Asynchronous Synthesis of Membrane Skeletal Proteins During

Asynchronous Synthesis of Membrane Skeletal Proteins During Terminal
Maturation of Murine Erythroblasts
By Manjit Hanspal, Jatinder S. Hanspal, Rajiv Kalraiya, Shiu-Chun Liu, Kenneth E. Sahr, Don Howard, and Jiri Palek
To study the changes in the synthesis of the major membrane
skeletal proteins, their assembly on the membrane, and their
turnover -during terminal red blood cell maturation in vivo,
we have compared early proerythroblasts and late erythroblasts obtained from the spleens of mice at different times
after infection with the anemia-inducing strain of Friend virus
(FVA). Metabolic labeling of these cells indicates striking
differences between early and late erythroblasts. In early
erythroblasts, spectrin and ankyrin are synthesized in large
amounts in the cytosol with proportionately high levels of
spectrin and ankyrin messenger RNA (mRNA). In contrast,
only small amounts of these polypeptides are incorporated
into the skeleton, which is markedly unstable. In late erythroblasts, however, the synthesis of spectrin and ankyrin and
their mRNA levels are substantially reduced, yet the net
amounts of these polypeptides assembled in the membrane
skeleton are markedly increased, and the membrane skele-
ton becomes stable with no detectable protein turnover. The
mRNA levels and the synthesis of the band 3 and 4.1 proteins
are increasedconsiderably in terminally differentiated normoblasts with a concomitant increase in the net amount and the
half-life of the newly assembled spectrin and ankyrin. Thus,
the increased accumulation of spectrin and ankyrin at the
late erythroblast stage is a consequence of an increased
recruitment of these proteins on the membrane and an
increase in their stability rather than a transcriptional upregulation. This is in contrast t o band 3 and 4.1 proteins, which
accumulate in direct proportion t o their mRNA levels and
rates of synthesis. These results suggest a key role for the
band 3 and 4.1 proteins in conferring a long-term stability t o
the membrane skeleton during terminal red blood cell differentiation.
o 1992by The American Society of Hematology.
T
into the membrane and band 3 is cotranslationally inserted
into the membrane lipid b i l a ~ e r . ~ - ~
Although avian erythroblasts exhibit many important
differences from enucleate mammalian erythrocytes, particularly the presence of intermediate filaments and the
nonerythroid isoform of a spectrin in the membrane skelesome of the key features of skeletal assembly have
been reproduced in mammalian erythroblasts. Notably,
studies of splenic erythroblasts of mice infected by Friend
leukemia virus have shown that band 3 synthesis is initiated
only after the cells have entered the pathway of terminal
differentiation, with a concomitant increase in the net
amounts of membrane-associated spectrin.’ An excessive
synthesis of spectrin, relative to the fraction of spectrin
assembled into the skeleton, has likewise been detected,
although there are considerable differences in the reported
ratios of the newly synthesized a and p spectrins: a high a to
p spectrin synthetic ratio has been detected in splenic
erythroblasts of anemic mice infected by the anemiainducing strain of Friend virus (FVA) or mice treated with
acetylphenylhydrazine,7.s while low ratios have been reported in erythroleukemic cell lines?JO
While most of the above studies focused on biosynthetic
changes during early events of erythroid maturation, leading to a formation of morphologically identifiable RBC
precursors, relatively little is known about the skeletal
protein synthesis and assembly at the terminal stages of
maturation, when the cell membrane properties approach
those of the circulating erythrocyte. For example, it is not
clear if the amounts of spectrin and ankyrin incorporated
into the skeleton remain proportional to the amounts
synthesized in the cytosol, and if the decline of spectrin and
ankyrin synthesis coincides with that of the band 3 and 4.1
proteins. Likewise, it is not clear at which stage of maturation RBCs acquire a stable skeleton with no detectable
turnover of skeletal proteins. To address these questions,
we studied the relative rates of skeletal protein synthesis,
assembly and turnover during in vivo terminal maturation
of splenic erythroblasts from FVA-infected mice. FVA
HE DIFFERENTIATION of early erythroid precursors into morphologically defined erythroblasts is
associated with marked changes in the expression and
synthesis of membrane proteins. The synthesis and assembly of membrane skeleton during erythroid differentiation
has been studied extensively in avian erythroblast^.^-^ In
avian erythroblastosis virus-transformed red blood cells
(RBCs), which are arrested at the early colony-forming
unit-erythroid (CFUe) stages of maturation, the synthesis
of both spectrin and ankyrin can be detected in the cytosol;
these newly synthesized proteins associate with the plasma
membrane but turn over rapidly. Band 3 synthesis is not
detectable in these cells.* In synchronized cohorts of
morphologically differentiated avian erythroblasts isolated
from chick embryos, the synthesis of band 3 has been
initiated and the skeleton assembles into a stable network
in which skeletal constituents accumulate during maturation? The synthesis of spectrin and ankyrin appears wasteful with large amounts of spectrin and ankyrin being
synthesized in the cytosol while the fraction of these
proteins assembled into the permanent skeletal network is
relatively small.’ In contrast to spectrin and ankyrin, the
majority of the newly formed 4.1 protein is incorporated
From the Department of Biomedical Research, St Elizabeth’s
Hospital of Boston, Tufts University School of Medicine, Boston, MA.
Submitted June 3,1991; accepted March 23,1992.
Supported by the National Institutes of Health Grants No. POI-HL37462 and ROI-HL-27215.
Presented in part in abstract form at the Thirty-Second Annual
Meeting of the American Society of Hematology in Boston, MA, 1990.
Address reprint requests to Manjit Hanspal, PhD, Department of
Biomedical Research, St Elizabeth’s Hospital of Boston, 736 Cambridge St, Boston, M A 02135.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement”in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 I992 by The American Society of Hematology.
0006-4971I92 /8002-OO23$3.OOt0
530
Blood, VOI 80, NO2 (July 15), 1992: pp 530-539
SYNTHESIS OF ERYTHROID MEMBRANE PROTEINS
causes mouse erythroid progenitor cells to proliferate
greatly, resulting in a marked increase in splenic nucleated
cell number. This acute erythroblast proliferation results in
virtually complete replacement of the other nucleated cell
types in the spleen with the erythroblasts thus making
spleen a highly enriched source of erythroblasts? Previous
studies have shown that FVA-infected splenic erythroblasts
represent a relatively homogeneous population of proerythroblasts. These cells undergo an average of two terminal
rounds of cell division during 48 hours of culture in the
presence of erythropoietin, producing late nonproliferative
erythroblasts almost ready to enucleate." In this study, we
have compared early proerythroblasts and nonproliferative
late erythroblasts obtained from spleens of mice at different
times after infection with FVA. We find striking differences
in the rates of synthesis, assembly, and turnover of spectrin,
ankyrin, band 3, and band 4.1 during terminal maturation.
