Experimental Hematology 31 (2003) 234–243
Mechanisms of anemia in SHP-1 protein
tyrosine phosphatase-deficient “viable motheaten” mice
Bonnie L. Lyonsa, Michael A. Lynesb, Lisa Burzenskia,
Melissa J. Joliata, Nacima Hadjoutb, and Leonard D. Shultza
a
The Jackson Laboratory, Bar Harbor, Me., USA;
Department of Molecular and Cell Biology, University of Connecticut, Storrs, Conn., USA
b
(Received 23 September 2002; revised 28 October 2002; accepted 4 November 2002)
Objective. Viable motheaten mice (abbreviated gene symbol mev) are deficient in SHP-1, a
critical negative regulator of signal transduction in hematopoietic cells. These mice exhibit severe immune dysfunction accompanied by hyperproliferation of myeloid cells, widespread inflammatory lesions, and regenerative anemia. The aim of this study was to investigate the
mechanisms underlying anemia in mev/mev mice.
Materials and Methods. Multiple hematologic parameters, osmotic fragility, and erythropoietin levels were measured to characterize the anemia in mev/mev mice. B-cell–deficient mev/mev
Igh-6null mice were generated to assess the role of anti-erythrocyte antibodies. Coombs assays
and flow cytometry were carried out for detection of anti-erythrocyte antibodies. Oxidant production by macrophages, glutathione levels, and lipid peroxidation products in erythrocytes
were measured, as was the impact of oxidant on the ultrastructure of mev/mev erythrocytes.
Erythroid maturation and erythrocyte plasma membrane integrity were assessed with flow cytometry by evaluating CD71 expression and annexin V labeling.
Results. The regenerative anemia of mev/mev mice was associated with erythrocyte changes
that were independent of the presence of anti-erythrocyte antibodies. Erythrocytes from mev/
mev mice had increased fragility and heightened susceptibility to oxidant damage. Macrophages from mev/mev mice demonstrated a higher basal level of oxidant production and enhanced production after stimulation. Oxidant damage in mev/mev erythrocytes was evidenced
by a significant elevation of lipid peroxidation and diminished levels of glutathione.
Conclusion. Our results support the hypothesis that as a consequence of severe inflammatory
disease, mev/mev erythrocytes are subject to exceptionally high oxidative stress resulting in oxidation of phospholipids in the erythrocyte membrane with subsequent hemolysis. © 2003 International Society for Experimental Hematology. Published by Elsevier Science Inc.
There are three fundamental mechanisms of anemia: blood
loss, accelerated rate of erythrocyte destruction, and impaired
erythrocyte production [1]. Accelerated destruction (hemolysis) can be caused by inherited disorders of the erythrocyte
membrane, abnormalities in hemoglobin synthesis, or specific enzyme deficiencies. Additionally, acquired hemolytic
anemias can be caused by pathogenic organisms, exposure
to abnormal levels of reactive oxygen species, or as a sequela of immune-mediated disease [2,3]. In this article, we
describe the nature of the hemolytic anemia in viable motheaten mice.
Offprint requests to: Leonard D. Shultz, Ph.D., The Jackson Laboratory,
600 Main Street, Bar Harbor, ME 04609; E-mail: [email protected]
The autosomal recessive viable motheaten (Hcphme-v)
mutation disrupts the structural gene (Hcph) for hematopoietic cell phosphatase. This gene, located on mouse chromosome 6, encodes a cytoplasmic protein tyrosine phosphatase
commonly called Src-homology 2-domain phosphatase-1
(SHP-1). SHP-1 is expressed primarily in hematopoietic
cells [4] and functions as a critical negative regulator of signal transduction in a number of immune and hematopoietic
signaling pathways [5,6]. Viable motheaten mice (abbreviated gene symbol mev) are deficient in SHP-1 and express
approximately 20% wild-type activity. These mice have a
regenerative anemia and severe immune dysfunction with
greatly increased numbers of myeloid cells resulting in multifocal pyogranulomas of the skin, lung, and elsewhere. Ho-
0301-472X/03 $–see front matter. Copyright © 2003 International Society for Experimental Hematology. Published by Elsevier Science Inc.
doi:10.1016/S0301-472X(02)01031-7
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
mozygotes (mev/mev) die as a consequence of acidophilic
macrophage pneumonia [7]. Additionally, mev/mev mice exhibit polyclonal B-cell activation with greatly elevated levels of circulating immunoglobulins and multiple autoantibodies[8].
