-
Proc. Natl Acad. Sci. USAVol. 80, pp. 2179-2183, April
1983Biochemistry
Quantitative correlation between proteolysis and macro-
andmicroautophagy in mouse hepatocytes during starvationand
refeeding
(liver/lysosomes/electron microscopy/protein
synthesis/cytoplasmic growth)
GLENN E. MORTIMORE, NANCY J. HUTSON*, AND CYNTHIA A.
SURMACZDepartment of Physiology, The Milton S. Hershey Medical
Center, The Pennsylvania State University, Hershey, Pennsylvania
17033
Communicated by Charles R. Park, January 4, 1983
ABSTRACT Cytoplasmic protein in hepatocytes is seques-tered and
degraded by two general classes of lysosomes, overt au-tophagic
vacuoles (macroautophagy) and dense bodies (microau-tophagy).
Volumes of the apparent space in each class that containthe
internalized protein, together with estimates of cytoplasmicprotein
concentration, were used as a basis for predicting rates ofprotein
degradation by the lysosomal system in livers of fed, 48-hr
starved, and starved-refed mice. Assuming that the turnoverof all
sequestered protein is equal to that previously determinedin overt
autophagic vacuoles (0.087 min'), we obtained closeagreement
between predicted and observed rates in the threeconditions
studied. The two autophagic components, though, ex-hibited
different patterns of regulation. Microautophagy followeda downward
course through starvation and into refeeding, a trendthat explained
fully the fall in absolute rates of protein degra-dation during
starvation. By contrast, macroautophagy remainedconstant throughout
starvation but was virtually abolished withrefeeding. Whereas
regulation of the latter can be explained largelyby immediate
responses to the supply of amino acids, present evi-dence together
with results of others indicate that microseques-tration could be
linked to functional and quantitative alterationsin the smooth
endoplasmic reticulum. Both types of regulationcontributed equally
to the marked suppression of proteolysis dur-ing cytoplasmic
regrowth.
Protein and other cytoplasmic constituents in hepatocytes
areinternalized within the lysosomal system in two general ways:(i)
by the formation of overt autophagic vacuoles, a highly vis-ible
mechanism involving the sequestration of organelles
withinsmooth-surfaced membranes (1-5), and (ii) by the engulfmentof
small bits of cytoplasm that subsequently appear in densebodies (1,
3-5), a microautophagic process that we believe ac-counts for basal
protein turnover (5). Although both are ongoingfunctions, we have
shown in perfused livers of fed animals thatthe addition or
deletion of amino acids can regulate overt au-tophagy rapidly over
a wide range without appreciably affectingmicroautophagy (3, 5). On
the other hand, the induction of cy-toplasmic growth, such as that
produced by starvation and re-feeding, is associated with decreases
in rates of protein break-down (6, 7) and sequestered protein (7)
to levels that can beexplained only by the suppression of both
autophagic functions.Nothing is presently known, however, of the
way microinter-nalization is regulated.The present investigation is
an attempt to quantify by stereo-
logic methods the contributions of each process to
proteinbreakdown in livers of fed, starved, and starved-refed mice
andto compare degradation rates predicted by this approach
withexperimentally determined rates. Close agreement was found
under all conditions, although activities of the individual
au-tophagic components were affected in distinctly independentways.
Rates of overt autophagy were interpretable from thegeneral state
of amino acid deprivation, whereas changes in mi-croautophagy
appeared to relate to intrinsic alterations withinthe cell.
EXPERIMENTAL PROCEDURESAnimals. Male, 7-week-old CD-1 mice
(Charles River Breed-
ing Laboratories), initially weighing 33-35 g, were used in
allexperiments. Except during periods of food withdrawal, theyhad
free access to water and Charles River chow and were main-tained in
an environmentally controlled room (light on 0700 to1900).
