S Y N T H E S I S O F P R O T E I N IN T H E P A N C R E A S

II. THE ROLE OF R/BONUCLEOPROTEIN IN PROTEIN SYNTHESIS

Bx V. ALLFREY, MARIE M. DALY, AND A, E. MIRSKY
(From the Laboratories of The Rockefeller Inslitute for Medical Research)
(Received for publication, June 12, 1953)
INTRODUCTION
The observations and experiments presented in this paper are concerned with
the role of the protein component of ribonucleoprotein in synthesis of protein.
This work shows that the nucleoprotein is part of the cell's equipment for
protein synthesis and provides evidence that the protein component of nucleo-
protein functions as precursor material for proteins synthesized in the cell.
It is well known that the uptake of isotopically labelled amino acids is more
rapid in the protein of the microsome fraction (a nucleoprotein-containing
material) of the liver than in other protein fractions of the tissue (1-3). Al-
though such high incorporations into microsomes suggest that they play a role
in protein synthesis, it is important to determine whether such activity is
limited to a replacement of microsome protein or whether it bears on the general
problem of protein synthesis in the cell. In the experiments about to be de-
scribed, using the pancreas, it has been possible to make this distinction and to
demonstrate that the microsome protein is concerned with the synthesis of the
secretory enzymes of the cell.
The pancreas is the tissue to which we have directed our attention. Our
reasons for selection of the pancreas should be given because in the numerous
investigations on protein synthesis that are now appearing from so many
different laboratories very little has been done on the pancreas. The exocrine
portion of this tissue is highly specialized for protein synthesis, much in the
way that striated muscle is specialized for contraction, so that cell components
concerned with protein synthesis would be expected to be more prominent in
the pancreas than in the generality of tissues. Protein synthesis is also a promi-
nent aspect of growth and for this reason growing cells and tissues are frequently
used as material for investigations of protein synthesis. Synthesis in a gland is,
however, probably less complicated by other processes which occur in cell
growth and division. Furthermore, in a digestive gland, such as the pancreas,
the investigator can readily control conditions which influence the rate of the
synthetic process. After a copious secretion there follows a period of active
synthesis to replenish the supply of secretory enzymes. By inducing or inhibit-
ing secretion the rate of synthesis can indirectly be influenced; and it can, in
157

The Journal of General Physiology

158 SYNTHESIS OF PROTEIN I N PANCREAS. I I

fact, be shown by isotope incorporation experiments that synthesis proceeds

more rapidly after secretion.
The pancreas we have used is that of the mouse. To obtain a sufficient quan-
tity of tissue it has been necessary to use about six animals for each experiment.
There is, however, a certain advantage gained by using a number of small
animals in place of a single large animal: experiments are more reproducible.
Another advantage is that the total weight of mice used ' in an experiment is
much less than the weights of rats, guinea pigs, or rabbits that would be re-
quired for an equal quantity of pancreas, because the size of the pancreas in
relation to total size of body is far greater in the mouse than in larger animals.
Results of experiments on the pancreas are more readily understood ff cor-
responding experiments on other organs, such as liver and kidney, are done at
the same time. Synthesis of protein in the liver is less active than in the pan-
creas, and in the kidney it is less than in the liver. It has been estimated that
plasma protein production in the rat liver is at the rate of 0.002 rag. protein per
1 rag. of dry liver per hour (4). Our experiments on the pancreas of the mouse
show that amylase synthesis is about 0.005 nag. per 1 rag. dry pancreas per
hour. This estimation is based on the amount of enzyme accumulated by the
gland after depletion by pilocarpine. Total protein synthesis in the pancreas,
if all secretory enzymes are considered, is, therefore, more than ten times
faster than in the liver. No process on a scale comparable to secretion of diges-
tive ferments by the pancreas or of plasma proteins by the liver is known to
occur in the kidney so that synthesis in its tissues is probably at a rate far less
than that of the liver. These marked differences in rates of protein synthesis
should constantly be borne in mind when comparing results of experiments on
the pancreas, liver, and kidney.
It was such comparisons with respect to ribonucleic acid content and capacity
for protein synthesis in different types of cells that led Brachet and Caspersson
to suggest that ribonucleic acid is in some way concerned with the process of
protein synthesis. A similar correlation between ribonucleic acid content and
protein synthesis has also been observed in microorganisms (24-26). Attempts
to proceed beyond this circumstantial evidence implicating ribonucleic acid in
protein synthesis have hardly been successful so far. Indeed, Caspersson's
contention that the quantity of ribonucleic acid in the pancreas varies during
the cycle of synthesis and secretion has been shown to be incorrect (5, 6). Even
so the mass of circumstantial evidence compels one to persist in searching for
the role of nucleic acid in protein synthesis. In this way we have been led to
examine the protein component of ribonucleoprotein and the part it plays in
protein synthesis.
The Nucleoproteln Pellet Prepared by Centdfugation at 40,000 R.P.~.IMost
of the ribonucleic acid of a tissue is in the so called microsome fraction, which
sediments at 40,000 ~.P.M. in a sucrose medium to form a pellet. In this paper
it will be referred to as the pellet material. In a study of the protein component
 V. ALL~FREY~ MARIE M. DALY~ AND A. E. MIRSKY 159

