N* Y. Acad. Soi® 39,

209-212 (1939)o
53*

S8rensen, S* P® L®, Studies on proteins.
berg.

Compt® rend® Lab, Carls-

12, (1917),

pij.® Schmidt, C® L® A®, The chemistry of the

amino acids and proteins*

Charles C.Thomas, Baltimore, (1938) p.
55® BrBnsted. J. N,.

597,

The acid—base functionof molecules and its

dependency on the electric oharge type®

J® Phys® Chem® 30,

777-790 (1926)®

56, Hill, T. H®, Theory of the isoelectric point®
417-428 (19i|2 ),
57*

Ibid 14 597-605 (1942),

San Clemente, C® L., and

J. Phys. Chem.

Ibid 46621-621; (1942).

1. P* Huddleson®, Electrophoretic studies

of the proteins of bovine serums with respect to
.Agr® -Exp.- "Sta* Bui. 182- 3“4 4 (1943) «
58®

J4.6

*

Slater, J. C«, Introduction to chemical physics®

Brucella* Mich.
- '■

McGr&w Hill Book

Co®. Hew York (1939) P 3&3«
59*

Fouss, R® M., and C. A. Kraus®, Properties of electrolytic solutions®
J® Am. Chem* Soo. 55, 1476, 1019, 2387, (1933).

60. Sookne, A. M®, and J® M® Harris.^ Electrophoresis studies of silk.
J® Research Nat® Bur. Standards 23. 299-308 (1939).
6l®

See reference

number

62.

Lewis, G. N®,and M. Randall®, The activity coefficient of strong
electrolytes®

63®

J®

Gorin, M. A®, and H. A.

1,page l6l*

Am* Chem* Soc* 43, 1112—1154 (1921).
Abramson and L. S. Moyer®, The effect of

phosphate buffers on the electric mobility of hemoglobin*
J. Am® Chem® Soc. 62, 643-^44 (194o).

-113-

614-* Caiman, R. K«, The effect of neutral 3:Its on the hydrogen ion
dissociation curves of proteins*

Cold Spring Harbor Symposia

on Quantitative Biology. 6, 1-8 (1938),

65* Smith, E. R 0 B«, The influence of method of preparation and of cations
on the isoelectric point of ovalbumin.

J. Biol. Chem, 113,

U73-U78 (1936).
66. Longsworth, L. G», The influence of pH on the mobility and diffusion
of ovalbumin.

Ann. li® Y. Acad. Sci. hi, 267-287 (I9J+I).

67® Long3woi*th, L. G®, and D* A® Maclnnes., Electrophoresis of proteins
by the Tiselius method-

68* Davies, C. W.,

Cheme Rev® 2 b , 271-287 (1939).

The conductivity of Solutions.

John ¥<‘iley and Sons

Inc., Herr York (1933) P 6 .

69® Shedlovsky, T®, Personal Communication.
70o

Henderson, L® J®, A diagramatic representation of equilibria between
acids and bases in solution.

J. Am. Chem. Soc. J>Qg 95^-960

(1908).
. 71.

The theory and practice --of the iodometrie -determination"' -

Washburn-,-

of nrsenious acid.
72.

Michaelis, L®, and

R« Kruger., Yfeitere ausarbeitung der indikatoren-

methode ohne puffer.
73®

Bio. Zeit. 119 307“327 (I92l)«

Cohn,E« J., The activity coefficients
phosphate solutions.
action®

lb .

J. Am. Chem. Soo. 30, 31“U6 (1908).

of the ions in certain

A contribution to the theory of buffer

J. Am. Chem. Soc. 1+9, 173“193 (1927)*

Clark, W. M., The determination of hydrogen ions.

Williams and Wilkins.