Starvation was induced by removing food at 0900, a pointthat served
as a timing reference for all subsequent procedures.To minimize
variability, large numbers of animals were em-ployed for the
determination of changes in liver protein contentduring starvation
and refeeding. At intervals, groups of micewere killed by
decapitation and the livers were removed for thedetermination of
protein (7, 8). Liver volumes were measuredby saline displacement
and densities were computed from theweights. Livers were also fixed
by vascular perfusion and thinsections were prepared for electron
microscopy as described (3,5, 9); standard stereologic procedures
were employed for mea-suring volume densities of lysosomal
components (3, 5, 9).
Determination of Hepatic Protein Synthesis and Degra-dation.
Rates of protein synthesis in liver were measured in sep-arate
groups of mice from the incorporation of L-[U-'4C]valine(295
mCi/mmol; 1 Ci = 3.7 x 1010 Bq; New England Nuclear)during 2-hr
cyclic perfusions. The details of this method as ameans for
determining resident protein synthesis in vivo havebeen fully
described and validated in a recent paper (7). Ratesof resident
protein breakdown were calculated by subtractingthe average rates
of change of total liver protein from corre-sponding rates of
synthesis (7).
Liver Volumes. Liver parenchymal volume and its majorfractions
were determined in order to provide a basis for com-puting absolute
lysosomal volumes (Table 2) and concentrationsof hepatocyte protein
(Table 3). The results and description ofthe methods used, except
those for extracellular volume, aregiven in Table 1 and the
preceding text. For convenience ofexplanation, extracellular volume
may be divided into threecomponents: (i) a nonvascular fraction,
composing 0.05 of totalliver volume (10), 92% of which is
continuous with the plasmacompartment (11); (ii) a residual blood
space of 0.03 (10); and
Abbreviations: ER, endoplasmic reticulum; DB, dense bodies; AV,
au-tophagic vacuoles.* Present address: Central Research, Pfizer
Inc., Eastern Point Rd.,Groton, CT 06340.
2179
The publication costs of this article were defrayed in part by
page chargepayment. This article must therefore be hereby marked
"advertisement"in accordance with 18 U. S. C. §1734 solely to
indicate this fact.
Dow
nloa
ded
by g
uest
on
June
24,
202
1
-
2180 Biochemistry: Mortimore et al.
(iii) an additional quantity of blood that is retained when
ani-mals, as in this study, are not exhaustively exsanguinated
(10).Williams and Woodbury have shown that the chloride space
ofthis extra blood is directly related to the difference between
theobserved extracellular chloride space and a basal value of
0.20,determined in exsanguinated rats (10). With their method,
weobtained differences (mean ± SEM) of 0.104 ± 0.009 and 0.102±
0.014 in seven and nine fed and 48-hr starved mice, re-spectively.
Results were higher in 24-hr refed mice (0.182 +0.024, n = 8), but
the distribution of [14C]inulin remained un-changed. Because
intracellular chloride might have increasedduring rapid cytoplasmic
regrowth, we used the average dif-ference in fed and starved
animals (0.103) for all groups. Thevolume of this extra space was
computed from the expression
[ (1 - 0.54) 1A (PCV)- + 1],
in which A is the above chloride space difference and 0.54 isthe
nominal chloride space of erythrocytes (12). Average he-matocrits
(PCV, packed cell volume) for the four groups in theorder listed in
Table 1 (left to right) were 0.365, 0.466, 0.425,and 0.386. Values
of total extracellular space in line E of Table1 are the sums of i,
ii, and iii above, expressed as fractions ofliver parenchymal
volume.
In tissue fixed for electron microscopy, the fractional
distri-bution of cellular components would be slightly increased
ow-ing to the washout of sinusoidal blood and the consequent
re-duction in extracellular volume. Values of hepatocyte
cytoplasmin Table 1 (line C, parentheses) were computed from a
reportedextracellular space of 0.159 (11).
Presentation of Data. Values expressed on the basis of
totalliver were normalized to an initial body weight of 34 g.
Wherepossible, results are presented as means ± SEM with numbersof
observations in parentheses.