Studies have shown that colony-forming unit erythroid
(CFU-E) progenitor cells in some patients with polycythemia
vera (PV) have reduced expression of SHP-1. This myeloproliferative disorder is characterized by hyperproliferation
of erythroid cells, myeloid cells, and megakaryocytes. The
erythroid progenitor cells of PV patients with reduced expression of SHP-1 have an enhanced proliferative capacity
and are hypersensitive to erythropoietin (Epo) [9]. Similarly, there are increased splenic CFU-E progenitors in mev/
mev mice, and these erythroid progenitor cells are also hypersensitive to Epo [10]. However, unlike PV, the increase
in CFU-E progenitors in mev/mev mice is not reflected by erythrocytosis, but paradoxically these mice develop a regenerative
anemia.
Humans and experimental animals with autoimmune disease and concurrent anemia commonly have a hemolytic
anemia secondary to anti-erythrocyte antibodies [2,11–13].
Anemia in mev/mev mice was also postulated to be a consequence of anti-erythrocyte antibodies [8]. However, the
variable presence of these antibodies in mev/mev mice suggests that other mechanisms contribute in significant ways
to the development of anemia. In this study, we investigated
the mechanisms by which erythrocytes are destroyed in mev/
mev mice. We hypothesize that as a consequence of persistent inflammatory disease, mev/mev erythrocytes are subject
to exceptionally high oxidative stress resulting in oxidation
of phospholipids in the erythrocyte membrane with subsequent hemolysis.
235
Hematology
Blood was collected from the retro-orbital sinus with EDTA-coated
capillary tubes and placed directly into Isoton II diluent (Coulter,
Miami, FL, USA). Hematologic values were determined using the
ADVIA120 hematology system (Bayer, Norwood, MA, USA). In
some studies, erythrocytes were enumerated using a Z1 Coulter
counter. Blood smears were air dried, fixed in methanol, and stained
with Wright-Giemsa. For demonstration of reticulocytes, equal volumes of blood and new methylene blue stain were incubated for 15
minutes at room temperature.
Flow cytometry
Flow cytometric analysis was performed using a FACSCalibur flow
cytometer (Becton Dickinson, Mountain View, CA, USA). Cells were
prepared in cold Hank’s balanced salt solution (HBSS, Sigma Chemical Corp., St. Louis, MO, USA) containing 5% fetal bovine sera
(FBS, Sigma) and 0.1% sodium azide (FACS buffer) and aliquoted at
106 cells per 100 L in wells of a 96-well microtiter plate. Nonspecific
binding of antibody was blocked by incubation with rabbit IgG at 1
mg/mL (Sigma). After blocking, cells were washed in FACS buffer,
optimized dilutions of antibodies were added, and the cells were
incubated for 1 hour at 4C. Cells were washed twice in FACS buffer
and resuspended in cold HBSS containing 0.1% sodium azide. At
least 10,000 cells were analyzed per sample. Annexin V and antibodies were obtained from BD PharMingen Inc. (San Diego, CA,
USA) as fluorescein isothiocyanate or phycoerythrin conjugates:
anti-IgM, clone R6-60.2; anti-Ig, polyclonal; anti-pan erythrocyte,
clone TER-119; anti-CD47, clone miap301; anti-MAC-1, clone
M1/70; anti-Gr-1, clone RB6-8C5, and anti-CD71, clone C2.
Materials and methods
Scanning electron microscopy
For ultrastructural studies, erythrocytes were fixed overnight in 2%
glutaraldehyde and 2% paraformaldehyde in cacodylate buffer (pH
7.3). Cells were washed with buffer, and a thin film of cells was pipetted onto poly-L-lysine coated coverslips. The coverslips were
air dried, coated with gold, and were imaged on a JEOL 35C scanning electron microscope at 30 kV. For some studies, erythrocytes
were incubated in 10 M H2O2 for 60 minutes prior to fixation.
Mice
Viable motheaten mice were raised by mating C57BL/6J- /Hcphme-v
(/mev) heterozygotes. Homozygous (mev/mev) offspring were identified at 4 to 5 days of age by the presence of multifocal alopecia. Heterozygotes (/mev) were identified by polymerase chain reaction
as previously described [14]. The targeted disruption of the membrane exon of the immunoglobulin gene (Igh-6tm1Cgn) (abbreviated gene symbol Igh-6null) results in a deficiency of mature B cells
and consequent agammaglobulinemia [15]. C57BL/6J-Igh-6null
mice were mated with C57BL/6J-/mev mice to produce mice
doubly homozygous for the mev mutation and the Igh-6null allele.
Mice transgenic for metallothionein, C57BL/6J-TgN(Mt1)174Bri
(abbreviated gene symbol Mt1), which carry 112 copies of the Mt1
transgene [16], mice homozygous for targeted mutation of the metallothionein 1 and 2 genes, 129S7/SvEvBrd-Mt1tm1BrMt2tm1Bri (abbreviated gene symbol Mt1 Mt2) [17], and NZB/B1NJ mice were
obtained from The Jackson Laboratory Animal Resources colony.