In early erythroblasts, the rapid synthesis of spectrin and
ankyrin is accompanied by a low incorporation of these
proteins into the skeleton, and the skeleton is unstable. In
contrast, in late erythroblasts, the synthesis of both spectrin
and ankyrin declines, yet the net amounts of spectrin and
ankyrin assembled into a permanent membrane skeleton
markedly increase. This more efficient use of newly synthesized spectrin and ankyrin parallels an increase in the
synthesis of the band 3 and 4.1 polypeptides during terminal erythroid maturation, implying that these two proteins
have a crucial role in the stabilization of the permanent
membrane skeleton.
MATERIALS AND METHODS
Mice and v i m . Eight- to 12-week-oldfemale Balb/c mice were
purchased from Charles River Breeding Laboratories (Wilmington, MA). The FVA was obtained from Dr Mark Koury (Vanderbilt University, Nashville, TN). The virus was maintained by
passage of infectious plasma in Balb/c mice?
Isolation of splenic erythroblasts. Mice were injected via tail vein
with approximately 104 spleen focus-formingunits12of FVA. At 15
or 25 days after infection, the mice were killed and their spleens
removed. A single cell suspension of spleen cells in Iscove's
modified Dulbecco's medium (IMDM GIBCO Laboratories,Grand
Island, NY) was made by passing through a polyethylene mesh
(spectramesh) of pore size 202 pm (Spectrum Medical Industries,
Inc, Los Angeles, CA). The cell suspension obtained from 15-day
FVA-infected mice spleens was subjected to unit gravity sedimentation through a linear gradient of 1% to 2% deionized bovine
serum albumin (BSA) in IMDM essentially as described by Sawyer
et al,13 with minor modifications. A continuous, linear gradient of
1%to 2% BSA (700 mL total volume) was formed in a 2 L glass
beaker, followed by a steep gradient of 80 mL from 0.2% to 1%
BSA. Spleen cells (6 to 7 x 108) were then loaded directly on the
top of the steep gradient. After 2 hours and 45 minutes of
sedimentation at 4"C, cells were collected in 40-mL aliquots from
the bottom using a peristaltic pump. The first eight aliquots
consisted almost entirely of immature erythroblasts.These aliquots
were pooled and used as early erythroblasts.
To obtain late erythroblasts, spleens were collected at 25 days
after infection. A single cell suspension in IMDM was made as
described above and then subjected to a discontinuous Percoll
gradient (Percoll; Pharmacia Fine Chemicals, Piscataway, NJ)
consistingof 45%, 65%, 70%, 77%, and 90%, as described earlier!
Fraction number 3 consisted almost entirely of polychromatophilic
53 1
and orthochromatic normoblastsas judged by stainingwith WrightGiemsa and benzidine-hematoxylin. This cell fraction was used as
late erythroblasts.
Peripheral blood of Balb/c mice infected with FVA for 25 days,
containing 12% to 15% reticulocytes, was used as a source of
reticulocytes that were used for steady-state protein and messenger
RNA (mRNA) quantitations. Synthetic studies were not performed in reticulocytes because a large number of animals would
have to be killed to obtain sufficient amounts of blood.
Labelingof cells. Both early and late erythroblastswere washed
twice in methionine-free Dulbecco's modified minimum essential
medium (DMEM; GIBCO) and resuspended in 10 mL of the
medium containing 20% fetal bovine serum (FBS) prewarmed to
37°C for 15 minutes. They were then cultured for different lengths
of time with [35S]methionine(30 pCi/mL; 1,OOO Ci/mmol; ICN
Biomedicals,Irvine, CA). For the pulse-chase experiment, further
incorporation of [35S]methionine was stopped by the addition of
unlabeled methionine (0.4 mmol/L) and the incubation was then
continued for different time periods. At the end of the labeling
period, 10 vol of 155 mmol/L choline chloride and 5 mmol/L
HEPES, pH 7.1, were added, and the cells were harvested by
centrifugation and washed once with ice-cold saline buffer (150
mmol/L NaCl, 5 mmol/L KCl, 1 mmol/L MgC12, and 1 mmol/L
CaC12).
Cell fractionation and preparation of membranes. [35S]methionine-labeled cells were treated with diisopropyl fluorophosphate
(Sigma Chemical Co,St Louis, MO) before lysing them in 4 vol of
lysis buffer containing 150 mmol/L NaCI, 10 mmol/L Tris-HCI,pH
7.2,s mmol/L MgC12,2 mmol/L EDTA, 0.25 mmol/L dithiothreitol, 1mmol/L phenylmethylsulphonylfluoride (PMSF), 1mmol/L
leupeptin, 10 p,g/mL aprotinin, and 1% Triton X-100, and separated into soluble and insoluble fractions. The insoluble fraction
was centrifuged at 80% for 5 minutes to remove nuclei that
sediments to the bottom of the tube in a tight pellet. During this
centrifugation, a majority of the skeletal residues remain in the
supernatant, although a small amount sediments and forms a loose
pellet on top of the tight nuclei botton. Therefore, care is taken to
remove the loose pellet with the supernatant and the resulting
suspension of skeletal residues is referred to as the membraneskeletal fraction.
Alternatively, cells were lysed under hypotonic conditions to
isolate plasma membranes by the procedure of Chan.I4 The cells
were suspended in hypotonic buffer (10 mmol/L Tris-HCI, pH 7.5,
10 mmol/L KCl, 1.5 mmol/L MgC12, 1 mmol/L PMSF, 1 mmol/L
leupeptin, and 10 pg/mL aprotinin) and then disrupted with 10
strokes of a tight-fitting Dounce homogenizer (Wheaton Scientific,
Millville, NJ). An appropriate volume of 2 mol/L sucrose was
added immediately to restore isotonicity. The homogenate was
layered over a sucrose-step gradient (3 ~0128%[wt/vol] sucrose, 1
vol 50% [wt/voI] sucrose; sucrose solutions made in 140 mmol/L
NaCI, 2.4 mmol/L MgC12, 5 mmol/L Tris-HCI, pH 7.4, 1 mmol/L
PMSF, 1mmol/L leupeptin, and 10 p,g/mL aprotinin) and centrifuged in a SW 50.1 rotor (Beckman Instruments, Inc, Fullerton,
CA) at 117,OOQg for 40 minutes. The membrane fraction at the
28%/50% sucrose interface and the soluble fraction on the top of
the sucrose gradient were collected. The membrane fraction was
diluted with 20 mmol/L Tris-HC1, pH 7.4, and centrifuged at
15,OOOgfor 15 minutes.