Table 1. Volume of liver parenchyma and its distribution in
fed,starved, and starved-refed mice
Starved Refed RefedVolume Fed 48 hr 12 hr 24 hr
Total volume, ml/34-g mouseLiver 1.670 1.039 1.843
1.771Parenchyma 1.577 0.946 1.750 1.678
Fraction of parenchymal volumeA. Hepatocytes 0.727 0.679 0.727
0.729B. Hepatocyte nuclei 0.071 0.071 0.071 0.071C. Hepatocyte
cytoplasm 0.656 0.608 0.656 0.658
(0.715) (0.678) (0.715) (0.718)D. Sinusoidal cells 0.045 0.075
0.041 0.042E. Extracellular 0.228 0.246 0.232 0.229
Liver volumes for the groups above (left to right) were
calculated bydividing weights in Fig. 1 by separately measured
densities, which were,respectively: 1.126 ± 0.007, 1.068 ±
0.024,1.145 ± 0.009, and 1.135 +0.019 g/ml (n = 6-8). Liver
parenchyma (liver minus connective tis-sue, large vessels, etc.)
makes up 92.4% of liver volume in 24-hr starvedrats (11). Because
-27% of liver volume in mice and rats is normallylost in 24 hr of
starvation, the volume of supporting tissue would beabout 0.093
ml/34-g mouse, a value assumed to remain fixed duringstarvation and
refeeding. The volume of sinusoidal cells was also as-sumed to
remain constant and, from Blouin et al. (11), was computedto be
0.086 ml/34-g mouse. Hepatocyte nuclear volume has been shownto
decrease in proportion to liver volume during starvation (13); the
dis-tribution below was based on data from ref. 11. Values in lines
A, D, andE total 1.0; A and C were determined by difference. The
distributionsin parentheses (line C) were computed for fixed tissue
(see text).
RESULTS AND DISCUSSIONAlterations in Liver Protein Content and
in Rates of Protein
Synthesis and Degradation During Starvation and Refeeding.The
results in Fig. 1 are in general agreement with our
earliermeasurements of responses to starvation and refeeding in
themouse (7). A closer approximation in this study of early
changesin liver protein indicates that the initial rate ofdepletion
is greaterthan that at 48 hr, an observation that may relate to the
morerapid decline in liver weight in the first 24 hr (7). The 36%
over-all loss of liver protein during 48 hr of starvation was
almostcompletely regained in the following 24-hr period of food
in-take. As emphasized in previous investigations (6, 7),
proteinregrowth was dominated by a dramatic down-regulation of
pro-tein degradation to rates approximately 10% of fed control
val-ues (7). While total protein synthesis per liver returned to
pre-starvation levels, it is evident that the observed suppression
ofproteolysis would have caused some cytoplasmic growth evenif
synthesis had not changed.
Volumes of Lysosoma-Vacuolar Components. In earlierstudies (5,
14), we found close agreement between predictedand observed values
of cytoplasmic protein internalized by ly-
z
H0
cr-w
-J-J4H
z
4
wCD
C/)
wI-
z
0-
HOURS
FIG. 1. Alterations in the content of liver protein (A) and in
ratesof protein synthesis and degradation (B) during starvation and
re-feeding in the mouse. Net rates of liver protein loss at 0 and
48 hr werecalculatedby least-squares regression from the first
three and last fourpoints; rates of gain at 12 and 24 hr of
refeeding represent the meanrate between 60 and 72 hr. Rates of
resident protein synthesis, deter-mined as previously detailed from
the 2-hr incorporation of L-[U-14C]valine during cyclic perfusion
(7), were 3.73 ± 0.26, 3.96 ± 0.20,3.56 ± 0.21, and 3.91 ± 0.16 mg
hr-1/g of liver, respectively, in fed,48-hr starved, 12-hr refed,
and 24-hr refed mice (four per group). Ratesper liver were obtained
by multiplying these values by the followingaverage liver weights
(g) for the same groups: 1.88 ± 0.03, 1.11 ± 0.03,2.11 ± 0.07,
and2.01 + 0.04 (8-12pergroup). Proteindegradationthenwas computed
as the difference between synthesis and net rates of pro-tein
change.