All mice were bred and housed at The Jackson Laboratory under
standard specific pathogen-free conditions and were provided pasteurized food and acidified water ad libitum.
Direct and indirect Coombs assays
The indirect Coombs assay was carried out on plasma. Blood was
obtained from the retro-orbital sinus of mice through heparin
coated microcapillary tubes. After removal of the cells by centrifugation, 25 L of the plasma was added to U-bottomed microtiter
wells containing 25 L of 1% BSA in phosphate-buffered saline
(PBS). The samples were serially diluted in 1% BSA/PBS. To each
well, 25 L of C57BL/6J erythrocytes [prepared from heparinized
blood in PBS as 1% (v/v)] were added. The microtiter plate was
tapped to mix the cells with the plasma dilutions, covered, and incubated at 4C for 2 hours. Agglutination was assessed as the inhibition of the formation of a tight cell button in the bottom of the
wells and is expressed as the last dilution to give a positive agglutination reaction. The direct Coombs assay was carried out with heparinized blood cells that were washed with 1% FBS/PBS. Fifty microliters of a 1% solution of blood cells in 1% FBS/PBS was added
to 50 L of rabbit anti-mouse Ig (heavy and light chain specific).
After mixing, the wells were incubated as described earlier and
scored for agglutination.
236
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
Osmotic fragility
Erythrocytes were collected from the retro-orbital sinus through
heparinized capillary tubes, washed with PBS, and resuspended to
1% (v/v). Aliquots (50 L) of the erythrocyte suspension were
added to 50 L of PBS that had been serially diluted with distilled
water in a 96-well V-bottomed microtiter plate. The cell suspensions were allowed to stand at room temperature for 30 minutes
and then centrifuged to pellet intact cells. A 50-L sample of the
supernatant was transferred to a flat-bottomed microtiter plate and
the optical density at 523 nm was determined with a Spectromax microplate reader (Molecular Devices, Menlo Park, CA, USA). Values were normalized to complete erythrocyte lysis in distilled H2O.
Oxidant production
Oxidant production by Mac-1 bone marrow cells was determined
by flow cytometry using dihydrorhodamine-123 (DH-123, Molecular Probes, Eugene, OR, USA) as previously described [18]. Briefly,
25 106/mL bone marrow cells were incubated with 1 mg/mL rabbit
IgG (Sigma) for 1 hour at 4C. Cells were rinsed, optimized dilutions
of cell lineage-specific antibodies were added, and the cells incubated for 1 hour at 4C. Cells were washed once in FACS buffer
and resuspended at 5 106/mL, and DH-123 was added for a final
concentration of 1 M. Replicate aliquots were stimulated with
phorbol 12-myristate 13-acetate (PMA, Sigma) at time zero, and
all aliquots were transferred to a 37C water bath. Cells were analyzed by flow cytometry at time zero and at 10-minute intervals for
1 hour. At least 10,000 cells were analyzed per sample following
incubation with antibodies for the cell lineage-specific markers
Gr-1 and Mac-1.
Erythrocyte glutathione levels
Blood was drawn into EDTA capillary tubes, centrifuged at 1,000g
for 5 minutes, and 20 L of packed erythrocytes was collected.
The cells were lysed in ice-cold HPLC-grade H2O, centrifuged at
10,000g for 15 minutes at 4C, and the supernatants (erythrocyte
lysates) collected and stored on ice. The erythrocyte lysates were
deproteinated and assayed for glutathione (GSH) according to the
manufacturer’s instructions using a glutathione assay kit (Cayman
Chemical, Ann Arbor, MI, USA). Optical density was measured at
405 nm in a Spectromax microplate reader.
Results
Viable motheaten mice display
significant hematologic anomalies
Evaluation of blood smears from mev/mev mice revealed the
presence of anisocytosis and polychromasia. These changes
in erythrocyte morphology and staining properties are reflective of reticulocytes, which are larger than mature erythrocytes and contain residual RNA (Fig. 1A and B). Specific
staining for reticulum confirmed the presence of greatly elevated numbers of reticulocytes in mev/mev smears (Fig. 1C
and D). Increased reticulocyte numbers are indicative of a responding bone marrow and a normal erythropoietic response.
These findings are consistent with a regenerative anemia.
Young adult mev/mev mice (5–7 weeks of age) do not exhibit
significant abnormalities in erythrocyte numbers or in hematocrit values (Table 1). However, these mice show erythrocyte anomalies consistent with a compensated regenerative
anemia. Erythrocytes from 5- to 7-week-old mev/mev mice
have increased mean cell volumes, decreased mean cell hemoglobin levels, and include a two-fold higher percentage
of reticulocytes when compared to littermate controls. By
12 weeks of age, mev/mev mice are unable to compensate for
their anemia as evidenced by a decrease in erythrocyte numbers despite a 10-fold increase in reticulocyte percentage.