Immunoprecipilation. Spectrin was immunoprecipitated from
the Triton-soluble and insoluble fractions of [35S]methioninelabeled cells as previously described.* Ankyrin, band 3, and band
4.1 proteins were immunoprecipitated by a modification of the
method described by Marcantonio and Hynes.15Ankyrin and band
4.1 were immunoprecipitated from the Triton-soluble and insoluble fractions, whereas band 3 was immunoprecipitated from the
HANSPAL ET AL
532
plasma membranes of [35S]methionine-labeled cells. The samples
were suspended in 0.5 mL of buffer A (0.1 mol/L Tris-HC1, pH 8.5,
0.15 mol/L NaCl, 5 mmol/L EDTA, 0.5% NP-40) containing 1%
sodium dodecyl sulfate (SDS), placed in a boiling water bath for 2
minutes, and then diluted to 5.0 mL with buffer A. Triton X-100
was then added to 5% final concentration. The 50-fold excess of
Triton X-100 forms mixed micelles with SDS, thereby preventing
antibodies from being denatured. Rabbit antihuman ankyrin,
antihuman 4.1, or antimouse band 3 IgG was then added and the
samples incubated overnight at 4°C with gentle shaking. (The
amounts of all three antibody preparations used for immunoprecipitation were optimized using 12sI-labeledhuman or mouse erythrocyte ghosts.) Thereafter, 100 pL of protein A-sepharose CL4B
(Pharmacia) (50 mg beads/mL of buffer A) was added and the
samples incubated for another 3 hours at 4°C with gentle shaking.
The protein A-sepharose beads were then collected and washed
four times with buffer A. The final pellet was resuspended in 70 pL
SDS sample buffer and boiled for 2 minutes. Beads were removed
by centrifugation and the supernatant was directly loaded on
SDS-polyacrylamide gels.
SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins
were separated by SDS-PAGE according to the buffer system of
Laemmli.16 The gels were processed for fluorography with En
3Hance (New England Nuclear, Boston, MA), dried, and exposed
to Kodak XAR-5 X-ray film (Eastman Kodak, Rochester, NY).
The autoradiograms were scanned at 570 nm with a Quick Scam
spectrophotometer (Helena Laboratories, Beaumont, TX) and the
area under each peak was integrated.
Normalization and quantitation. All results on the synthesis and
assembly of membrane skeletal proteins as well as their steadystate protein and mRNA contents were normalized to cell number
(determined by multiple measurements in a hemacytometer). An
equal number of early and late erythroblasts were pulse-labeled
with a given amount of the isotope. Percent incorporation of
radioactivity in a specific protein was determined in each experiment to ensure proper metabolic labeling. This was performed as
follows. Total incorporation of radioactivity was determined in a
small amount of the Triton-soluble and insoluble fractions or
plasma membranes and cell lysates (before immunoprecipitation)
by TCA precipitation. Incorporation of radioactivity in a specific
protein after immunoprecipitation was determined by spotting 5
pL of the immunoprecipitated SDS-solubilized sample on a DE51
filter disc. The filter was then washed, dried, and counted in a
scintillation counter.
Quantitation of the synthesis and assembly of a given protein was
determined after immunoprecipitation by either (1) scanning the
autoradiograms (exposed for the same length of time for early and
late erythroblasts) at 570 nm and measuring the area under each
peak, or (2) excising the immunoprecipitated protein band out of
the gel, solubilizing in aquasol (New England Nuclear), and
counting radioactivity in a scintillation counter.
Negative staining electron microscopy. To prepare specimens for
electron microscopy, aliquots of cells were lysed with a buffer
containing 5 mmol/L sodium phosphate, pH 7.4,5 mmol/L MgC12,
and 0.1% Triton X-100, quickly fixed with 1% glutaraldehyde at
25°C for 10 minutes and washed two times with 5 mmol/L
Tris-HC1, pH 7.4. The samples were applied to a carbon-coated
grid, negatively stained with 1% uranyl acetate solution, and
examined in a JEOL JEM-100 S electron microscope as described
previously (JEOL, USA Inc, Peabody, MA)."
Preparation of cDNAprobes. A 300-bp mouse ci spectrin cDNA
fragment was amplified from a murine erythroleukemia (MEL)
library by the polymerase chain reaction (PCR) technique using
oligonucleotide primers flanking repeat unit 16 of the ci spectrin
cDNA.18 The mouse p spectrin cDNA probe was prepared from
the plasmid pp58 (kindly provided by Dr Peter Curtis, The Wistar
Institute of Anatomy and Biology, Philadelphia, PA) containing a
720-bp insert spanning repeat units 8, 9, and 10 of the f3 spectrin
cDNA.I9 A 350-bp mouse ankyrin cDNA fragment was isolated
from a MEL library by screening with a human ankyrin cDNA
fragment20 spanning the region encoding the spectrin binding
domain. A mouse band 3 cDNA fragment was prepared from the
plasmid ~ R K 0 0 2 ~
(kindly
l
provided by Dr Ron Kopito, Stanford
University) containing a 3.9-kb full-length cDNA. The human 4.1
cDNA probe was prepared from the plasmid A HE-4.1@ containing a 2.0-kb insert.
Smaller cDNA inserts of each of these subclones were cloned
into pGEM4Z or pGEM7Z (Promega Biotec, Madison, WI)
vectors, containing SP6 and T7 RNA polymerase initiation sequences. [32P]-labeledantisense RNA probes were generated using
SP6 or T7 RNA polymerase (Promega) in the presence of [32P]
CTP (New England Nuclear).
RNA isolation and SI nuclease protection analysis. Total cellular
RNA was prepared by the guanidinium/cesium chloride method.23
Dried total RNA from an equal number of early or late erythroblasts was dissolved in 30 pL of hybridization buffer (80%
formamide, 40 mmol/L piperazine-N-N'-bis 2-ethane sulfonic
acid, pH 6.7, 0.4 mol/L NaCl, 1 mmol/L EDTA) containing the
appropriate antisense probe (5 X l@cpm), heated at 95°C for 5
minutes, and incubated at 52°C overnight. The ratio of probe to
mRNA was determined to be in substantial probe excess by
demonstrating a commensurate increase in protected probe with
increase in added RNA (data not shown). After hybridization, 250
pL of S1 digestion buffer (50 mmol/L NaOAc, pH 4.5,400 mmol/L
NaCl, 4.5 mmol/L ZnS04, 20 pg/mL salmon sperm DNA) containing 800 U of S1 nuclease (US Biochemicals, Cleveland, OH) was
added and incubated at 37°C for 30 minutes. The reaction was
stopped by the addition of 100 p,L of stop buffer (3 mol/L
NHdOAc, 100 mmol/L EDTA, 200 pg/mL yeast tRNA). The
samples were ethanol precipitated, washed twice with 70% ethanol, dried, and dissolved in the loading buffer. The protected
fragments were denatured and analyzed on 6% polyacrylamide/7
mol/L urea gels.