Proc. Natl. Acad. Sci. USA 80 (1983)
Dow
nloa
ded
by g
uest
on
June
24,
202
1
-
Proc. Natl. Acad. Sci. USA 80 (1983) 2181
-sosomes in perfused livers of fed rats when we used the
ag-gregate volumes of AVd (degradative autophagic vacuoles) +total
DB (dense bodies) (for description, see legend to Table 2)in the
calculations. Because these were acute studies in whichovert
autophagy but not the total volume ofDB was altered, wecould draw
few conclusions concerning the nature of microau-tophagy or its
regulation. Two relevant points, however, didemerge from
experiments in which macroautophagy was vir-tually eliminated by
amino acid additions (5, 14). First, the per-sistence of type A DB
(for description, see legend to Table 2)clearly suggested that
microautophagy is an ongoing functioneven when macroautophagy is
suppressed. Second, the appar-ent space of sequestered protein was
shown to approximate thetotal volume of DB. Although the observed
volume-of type ADB constitutes =30% of the total DB in livers of
fed animals(3, 5, 9), it must be recognized that this percentage
greatlyunderestimates their actual volume. This is so mainly
becausethe electron-lucent zone that, marks these particles is
consid-erably smaller than the surrounding electron-dense region
(seebelow) and would not be visualized in the majority of
randomsections. But how the sequestered protein is distributed
withinthe population of DB or a given component of it is not at
allclear. Because space would be required for products of
diges-tion and intrinsic lysosomal proteins, it follows from the
secondpoint above that the concentration of internalized protein
mustbe higher than that in cytoplasm. This is not an
unreasonablepossibility, because in the process of formation of AVi
(initialautophagic vacuoles), cytoplasm is believed to lose as much
ashalf of its water content immediately after sequestration
(5).
In contrast to the above amino acid effects, which were
largelyrestricted to macroautophagy (5, 14), longer-termed
metabolicalterations induced changes in both micro- and
macroauto-phagic components. As shown in Table 2, absolute volumes
oftotal DB decreased progressively during starvation and
refeed-ing. In relative terms, the volume of type A DB fell to a
greaterextent than did total DB, decreasing 79% as compared with
49%for the total by 12 hr of refeeding. Similarly, the volume
ratioof type A to R fell by nearly 50% after 48 hr of starvation,
ad-vancing to 67% in the initial periods of refeeding. This
apparentshift in the distribution of the two classes of particles
suggests
that microautophagy was selectively reduced. We consideredthe
possibility that only the electron-lucent zones had becomesmaller.
However, no differences were found in zone-to-par-ticle volume
ratios, which averaged 0.14, 0.13, and 0.14, re-spectively, in the
fed, 48-hr starved, and starved-refed (12 +24 hr) groups. From
these considerations as well as the generalappearance of the
particles, we conclude that the overall struc-ture of type A
lysosomes was not affected.
Macroautophagy followed a distinctly different course (Table2).
In general agreemenrwith others (15, 16), absolute volumesof AVd
remained comparatively constant during starvation butdecreased by
more than 90% after refeeding. We should per-haps point out that
the doubling of the fractional volume of totalAV by 48 hr of
starvation noted in Table 2 was not an intrinsiclysosomal
alteration, but rather the consequence of a reductionin cytoplasmic
volume.
Uptake and Degradation of Cytoplasmic Protein by Lyso-somes. In
predicting the quantities of protein sequestered bydense bodies
during starvation and refeeding, we consideredtwo possibilities for
computing the hypothetical protein spaceand its decrease. The first
was the observed volume of DB it-self. The second, designated DBp,
was the volume of DB in fedanimals, multiplied by the relative
volume of type A particlesin each state. We reasoned that if type A
lysosomes are in factmarkers for microautophagy, the apparent space
should de-crease in proportion to the fall in type A particles seen
in Table2. Calculated values for DBp are given in the legend to
Fig. 2.They differ from absolute volumes of DB in Table 2 only
instarved and starved-refed states, in which the numbers reflectthe
shift in DB distribution from type A to type R.