of the pellet material it is important to know whether the various components of

the pellet have merely been brought together by centrifugation at a certain
speed for a certain time or whether the components are associated in sulficienfly
definite form to permit us to presume that they were also associated with each
other in the living cell. The properties of the pellet material, to be described
now, indicate that protein and nucleic acid are combined in a definite complex.
Most of these experiments have been done on pellet material of the liver
rather than on that of the pancreas because of the intense activity in the latter
of ribonuclease, which alters the composition of the pancreas pellet from the
moment the tissue is disintegrated. A study of the action of ribonuclease on the
pancreas pellet shows that combination of protein with the nucleic acid pro-
tects the latter to some extent from the enzyme; decomposition of free nucleic
acid at pH 6.8 proceeds five times as fast and at pH 7.6 twice as fast as does
decomposition of the nucleic acid of the pellet; but ultimately all the nucleic
acid of the pellet is broken down. This is in marked contrast to the combination
of protein and ribonucleic acid in the tobacco mosaic virus; in this case com-
bination completely protects the nucleic acid from the action of ribonuclease.
In the pancreas pellet combination with protein protects the nucleic acid no
more than do the salt-like combinations formed when histone is added to
ribonucleic acid, for experiments show that the breakdown of ribonucleic acid
in presence of an excess of thymus histone proceeds at about the same rate as
it does in the pancreas pellet at pH 6.8 and pH 7.6. (This comparison should
not be taken to mean that the protein of the pellet is of a histone-type. At a
later point in this paper a discussion of this problem is given.)
Electrophoretic experiments with pellet material show that in it protein and
ribonucleic acid are combined in such a way that they do not migrate separately
in an electric field. In Fig. 1 the results are given of a zone electrophoretic
experiment on liver pellet material. The distribution of nucleic acid after electro-
phoresis in glass powder was followed by its characteristic absorption maxi-
mum at 258 m#. Since nucleic acids have an absorption minimum at 230 m~,
where protein solutions have high extinctions, it was possible to measure the
protein distribution after electrophoresis by measuring E2s0. (This was supple-
mented by nitrogen analyses as explained in detail in the experimental section.)
In the figure both E ~ and E2s0 are plotted against the distance of migration
in centimeters. The pellet material shown in Figure 1 A was prepared in 0.25
sucrose and the electrophoresis was done in 0.25 ~ sucrose. For comparison
the migration of ribonucleic acid itself under similar conditions is given, show-
ing clearly that the nucleic acid of the pellet material remains firmly attached
to protein in the electric field.
That sucrose maintains the integrity of the pellet material is shown when the
electrophoresis of material prepared in 0.25 ~ sucrose is carried out in the
absence of sucrose. In Figure 1 B it is seen that in absence of sucrose there is a
tendency for the nucleic acid of the pellet to move away from the protein.
160 SYNTHESIS OF PROTEIN IN PANCREAS. H

6-'1c m . Liver A. pH 7.0

10.0 peilet 0.251~I5ucpose
buffeped with
00
phosphate
Yeast ~aNA
6.0-
L £~5Bm~\ " ~i
4.0 - ~

2.0~ L :/
0 I "1 I I -I" - t ~ "
8.0 -6 ~3o
/~LLve~ ~. pH 7.0
-°'2 * \ pellet pho~phat,
e buffe~

4.0-

0-13 -5 0 +5 +18 +21 +29 *37 +45

Mi~paZion dis%ante in cm.
FIG. 1 A and 1 B. Fig. 1 A shows zone electrophoresis pattern of mouse liver pellet
material after 100 hours' migration through glass powder in 0.25 M sucrose buffered
with phosphate at pH 7.0. Nucleic acid distribution indicated by o~. Protein dis-
tribution shown by e~a0. The dotted curves show the electrophoretic pattern of yeast
RNA under similar conditions. Fig. 1 B shows electrophoretic pattern of mouse liver
pellet material prepared in 0.25 ~ sucrose but placed in 0.2 ~ phosphate buffer at
pH 7.0 for electrophoresis.

When pellet material of pancreas, rather than that of liver, is subjected to

electrophoresis the nucleic acid moves out far ahead of the protein, but when
this fast moving nucleic acid is examined it is found to have been decomposed in
 V. ALLFREY~ MARIE M. DALY, AND A. E. MIRSKY 161

the long period required for this experiment by the ribonuclease present in
the preparation.
When the pancreas pellet material is prepared, if done quickly, only a small
part of its nucleic acid is broken off by ribonuclease. The high percentage of
ribonucleic acid in this material is itself evidence that the combination of
nucleic acid and protein is not an artifact formed during the preparation. Con-
tent of nucleic acid, varying slightly in different preparations, is about 23
per cent of the lipid-free pellet material. To this can be added the nucleic acid
split off, about 10 per cent of the total. This amount is found by measuring the
acid-soluble nucleotides present in the supernate after centrifuging at 40,000
R.P.K. and subtracting therefrom the amount of acid-soluble nucleotide found
in pancreas treated at once with cold 2 per cent perchloric acid. Correcting in
this way for decomposition occurring during preparation, the pellet material
of the pancreas contains about 25 per cent ribonucleic acid. It hardly seems
likely that protein capable of combining with so much nucleic acid at pH 6.8,
the pH of the sucrose preparation, is not the protein that was combined with
nucleic acid in the intact pancreas but was merely picked up in the course of
preparation.
When the liver pellet is prepared two of the main conditions of the prepara-
tion, duration and speed of the high speed centrifugation, can be varied con-
siderably so that several pellet samples can be prepared from the same mass of
tissue and yet the samples with different sedimentation characteristics do not
differ in the proportions of nucleic acid and protein. Experiments along these
lines, done by two different investigators on rat liver, indicate that the nucleic
acid and protein components of the pellet material have not become associated
during preparation, hut were originally combined in the cells of the liver (3, 7).
Constancy of the proportions of nucleic acid and protein in the liver pellet
is also shown when this material (prepared from mice) is suspended in a sucrose
solution buffered at pH 9.0 with borate and then centrifuged at 40,000 R.~'.~.
We found that although some 20 per cent of the material did not sediment, the
ratio of nucleic acid to total nitrogen in the pellet was 0.10, the same as in the
original pellet, which had been prepared at pH 6.8.
Another observation showing the definiteness of the combination between
nucleic acid and protein in the pellet comes from a study of this material in
the pancreas of fasting mice. After a 4 day fast the ribonucleic add content of
the pancreas decreased 39.6 per cent. Pellet materials prepared from the fasted
mice and from control mice had, however, about the same composition. Lipid-
free material of control mice had 17.35 per cent nitrogen and 2.3 per cent phos-
phorus (estimated both as total phosphorus and as nucleic acid phosphorus by
determination of the ultraviolet extinction coefficient in the material released
by 10 per cent perchloric acid at 70°C.); that of fasted mice had 17.4 per cent
nitrogen and 2.2 per cent phosphorus.
In Table I the ribonucleic acid contents of the pancreas, liver, and kidney of
162 SYNTHESIS OF PROTEIN IN PANCREAS. H

mice are given along with the nucleic acid contents of pellet materials of these
tissues. The correspondence between RNA contents of these tissues and their
activities in synthesizing protein is an example of the correlation first noted by
Brachet (8) and Caspersson (9) and more recently observed in microorganisms
(24-26). The data in this table show that there is also a correlation between the
percentage of nucleic acid in the pellet material of a tissue and the activity of
the tissue in protein synthesis.
The high nucleic acid content of the pancreatic pellet material raises the
question of whether the protein of this pellet is especially rich in basic amino
acids. A brief discussion of this matter must now be given, for it has, indeed,
already been claimed that in such cells as those of the pancreas which contain
high concentrations of ribonucleic acid there are proteins rich in diamino acids
associated with the RNA, and much has been said about the production of such
proteins by the nucleolus and the "nucleolus-associated chromatin" and
TABLE I
Ribonucleic Acld Contents of Mouse Tissue and PelletMalerials
R N A phosphorus
Tissue
Tissue Pellet material