RESULTS
Decline in the synthesis of spectrin and ankyrin during
terminal maturation contrasts with the increased assembly of
these proteins into the membrane skeleton. The maturation
of erythroblasts from the proerythroblast stage to the stage
of late orthochromatic normoblast is accompanied by dramatic changes in the synthesis and assembly of a spectrin, p
spectrin, and ankyrin (Fig 1).In early erythroblasts, a large
amount of a and p spectrin polypeptides (with excess a
spectrin over p spectrin) are synthesized in the cytosol, with
only a small fraction of the newly synthesized spectrin being
incorporated into the membrane skeleton. In these cells,
only 5% to 7% of the newly synthesized spectrin is
incorporated into the membrane skeleton. In contrast, the
synthesis of a spectrin and p spectrin in late erythroblasts is
markedly decreased, yet the assembly of the spectrin
polypeptides into the membrane skeleton is markedly
increased. In late erythroblasts, the fraction of newly
synthesized spectrin associated with the skeleton is increased to 20% to 25% (Table 1).The changes in synthesis
and assembly of ankyrin during erythroid maturation are
similar to those of spectrin; the synthesis of ankyrin declines
with erythroblast maturation (Fig l), but the fraction of
SYNTHESIS OF ERYTHROID MEMBRANE PROTEINS
533
A
Early erythroblasts
Soluble
Skeletal
90
30 60
30
60
Late erythroblasts
Soluble
Skeletal
30
90
60
30
90
Spectrln
.-
60
90
9
Ankyrin
B
Spectrin
soluble
skeletal
Y
500
Fig 1. Synthesis and assembly of spscMn and
ankyrin in early and late erythroblasts. (A) An equal
number (1 x 10) of early and late erythroblasts were
labeled with PSImethionine for the times indicated
(30 to 90 minutes), lysed in a buffer containing Triton
X-100. and separated into skeletal and soluble fractions. Aliquots of each sample were immunoprecipitated with either antispectrin or antlankyrin IgG and
the immunoprecipitates were separated on SDSpolyacrylamide gels. The gels were processed for
fluorography. [E) The autoradiograms shown in (A)
were scanned and the area under each peak was
integrated. [-----) Early erythroblasts; (-) late erythroblasts. The area on the left-hand axis represents
the relative area. Note that our antibody precipitates
1.1 times more a spectrin than @ spectrin from
lnl-labeled mouse erythrocyte membranes. This correction factor for @ spectrin was taken into account in
the quantitations shown in (E). The results show that
the synthesis of a spectrin, @ spectrin, and ankyrin
declines, whereas their assembly into the membrane
skeleton increases during terminal maturation.
I
l
l
l
l
r
'0
Ankyrin
- 0
0
newly synthesized ankyrin incorporated into the membrane
skeleton increases from 16% in early erythroblasts to 37%
in late erythroblasts (Table 1). These results indicate that
newly synthcsizcd spectrin and ankyrin become more efficiently assembled into the membrane skeleton in RBCs in
the late stages of maturation. This is consistent with an
increased rccruitmcnt of these polypeptides to the peripheral membrane skeleton as part of the process of terminal
RBC maturation.
The assembled spectrin and ankyrin of early erythmblasts
exhibit a rapid tumover, but these proteins become stable in the
Table 1. The Fraction of Newly Synthesized SpscMn, Ankyrin, and
Protein 4.1 Assembled Into the Membrane Skeleton in Early and Late
Erythroblasts
Late
Erythroblasts
Early
Erythroblasts
(XI
(%)
~~
Sk altotal a
Sk @/total@
Sk Ankltotal Ank
Sk 4.1ltotal4.1
5
7
20
3
20
19
25
37
10
The data were derived from the experiments in Figs 1and 3 as well as
similar experiments, using the 60-minute time point for all calculations.
These experiments were performed four times with similar results.
Abbreviation: Sk, skeletalfraction.
20
60
00
soluble
,
/
40
lime (min)
40
100
Y
60 00 100
l i m e (min)
loor
- 0
0
20
60
00 100
Time (min)
skeletal
250
0
40
20
40
60
80 100
Time (min)
skeletons of late erythmblasts. Consistent with previous
reports,RaZ4we find that the newly synthesized cytosolic
spectrin and ankyrin undergo a rapid degradation irrespective of the stage of erythroblast maturation (Fig 2). However, the turnover rates of these polypeptides in the
assembled membrane skeleton change strikingly during
terminal differentiation. In early erythroblasts, spcctrin and
ankyrin associated with the skeleton turn over rapidly, with
a half-life similar to that of the cytosolic proteins (Fig 2). It
should be noted that these half-lives are considerably
shorter than the doubling time (20 to 24 hours) for early
erythroblasts, so it is unlikely that this instability could be
linked to the fact that these cells are proliferating during
the chase period. In a striking contrast, the skeletalassociated a spectrin,
spectrin, and ankyrin of late
erythroblasts are stable with no detectable turnover.
The steady-state levels of spectrin and ankyrin in early
and late erythroblasts correspond to the above changes in
the assembly and turnover of these proteins. In the cytosol
of both early and late erythroblasts, there is no net
accumulation of either spectrin or ankyrin. In the plasma
membranes of early erythroblasts, the steady-state levels of
spectrin and ankyrin are about 50% of the values of mature
erythrocytes, whereas at the late erythroblast stage. the
steady-state levels of both proteins are increased to about
HANSPAL ET AL
534
Spectrin
400-
A
II
Soluble
\\\
Q.,
320d
Time of chase (min)
Time of chase (min)
Ankyrin
------
-early
0'
20 4 0
60
80
Time of chase (min)
100
20 40 60 80
Time of chase (min)
75% to 80% of the amounts present in mature erythrocytes
(Table 2).
The increased assemb3,of spectrin and ankyrin into a stable
membrane skeleton in late erythroblasts is concomitant with
increased synthesis of the band 3 and 4.1 proteins. In
contrast to the decline observed in spectrin and ankyrin
synthesis during erythroid maturation, the synthesis of both
band 3 and 4.1 proteins increases (Fig 3). Band 3 and 4.1
proteins are synthesized in both early and late erythroTable 2. Quantitation of Steady-State Protein Levels During
Erythroid Maturation
Early
Erythroblasts
W)
a Spectrin
p Spectrin
Ankyrin
Band 3
Band 4.1
51
56
49
51
41
Late
Erythroblasts
Reticulocytes
(Oh)
( 0 4
75
76
83
105
106
98
90
101
03
81
The steady-statelevels of a,p spectrin, ankyrin, band 3,and band 4.1
proteins were determined in the plasma membranes of early erythroblasts, late erythroblasts, and peripheral blood reticulocytes. Total cell
numbers were determined through duplicate counts in a hemacytometer before isolating plasma membranes as described in Materials and
Methods. Cell equivalents (1 x 1Os) of plasma membranes were electrophoresed on an SDS-polyacrylamidegel and transferred to nitrocellulose. The nitrocellulose filter was probed with specific antibodies and
the immunoreactive polypeptides were detected by incubation with
'251-protein A followed by autoradiography. The resulting autoradiograms were scanned and the area under each peak was integrated. The
values in the table represent the levels of these polypeptides relative to
their value (taken as 100%) in mature erythrocyte membranes. The
values shown in the table represent the mean of three independent
measurements.