This inability to balance erythrocyte loss with erythrocyte production is not due to reduced Epo levels, as mev/mev mice have
a 50-fold increase in plasma Epo when compared to littermate
control mice (250 mU/mL for mev/mev vs 5 mU/mL for littermate control; pooled samples of six mice for each genotype).
Assay for thiobarbituric reactive substrates
Measurement of thiobarbituric reactive substrates (TBARS) was done
essentially as described by Jain [19]. 1,1,3,3-tetramethoxypropane
[malondialdehyde (MDA)] was used as a standard for these assays.
Optical density measurements were made at 532 and 600 nm using a
Spectromax microplate reader. Quantification was based upon a molar
extinction coefficient 1.56 105M1cm1.
Epo levels
Serum Epo levels were determined using a radioimmunoassay as
previously described [20].
Statistical analysis
Data were analyzed using the Prism Statistical Software package
(GraphPad Software, San Diego, CA, USA). Unpaired t-tests were
done to assess statistical significance. p 0.05 was considered statistically significant.
Figure 1. Blood cell morphology. (A,B) Wright-Giemsa stained blood
smears. (A) /? erythrocytes exhibit normal size and staining. (B) mev/mev
erythrocytes exhibit polychromasia and anisocytosis. (C,D) New methylene blue stained blood smears. (C) Normal numbers of reticulocytes in /?
blood. (D) mev/mev blood film shows increased numbers of reticulocytes.
Reference bar 20 m.
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
237
Table 1. Erythrocyte values for mev/mev and littermate control mice
Age (weeks)
5–7
5–7
12
12
Genotype
RBC 109/mL
Hematocrit (%)
Mean cell volume (fL)
Mean cell hemoglobin (g/dL)
Reticulocyte (%)
/?
mev/mev
/?
mev/mev
9.3 0.2
8.9 0.2
9.5 0.3
6.9 1.0*
47.6 1.0
49.2 1.6
46.9 2.0
44.8 6.3
51.2 0.3
55.4 1.0†
49.2 1.0
65.2 2.5‡
30.1 0.3
27.3 0.3‡
32.4 0.5
25.0 3.6
7.9 1.4
17.6 2.8*
3.3 0.1
30.4 5.8‡
Data are given as mean SEM of values for 4–8 individual mice.
Comparison of age-matched mev/mev and /? blood values show statistically significant differences: *p 0.05; †p 0.01; ‡p 0.001.
Anti-erythrocyte antibodies in mev/mev mice
Viable motheaten mice develop an autoimmune disease
characterized in part by the presence of high levels of circulating autoantibodies [8]. To determine if autoantibody directed against erythrocyte plasma membrane determinants
was present in mev/mev mice, indirect and direct antiglobulin (Coombs) assays were carried out. The indirect Coombs
assay, which detects anti-erythrocyte antibodies in the plasma,
was positive in 11 of 11 mev/mev mice tested. Younger mev/mev
mice had the lowest titers, whereas most mev/mev mice over 70
days of age exhibited higher titers (Fig. 2A). Anti-erythrocyte
autoantibodies were not detected in plasma from any littermate control mice tested (data not shown). The direct Coombs
assay, which detects endogenous antibodies attached to the
erythrocyte membrane, was negative in all mev/mev and littermate control mice tested (data not shown). We also used flow
cytometry to detect any erythrocyte-bound immunoglobulin.
Erythrocytes from mev/mev mice did not display bound IgM
or polyclonal Ig at levels higher than littermate controls (Fig.
2B). Blood from an aged NZB mouse was used as a positive
control [21].
Absence of immunoglobulin production
does not prevent anemia in mev/mev mice
As another measure of the contribution of anti-erythrocyte antibodies to the anemia of mev/mev mice, we examined the hematologic characteristics of mev/mev mice that also were homozygous for a targeted disruption of the membrane exon of
the immunoglobulin gene (Igh-6tm1Cgn) (abbreviated gene
symbol Igh-6null). This targeted mutation results in B-cell deficiency [15]. As previously reported, Igh-6 null mice express
little or no serum immunoglobulin with the exception of
small amounts of IgA [22]. In our laboratory, adult Igh-6 null
mice (14–30 weeks old) had less than 3.1 g/mL of serum Ig
by enzyme-linked immunosorbent assay (data not shown).
This targeted mutation was combined with the mev mutation
by selective breeding. Mice homozygous for both mev/mev
and the Igh-6null allele showed significant changes in erythrocyte parameters similar to age-matched mev/mev mice (Table 2). Erythrocytes from 8-week-old mev/mev Igh-6null mice
had increased mean cell volumes, decreased mean cell hemoglobin levels, and a higher percentage of reticulocytes than
did age-matched /mev Igh-6 null controls (Table 2). These results are consistent with a compensated anemia, as is seen in
young mev/mev mice. This confirms that the presence of antierythrocyte autoantibody is not essential for the development of anemia in mev/mev mice.