Quantitative predictions of internalized protein, based
onabsolute volumes of DBp + AVd and averages of minimal andmaximal
estimates of cytoplasmic protein concentration in Ta-ble 3, were in
fact remarkably similar to experimentally de-termined values of
degradable intralysosomal protein in liverhomogenates reported
earlier (7). The latter results, measuredin control fed and 48-hr
starved then 12-hr refed mice (7) andcorrected for a known
underestimation of -10% (14), were 1.89and 0.24 mg of protein per
liver (34-g mouse). Predicted values,computed from DBp + AVd as
described above, were 2.01 and
Table 2. Lysosomal volumes in hepatocytes of fed, starved, and
starved-refed mice
Volumes Fed Starved 48 hr Refed 12 hr Refed 24 hrFractional
volumes, ,Al/ml of cytoplasm
AVi + AVd 1.55 ± 0.58 3.01 ± 0.62 0.10 ± 0.09*11 0.13 ±
0.07*¶Type A DB 1.18 ± 0.26 0.98 ± 0.27 0.22 ± 0.03tt 0.30 ±
0.15*Type R DB 3.18 ± 0.99 4.94 ± 0.82 1.80 ± 0.35§ 1.38 ±
0.43§
Absolute volumes, /ul/liver of 34-g mouseAVd 0.839 (100) 0.927
(10) 0.060 (7) 0.075 (9)Total DB 4.916 (100) 3.797 (77) 2.528 (51)
2.024 (41)Type A DB 1.331 (100) 0.629 (47) 0.275 (21) 0.361
(27)Type R DB 3.585 (100) 3.168 (88) 2.253 (63) 1.663 (46)
AVi and AVd denote initial (sequestrational) and degradative
forms of overt or macroautophagy (5, 9),with AVd making up -48% of
AVi + AVd over the full range of regulation (5). After a lag of 7-8
min,AVi are transformed into AVd, at which time intravacuolar
digestion begins (5). DB consist of all re-maining lysosomes (1-5).
On the average they are smaller than AVd and can readily be divided
into twogroups according to whether the profiles contain a sharply
demarcated, electron-lucent zone (type A) ornot (type R). Because
these zones are usually filled with glycogen or other granular
material, togetherwith occasional membrane remnants (3, 5, 9), type
A DB may be considered autophagic. Some may containendocytosed
material, but the effect on volume measurements is probably
negligible. Fractional cyto-plasmic volumes were converted to
absolute volumes by multiplying the values by fractional
parenchymalvolumes of hepatocyte cytoplasm in livers fixed for
electron microscopy (Experimental Procedures) andby corresponding
parenchymal volumes (Table 1). Each group consisted of three to
five livers. Numbersin parentheses are percentages of fed control
values. Probabilities that differences could have arisen bychance
are as follows. Versus fed: *, P < 0.05; t, P < 0.01; versus
starved: t, P < 0.05; §, P < 0.02; ¶, P
-
2182 Biochemistry: Mortimore et al.
Table 3. Distribution of protein in livers of fed, starved,
andstarved-refed mice
Starved Refed RefedProtein Fed 48 hr 12 hr 24 hr
Liver protein, mg/ml of parenchymaA. Total 274.6 291.8 190.3
244.4B. Hepatocyte 240.7 248.5 152.9 209.5
Hepatocyte protein, mg/mlC. Per cell volume 331.1 366.0 210.3
287.4D. Per cytoplasmic
volume 366.9 408.7 233.1 318.4
Results in line A were determined by multiplying observed liver
pro-tein by the ratios of mean liver weight to parenchymal volume
(Fig. 1and Table 1). Protein concentrations (mg/g) in the four
groups were(left to right) 230.3 ± 6.5, 248.7 ± 4.0, 157.8 ± 2.7,
and 204.0 ± 3.1 (8-12 livers per group). Line B.then was obtained
by subtracting from Athe following sources of nonhepatocyte
protein: (i) sinusoidal cell pro-tein, estimated as -45 mg/ml of
parenchyma (17, 18), and (ii) extra-cellular protein, computed from
the content of erythrocytes and com-bined plasma and nonvascular
spaces (see Experimental Procedures).Erythrocyte protein was taken
as 336 mg/ml of packed cells [280 mgof hemoglobin (personal
communication from Richard Lee, Charles RiverBreeding
Laboratories), plus a 20% allowance for other proteins]; pro-tein
in the noncellular space was assumed to be similar to that of
plasma,64 mg/ml. Total nonhepatocyte protein then was calculated
from theparenchymal distributions of these components. Lines C and
D rep-resent two estimates of hepatocyte protein concentration made
by di-viding line B by appropriate distributions in lines A or C of
Table 1. Cis minimal and assumes a uniform distribution of protein
in cell water.However, because the nucleus is very likely lower in
protein then cy-toplasm, a maximal estimate D was computed on the
basis that proteinwas distributed only in cytoplasmic water;
0.24 mg, respectively. By contrast, the use of total DB as
anestimator of the protein space gave a value for the
starved-refedrat livers that was 0.59 mg-2.5 times higher than the
ob-served.