Pancreas 1.20 2.3 (2.5)*

Liver 0.389 1.11~
Kidney O.237 0.6So~t
* Total phosphorus.
Orcinol.
their role in protein synthesis in nucleus and cytoplasm. In their first claims
about these proteins Caspersson and his colleagues referred to them as of
"histone type" because it was supposed by them that the ultraviolet absorption
spectrum of a histone differsfrom that of other proteins, being shifted somewhat
to longer wave lengths and this seemed to account for the absorption at these
wave lengths which they reported in regions of cells rich in ribonucleic acid
(10). W h e n the absorption spectra of satisfactory histone preparations were
determined, it was found, however, that the maximum of the ultraviolet ab-
sorption spectrum of a histone does not differ from that of the generality of
proteins (II). W h e n this became known the term "histone type" protein was
dropped and these proteins associated with ribonucleic acid were referred to as
proteins rich in diamino acids. Histones are in fact rich in diamino adds, but
the evidence that the proteins associated with high concentrations of ribo-
nucleic acid are rich in the basic amino acids is something one searches for in
vain in the numerous papers from Caspersson's laboratory in which reference is
frequently made to proteins rich in diamino acids (see also the recent book by
Caspersson (9) and the book by Danielli (12)).
 V. ALL~REY~ M A R I E M. DALY, AND A. E . MIRSKY 163

It is, of course, possible that the proteins associated with ribonudeic acid do
have a high concentration of basic amino acids and for this reason the quantities
of these amino acids in the proteins of pellet materials prepared from liver and
pancreas were determined. The analytical results are given in Table II along
with similar analyses of a histone and of mixed cytoplasmic proteins of the
pancreas and liver. The diamino acids of the pellet proteins are no higher than
in mixed cytoplasmic proteins and far lower than in a histone. The lack of an
especially high content of diamino adds in pancreas pellet protein is noteworthy
because this protein is associated with an exceedingly high concentration of
ribonucleic acid. Absence of especially high concentrations of the diamino adds
in liver and pancreas pellet materials is of interest in connection with what is
known about the composition of the plant viruses, a group of ribonucleoproteins
with nucleic acid contents varying from 5 to 41 per cent (13). The nitrogen
contents of these nucleoproteins have been determined and if an accumulation
of nucleic acid were correlated with a high content of diamino acids this would
TABLE II
Basic Ami~o Acid Composition of Pellet Ptotdns, Mital Cytoplasmic Proteinr, and ttlstone
Moles mlno scid per 100 moles nitrogen
Amino acid P~nct~s Liver Liver
' Liver pellet Pltllcre~s histone
pellet (calf)

Histidine 1.21 1.57 1.69 1.81 1.15

Lysine 4.50 4.42 5.08 5.20 6.25
Argin~ue 2.85 3.51 3.22 3.65 6.55

probably be shown by a high nitrogen content. The analyses show no such

correlation. The nitrogen content of the nucleoprotein with 41 per cent nucleic
acid is actually less than that of a nucleoprotein with only 5 per cent nucleic
acid. Even if phosphoric acid groups of nucleic acid combine with amino groups
of the diamino acids of the plant virus nucleoproteins and of the proteins in
pancreas pellet material there is no need to suppose that the diamino acid
contents of these proteins are unusually high.
Comparison of the Rates of Uptake of Glycine Conta~'ning N ~5 by Pancreas,
Li~er, and Kidney.--In this way some insight can be gained concerning protein
metabolism and protein synthesis. In Table H I the results are given of rates of
incorporation of N Is, administered to mice as glydne, into the mixed proteins
of the pancreas, liver, and kidney, into the protein of the pellet materials pre-
pared from these tissues, into the protein of the supernatants remaining after
sedimentation of the pellets at 40,000 R.l'.~., and into the ribonucleic acid of
the pellet materials. The figures for liver proteins in Table III showing that
rate of incorporation of N lb into protein of the pallet is markedly greater than
that into other liver protein fractions confirm the results obtained by previous
investigators (1-3).
164 SYNTHESIS OF PROTEIN" IN" PANCREAS. H

Comparison of rates of N Is uptake by proteins of the three tissues studied

shows that they follow the sequence: pancreas, liver, kidney--the first being
about twice that of the second and the second twice that of the third. Uptake
by proteins of the pellet material is likewise highest in the pancreas, less in the
liver, and lowest in the kidney. This is also the sequence of rates of protein
synthesis in the three tissues. The high rates of protein synthesis in liver and to
even a greater extent in pancreas are due to the massive protein secretion by
these tissues. There is no simple way of collecting the protein secreted by the
liver and it is not at present possible, therefore, to distinguish sharply in this
tissue between protein synthesized as part of the internal metabolism of the
cell and protein synthesized for secretion. Since the secretion of the pancreas
flows out through a duct it can readily be collected. Examination of the pan-
creatic secretion shows that no nucleoprotein is present. The nucleoprotein of

TABLE HI
Incorporation of Glycine-N1~into Proteins and Ribonudeic Acid of Cell Ftaagons
N It excess 30 rain. after glycine administrstion in
Tissue
Mixed Pellet Supernate RNA of pellet
protein protein protein
atom #er cen# atom per cent atom per cent atom #er cent
Pancreas 0.105 0.183 0.067 0.006
Liver 0.059 O.107 O. 055 0.038
Kidney 0.025 0.046

E a c h animal received 30 rag. NlS-glycine (33.3 a t o m per cent excess) in 0.2 ml. H20 in-
traperitoneally a t zero time.
All animals fasted 20 hours prior to Nn6-glycine administration.