Figure 2. Anti-erythrocyte antibodies. (A) Plasma from 11 mev/mev mice were
collected at 20 through 80 days of age and tested by the indirect Coombs
assay. No anti-erythrocyte antibodies were detected in 11 age-matched littermate controls (data not shown). (B) Peripheral blood from NZB, /?, and
mev/mev mice was incubated with fluorescent labeled anti–TER-119 and
polyclonal anti-Ig antibodies and analyzed by flow cytometry. Unlabeled
cells from a /? mouse showed no autofluorescence and were used to set
the quadrant markers (FL1 FL2). The percentage of double-positive cells
is indicated in the upper right quadrant. Data are representative of four independent assays.
238
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
Table 2. Effect of immunoglobulin deficiency on erythrocyte values
Age (weeks)
8
8
Genotype
RBC 109/mL
Hematocrit (%)
Mean cell volume (fL)
Mean cell hemoglobin (g/dL)
Reticulocyte (%)
/mev Igh-6null
mev/mev Igh-6null
8.6 0.2
9.2 0.5
43.8 0.6
53.4 2.4*
51.4 1.1
57.9 1.2*
32.0 0.3
28.5 0.2†
3.5 0.3
11.0 3.2
Data are given as mean SEM of values for 3–5 individual mice.
*p 0.05; †p 0.01.
Evidence of membrane damage in mev/mev erythrocytes
Normal shape and deformability are critical to erythrocyte survival in vivo. Because defects in the structural integrity of the
erythrocyte membrane render them susceptible to premature
removal from circulation, we examined erythrocytes from mev/
mev mice by electron microscopy. Erythrocytes from littermate
control mice had a normal discoid shape, whereas erythrocytes
from mev/mev mice showed mild deformation of the erythrocyte membrane (Fig. 3A and B). Moderate deformation of
the erythrocyte membrane was evident in erythrocytes held
for 1 hour at room temperature prior to fixation (Fig. 3C and
D). Blood samples from mice of both genotypes contained
echinocytes, but echinocytes were more numerous in mev/mev
blood. These membrane alterations were exacerbated by incubation of the erythrocytes with H2O2, a strong oxidant, prior
to fixation. Blood from littermate control mice subjected to
H2O2 incubation resulted in a small number of echinocytes,
whereas H2O2-treated mev/mev blood contained a large pro-
Figure 3. Erythrocyte ultrastructural features. Erythrocytes were fixed
immediately following collection or following 1-hour incubation at room
temperature and examined by scanning electron microscopy. Most wildtype erythrocytes had a normal discoid appearance (A), whereas many mev/
mev erythrocytes showed mild deformation of the erythrocyte membrane
(B). These differences in membrane morphology were exacerbated by an
hour-long incubation at room temperature (C, wild-type and D, mev/mev).
Reference bar 10 m.
portion of echinocytes (Fig. 4A and B). Erythrocytes from
/?Igh-6null mice showed minimal morphologic changes,
whereas blood from mev/mev Igh-6null mice contained a large
proportion of echinocytes similar to the numbers observed in
mev/mev blood (Fig. 4C and D). Because metallothionein is
an important antioxidant in erythrocytes, we also explored
the impact of overexpression and targeted disruption of met-
Figure 4. Erythrocytes treated with H2O2 developed membrane anomalies. (A) Erythrocytes from /? and (C) /? Igh-6null mice were minimally
affected by H2O2. A marked increase in echinocyte (e) formation was
observed in blood from (B) mev/mev or (D) mev/mev Igh-6null mice incubated with H2O2. Metallothionein levels also affected H2O2 susceptibility.
Erythrocytes from Mt1 transgenic mice were only slightly affected by
H2O2 (E), whereas erythrocytes from mice with a targeted disruption of
metallothionein 1 and metallothionein 2 were completely destroyed by oxidant treatment (F). Reference bar 10 m.
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
239
allothionein gene expression on oxidant-induced echinocyte
formation. Erythrocytes from metallothionein-1 (Mt1) transgenic mice appeared resistant to oxidant induced injury (Fig.
4E), whereas erythrocytes from mice with a targeted disruption of the Mt1 and Mt2 genes were exquisitely sensitive to
oxidant-induced injury as demonstrated by the absence of intact erythrocytes (Fig. 4F).