Having tested the predictability of internalized protein,
weextended. our calculations to proteolytic rates. For this we
as-sumed that the turnover constant of the sequestered protein
isthe same for microautophagic components as we have dem-onstrated
in macroautophagy, 0.087 min- (5). Several lines ofevidence support
this assertion (5,. 7, 14), but the major ar-gument rests on the
direct proportionality that has been dem-onstrated between
degradable intralysosomal protein and ratesof protein degradation
over 95% of the proteolytic range frommaximum to slightly above
zero (7). While this relationship doesnot, of itself, exclude other
possible proteolytic mechanisms(14), it does strongly indicate that
the velocity of degradationof internalized protein is a linear
function of the quantity pres-ent (7, 14).
Fig. 2 shows that relative rates of resident protein
degra-dation predicted from AVd + DBp and the above rate
constantwere virtually identical to observed values up to 60 hr.
whereaspredictions based on AVd + total DB deviated from them
ap-preciably. The apparent discrepancy noted at the 72-hr pointwill
be discussed below. Relative agreement was not affectedby more than
±4% when maximal and minimal rather than av-erage estimates of
cytoplasmic protein were employed in thecalculations. As shown in
the legend to Fig. 2, the predictedabsolute rate of protein
degradation in fed animals averaged92% of the observed (88%
minimal, 96% maximal).
Regulation of Microautophagy. Whereas overt autophagy issubject
to direct regulation by amino acids, insulin, and glu-cagon (19),
microautophagy is not (5, 20). What emerged un-expectedly then was
the pronounced decline in this processduring starvation and
refeeding (Table 2). The fact that overt
10O0
Z 800
c< 60cubcDwaZ 40w
I-L 20- 0
0 12 24 36 48 60 72HOURS
FIG. 2. Correlation between predicted and observed rates of
he-patic protein degradation during starvation and refeeding.
Observedrates (e), expressed as. a percentage of fed control
values, were takenfrom Fig. 1; predicted rates, similarly
expressed, were computed fromthe absolute volumes of AVd + DBp (0)
orAVd + totalDB (n) and mul-tiplied by (i) the average cytoplasmic
protein concentration in hepa-tocytes (mean of lines C and D, Table
3) and (ii) the assumed first-orderturnoverconstant of internalized
protein, 5.22 hr-' (5). Volumes of to-tal DB are given in Table 2;
volumes of DBp for fed, 48-hr starved, and12- and 24-hr
starved-refed mice were calculated to be 4.916, 2.319, 1.016,and
1.334 1.l per liver (34-g mouse), respectively (see text). The
ob-served fed control rate (100%) was 11.4 mg hr-1/34-g mouse;
corre-sponding predicted rates were 11.0 (mximum), 10.5 (average),
and 10.0(minimum). The prediction shown by A is discussed in the
text.
autophagy continued at postabsorptive levels in the absence
offood intake is in agreement with observations of Pfeifer (13)
andcan be attributed in large part to the maintenance of
relativelynormal plasma amino acid concentrations (21). Its
sharpsuppression with refeeding would be the expected result of
anincrease in amino acid and insulin levels during intestinal
ab-sorption (5, 21). But reasons for the decrease in
microautophagyare not as apparent. Novikoff and Shin (4) have
called attentionto the probable role of smooth ER in isolating
small bits of cy-toplasm, and we have encountered type A lysosomes
most fre-quently in areas filled with dilated smooth ER. Because
thequantity of smooth ER per cell falls appreciably during
star-vation (22), one might expect to see a progressive decline
inmicroautophagic activity with starvation if in fact there is a
con-nection between smooth ER and this process.