the pellet material is, therefore, definitely not one of the substances synthesized
to be itself secreted by the gland. With this possibility excluded, there are
grounds for considering the pellet material to be concerned with the synthetic
process itself. In a tissue, such as the pancreas, highly specialized for protein
synthesis a protein fraction with both an unusually high content of ribonucleic
acid and an unusually rapid uptake of Na~-containingglycine must be intimately
involved in synthesis of the secretory proteins. Experiments designed to show
the part played by the pellet material will presently be described.
In Table III figures are given for the uptake of N 15by the nucleic acid of the
liver and pancreas pellet materials. Quite the contrary to the results obtained
for uptake by the proteins of these tissues, rate of incorporation of N 16 into
ribonucleic acid is far higher in the liver than in the pancreas. Quantity of
ribonucleic acid in a pellet seems to be correlated with uptake of N 1~ by its
protein (as can be seen by comparing the data in Tables I and III) but uptake
of glycine-N 1~into the purines of the nucleic acid is not correlated with uptake
 V. ALLZFREY, MARIE M. DALY, AND A. E. MIRSKY 165

into the protein attached to it or into other proteins of the cell. These observa-
tions are similar to those made by Abrams in experiments in which he studied
the effect of x-rays on the incorporation of N15-glycine into proteins and also
into adenine and guanine of RNA and DNA of the intestinal epithelium (14).
The rates of incorporation into RNA and DNA were markedly reduced by
x-rays but incorporation into protein was relatively unaffected.
When an animal is fasted, the acinar cells of the pancreas synthesize and
store digestive ferments which are rapidly secreted after ingestion of food.
The presence of the accumulated enzymes in the fasted pancreas is recognized
by microscopic observation of zymogen granules or by enzymatic assays of the
tissue. Mter secretion both microscopic observation and chemical assay show
that the stores of enzymes are depleted. When digestion is finished the stores
of enzymes in the gland are built up again. Even in the "resting" pancreas, the
pancreas of a fasting animal, a slow, steady secretion goes on. The amount of
this varies in different animals. In the rabbit, for which precise measurements
are available, totai enzyme content of the constant secretion of the resting
pancreas is about ~ that of an actively secreting gland (15). Since the quantity
of enzymes in a resting pancreas remains fairly constant, synthesis is constantly
in progress to replenish the steady secretion that occurs in periods of rest.
Such a steady state, in which the rate of synthesis of enzymes is balanced by
their rate of secretion is well suited for tracer studies which test the relation-
ship between a precursor and a product (16). The following experiments were
designed to test the hypothesis that the protein of the pancreas pellet serves as
precursor material for the synthesis of the secretory proteins. At different
intervals after a single massive injection of N15-glycine, measurements were
made on the uptake of N ~6by the mixed protein of the tissue, the protein of
the pellet, and the protein remaining in the supernate after centrifugation at
40,000 R.r.~. The data for the time course of N ~5 uptake are given in Table
IV and plotted in Fig. 2.
A comparison of the curves for pallet protein and for mixed tissue protein
shows that the pellet fulfills two requirements of a precursor, that its N Is up-
take be higher than that of the product before the latter reaches its maximum
incorporation and that the isotope content of the precursor be lower than that
of the product after that time.
This relationship between the pellet protein and the total mixed protein of
the tissue indicates that the pellet protein probably lies on the direct pathway
of protein synthesis. It does not follow, however, that the pellet protein is
simply the next to the last step in the synthesis of an enzyme. Indeed, the latter
viewpoint is most unlikely since the pellet protein prepared in our experiments
is the mixed protein of the microsome fraction of the tissue. Recent electron
micrographs of microsomes (17) reveal a fairly definite structural entity, a
particle about 10 mtt in diameter. Our experiments indicate that these particles
166 SYNTHESIS OF PROTEIN IN PANCREAS. II

are a site of protein synthesis. H this is true one would expect that the micro-
some or pellet protein is a combination of synthetic machinery and many stages
in the synthesis of the tissue proteins. The rates of uptake of NlS-giycine by
individual proteins in the pellet m a y be expected to vary considerably and,
indeed, part of the pellet protein may be relatively inert. There are two lines of
experimental evidence which indicate that the pellet proteins differ widely in

TABLE IV
Time Course of Itu= I ~ oora~ion into Proteins of Cell Fracl~ons
N~Jexcessin cellproteins
Tissue Time
Mixedtissue Pe/letprotein Supernate
protein protein
atom per ¢¢~1 per c~n~ a/ore ~er ¢~n4
Pancreas 0.5 0.105 0.183 0.O67
1 0.278 0.440 0.255
1.5 0.373 0.532 0.545
2 0.418 0.438 0.587
3 0.362 0.339 0.416
4 0.308 0.298 0.377
5 0,278 0.254 0.378

Liver 0.5 0.059 0.107 0.049

1 0.110 0.210 0.100
1.5 0.133 0.223 0.122
2 0.173 0.259 0.157
3 0.205 0.270 0.189
4 0.186 0.243 0.176

Kidney 0.5 0.025 0.046

1 0.051 0.087
2 0.099 O. 128
3 0.110 0.162
4 0.111 0.150
Each animal received 30 nag. of N1S-glycine (33.3 atom per cent exces~ in 0.2 ml. H=O
intraperitoneally at zero time.
All animals were fasted 20 hours prior to N15-glycine administration.

their capacities for isotope incorporation. The first indication that part of the
protein of the pellet is relatively inert is given by a comparison of the "time-
N ~6 uptake" curves for supemat'e and pellet proteins. Since the supemate con-
tains about ~ of the protein nitrogen of the tissue and about ~ of its amylase
and protease, uptake into it m a y be considered to give a better indication of
what the uptake is in the secretory proteins than is given by the mixed proteins
of the whole tissue. (Experiments are now being undertaken in which a secre-
tory protein is isolated.) The curves indicate that the rate of N 15 incorporation
 V. ALLI~REY~ MARIE M. DALY, A N D A. E. MIRSKY 167

by pellet protein far exceeds that of the supemate protein up to 1 hour after
glycine administration. The maximum uptake into the pellet occurs at 1 ~
hours, that of the supernate occurs ~ hour later. However, it is significant that
the maximal concentration of isotope in the supemate exceeds that of the
pellet proteins. Although many complicating factors may exist/this would be
expected if part of the peUet were relatively inert.
There is a second and more direct test of this hypothesis; namely, the frac-
tionation of the pellet proteins. In preliminary experiments this fractionation

0.70C Pancr, e a z Live~ Kidney

- - P e l l e t p~mmein
0.~ . ps~,ein
.+..,,+5ul:mc+na+t¢
~0.~

o.+z
\ pr.o'Ceir,,.