Increased osmotic fragility of mev/mev red blood cells
Because subjecting mev/mev erythrocytes to exogenous oxidant resulted in membrane alterations far more dramatic
than those observed with wild-type erythrocytes, we evaluated the osmotic fragility of erythrocytes from mev/mev and
littermate control mice. Whole blood was incubated in serial
dilutions of increasingly hypotonic saline to assess fragility.
Erythrocytes from mev/mev mice were significantly more fragile than control erythrocytes at both 60% and 50% (v/v) PBS.
At 40% and 30% (v/v) PBS, erythrocytes from mev/mev and
control mice were equally sensitive to hypotonic lysis (Fig. 5).
Oxidant production by Mac-1
bone marrow cells in mev/mev mice
An important mechanism by which myeloid cells kill invading microbial organisms is dependent upon the generation
of reactive oxygen intermediates. The overproliferation and
chronic activation of myeloid cells in mev/mev mice suggest
that these cells might be a potent source of exogenous oxidant. In order to examine this possibility, we labeled bone
marrow cells from mev/mev and littermate control mice with
DH-123, a cell permeant and nonfluorescent probe that is
converted to fluorescent rhodamine-123 in the presence of
oxidants (Fig. 6) [18]. Mac-1 bone marrow cells from mev/
mev mice had significantly elevated basal levels of oxidant
production compared to littermate control mice. After stimulation with PMA, Mac-1 cells from both mev/mev and littermate control mice demonstrated a substantial increase in
oxidant production; however, oxidant production from PMA-
Figure 5. Evaluation of erythrocyte osmotic fragility. Erythrocytes from
11- to 14-week-old mev/mev and /? littermate control mice were exposed
to increasingly hypotonic saline solutions. Hemoglobin release was measured as an indicator of erythrocyte lysis. Values are expressed as a fraction of the distilled H2O-lysed cells from the same sample (taken as 100%
lysis) and represent the mean of five mice SEM. *p 0.005.
Figure 6. Oxidant production in Mac-1high, Gr-1low bone marrow cells.
Bone marrow cells from 14-week-old mev/mev and /? littermate mice were
labeled with dihydrorhodamine and then cultured in the presence (filled
symbols) or absence (open symbols) of PMA. Cells gated for Mac-1high
expression were analyzed for fluorescence over time. Data are expressed as
the mean fluorescence intensity. Data are representative of three independent experiments. Square /?; circle mev/mev.
stimulated mev/mev Mac-1 cells was markedly elevated
compared with PMA-stimulated control Mac-1 cells.
Erythrocytes from mev/mev mice
have higher levels of lipid peroxides
The breakdown of polyunsaturated fatty acids in the lipid
bilayer of membranes produces MDA, which is a sensitive
indicator of lipid peroxidation. To assess oxidant damage in
mev/mev erythrocytes, we used the TBARS assay in which
MDA reacts with thiobarbituric acid to form a colored complex that is used to quantify the level of lipid peroxidation.
Erythrocytes from mev/mev mice had a significantly higher
level of lipid peroxides compared with erythrocytes from control mice (Fig. 7).
Glutathione levels in mev/mev erythrocytes
The observation of increased oxidant production in mev/mev
bone marrow cells and lipid peroxidation of mev/mev erythrocyte membranes suggested that accumulating oxidant damage might affect the intracellular pool of antioxidant. To evaluate erythrocyte antioxidant levels, total glutathione (GSH)
Figure 7. Lipid peroxidation is higher in mev/mev erythrocytes than in control erythrocytes. Erythrocytes from mev/mev and littermate control mice were
harvested for analysis in the TBARS assay. Values presented represent the
means of triplicate values for each of five mev/mev and littermate controls SEM. *p 0.001.
240
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
was measured. GSH levels were significantly depleted in
mev/mev erythrocytes compared to littermate control erythrocytes (4.15 0.51 M GSH/mg protein in mev/mev erythrocytes vs 6.52 0.57 in normal erythrocytes, p 0.05).
Externalization of phosphatidylserine
on immature erythrocytes of mev/mev mice
To determine if the decreased survival of erythrocytes in mev/
mev mice affected mature or immature erythroid populations
disproportionately, we examined erythrocytes from mev/mev
and littermate control mice for phosphatidylserine (PS) exposure. Externalization of PS on the outer leaflet of the plasma
membrane has been proposed as an important signal for
erythrocyte removal from the circulation [23,24]. Peripheral
blood was immunolabeled with the erythroid-specific TER119 antibody and the transferrin receptor (CD71) antibody.