By 12 hr of refeeding, hepatocyte volume was fully restored,and
large areas of cytoplasm contained newly synthesized gly-cogen. The
most striking difference between starved-refed an-imals and the
other groups was a sharp change in the distri-bution of type A
particles. They were not found at the edge ofthe zones of glycogen
as we typically see them, but instead werelocated almost
exclusively in areas that had the general ap-pearance of cytoplasm
in starved animals-i.e., glycogen wassparse and the smooth ER was
vesiculated (Fig. 3). Internalizedglycogen was only occasionally
noted in type A bodies, but sev-eral images contained ribosomes
(Fig. 3).The apparent partition of microsequestrational activity
dur-
ing cytoplasmic regain has two significant implications.
First,if the only portions of cytoplasm susceptible of being
inter-nalized were those that were carried over from starvation,
itcould mean that we had overestimated predictions of
proteindegradation during refeeding. When rates at 60 and 72 hr
wererecalculated on the basis of the protein content in 48-hr
starvedlivers (Fig. 2, ), the original discrepancy at 72 hr between
ob-
STARVATRON REFEEDING
0 I~~~~~~-
v \ \~~~~~~~~~~~~~~~~~~~~
A~ ~ I ,
Proc. Natl. Acad. Sci. USA 80 (1983)
tr
Dow
nloa
ded
by g
uest
on
June
24,
202
1
-
Proc. Natl. Acad. Sci. USA 80 (1983) 2183
K..|2'.~^~~~~~~~ g~ ~~bcqW,' '-A
~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~V
' .....
FIG. 3. Early type A dense bodies in hepatocytes of 48-hr
starved,12-hr refed mouse livers. (a) View at low magnification (x
10,000) oftwo type RDB (lines) and one early typeA DB (arrowhead),
also shownin the Inset (x20,000). The granular material within the
vacuole ap-pears to have been sequestered in situ. Residual clumps
of glycogen, gl,which were intentionally unstained, show as clear
areas; bc, bile can-aliculus. (b) Type A DB with vacuole containing
ribosomes in a bed ofgranular material. The enclosing double
membrane appears to havebeen formed from smooth ER and may be in
the process of fusing (ar-rowhead). (x 38,000.) (c and d) Sections
at two depths of an unusual DBprofile suggesting internalization of
ribosomes and cytoplasmic matrixin situ. The flap-like membrane
extension could represent smooth ER,an invagination of the
lysosomal membrane (23), or both. Note the densematerial between
the membranes at the tip of the extension. d also showsnumerous
dilated tubules of smooth ER. (x30,600.)
served and calculated rates disappeared; little effect was
notedat 60 hr, where agreement was already close. Second, it
sug-gests that a large fraction of the total suppression of
proteolysisduring cytoplasmic regrowth might be the consequence of
growthitself. Resynthesis of RNA, glycogen, and ER occurs very
earlyin refeeding (21, 22, 24), and for this reason alone,
numerouspossibilities exist for a transient suppression of
microauto-phagy. In view of the probable involvement of smooth ER
inthe formation of type A particles during the postabsorptive
pe-riod and starvation, the question of its role during refeeding
isparticularly relevant. Unfortunately, because glycogen is
closelyassociated with smooth ER (22, 25), its visualization is
difficult
in refed states, in which the content of glycogen is high.
How-ever, the characteristic branched tubular network can be
ap-preciated, as we have observed, when overlying glycogen is
re-moved during perfusion with glucagon. What the
ultimateexplanation for the microautophagic suppression will
eventuallybe is a matter for speculation at present, but available
evidenceindicates that it could relate in some way to the
restructuringof smooth ER or even to its physical relationship with
glycogen.