100
Fig 2. Turnover of spectrin and ankyrin. Early and
late erythroblasts (1 x 1 0 each) were labeled with
[%]methionine for 15 minutes and then chased with
cold methionine for different time periods (0 t o 90
minutes). Thereafter, the cells were lysed and separated into soluble and skeletal fractions. Equal volumes of each sample were Immunoprecipitated for
spectrin or ankyrin and the immunoprecipitateswere
analyzed by SDS-PAGE. The gels were processed for
fluorography and the resulting autoradiograms were
scanned and the area under each peak Integrated.
The skeletal-associated spectrin and ankyrin tum
over rapidly in early erythroblasts, but are stable in
late erythroblasts. The cytosolic spectrin and ankyrin,
however, undergo rapid degradation in both aarly
and late erythroblasts. This experiment was repeated
three times with similar results.
blasts. In striking contrast to spectrin and ankyrin, however,
the synthesis of band 3 is 1.5- to 2.0-fold higher in late
erythroblasts as compared with the early erythroblasts. The
synthesis and assembly of protein 4.1 in the membrane
skeleton increases similarly with cell maturation (Fig 3).
Immunoprecipitation of protein 4.1 detects four newly
synthesized species, presumably the isoforms of protein 4.1,
in both early and late erythroblasts. The synthesis of all four
4.1 proteins increases at the late erythroblast stage. Immunoprecipitation with preimmune serum did not detect these
proteins (data not shown). The major band (indicated by
arrow in Fig 3) comigrates with the protein 4.lb isoform
(data not shown). When comparing the relative amounts of
newly synthesized spectrin and ankyrin assembled into the
membrane skeleton to the synthesis of band 3 and 4.1
proteins (Fig 4), it is evident that the increased skeletal
assembly of both spectrin and ankyrin in late erythroblasts,
despite their reduced synthesis, is related to an increase in
the synthesis of band 3 and 4.1 proteins. Moreover, the
increased stability of the skeletal-associated spectrin and
ankyrin at the late erythroblast stage suggests that the band
3 and 4.1 proteins confer a long-term stability to the
membrane skeleton.
Further, the assembly of the skeletal components into a
two-dimensional membrane skeletal network in the erythroblasts was examined ultrastructurally by negative staining
electron microscopy (Fig 5). Consistent with previous
reports," the skeletal network of mature erythrocytes
consist primarily of a hexagonal lattice of junctional F-actin
complexes crosslinked by spectrin tetramers that are often
decorated with globular structures containing ankyrin or
ankyrin/band 3 complexes (Fig 5). In late erythroblasts, we
have detected the structural elements consisting of (1)
SYNTHESIS OF ERYTHROID MEMBRANE PROTEINS
Band 3
535
B
A
30
120
60
P
1OOOr
Early
erythroblasts
Late
erythroblasts
9’
0:
Fig 3. Synthesis and a s ” bly of band 3 and 4.1 proteins in
early and late erythroblasts. (A)
Band 3 protein was immunoprecipitated from plasma membranesof ~S]methionine-labeled
(30, 60. and 120 minutes) early
and late erythroblasts (1.5 x 1W
each) and analyzed by SDSPAGE. (6) The autoradiogram
shown in (A) was scanned and
the area under each peak integrated. The synthesis of band 3
protein is increased at the late
stage of maturation. (C) Band 4.1
protein was immunoprecipitated
from theskeletal and solublefractions of InSlmethionine-labeled
early and late erythroblasts
(1 x 1W each), as described in
Materials and Methods. Band 4.1
corresponding to 78 Kd (indicated by arrow) was quantitated
by scanning the autoradiogram
and the result is shown in (D).
The synthesis and assembly of
band 4.1 protein is increased at
the late erythroblast stage.
Protein 4.1
C
90
60
30
60
80
ab Id0 4 0
90
-
Early
erythroblasts
Late
erythroblasts
io
Time (min)
skeletal
soluble
30
2b
4.1
4
I
--
I-
-4.1
.A-
D
P
P
$0 i 0 lA0
Time (min)
flexible fibres of spectrin decorated with globular structures
at the ankyrin binding site, and (2) the junctional complexes
containing short F-actin filaments (Fig 5B). However, these
structural elements are not organized into a continuous
meshwork of hexagonal lattice as seen in the skeletons
derived from marure erythrocytes. In contrast, in early
erythroblasts, we have detected very few spectrin fibres and
short F-actin-containing junctional complexes (Fig 5A).
Most of the obvious structures are presumably those from
the cytoskeletal rather than the membrane skeleton. These
results are consistent with the above biochemical observation that, despite an active synthesis of spectrin and ankyrin
in early erythroblasts, these proteins are not assembled
stably onto the membrane forming the underlying skeletal
network. Instead, the subsequent increased synthesis of
band 3 and 4.1 proteins at the late erythroblast stage may be
required for the assembly of spectrin-actin-protein 4.1
meshwork.
Asynchronous changes in the mRNA levels of spectrin,
ankyrin, band 3, and hand 4.1 proteins during terminal
maturation. The changes in the mRNA levels of spectrin,
ankyrin, band 3, and band 4.1 proteins during erythroid
IbO
L
0
20 40 60 80 100 120
Time (min)
maturation correspond to the changes observed for the
synthetic rates of these protcins. The mRNA levels of a
spectrin, B spectrin, and ankyrin are highest in early
erythroblasts, and decline with maturation until virtually
undetectable in reticulocytes (Fig 6). SI nuclease protection analysis of ankyrin mRNA using a probe within the
spectrin binding domain of ankyrin cDNA reproducibly
detected two protected fragments at all stages of maturation, suggesting that there may be an alternative splicing in
the spectrin binding domain of mouse ankyrin mRNA. In
contrast to spectrin and ankyrin, the mRNA levels of band
3 and 4.1 proteins increase from early erythroblast to the
late erythroblast stage and then decline in reticulocytes (Fig
6). SI nuclease protection analysis of protein 4.1 mRNA in
FVA-infected murine cells used a human erythroid protein
4.1 cDNA probe because of the unavailability of mouse 4.1
probe. The same human 4.1 cDNA probe was used in
Northern blot analysis of total RNA from an equal number
of early erythroblasts, late erythroblasts, and reticulocytes.