CD71 is expressed by erythroid precursors and reticulocytes
but is not expressed on mature erythrocytes [25]. The cells
then were incubated with annexin V, which binds exposed
PS. After gating on the TER-119–positive erythrocyte lineage cells, cells from mev/mev and littermate control mice
were analyzed for anti-CD71 and annexin V binding. Peripheral blood from mev/mev mice had a higher percentage of immature erythrocytes, and an increased percentage of these erythrocytes had PS on the external leaflet of the plasma membrane
compared with littermate control blood (2.82% 0.82% mev/
mev erythrocytes were CD71AnV vs 0.55 % 0.05% on
control erythrocytes, p 0.05) (Fig. 8). Intriguingly, mev/
mev mice also had a lower percentage of mature erythrocytes with PS exposure (0.86% 0.34% mev/mev erythrocytes were CD71AnV vs 2.45% 0.10% on control
erythrocytes, p 0.05).
CD47 expression on mev/mev erythrocytes
Retention of erythrocytes in the circulation is associated
with expression of CD47 on the plasma membrane [26]. This
integrin-associated protein is expressed on the external leaf-
Figure 8. CD71 expression and annexin V binding in TER-119 gated
erythroid cells. Peripheral blood from 16- to 22-week-old mev/mev and littermate control mice was labeled with anti–TER-119, anti-CD71, and annexin
V. After gating on the TER-119 erythroid cells, the levels of CD71 and
annexin V were assessed. Dot plots are from individual mice of each genotype.
Data are representative of five mice analyzed per genotype.
let of the erythrocyte plasma membrane and interacts with
macrophage signal regulatory protein- (SIRP-) on phagocytic cells. Ligation of the inhibitory SIRP- by CD47 on the
erythrocyte surface prevents engulfment by phagocytes [26].
This inhibitory signal is mediated by SHP-1. Binding of
SHP-1 to SIRP- results in the interruption of signaling
through tyrosine kinase-dependent activation pathways. To
determine if interaction between SIRP- and CD47 contributed to the anemia of SHP-1–deficient mev/mev mice, we examined TER-119 cells for CD47 expression. Levels of CD47
were found to be higher on mev/mev erythrocytes than on littermate control erythrocytes (Fig. 9).
Discussion
SHP-1 deficiency in motheaten and viable motheaten mice
results in a wide variety of immune defects. Extensive studies of the immune system in these mutant mice have provided valuable insight into intracellular signaling pathways
in the hematopoietic system [8,27,28]. However, the mechanisms underlying the progressive anemia in these mice
have not been clearly established [7].
Previous reports have suggested that abnormalities in erythrocyte proliferation and differentiation are linked to decreased
SHP-1 activity [10,29]. Reduced SHP-1 expression in erythroid progenitor cells has been demonstrated in a subset of
patients with PV, a clonal myeloproliferative disorder that
results in erythroid cell hyperproliferation [9]. However, an
earlier report failed to find alterations in SHP-1 protein expression in granulocytes from PV patients [30]. Although it
is difficult to assess the role of SHP-1 in the pathogenesis of
PV, nevertheless, mev/mev mice and PV patients exhibit similar erythroid defects and may share similar mechanistic anomalies in the regulation of erythropoiesis. SHP-1 has been
Figure 9. CD47 expression on mev/mev and littermate control erythrocytes. Whole blood from mev/mev and littermate control mice was labeled
with fluorescent anti-CD47 and analyzed by flow cytometry. Open histogram shows unlabeled littermate control erythrocytes. Values represent the
mean fluorescent intensity. Data are expressed as mean SEM of values
for three individual mice. *p 0.05.
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
shown to be important in the regulation of erythropoiesis by
inhibiting signaling through the Epo receptor [31]. SHP-1–
deficient mev/mev mice have a dramatic increase in the CFU-E
precursor population of the spleen. These precursor cells are
much more sensitive to Epo, with a subpopulation of these
cells exhibiting Epo-independent proliferation [10]. The increased numbers of erythroid progenitors accompanied by
increased responsiveness to Epo would predict that mev/mev
mice develop an erythrocytosis, but the cellular composition of
the blood in these mice instead reflects a progressive anemia.
The severity of this anemia suggests that it is not solely
a consequence of the low levels of anti-erythrocyte autoantibodies that can be detected in the plasma of mev/mev mice
by indirect Coombs assay. We found no evidence that these
antibodies bind in significant amounts to mev/mev erythrocytes in vivo. Neither direct Coombs assays nor flow cytometric analysis of TER-119 erythroid cells detected antierythrocyte antibodies on the surface of mev/mev erythrocytes
at levels in excess of normal littermate control levels. In contrast, autoimmune hemolytic anemia in NZB mice is associated with significant amounts of anti-erythrocyte antibodies
bound to the surface of circulating erythrocytes. Additional
evidence that autoantibodies are not responsible for anemia
in mev/mev mice is that anemia persists despite the absence
of antibodies in mev/mev Igh-6null mice.