The authors thank Ann Bryan, Chih-Ying Ho, -Carolyn Lloyd,
DanMcBride, Kathleen Shiffer, and John Wert for their skillful
technicalassistance; Dr. Bryce L. Munger for use of electron
microscopic facilitiesin the Department of Anatomy; and Bonnie
Merlino for typing themanuscript. This work was supported by U.S.
Public Health Service GrantAM-21624 and a grant from the American
Diabetes Association to N.J.H.
1. de Duve, C. & Wattiaux, R. (1966) Annu. Rev. Physiol. 28,
435-492.
2. Ericsson, J. L. E. (1969) in Lysosomes in Biology and
Pathology,eds. Dingle, J. T. & Fell, H. B. (North-Holland,
Amsterdam), Vol.2, pp. 345-394.
3. Mortimore, G. E. & Schworer, C. M. (1977) Nature (London)
270,174-176.
4. Novikoff, A. B. & Shin, W.-Y. (1978) Proc. Nati. Acad.
Sci. USA75, 5039-5042.
5. Schworer, C. M., Shiffer, K. A. & Mortimore, G. E. (1981)
J. Biol.Chem. 256, 7652-7658.
6. Conde, R. D. & Scornik, 0. A. (1976) Biochem. J. 158,
385-390.7. Hutson, N. J. & Mortimore, G. E. (1982)J. Biol.
Chem. 257, 9548-
9554.8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. &
Randall, R. J.
(1951) 1. Biol. Chem. 193, 265-275.9. Schworer, C. M. &
Mortimore, G. E. (1979) Proc. Natl. Acad. Sci.
USA 76, 3169-3173.10. Williams, J. A. & Woodbury, D. M.
(1971) J. Physiol. (London)
212, 85-99.11. Blouin, A., Bolender, R. P. & Weibel, E. R.
(1977) J. Cell Biol.
72, 441-445.12. Roughton, F. J. W. (1964) in Handbook of
Physiology: Respira-
tion, eds. Fenn, W. 0. & Rahn, H. (Am. Physiol. Soc.,
Washing-ton, DC), Sect. 3, Vol. 1, pp. 767-825.
13. Pfeifer, U. (1973) Virchows Arch. B 12, 195-211.14.
Mortimore, G. E. & Ward, W. F. (1981)]. Biol. Chem. 256,
7659-
7665.15. Pfeifer, U. & Bertling, J. (1977) Virchows Arch. B
24, 109-210.16. Khairallah, E. A. (1978) in Protein Turnover and
Lysosome Func-
tion, eds. Segal, H. L. & Doyle, D. J. (Academic, New York),
pp.89-104.
17. Munthe-Kaas, A. C., Berg, T. & Seljelid, R. (1976) Exp.
Cell Res.99, 146-154.
18. Knook, D. L. & Sleyster, E. C. (1980) Biochem. Biophys.
Res.Commun. 96, 250-257.
19. Mortimore, G. E. & Schworer, C. M. (1980) in Protein
Degra-dation in Health and Disease, Ciba Foundation 75
(Elsevier/North-Holland, Amsterdam), pp. 281-305.
20. Amenta, J. S. & Brocher, S. C. (1981) Life Sci. 28,
1195-1208.21. Waterlow, J. C., Garlick, P. J. & Millward, D. J.
(1978) Protein
Turnover in Mammalian Tissues and in the Whole Body
(North-Holland, Amsterdam), pp. 117-176; 697-746.
.22. Cardell, R. R., Jr. (1977) Int. Rev. Cytol. 48, 221-276.23.
Marzella, L., Ahlberg, J. & Glaumann, H. (1980) Exp. Cell
Res.
129, 460-465.24. Fawcett, D. W. (1955) J. Natl. Cancer Inst. 15,
1475-1503.25. Porter, K. R. & Bruni, C. (1959) Cancer Res. 19,
997-1009.
Biochemistry: Mortimore et al.
Dow
nloa
ded
by g
uest
on
June
24,
202
1