3
~'"o
o~ 0200

0.100

2 3 4 5I 0
2 3 ~ 0 I 2 5 4 5
Tim~ in h o u r s
FIo. 2. Time course of N ~ incorporation into proteins of cell fractions of the pan-
cress, liver, and kidney. Ordinate, atom per cent N ~s excess in protein fraction.
Abscissa, time in hours after intraperitoneal injection of 30 rag. N~5-glycine (33.3
atom per cent excess) into each animal. All animals fasted 20 hours prior to glycine
administration.

has been accomplished by incubating the pellet with ribonuclease and subse-
quently centrifuging at 40,000 g.P.~. After treatment with ribonudease about
20 per cent of the total N of the pellet remains in solution; the remaining 80
per cent is sedimentable. The N 15concentrations of the different pellet fractions
are quite different. In one experiment the over-all N 1~ concentration in the
pellet proteins was 0.545 atom per cent excess. The sediment obtained after
treatment with ribonuclease had an isotope concentration of 0.405 atom per
cent excess. The non-sedimentable protein had an N ~5 concentration nearly
three times as high, 1.189 atom per cent excess. This figure is nearly twice the
maximal uptake observed in the pancreas supemate protein and further
strengthens the case for part of the pellet protein serving as precursor material
168 SYNTHESIS OJF P R O T E I N IN PANCREAS. II

for the secretory proteins of the cell. From the over-all N 15 concentration of
the pellet proteins and the Na5 contents of the pellet fractions it is possible to
calculate the relative amounts of each of the fractions. Such a calculation shows
that the non-sedimentable portion is 20 per cent of the total pellet protein; in
complete agreement with direct chemical analysis of the fractions.
The experiments just described have tested conditions in the pancreas of a
fasting animal, in which a "steady state" prevails, balancing synthesis and secre-
tion. Further evidence for a role of pellet protein in enzyme synthesis is obtained
when rates of uptake in the steady state are compared with N Is incorporations
by an actively secreting pancreas or by a pancreas in process of building up its
enzyme reserves.
The pancreas can be made to secrete by the ingestion of food, or by the injec-
tion of pilocarpine. Under these conditions the stored enzymes are secreted and
the enzyme content of the tissue is diminished. If the pancreas is no longer
stimulated to secrete, it gradually falls up with enzymes. It is important to
know whether the enzymes accumulate in the pancreas simply because secre-
tion slows down, or whether they accumulate also because synthesis is accel-
erated. Long term isotope incorporation experiments show that the total syn-
thesis in an active pancreas is, in fact, greater during a prolonged period of
synthesis and secretion than it is in a resting pancreas. In some of our experi-
ments described in a recent paper (18) mice received injections of N1S-glycine
every I ~ hours for 8 hours. One group was fed just before the experiment
began and received no more food, so that in these animals the enzyme content
of the pancreas gradually increased during the period of the experiment. A
second group was fed continuously throughout the experiment providing several
cycles of pancreatic synthesis and secretion. The mixed cytoplasmic proteins
were examined for their N 15contents 2 hours after the final injection of glycine.
N ~5 contents of liver proteins of the two groups of animals were the same, but
in the pancreas they were quite different--0.733 atom per cent excess in those
fed before the experiment and 0.330 in those fed continuously (and, it may be
added, 0.423 in steady-state animals that were fasting both before and during
the experiment). The figures show that more protein was passing through the
glands of the continuously fed mice and, therefore, that more synthesis occurred.
The duration of the experiment was the same in the two groups and it follows
that the rate of synthesis is accelerated by secretion.
The effect of secretion on N 16uptake is also evident in short term experiments.
In the resting pancreas the N ~6concentration of supernate proteins 30 minutes
after glycine injection is 0.067 atom per cent N ~5 excess. In the secreting pan-
creas of animals fed immediately before the experiment the corresponding
figure is almost 42 per cent higher, 0.093 atom per cent excess. (No difference is
found in the N ~5 contents of mixed tissue or supernate proteins in the livers of
fed and fasted animals.) Similarly the N a5 concentration in the pancreas pellet
protein of animals fed before the experiment is 0.238 atom per cent excess while
 V. ALL]~REY, MARIE M. DALY~ AND A. E. MIRSKY 169

that of the pellet from fasting animals is only 0.183. (Again N Is uptakes into
liver pellet proteins are identical in the two groups.) Thus, secretion accelerates
NlU-glycine incorporation into both pellet and supernate proteins. To the ex-
tent that such incorporation represents synthesis, secretion accelerates syn-
thesis. It should be pointed out that this difference in N Is incorporation into
the pancreas pellet proteins of fed and fasted animals is a direct demonstration
that the pellet protein is concerned, not only with its own replacement or
turnover, but with the synthesis of other proteins in the cell. The increased
synthesis of digestive enzymes which follows secretion is matched by a higher
rate of N ~6incorporation into the pellet. No similar differences are observed in
liver pellet proteins, This illustrates the advantage of dealing with a gland like
the pancreas in the study of the general problem of protein synthesis.
It is known that the administration of pilocarpine to a fasted animal results
in copious pancreatic secretion. The effect of pllocarpine injection on N 15-
glycine incorporation is currently under investigation and preliminary results
are compatible with those obtained in feeding experiments. A related problem
has to do with the inhibition of pancreatic secretion by drugs like atropine.
This, too, is being investigated using tracer techniques.
Table IV also includes data for the time course of N 16incorporation by mixed
tissue protein, supernate and pellet proteins of the liver, and for the mixed tissue
and pellet proteins of the kidney. This information is plotted in Fig. 2. In all
cases the N lu uptakes into pellet proteins are far greater than the uptakes into
supernate or mixed tissue proteins. The over-all protein synthesis in both liver
and kidney is quite small compared to that which occurs in the pancreas, and
the N ~6 concentrations observed reflect these differences. A comparison of the
time-N 1~uptake curves of liver and kidney proteins with those obtained from
pancreas proteins reveals a major difference; at no time in the period investi-
gated did the N is concentration in the liver or kidney pellet fall below the 1kus
concentration in the corresponding mixed tissue or supernate proteins. This
difference between the pattern of pancreas protein incorporations on one hand
and of liver and kidney uptakes on the other is reasonable if one considers the
comparatively small amount of protein synthesis occurring in the latter tissues,
and the likelihood that many of the proteins in the mixed tissue or supernate
fractions of liver and kidney are relatively inert. Therefore~ although the role
of the pellet in protein synthesis in liver and kidney is probably the same as it
is in the pancreas the detection of its function is not so readily accomplished.
An interesting point emerges when one measures the effect of fasting on
glycine-N~5uptake in the liver. This is shown in Table V. It can be seen that
fasting for periods up to 96 hours leads to a progressive dlrn{nution in the rate
of N lu incorporation into the mixed tissue, supernate and pellet proteins, and
also into the RNA. No such striking effects of fasting are seen in pancreas
proteins, although some decline in RNA-N 16 concentrations is observed.
The evidence given thus far points to a direct role of the pellet protein in
170 SYNTHESIS OF PROTEIN IN PANCREAS. H