The results (data not shown) were identical to those
obtained by the SI nuclease protection analysis, ie, the
protein 4.1 mRNA level increased from early crythroblast
HANSPAL ET AL
536
Soluble
Skeletal
1
500
4001
d
1601
I
Band 3 ( membranes)
400
300
200
;::I 1
~
100
0
$l
100
*I
0
Fig 4. A comparison of the synthesis and assembly of spectrin,
ankyrin, band 3, and band 4.1 proteins in early ( 0 )and late (8)
erythroblasts. The synthesis and assembly of proteins is obtained
from tha experiments in Figs 1 and 3 as well as similar experiments,
using the m-minute time point. The data are presented as mean
values f SD observed in three separate experiments. Note that the
increased membrane assembly of ankyrin at the late erythroblast
stage is directly proportional to the increased synthesis of band 3,
while the 2.5- to 3.0-fold increase in the membrane assembly of
spectrin is related to a proportionalincrease in the amount of band 3
and 4.1 proteins at the late erythroblaststage.
to the late erythroblast stage and then declined slightly in
reticulocytes.
The observed decrease in the amount of spectrin and
ankyrin mRNAs during maturation in conjunction with the
several-fold increase in their polypeptide levels indicates
that the extent of accumulation of these proteins is not
primarily determined by the abundance of their mRNA. In
contrast, the accumulation of band 3 and 4.1 proteins is
primarily determined by their mRNA levels during terminal maturation.
DISCUSSION
In this report, we have compared the changes in the
synthesis of the major membrane skeletal proteins, their
assembly in the membrane, and their turnover rates during
terminal erythroblast maturation in vivo, using highly pure
(greater than 95%) erythroid cells at the early erythroblast
(proerythroblast) stage and at the late erythroblast (polychromatophiliclorthochromatic normoblast) stage obtained from spleens of mice at different times after infection with FVA. Late erythroblasts were also obtained by
48-hour in vitro culture of early erythroblasts, as described
previously by Koury et al.' Late erythroblasts obtained
either by in vivo or in vitro maturation of early erythroblasts
are similar in terms of their morphology and rates of
synthesis of membrane skeletal proteins (data not shown).
In this report, we show that in early erythroblasts, large
amounts of spectrin and ankyrin polypeptides are synthesized in the cytosol but only a small fraction of the newly
synthesized polypeptides are incorporated into the membrane skeleton. In contrast, in late erythroblasts, the
synthesis of both a and p spectrins as well as ankyrin
decreases by about 1.5- to 2.0-fold, yet the net assembly of
all three polypeptides in the membrane skeleton increases
by 1.5-fold (ankyrin) to 2.5-fold (a and f3 spectrins).
Furthermore, the rates of turnover of these polypeptides in
the assembled membrane skeleton decreases during erythroid maturation. In early erythroblasts, skeletal-associated
spectrin and ankyrin turn over rapidly, whereas in the
membrane skeleton of late erythroblasts these proteins are
stable with no detectable turnover. The steady-state levels
of spectrin and ankyrin in the plasma membrane are
proportional to the changes in their assembly and turnover,
increasing from about 50% of the values of mature erythrocytes in early erythroblasts to about 75% to 80% in late
normoblasts. We further show that the synthesis of the
band 3 and 4.1 proteins increases at the late stages of
terminal maturation and that the increased assembly of
spectrin and ankyrin on the membrane, despite their
reduced synthesis, parallels an increase in the synthesis of
the band 3 and 4.1 proteins. A concomitant increase in the
half-life of the skeletal-associated spectrin and ankyrin
suggests a key role for the band 3 and 4.1 proteins in
conferring a long-term stability to the membrane skeleton
during terminal maturation.
These observations are in concert with previous studies
showing direct biochemical interactions among these proteins. The cytoplasmic domain of band 3 provides attachment sites for ankyrin that, in turn, via their binding to the p
subunit of spectrin, stabilize spectrin tetramers on the
membrane, whereas protein 4.1 directly binds both to
spectrin and glycophorin C.25 However, because lower
levels of both the band 3 and 4.1 proteins are also
synthesized at the early erythroblast stage, it should be
cautioned that the increased synthesis of these proteins
during terminal maturation may not be the only determinant of the stabilization of spectrin and ankyrin at the late
erythroblast stage. The hypothesis that band 3 andlor band
4.1 may play a role in the stabilized assembly of spectrin
during terminal maturation was also suggested by previous
studies of FVA-infected murine erythroblasts' and avian
erythroblasts.',* However, the major new aspects of the
present study include the demonstration of a progressive
decrease in the synthesis of spectrin and ankyrin concomitant with an increase in the stability of skeletabassociated
spectrin and ankyrin with terminal maturation in vivo.
SYNTHESIS OF ERYTHROID MEMBRANE PROTEINS
537
..,..a,-
-.-.*-*-*
Fig 5. Negative staining electron microscopy of Triton-treated erythroblasts. The skeletons derived from Triton-treated early erythroblasts
(A). late erythroblasts (E). and erythroeytes (C) were fixed with glutmldehyde, applied t o thin arbon-coated grids, stained with uranyl acetate,
air-dried, and examined by transmission electron microscopy. Note the presence of spactrin fibres ( a r r o w ) and junctional complexes
fanowheads) in the membrane skeleton of late erythroblasts and mature erythrocytes.
The conclusion regarding the proposed role of the band 3
and 4.1 proteins in the stabilization of the membrane
skeleton is based on the temporal sequence of the synthesis,
assembly, and turnover of membrane and skeletal proteins.
However, it should be pointed out that a subset of patients
with hereditary elliptocytosis who lack the protein 4.1 (due
to a genetic defect involving a deletion including the
erythroid initiation
possess a full complement of
spectrin, as suggested by normal spectrin to band 3 ratios.
Thus, the spectrin-ankyrin-band 3 association is likely to
represent a critical site involved in the assembly of spectrin
on the membrane.
The mature RBC contains about 1.2 x 106 copies of the
band 3 protein.2nHence, based on our data on the steadystate levels, approximately 0.6 x 106 molecules of the band
3 protein are present in early erythroblasts compared with
about 1.0 x 106 molecules in late erythroblasts. Previous
studies have shown that at any given time, only 10% to 20%
p Spectrin
01Specfrin
c
1
2
3
c
1
Ankyrin
2
3
c
788720412354-
374314Fig 6. Qurntitaion of spectrln, ankyrin, band 3, and band
4.1 mRNA levels during erythroid
maturation determined by quantitative SI nuclease protection assay. For each assay, 7 x 1 0 cell
equivalents of total cellular RNA
from early erythroblasts (lanes
1). late erythroblasts (lanes 2).
and peripheral blood reticulocytes (lanes 3) were probed with
p*P]-labeledantisense RNAs spacific for a spectrin, f3 spectrin,
ankyrin, band 3, and band 4.1
proteins. The size in base pairs of
each probe and the protected
fragment is shown. Lanes C represents the no RNA control. The
mANA levels of a spechin, f3 spectrin, and ankyrin decline gradually during erythroid maturation,
whereas those of the band 3 and
4.1 proteins increase from the
early erythroblast t o the late
erythroblast stage and then decline in reticulocytes.