Baltimore (1928) p. 217, 678®
75®

Robinson, H. W®, The influence of neutral Salts on the pH of phosphate
buffer mixtures*

J. Biol. Chem. 82, 775“802 (1929)•

-119-

76. Cohn, £. J®, F. P. Heyroth and M. p. Menken., The dissociation constant
of acetic acid and the activity coefficients of the ions in certain
acetate solutions.
77*

J. Am. Chem. Soc. 50, 696-7li; (1928),

Green, A® A., The preparation of acetate and phosphate buffer solutions
of known pH and ionic strength.

J. Am. Chem. Soc. 55, 2551-2336

(1933).

78® Perlman, E., J® G« Bui Iowa, and R. Goodkind., An immunological and
electrophoretic comparison of uhe antibody to C Polysaccharide
and the C reactive protein of acute phase serum* J. Exp. Med.
77

79®

Q7_nn

fial.x'i

MacInnes, D. A®, The principles of electrochemistry.
Publishing Corp.

Reinhold

Hew York. (1939) P® UO.

8O 0 Getman, F. H*, and F. Daniels., Outlines of theoretical chemistry.
John Wiley and Sons Inc. New York. (1937) PP« 365-3^7*
81.

Shedlovsky, T*, The electrolytic conductivity of some uni-univalent
electrolytes in water at 25®«

82.

J. Am. Chem. Soc 5^4, li4.ll (1932).

Longsworth, L. G., T. "Shedlovsky, and DV A. Maclhnes, Electrophoretic
patterns of normal and pathological human blood plasma and serum.
J. Exp. Med. 70, 399-1)13 (1939)®

83®

Jones, G., and B. C. Bradshaw., A redetermination of the conductance of
standard potassium chloride solutions in absolute units.
Chem. Soc.

Sl u

J* Am.

1780-1800 ( 1933) *

Daniels, F®, H. J. Mathews., and J® W. Williams.* Experimental physical
chemistry.

MoGraw Hill Book Co., New York (19)4.1) Third edition®

P 3^9.
35.

Pierce, W. C®, and E. L. Haenisch., Quantitative analysis.
and Sons Ino® New York (I9l|-0)j Second edition. P. 3&®

Jchn Wiley

-1 2 0 -

96.

Handbook of Chemistry and Physics.

Chemical Rubber Publishing Co.,

Cleveland Ohio. (I9I4.I) Twenty fifth edition P. 1287.
87.

Broazeals, Tft. M®, A vacuum tube alternating current bridge detector.
Review of Scientific Instruments. 8, 250-251 (1936).

33.

German, R« L., A bridge balance indicator.

Review of Scientific

Instruments 8, 327-329 (1937).
39.

Clark, J. C«, Associate Professor of Physics.
East Lansing, Michigan.

90.

Mich.State College,

Personal Communication.

Dorman, F« G.» Theorie der membrang-leichgewichte und membranpotentiale
bei vorbn.Tjden.sein von nicht dialysiorenden slsoktrolyteu.
beitrag zur physikalisch-ohemischen physiologie.

Ein

Z. Elektrochem.

17, 572-581 (19H)®
91®

Longsworth, L. G., Personal communication.

92.

Longsworth L. G*, R« K* C&nnan, and D» A. Maclnnes., An electrophoretic
study of the proteins of egg white., J. Am® Chem. Soc. 62

2580=2590 ( 19lj.o)«
95®

Stenhagen, E.y Electrophoresis of human blood plasma.

Bioohem. J.

32 71^-718 (1958)a
9I4.. Smith, E. R., A new design of apparatus for the moving boundary method
of determining transference numbers.

J» Am. Chem. Soc* 5£

190U-1906 (1928).
95.

Longsworth, L. G., and D. A® Mac Innas., The interpretation of simple
electrophoretic patterns.

96.

J. Am. Chem. Soc* 62 705-711 (19^4-0 ).

Longsworth, L. G., and D, A® Maclnnes., Transference numbers.

Chem.

Rev. 11, 171-230 (1952).
97.

Miller, W® L., The theory of the direct method of determining transport
numbers. Z. physik Chem. 69 i+36-iji+i (1909)•

-1 2 1 -

98o

Tiselius, A., and E* A. Kabat, An electrophoretic study of immune
sera and purified antibody preparations,

J, Exp. Med. 69

119-131 (1939).