the synthesis of other proteins. In a fine investigation Siekevitz has shown that
the incorporation of C14-alanine into the microsome fraction of liver represents
peptide bond formation and, furthermore, that the energy required for this
incorporation is derived from oxidative phosphorylation (19). Our work in-
dicates that this incorporation of amino acids into the microsome is a stage in
the synthesis of the other proteins of the cell. The evidence presented in this
paper is not taken to mean that all the steps in protein synthesis occur in the
microsome, but rather that one of the important stages in the process directly
involves the ribonucleoprotein of the microsome fraction.
The Role of Ribonucleic Acid in Protein Synthesis.--Siekevitz (19) has de-
scribed an in vitro system of microsomes + mitochondria which will incorporate

TABLE V
Effezt of Fa~t,ing on 2W6 Incor ~ora~ions into Proteins and R N A of Cell Fractions
Nu excess30 rain. after glycine administration
Tissue Duration
of fast Mixed Pellet Supernate
protein protein protein RNA

atom per cent atomper cent atom per cent atom per ¢etct

Liver 20 0.059 0.120 0.055 0.038

45 0.052 0.093 0.051
72 0.041 0.074 0.040 0.020
96 0.037 0.061 0.039 0.019

Pancreas 20 0.105 O. 228 0.067 0.006

45 O. 138 0.256 0.100 0.010
72 0.114 0.231 0.072 0.004
96 0.102 0.237 0.072 0.003

Each animal received 30 rag. of NlS-glycine (33.3 atom per cent excess) in 0.2 ml. H~O
intraperitoneally at zero time.

radioactive alanine, and he has presented evidence for such incorporation being
a measure of peptide bond formation. It occurred to us that such a system is
well suited to a study of the role of ribonucleic acid in protein synthesis. More
specifically it permits a direct test of the dependency of amino acid incorpora-
tion upon the presence or intactness of the nucleic acid in the system. In pre-
liminary experiments using essentially the same procedures described in detail
by Siekevitz (19) a rat liver microsome fraction was preincubated with added
crystalline ribonuclease for 15 minutes at 37 °. A control portion of the micro-
some preparation was incubated under the same conditions without added
enzyme. Following incubation these two fractions were added to sucrose
suspensions of mitochondria (which had been kept at 0 °) and the mixtures in-
cubated at 37 ° for 30 minutes in the presence of DL-alanine-l-C14. The specific
activity of the "RNAase-treated" microsome protein was only 46 per cent of
 V. ALLPREY~ MARIE M. DALY~ AND A. E. MIRSKY 171

that observed in the control preparation. I t is evident that preliminary treat-

ment with RNAase impaired the ability of the system to incorporate alanine.
A similar impairment of amino acid incorporation by tissue homogenates is
observed when they are incubated in the presence of RNAase. T h a t the in-
corporations measured do not represent an adsorption of isotopic amino acid is
evident from the dependence of the system on the presence of an oxidizable
substrate, for in the absence of a-ketoglutarate no appreciable incorporation of
amino acid is observed. T o the extent that the incorporation of C14-alanine by
the microsome protein represents protein synthesis, the above experiments
show that synthesis is dependent upon the presence of ribonudeic acid in the
microsomes.

EXPER~r~F~NTAL
Adult animals of the Rockefeller Institute colony of Swiss mice were used in these
experiments. Animals which were used for studies of the incorporation of NtLglycine
were first fasted for the periods of time indicated in the tables then injected with
glycine and killed at designated intervals after the injection. In experiments designed
to study the effect of feeding on isotope uptake the mice were allowed to feed for
30 minutes starting 20 minutes before the injection and were killed 30 minutes after
the injection.
Mice used for other studies of the pellet material were fasted only when specifi-
cally indicated.
In the isotope experiments injections of Nl~-glycine were given intraperitoneally.
The animals received 1 rag. glycine (33.3 atom per cent Nas excess) per gin. of body
weight.
Animals were killed by decapitation following ether anesthesia; the organs were
removed and placed immediately in a container chilled in ice or in an ice cold sucrose
solution.
Preparation of the Pellet Mater/aL--Samples of pancreas were homogenized for
3 ~ minutes in a glass tube fitted with a teflon pestle in I gin. portions with 7 ml. of a
solution containing sucrose 0.25 M, citrate 0.01 M, and soy bean trypsin inhibitor
0.1 mg./ml. Citrate was added to inhibit desoxyribonuclease activity, and the soy
bean inhibitor to eliminate the activity of traces of active trypsin which might be
present.
Small samples of liver tissue were homogenized in 0.25 ~ sucrose in the same manner
as the pancreas, but when larger quantities of tissue were used, it was more convenient
to homogenize the tissue in a small metal blendor run at 35 v. for 2 ~ minutes as sug-
gested by Petermann et al. (20).
Kidney samples were homogenized first in the blendor for 2 ~ minutes and then
for a few seconds in the teflon homogenizer to break up bits of tissue that were not
disintegrated in the blendor.
Microscopic examination of these homogenates showed that few of the ceils were
unbroken. For preparation of the pellet material the homogenates were centrifuged
at 5000 X g for 10 minutes, the residues were washed once with 0.25 M sucrose, and
the combined supernatants and washings were centrifuged again for 10 minutes at
172 S Y N T H E S I S OF P R O T E I N IN" P A N C R E A S . I I