Band 3
c
1
551.
500.
2
Band 4.1
c
1
3
590510-
2
3
1
2
3
HANSPAL ET AL
538
of the band 3 protein on the membrane binds ankyrin and
that the unbound fraction of the band 3 protein may be in
equilibrium with the bound f0rm.2~The ability of the band 3
protein to bind ankyrin may be related to the degree of
oligomerization, with the band 3 dimers being detected in
the skeleton free fraction and the higher oligomers being
tightly associated with the skeleton.28 In light of these
observations, it is not unreasonable to assume that despite
the presence of the band 3 protein in early erythroblasts,
band 3 may still represent the critical factor that limits
ankyrin assembly on the membrane. Moreover, our data on
the steady-state levels indicate that the relative abundance
of band 3 to ankyrin to spectrin remains approximately the
same in early and late erythroblasts (band 3:AnkaSp:pSp ::
10:1:22 in early erythroblasts and 10:1:1.8:1.8 in late
erythroblasts) based on previously reported results that
there are 200,000 copies of each of the two spectrin chains
and 100,000 copies of ankyrin in the mature RBC.25This
observation is consistent with the assumption that the band
3 protein may limit ankyrin assembly on the membrane.
Analysis of mRNA levels shows that the changes in the
amounts of spectrin, ankyrin, band 3, and band 4.1 mRNA
during erythroid maturation are similar to the changes in
the synthetic rates of these proteins. These results suggest
that the increased accumulation of the spectrin and ankyrin
at the late erythroblast stage is due to an increased
incorporation of these proteins in the membrane and an
increase in their stability rather than a transcriptional
upregulation. In contrast, the accumulation of the band 3
and 4.1 proteins during terminal maturation is directly
related to their mRNA levels and synthesis. We are
currently investigating whether these changes in mRNA
levels of membrane skeletal proteins correspond to the
changes in the rates of transcription or to their rates of
turnover.
The biogenesis of the erythroid membrane skeleton has
been studied in various RBC models. The results presented
here show several similarities to the previous reports. For
instance, (1) the unequal synthesis of a and p spectrin
polypeptides (increased synthesis of a spectrin over p
spectrin) is in agreement with the previously reported
results in explanted chicken embryo erythroblast^,^^ erythroid cells obtained from the spleens of rats with hemolytic
anemia induced by phenylhydrazine,8 as well as FVAinfected murine erythroblasts undergoing maturation in
vitro.7 However, in MEL cells, p spectrin synthesis exceeds
that of a s p e ~ t r i n . ~(2)
J ~Ankyrin synthesis being in excess
of the amount assembled on the membrane is similar to
avian30and human 31 erythroid cells. (3) The observation of
asynchronous membrane protein synthesis with the synthesis of the band 3 and 4.1 proteins continuing after the
synthesis of spectrin and ankyrin has been downregulated,
as has been shown previously by in vivo pulse-labeling
studies in mice.32
Despite the above similarities, the present data shows
several important differences from the previous findings.
(1) While we and 0 t h e 1 - sshow
~ ~ ~ that
~ the newly synthesized
a and p spectrin polypeptides assemble in equimolar
amounts on the membrane and membrane skeletons, Koury
et a17 have shown that in FVA-infected splenic erythroblast
membranes, the newly synthesized p spectrin is assembled
in twice the amounts of a spectrin. The excess p spectrin
over a spectrin in the membrane would imply that some p
spectrin is assembled as homodimers. However, Woods and
L a z a r i d e ~have
~ ~ shown that in avian erythroid cells spectrin
is assembled into the membrane skeleton in the form of a,p
heterodimers, while the homo-oligomers are degraded
rapidly in the cytosol. Moreover, we have recently shown
that the exposure of FVA-infected murine erythroblasts to
erythropoietin in vitro results in an increased synthesis of p
spectrin in the cytosol with a concomitant increase in the
amount of both a and p spectrin polypeptides assembled
into the membrane skeleton without altering their stoichiometric a~sembly.3~
Furthermore, because the autoradiograms of membrane proteins were not shown in the above
study; it is possible that the unequal assembly of a and p
spectrin polypeptides reported in FVA cell membranes7 is
due to unequal immunoprecipitation of the two polypeptides. (2) Unlike avian erythroid cells, in which protein 4.1
is assembled on the membrane soon after its synthesis with
only small amounts detectable in the cyt0~01,4~~
in murine
erythroid cells (present study), as well as in rat erythroblasts and retic~locytes,3~
protein 4.1 is synthesized in large
excess in the cytosol, with only a small amount being
incorporated into membrane skeleton. (3) In previous
studies using splenic erythroblasts undergoing terminal
maturation in vitro, the synthesis of the band 3 protein was
not detected in early erythroblasts and the synthesis of
protein 4.1 was not ~ t u d i e d .However,
~
the membraneassociated spectrin was shown to be stable in early erythroblasts. These observations are in contradiction with the
previous studies in avian erythroblasts that have suggested
that the skeletal-associated peripheral proteins become
stabilized only after the synthesis of the band 3 protein is
initiated.2 We have shown here that skeletal-associated
spectrin and ankyrin turn over rapidly in early erythroblasts
that synthesize small amounts of the band 3 and 4.1
proteins. In late erythroblasts, however, the increased
synthesis of the band 3 and 4.1 proteins is accompanied by
formation of a stable membrane skeleton.
Taken together, our data indicate that, while the mRNA
levels and the net synthesis of a spectrin, p spectrin, and
ankyrin decline during erythroid maturation, the recruitment of these proteins into the membrane skeleton and
their stability increases coincident with increased synthesis
of the band 3 and 4.1 proteins, leading to net accumulation
of these proteins in the developing erythroid membrane.
Furthermore, our data suggest, but do not prove, that
increased recruitment of a spectrin, p spectrin, and ankyrin
into a stable peripheral membrane skeleton is mediated by
the band 3, which may play a critical role in the formation of
the stable skeletal network in the terminally differentiated
erythroid cells.
ACKNOWLEDGMENT
We thank Laura Derick for her patience and skill in preparing
negatively stained electron micrographs, and Gail Gao for technical assistance in the measurements of mRNA levels. We also thank
Joan Joos, who did the artwork, and Loretta Wencis, who typed the
manuscript.