99«

Svedberg, T®, and K. 0* Pedersen., The ultra-centrifuge.
University Press.

100*

Oxford

London (I9U 0 ). p. 296.

Longsworth, L. G», and D« A* Maclnnes., An electrophoretic study of
mixtures of ovalbumin and yeast nucleic acid. J. Gen. Phys,

25. 507-516 (I9i|2)e
101.

Denison, R. B., and B, D. Steele,
von uberfflhrungszahlen-

102,

Uber eine methode tur genauen me3sung

?.~ physik. Chem.

57 110-127 (I907).

Longsworth, L® G., An application of moving boundaries to a study of
aqueus mixtures of hydrogen chloride and potassium chloride.
J® Am. Chem. SoG. 52 1897-1910 (1930).

103*

Henry, D. C«, and J. Brittain., A comparison of the results of
measurements by the transport and moving bound? *y methods,
and a theory of the latter method Trans. Faraday Soc. 29
798—8ll+ ■{1939}® ■■■■■■■■■• ■■ ..

■

■

-

■

■

'”

id**

Tiselius, A., and Horsfall., Ark 0 Kemi. Minera Geo. 13& No. 18 (1939)®

105.

Svensson, H., Fractionation of ss-'um with ammonium sulfate and water
dialysis, studied by electrophoresis* J* Biol. Chem. 139

805-825 (191*1)©
106*

Longsworth, L. G., and D. A. M a c r o s . , An electrophoretic study of
nephrotic sera and urine. J* Exp® Med, JQ 77-92 (19I1O)®

107*

Moore, D. H., and J. Lynn*
human plasma*

108.

Electrophoretic measurements on normal

J® Biol® Chem* 11*1 819-925 (I9I4I )®

Sorensen, S. P. L. quoted by Longsworth, -L® G* and

D« A. Maclnnes (106)*

-32JU

109®

Tiselius, A,, Bleotrophoretische messungen am eiweis:', Kolloid Z.
85 129-137 (1938)-

110®

McFa.rj.ano, A. S«, T* 3 elsct n o
Tran 3* Faraday Soc*

111«

double layer and virus stability.

56 257-26I4. (lgl+O).

Shedlovsky, T» , and. J• E* Smadole, Electrophoretic Studies on
elementary bodies -of vaccinia.

112®

J, Exp. Med, 72 511-521 (I9I4.0).

Jameson, E*, and Alvarea-Tostado., Qualitative separation of proteins
of rat serum by electrophoresis.

Proo. Soc. Exp, Biol, Med.,

I4O l4.
76-i4.78 (1939).
113-

Scudder, J. , Symposium on fluid and electrolyte needs of surgical
patientsj studies in blood preservation; stability of plasma
proteins.

Ill;®

Luetscher, J. A© Jr,, Electrophoretic analysis of plasma and
proteins.

115.

J. Ann. Surg. 112 502-519 (l9lj.O)*

J. Clin. Inv. 19 313-320 (I9U 0 )., 20 99 (19l+l)«

Kekwiok, R. A . , The electrophoretic analysis of normal human serum,
Bioohem, J. 33 1122-1129 (1939)*

116.,

Blix,-Gr,, .Quantitative, determination -of oQ
normal serum and pneumonia serum.

117.

Cohn, E. J*, The properties and functions
with a consideration of the methods
and purification*

118.

@ . -and ..^globulin-

Hygiea

in- ■-

100 77-81+ (1938),

of the plasma proteins
for their separation

Chem. Rev, 28^ 395-1+17 (191+1).

Kekwick, R. A,, The serum proteins in multiple myelomatosis.
Bioohem. J, 3i+ 121(8-1257 (191+0).

119.

Blix, G., Quantitative bestimmung von elelrtrophoretisch getrennten
serum globulinen*

120®

Z. ges. exp. Med. 105 595-598 (1939)®

Green, A. A®, The amphoteric properties of certain globulin fractions
of normal hors© serum.