5000 X g, and the final supernatant centrifuged in the Spinco model L ultracentrifuge
for 30 minutes at 40,000 a.p.m (105,400 X g). The pellet was washed by resuspension
in sucrose and centrifuged for 1 hour at 40,000 a.P.m
In the isotope incorporation experiments the pellet material was then treated with
ribonuclease in order to separate the nucleic acid from the protein. The pellet was
suspended in water with the addition of a little alkali or buffer to bring the pH to 7.5
and treated with ribonudease for 1 hour at room temperature. The pellet proteins
were precipitated with trichloroacetic acid and washed several times with trichloro-
acetic acid, hot alcohol, and ether and dried at 110°C. for 10 minutes. When the
uptake of N 16 into the RNA of the pellet material was to be measured the trichloro-
acetic acid supernatant, which contained the ribonudeotides, was saved. The nudeo-
tides were separated from the small amount of protein remaining in solution by ad-
sorption on a column of dowex-2 (chloride form). Before adsorption, the solution
was extracted with ether to remove the trichloroacetic acid and adjusted to pH 9
with alkali. The solution was passed through the column, the column washed with
water and sodium acetate 0.01 N at pH 6, and then nucleotides were duted with 1 N
HC1.
The isotope incorporation into the mixed proteins of the tissue and into the pro-
reins of the supernatant obtained after the first centrifuging at 40,000 R.P.m was also
measured. Samples of the homogenate and supernatants were precipitated with
trichloroacetic acid, washed with trichloroacetic acid, alcohol, and ether, and dried
at 110°C. for 10 minutes. The samples of supematant proteins from the experiments
described in Table IV, however, were prepared in a slightly different manner. Before
precipitation of the proteins with trichloroacetic acid the supernatants were brought
to pH 5.4 by the addition of a little acetic acid in order to remove small amounts of
nucleoprotein which did not sediment at 40,000 a.P.g.
2g~6Analyses.--The N t5 concentrations of the proteins and l~q'A were determined
in the mass spectrometer (Process and Instruments Company model) after the usual
conversion of organic nitrogen to gaseous N2 by Kjeldahl digestion and treatment with
hypobromite. N 15 concentrations of samples were determined to within 4-2 per cent
(average deviation) and compared with tank nitrogen as standard.
Ct~emic~l Analyse~.--Nitrogen, phosphorus, and RlgA contents of the peUet mate-
rials were determined on samples prepared as described above except that after the
final centrifugation at 40,000 a.P.M, the pellets were washed immediately with hot
95 per cent alcohol several times, then with ether, and finally dried at 110°C. for 10
minutes.
(a) Nitrogen and P~sp~rus.--After Kjeldahl digestion of the samples the nitro-
gen content was determined by nesslerization and the phosphorus by Allen's method
(21).
(b) RgVA.--Samples of pellet material were washed once with cold 2 per cent per-
chloric acid and then heated at 70°C. for 20 minutes to extract the nucleic acid.
I ~ A was estimated by the orcinol reaction, or by the ultraviolet extinction coet~cient
of the perchloric acid extract. For the pancreas pellet material values for RNA con-
tent obtained in this way are in good agreement with the total phosphorus. In the
liver pellet preparations there was a certain amount of DNA P (10 per cent of the
RNA P): therefore, the orcinol figures for liver pellet are reported in Table III.
 V. ALLFRE¥~ MARIE M. DALY~ AND A. E. MIRSKY 173

Amino Acid Composition.--The pellet materials were treated with ribonuclease,

and then precipitated and washed with trichloroacetic acid, alcohol, and ether. The
proteins were hydrolyzed by refluxing with 6 N HC1 for 18 hours, and the basic amino
acids determined by the ion exchange procedure of Moore and Stein (22).
Zone Ele,~rophoresis.--The pellet material prepared from mouse liver was resus-
pended in 0.25 M sucrose to give a final nitrogen concentration of 5 mg./ml. To 4 ml.
of this suspension was added 1 ml. of 1 u phosphate buffer at p H 7.0. The mixture
was carefully poured into a small rectangular well 30 cm. from one end of a moist
glass powder slab 70 cm. long, 8 on. wide, and 1 on. deep. The preparation of this
glass powder and the technique of applying a potential across it are described by
Taylor, du Vigneaud, and Kunkel (23). In the experiment pictured in Fig. 1 A the
glass powder was suspended in a mixture of 9 parts 0.25 u sucrose and 1 part 1 u
phosphate buffer at p H 7.0. The same mixture was used in the electrode vessels. The
potential applied across the electrodes was 320 v.--the current passing in the system
was 10 milliamperes.I)uration of the experiment was 100 hours. In the experiment
shown in Fig. 1 B the glass powder was suspended in 0.2 # phosphate buffer at p H 7.0.
The applied potential was 320 v.--the current was 12 milliamperes. Duration of the
experiment was 100 hours. For purposes of comparison an electrophoretic pattern
is shown for a solution of yeast ribonucleic acid (5 mg./ml.). This was a sample of
commercial yeast RNA which was purified by acid precipitation and subsequent dialy-
sis. In this case 5 ml. of RNA solution was placed on a starch slab, 70 on. long, 10
cm. wide, and I on. deep. The applied potential was 320 v.--the current passing in the
system was 11 mRliamperes. Duration of the experiment was 100 hours.
Following electrophoresis, the slabs were cut at 2 cm. intervals to give 35 zones 2
on. across by 1 cm. deep by 8 cm. wide. The glass powder in each zone was suspended
in 8 ml. cold HsO and the suspension was filtered through a coarse sintcred glass filter
to remove the glass powder. The filtrate was then analyzed in one of two ways: (a)
When the filtrate was clear, extinctions at 258 n ~ and at 230 m~ were determined
in the Beckman spectrophotometer. These were used as an indication of nucleic acid
and protein distribution. (b) When the filtrate was turbid (this was always the case
for zones near the origin), nitrogen analyses were performed. In order to maintain the
single parameter, E~0, throughout the plot of the electrophoresis pattern, the nitro-
gen content thus determined was converted to F-~0 in the following way. The ratio
F_~0/N was determined for zones adjacent to the turbid areas but which gave clear
filtrates. This ratio was then used to compute the equivalent E~30 for turbid zones.
The values plotted near the origins of the curves in Fig. 1 were obtained in this way.
The values of ~ for turbid solutions were obtained by heating in 5 per cent tri-
chloroacetic acid for 15 minutes at 90°C. to remove protein and release the nucleic
acid. After centrifugation E~m was determined on the clear supernate. These values
were then corrected for an increase in the Esm mused by a reaction between sucrose
and hot 5 per cent trichloroacetic acid.
SUMMARY
1. T h e ribonucleoprotein of the microsome fraction which sediments a t
40,000 R.P.M. as a pellet (and which is referred to as the pellet n~z~riaZ) has
been studied with reference to its role in protein synthesis in the pancreas.
174 SYNTHESIS OF PROTEIN IN PANCREAS. II