SYNTHESIS OF ERYTHROID MEMBRANE PROTEINS
539
REFERENCES
20. Lambert S, Yu H, Prchal JT, Lawler J, Ruff P, Spiecher D,
1. Lazarides E From genes to structural morphogenesis: The
genesis and epigenesis of a red blood cell. Cell 51:345,1987
Cheung MC, Kan YW, Palek J cDNA sequence for human
erythrocyte ankyrin. Proc Natl Acad Sci USA 871730,1990
2. Woods CM, Boyer B, Vogt PK, Lazarides E Control of
erythroid differentiation: Asynchronous expression of the anion
21. Kopito R, Lodish HF: Primary structure and transmemtransporter and the peripheral components of the membrane
brane orientation of the murine anion exchange protein. Nature
skeleton in AEV- and S13-transformed cells. J Cell Biol 103:1789,
316234,1985
1986
22. Conboy J, Kan YW, Shohet SB, Mohandas N Molecular
3. Cox JV,Stack JH, Lazarides E Erythroid anion transporter
cloning of protein 4.1, a major structural element of the human
assembly is mediated by a developmentally regulated recruitment
erythrocyte membrane skeleton. Proc Natl Acad Sci USA 83:9512,
onto a preassembled membrane cytoskeleton. J Cell Biol105:1405,
1986
1987
23. Maniatis T,Fritsch EF, Sambrook J Molecular Cloning: A
4. Staufenbiel M, Lazarides E Assembly of protein 4.1 during
Laboratory Manual. Cold Spring Harbor, NY,Cold Spring Harbor
chicken erythroid differentiation.J Cell BiollO21157,1986
Laboratory, 1989, p 196
5. Granger BL, Lazarides E Membrane skeletal protein 4.1 of
24. Blikstad I, Nelson WJ, Moon RT, Lazarides E Synthesis
avian erythrocytes is composed of multiple variants that exhibit
and assembly of spectrin during avian erythropoiesis. Stoichiomettissue-specificexpression. Cell 37595,1984
ric assembly but unequal synthesis. Cell 321081,1983
6. Moon RT,McMahon Ap: Generation of diversity in noneryth25. Palek J, Lambert S: Genetics of the red cell membrane
roid spectrins. J Biol Chem 265:4427,1990
skeleton.
Semin Hematol27290,1990
7. Koury MJ, Bondurant MC, Rana SS: Changes in erythroid
26. Conboy J, Mohandas N, Tchernia G, Kan YW: Molecular
membrane proteins during erythropoietin-mediatedterminal differbasis of hereditary elliptocytosis due to protein 4.1 deficiency. N
entiation. J Cell Physiol133:438,1987
En@J Med 315:680,1986
8. Hanspal M, Palek J Synthesis and assembly of membrane
27. Conboy J, Kim R, Agre P, Tchemia G, Kan YW, Mohandas
skeletal proteins in mammalian red cell precursors. J Cell Biol
N: Novel alterations of erythroid protein 4.1 gene expression in
105:1417,1987
patients with hereditary elliptocytosis. Blood 74181a, 1989 (abstr,
9. Pfeffer SR,Huima T, Redman C M Biosynthesis of spectrin
and its assembly into the cytoskeletal system of Friend erythroleuSUPPI 1)
kemia cells. J Cell Biol103:103,1986
28. Casey JR, Reithmeier RAF: Analysis of the oligomeric state
10. Lehnert ME,Lodish HF: Unequal synthesis and differential
of band 3, the anion transport protein of the human erythrocyte
membrane, by size exclusion high performance liquid chromatogradegradation of a and fi spectrin during murine erythroid differentiation. J Cell Biol107413,1988
phy. J Biol Chem 26615726,1991
11. Koury ST, Koury MJ, Bondurant M C Morphological changes
29. Thevenin BJ-M, Low PS: Kinetics and regulation of the
in erythroblasts during erythropoietin-inducedterminal differentiankyrin-band 3 interaction of the human red blood cell membrane.
ation in vitro. Exp Hematoll6758,1988
J Biol Chem 265:16166,1990
12. Axelrad AA, Steeves RA: Assay for Friend leukemia virus:
30. Moon RT, Lazarides E Biogenesis of the avian erythroid
A rapid quantitative method based on enumeration of macroscopic
membrane skeleton: Receptor-mediated assembly and stabilizaspleen foci in mice. Virology 24:513,1964
tion of ankyrin (Goblin) and spectrin. J Cell Biol98:1899,1984
13. Sawyer ST, Koury MJ, Bondurant MC: Large-scale procure31. Hanspal M, Yoon S-H, Yu H, Hanspal JS, Lambert S, Palek
ment of erythropoietin-responsiveerythroid cells: Assay for biologJ, Prchal J Molecular basis of spectrin and ankyrin deficiencies in
ical activity of erythropoietin. Methods Enzymol147340,1987
severe hereditary spherocytosis: Evidence implicating a primary
14. Chan CNL: Changes in the composition of plasma memdefect of ankyrin. Blood 77165,1991
brane proteins during differentiation of embryonic chick erythroid
32. Chang H, Langer PJ, Lodish H F Asynchronous synthesis of
cell. Proc Natl Acad Sci USA 741062,1977
erythrocyte
membrane proteins. Proc Natl Acad Sci USA 73:3206,
15. Marcantonio EE, Hynes RO: Antibodies to the conserved
1976
cytoplasmic domain of the integrin a I subunit react with proteins
33. Woods CM, Lazarides E Spectrin assembly in avian erythin vertebrates, invertebrates, and fungi. J Cell Biol106:1765,1988
roid development is determined by competing reactions of homo16. Laemmli UK Cleavage of structural proteins during the
and hetero-oligomerization. Nature 321235,1986
assembly of the head of bacteriophage T4. Nature 227680,1970
34. Hanspal M, Kalraiya R, Hanspal J, Sahr K, Palek J
17. Liu S-C, Derick LH, Palek J Visualization of the hexagonal
Erythropoietin enhances the assembly of a$ spectrin hetlattice in the erythrocyte membrane skeleton. J Cell Biol 104:527,
erodimers on the murine erythroblast membranes by increasing i3
1987
spectrin synthesis.J Biol Chem 266:15626,1991
18. Curtis PJ, Palumbo A, Ming J, Fraser P, Ciot L, Meo P,
35. Hanspal M, Hanspal JS, Palek J The biogenesis of memShane S, Rovera G: Sequence comparison of human and murine
brane skeleton in mammalian red cells. Cellular and molecular
erythrocyte a-spectrin cDNA. Gene 36:357,1985
biology of normal and abnormal erythroid membranes, in Cohen C,
19. C i d L, Laurila P, Meo P, Krebs K, Goodman S, Curtis P J
Palek J (eds): UCLA Symposia on Molecular and Cellular Biology,
Cloning and nucleotide sequence of a mouse erythrocyte p-spectrin cDNA. Blood 70915,1987
New Series, ~01118.New York, NY,Liss, 1990,p 145