J. Am® Chem® Soo.

80 1108—1115 (1931)®

-123 -

121.

Shedlovsky, T®, and J, Scudder®, A comparison of erythrocyte

sedimentation rates and electrophoretic patterns of normal
and pathological human blood*

122®

J. Exp. Med® 75 119-126 (192+2)®

Shapiro, S®, V® Ross, and D® H» Moore®, A viscous protein obtained
in large amount from the serum of a patient with multiple
myeloma® J® Clin® Inv® 22 137-U+2 (19U3).

123.

ui'fty» S» J<*» and E* S® Guzman Barron®, The electrophoretic analyses
of the serum proteins in diseases of the liver®

J® Clin® Inv®

22 191-200 (19U 3).
4.104.0

Tiselius, A®, ana E® Rabat®, An electrophoretic study of immune sera
and purified antibody preparations®

J® Exp® Med® 69 119-131

(1939).
125®

Kekwiok, R.® A®, and B® R® Record, Some physical properties of
diphtheria antitoxic horse sera®

Brit® J® Exp® Path® 22 29-14+.

(I9I4I).

126®

Moore, D. H®, J® van der Scheer and R® W* G. Wyckoff •, An electrophoretic
study of antipneumoccal horse sera.

J® 1mm® 38 221-229 (I9I+.O).

127® Fell, N® E®, G® S t e m and R.® D. Coghill®, A physical-Chemioal study
of normal and immune horse sera®

128®

Jameson E®, and C. Alvarez-Tcstado®, A study of b lo o d s c n m proteins

by electrophoresis*
1?9«

J. Imm* 39 223-21+5 (191+0) •

J® Phys® Chem* i;.3 1165-1172 (1959).

Eili, R® Mo, and V® Trevcrrow®, Normal variation in the concentration

of fibrinogen, albumin, and globulin in blood plasma.

J® Phys®

Chem® k 6 1117-1129 (19142)®
130 .

Cohn, E. J®, T® I.® McMeekin, J. 1® Oneley, J® K® Newell and T7. L.
Hughes®, Preparation and properties cf serum and plasma proteins

I.

J® Am. Chem® Soc. 62 3586-3393 (19U0).

3yl« McMeekin,
II.
332.

T® I** Preparation and properties of serum and plasma proteins
J. Am. Chem. Soc. 62 3393-3396 (I9I4.O).

Cohn, E. J., J. A. I.uetscher Jr., J. L. Oncley, S.

H. Armstrong Jr.,

and B* I)« Davis®, Preparation and properties of serum and plasma
proteins. III. J* Am. C-bem. S00^396-31+00 (191+0).
333®

Seudder, J., Electrophoretic patterns of Serum and plasma,
substitutes and blood transfusions.
Charles C. Thomas.

3J'1+.

Fabat, E» A

335®

Mudd, S., and W k Thalhimer.

Baltimore (19L<£) p 126-139.

F® Iandow and D. F* Moore®, Electrophoretic patterns of

concentrated cerebrospinal fluid.
b$

Elood

Prcc. See. Ezp. Biol. Med.

260-263 (191(2 ).

Hesselvik, I.., Electrophoretical investigation 011 proteins of eye
lens and vitreous body®

Acta Physiol. Scand. 101 151-156 (1939)*

336®

Blioc, G. , Glueoproteins .

337*

Hesselvik, I,., Electrophoretical study of normal and pathological
body fluids.

Acta Physiol. Scand, {01.29-1(2 (L-'i+O).

Acta Med. Scand. 101 l+6l-l(6Lj. (1939)*

w-. I 36® ■ Tiselius, "A®, and EV A.- Fabat.-,' 'Electrophoresis on Immune serum.""' "'”

Science 87 1+16-1+17 (1938)®
339*

Vah der Scheer, J. J®, P. Lags din and R. F. G. V/yckoffElectrophoretic
and ultracentrifugal analyses of antipneumoc oecal horse sera.
J. Imm. 1+1 209-225 (191(1)®

3J...0* Van der Scheer, J., and R® W. G. Fyckoff., An electrophoretic study of
tetanus antitoxin sera,
ll+l.