2. In pellet material nucleic acid and protein form a definite complex as

shown by its electrophoretic behavior and unchanging composition under
various conditions.
3. Protein of pellet material is not especially rich in the diamino acids.
4. Evidence is brought forward indicating that the protein component of
pallet material takes part in the general process of protein synthesis in the cell.
(a) The well known correlation between quantity of RlgA and rate of pro-
tein synthesis in a tissue implicates the protein of the pellet material, for most
of the RNA in the pancreas and other tissues is in this material.
(b) Uptake of isotopically labelled glycine by the pellet material, confirming
results of previous workers, is for short periods greater than in other protein
fractions.
(c) Comparing the pellet materials of pancreas, liver, and kidney--three
tissues with vastly different rates of protein synthesis, in the sequence given--
there is a correlation between the quantity of RNA in the pellet and the rate
of protein synthesis in the tissue; a similar correlation between quantity of
RNA in the pellet material and rate of N15-glyeine uptake by the protein com-
ponent of the pellet; and finally, the level of uptake by total protein varies with
the tissue and is related to the uptake of N16-glycine by protein of the pellet.
5. In the pancreas a distinction can be made between proteins synthesized
for secretion and the nucleoprotein of the pellet (not found in the secretion)
which, however, takes part in the synthetic process, as shown by the fact that
the N 16 uptake by protein of the pellet is increased when the synthesis of di-
gestive enzymes is stimulated by secretion.
6. The time course of N 16uptake by proteins of the pancreas indicates that
pellet protein serves as precursor material in the synthesis of the secretory
proteins.
7. Rate of uptake of Na-glycine by the purines of RNA of the pellet material
is not correlated with uptake by the protein.
8. The uptake of CX4-alanine by an in dtro system of microsomes q- mito-
chondria is impaired by preincubation of the microsomes with ribonuclease.
This is direct experimental evidence for the dependence of protein synthesis
upon the presence or intactness of ribonucleic acid in the microsomes.
REFERENCES
1. Borsook, H., Deasy, C. L., Haagen-Smit, A. J., Keighley, G., and Lowy, P. H.,
I. Biol. Clm~., 1950, 187, 839.
2. Hultin, T., Exp. Cell Research, 1950, 1, 376.
3. Keller, E. B., Fed. Proc., 1951, 10, 206.
4. Miller, L. L., Bly, C. G., Watson, M. L., and Bale, W. F., J. Exp. Med., 1951,
94, 431.
5. Daly, M. M., and Mirsky, A. E., J. Gen. Physiol., 1952, 36, 243.
 v. ALLFREY~MARIE M. DALY, AND A. E. MIRSKY 175

6. Rabinovitch, M., Valeri, V., Rothschild, H. A., Camara, S., Sesso, A., and Jun-
queira, L. C. U., ]. Biol. Chem., 1952, 198, 815.
7. NovikotI, A. B., Podber, E., Ryan, J., and Noe, E., .?'. Histochon. and Cytochem.,
1953, 1, 27.
8. Brachet, J., Embryologie Chimique, Editions Desoer, Paris, Masson et Cie, 1947.
9. Caspersson, T., Cell Growth and Cell Function; A Cytochemical Study, New
York, W. W. Norton, 1950.
10. Caspersson, T., LandstrSm-Hyd6n, H., and Aquilonius, L., Chromosoma, 1941,
2, 111.
11. Mirsky, A. E., and Pollister, A. W., .l. Gen. Physiol., 1945, 30, 117.
12. Daniel]i, J. F., Cytochemistry, A Critical Approach, New York, John Wiley
and Sons, Inc., 1953.
13. Bawden, F. C., Plant Viruses and Plant Diseases, Waltham, Massachusetts,
Chronica Botanica Co., 3rd edition, 1950, 185.
14. Abrams, R., Arch. Biochem., 1951, 30, 90.
15. Baxter, S. G., Am. ]. Physiol., 1931, 96, 349.
16. Zilversmit, D. B., Entenman, C., and Fishler, M. C., J. Gen. Physiol., 1943, 26,
325.
17. Smellie, R. M. S., MeIndoe, W. M., Logan, R., and Davidson, J. N., Bioche~.
J., 1953, 54, 280.
18. Daly, M. M., AUfrey, V. G., and Mirsky, A. E., J. Gen. Physiol., 1952, 36, 173.
19. Siekevitz, P., J. Biol. Chem., 1952, 195, 549.
20. Petermann, M. L., Mizen, N. A., and Hamilton, M. G., Fed. Proc., 1953, 12, 254.
21. Allen, R. J. L., Biochem. J., 1940, 34, 858.
22. Moore, S., and Stein, W. H., Jr. Biol. Chem., 1951, 192, 663.
23. Taylor, S. P., du Vigneaud, V., and Kunkel, H. G., J. Biol. Chem., in press.
24. Caldwell, P. C., Mackor, E. L., and Hinshelwood, C., J. Chem. Soc., 1950, 3151.
25. Priee, W. H., J. Gen. Physiol., 1952, 35, 741.
26. Northrop, J. H., J. Gen. Physiol., 1952, 36, 581.