Proc. Soc, Exp. Biol, Med. 1(3_ 1+27-1+28 (I9I+0)

Van der Scheer, J®, P., W. G. Wyckcff and F . E. Clarke., The electrophoretic
analysis of tetanal antitoxic horse-eera.
(1921)®

J. Imm. 1+0, 173-177

-125-

llj2.

Van der Scheer J», R. W©

G. Wyckoff and F. H. Clarke., The electro­

phoretic analysis of several harper immune horse sera.

J. Imm,

39 65-71 (I9ii0).
3ll3©

Rothen, A«, Purified diphtheria antitoxin in the ultracentrifuge and
in the electrophoresis apparatus.

J. Gen. Physiol. £5 J+87-U96

(19^2).
Parfentiev, I. A.* U. S© Patents Ko.
Ill5*

2065196 and. Ho. 2123198©

Van der Scheer, J., p.. W. G-. Wyckoff and F. H© Clarke., An electro­
J. Imm. h i 3i+9-36o (19U ) .

phoretic examination of digested sera.
Jli6.

Siebert, F. B«, and J. F s N«l«cn*, Proteins of
Chem. Soo.

3*77®

Tuberculin.

65 272-278 (19^3)*

Siebert, F© P., and J. W .

-^e1son,, Electrophoretic study of the

blood protein response in tuberculosis*
2 9 -3 6

3i;8«

J. Biol. Chem. lljg

(19h2je

Siebert, F© E., and D. W* hat son*, Isolation of the polysaccharides
and nucleic acid of tuberculin by electrophoresis.
■ Chem. --liiO -55-69-(X9i|l)«

3l;9©

J. Am.

'■

J. Biol.

"

"

Pi 1liner, I®, ii. E. Ecker, J. L« Oncley and E. J. Cohn©

'

The preparation

arid physicochemical characterisation of the serum protein components
of complement©
350.

J. Exp. Med. 7 k 297-308 (19^1).

Quensel, I. B. £„, and T. Svedberg,, Sedimentation and electrophoresis
of the tobacco mosaic virus protein.

J. Arc. Chem. Soc.

58

1863“1867 (1936)©
351©

Pfankuck, E*, Clevage of the virus proteins of the tobacco mosaic
variety®, Biochem. Z© g()6 125-129 (19^0)®

352.

MeFarlane, Physical properties of bushy stunt virus protein©
J.

32 1607-1613 (1938).

Biochem

- 126-

155.

Douglas, S. R., and W. Smth., Cataphoresis experiments with the
virus of vaccinia.

I5h,

Brit. J. Exp. Path.

Lee, C. D., and H. L. Wilcke.,

9 215-215 (1928).

The migration of the etiologic agent

of i owl leukosis when subjected to electrophoresis. J. Am.
Vet. Med, Ass. 9^ 176-,86 (1939).
155*

Smadel,

J, E., and T . S-ho-'lovsky,, Antigens of vaccina. Ann. U. Y.

Acad. Soi. XLIII 35-1*6 (9l*2),

156. Anramson, II* A*, D. H* Moore and 11. G. Gebhror©, A general electro­
phoretic pattern in extracts of pollens causing hay fever.
J. Phys. Chem. j*6 1129-1139 (191*2).
*i v n

I•

Ringer, VI. S., Vfeitere studio l am pekelharingschn pepsin.
Physiol. Chem.

150.

95 195*258 (1915).

Tiselius, A., G. E* Hanschen and It Svenson., Electrophoresis of
pepsin.

159.

Z.

Biochem. J. 3j~ 1834-1818 (1938).

Herriot, R. M., V. Desreux and J.
pepsin.

Northrop., Electrophoresis of

J. Gen Physiol. £3 439 "447 (1948) *

166. Theorell, H., Ein neuer kataphoresearyarat f\ir untersuchungszwecke•
Biochem. Z. 275 1-36 ( 1 9 3 5 ) Ein kataphoreseapparat fur
.paraparative zwecke. Biochem, Z. 2?8 291-297 (1935).

161.

Theorell, H., Reines cytochrome C. Biochem. Z. 279 463-464 (1935).

162. Vahlquist, B.,

The esterase activity of human blood plasma.

Skand.

Arch. Physiol. 72. 133-16© (1935)«

163. Rothen, A., Moelcular -weight and electrophoresis of crystalline
rihonuclease.

161*.

J. Gen. Physiol. 21* 203-211 (1940~4l)•

Bate-Smith, E. C., Cataphoretlc behaviour of zymohexass.
J.

Biochem.

31* 1122 (19i{.0).

165. lie1nick, J. L., and R. G* Stem., Carboxylase, Enzycnologia 8
129-151 (194.0).

-127-

166.

K u n rtz , J . ,

F o rm a tio n o f new c r y s t a l li n e enzymes fro m c h y m o try p s in .

J . Gen. P h y s io l.
10/»

22 207-237 (1 9 3 8 ).

S h ip le y , R« A ., K. G. S te rn , and. A . W h ite .,
a n t e r io r p i t u i t a r y p r o te in s .

l6 0 .

J . Exp. Med. 69 785-900 (1 9 3 9 ).

L i , C. H ., W. R. Lyons and H. M» E va n s., S tu d ie s on p it u it a r y / la c to g e n ic
hormone.

169.

E le c tro p h o re s is o f

J . Gen. P h y s io l.

S h e d lo v s k y , T.#
Chow.,

£3

k33-k3& (1 9 ^ 0 ).

A . Rot hen , R. 0 . Oreep, H.

The i s o l a t i o n in pure fo rm o f

B. Van Dyke, and B. F.
th e i n t e r s t i t i a l c e l l -

s t im u la t in g ( l u t e i n i z i n g ) hormone o f th e a n t e r io r lo b e o f th e
p i t u i t a r y g la n d .
170.

Chow, B. F., II. B. Van

S cience 9_2 178-130 ( I 9I4.O).
Dyke, R. 0 . Graep, A. R othen, and T . S h e d lo v s k y .,

C hem ical a s p e c ts o f m e ta k e n trin ( i n t e r s t i a l c e l l S t im u la t in g
horm one).
171«

J . B i o l . Chein. II4O XXVI-7DCVII ( I 9I4I ) .

L i , C. H ., H. M. Evans and D. H» W onder., E le c tr o p h o r e tic hom ogeneity
o f P regnant mare serum.

G o n a d otrop h in. J . Gen. P h y s io l, 23

733-739 ( 1939-i+O ).
172.

G u rin , S., C. Bachman* and D, >V. Wilson.,. , h o ia o g a n itv ,o f g o n a d o tro p ic
hormone p re p a ra tio n s is o la te d fro m pregnane;/ u r in e .

J . A31.

Chern. Soc. 6 l 2251 (1 9 3 9 ).
173.

Colin, M ., G. W. I r v i n g

J r ; and V . du V ig n e a u d .,

The a m photeric n a tu re

o f th e p re s s o r p r in c ip le o f th e p o s t e r io r lobe
g la n d . J . B i o l. Chera*
171+.

o f th e p i t u i t a r y

137 635~^U2 (^ 9 U l)»

du V ig n e a u d , V ., G. W. I r v in g J r . , H. M. D yer, and R. R. S e a lo c k .,
E le c tr o p h o r e s is o f p o s te r io r p i t u i t a r y gland p re p a ra tio n s .
J . B i o l. Chem. 123

U5-55 (1 9 3 8 ).

-128-

175*

Irving, G. W. Jr., and V. du-Vigneaud.,

The differential migration

of tne pressor and oxytocic hormones in electrophoretic studies
of the untreated pressjuice of the posterior lobe of the pituitary
gland.
1 7 6 .

J. Eiol. Chem.

123 1*85-’
489 (1938).

Heiuelbergar, M., and K. 0. Pedersen.,

The molecular weight and

isoelectric point of thyroglobulin J. Gen. Physiol.
19 95-108. (1935-36). '
177.

Landsteiner, K,, L. G. Lon rsworth, and J. Van dor Scheer.,
Electrophoresis experiments 'with egg albumins and hemoglobins.
Science

l?a*

88_83-85

(1938).

Herbert, D., H, Gordon, V. Subrahmanyan and D. E. Green., Zymohexase.
Biochem. J. 3h 1108-1123 (191+0).

179.

Fishman, M«, and L. S. Hover.,

Electrophoresis of the chlorphyll-

protein complex. Science 95 128-129 (191*2),
180,

Harx'is, M.,
Standard

181.

The isoelectric point of wool. J. Research Nat. Bur.
8

779-782 (1 9 3 1 ).

Sookne, A. H . , and M.. Harris.

Electrophoretic studies of wool. ■

J. Research Hat. Bur. Standards 23 lf.71"U?7 (1939).
182.

Wilkersori, V. A.,

The chemistry of human epidermis II.

The isoelectric

points of the stratum corneum, hair, and nails as determined by
electrophoresis.'
I8 3 .

Wilker 3on, V. A.,

'J. Biol. Chem. 112 329-335 (1935-36).

The Chemistry of"-human skin.

IV

The elactrokinetic

effect of various ions upon suspended particles of stratum corneum.
J. Gen. Physiol. 23 165-170 (1939-^0).

181*
. Beek, J, Jr., and A. M. Sookne.,

Electrophoresis of collagen. J.

Research Nat. Bur. Standard 23 2?l-273 (1939).

185.

Martin, W. II., Electrokenetic properties of proteins,

I Isoelectric

point and solubility of wheat proteins in aqueus solutions of
alcohol. J. Phys. Chem. 35 2065—2090 1,1931).

-1 2 9 -

186.

Kemp, I.,

187•

Williams,

Trans. Faraday Soc., 52 837 (1936).
J.

iff.,

prolamines.
Biology.
138.

and C. C. Watson.,

The physical chemistry of the

Cold Spring Harbor Symposia on Quantitative

6 208-217 (1938).

Andrews, A. C.,

The physical chemistry of secalin.

Electrophoresis

and diffusion constant studies of the prolamine of rye.
A®. Chem. Soc.

189. Me Hand e r , 0 .,

J.

62 942-948 (1940).

Elektrophoretische untersuehung von casein.

Biochem.

Z. 300 240-21+5 (1939).
HO.

Hall, J.

The physical chemical investigation of certain nucleo-

proteins.
Soc.
191®

63

II®

General electrophoretic behavior.

J. Am. Chem.

794-798 (1941).

Stahl, Iff. H., R. B. Pennell and I. F. Huddleson.,

A study of the protein

nucleates of the species of the genus Brucella.

Tech. Bull. 168

Hich. Agr . Exp. Sta. (1937)®
1.92«

Pennell, R. B., and I. F. Huddleson.,

Chemical constitution and biological

properties of the endo-antigen of the Brucella group of microorganisma.
193®

Tech. Bui. 156® Mich. Agr. Exp. Sta. (1937)®

V a n der Scheer J.,

R. W. G. Wyckoff, and F. L. Ciarke., An electro­

phoretic study of heated horse serum. J. Imm. 40 39-45 (1941)*

194®

Hardb., C. R., I. F. Huddleson and C. D. Ball.,

The protective action

of glucose of bovine plasma against heat coagulation.

To appear

in Science.
195.

Huddleson.

I. F.,

infection.

The immunization of guinea pigs against Brucella

Mich® Agr. Exp. Bui.

182 45-56 (1943)®