Erythrocytes are subjected to wide fluctuations in oxygen
exposure, ranging from approximately 40 mmHg oxygen pressure (Po2) in the tissues to 104 mmHg Po2 as they pass through
the lung vasculature [32]. Normally cellular antioxidant systems offset the deleterious effects of oxygen. Erythrocytes use
several antioxidant mechanisms, including glutathione [33],
the principal nonprotein thiol in the cell, and metallothionein
[34], the principal thiol-containing protein in the cell. Moreover, a number of anti-oxidant enzyme systems limit erythrocyte oxidative damage.
The presence of echinocytes in the blood of mev/mev mice
is consistent with oxidative damage to the plasma membranes
[35]. Erythrocytes exposed to room air at room temperature
experience oxidation by dissolved oxygen from the atmosphere. Erythrocytes from mev/mev mice developed echinocyte
morphology during a 1-hour incubation, and the mev/mev erythrocyte population was highly enriched for echinocytes following incubation with exogenous H2O2.
The severe morphologic changes observed following exposure of mev/mev erythrocytes to hydrogen peroxide suggested increased fragility of mev/mev erythrocytes, a conjecture that was borne out by the increased sensitivity of mev/
mev erythrocytes to hypotonic lysis. Exposure of erythrocytes to hypotonic saline at 60% of normal tonicity caused
dramatic hemolysis of mev/mev erythrocytes, whereas littermate control erythrocytes demonstrated little hemolysis. This
fragility may be a consequence of chronic exposure to exogenous oxidant produced by the persistent activation of myeloid
lineage cells in mev/mev mice. Macrophages from mev/mev
mice have an elevated level of oxidant production compared
241
to normal macrophages, and they respond to phorbol ester
activation with a more robust oxidative burst. These observations are consistent with previous studies demonstrating
that mev/mev neutrophils display higher than normal levels
of oxidant production [27].
Our findings suggest that the mev/mev erythrocyte plasma
membrane may be damaged as a consequence of exposure to
exogenous oxidants. Erythrocyte plasma membranes from mev/
mev mice contained markedly elevated levels of lipid peroxides
(as measured by the TBARS assay) compared with littermate
control erythrocytes. Reflective of persistent severe oxidative
stress, the primary nonprotein thiol antioxidant in cells, glutathione (GSH), was significantly depleted in mev/mev erythrocytes.
Identification of erythrocytes for clearance from the circulation by splenic macrophages is mediated by reduced expression of CD47 on the erythrocyte plasma membrane. This
integrin-associated protein functions as a marker of self on
erythrocytes [26]. Binding of CD47 to macrophage SIRP-
inhibits erythrocyte phagocytosis [26]. Signaling through
SIRP- is mediated by SHP-1 [36]. SHP-1–deficient mev/
mev mice have an accelerated rate of clearance of opsonized
erythrocytes; however, there is no difference in the clearance of unopsonized erythrocytes in mev/mev mice compared
with that in wild-type mice [36]. These data suggest that
even though the signaling through the CD47/SIRP- complex is attenuated in mev/mev mice, it still is sufficient to prevent phagocytosis of erythrocytes in the absence of other prophagocytic stimuli. Our observation that mev/mev erythrocytes
express elevated levels of CD47 may reflect the fact that there
is a larger percentage of immature erythrocytes in mev/mev
mice, as aged erythrocytes demonstrate reduced CD47 expression [37].
The presence of anisocytosis and polychromasia in blood
smears from mev/mev mice is consistent with an increased
proportion of immature erythrocytes in mev/mev blood. This
is supported by the increased number of reticulocytes in
mev/mev blood smears and by the increased frequency of
TER-119 erythrocytes that coexpress CD71, a marker of
immature erythrocytes. The mechanism by which immature
mev/mev erythrocytes are removed from the circulation appears to be associated with PS exposure, because these cells
show elevated annexin V binding compared to littermate control erythrocytes. Alternatively, a study has demonstrated that
annexin V will also bind to MDA on the plasma membrane
[38]. This product of lipid peroxidation of the plasma membrane was shown to be elevated in mev/mev erythrocytes by
the TBARS assay. Recognition of membrane damage in mev/
mev erythrocytes and consequent removal by phagocytes may
be MDA dependent. These data suggest that in mev/mev mice,
either immature erythroid cells are more susceptible to oxidant
induced membrane damage or cumulative membrane damage
prevents most mev/mev erythroid cells from reaching maturity.
Our results support the hypothesis that mev/mev erythrocytes are exposed to a highly oxidative environment sec-
242
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
ondary to myeloid proliferation and that this environment
causes oxidative damage with subsequent removal of erythrocytes from the circulation. Other anemias, such as sickle
cell, thalassemia, malaria, and erythropoietic protoporphyria have a hemolytic component secondary to oxidative
damage [3]. Moreover, exposure to a wide variety of oxidative toxicants can result in anemia [39]. Viable motheaten
mice may prove to be a useful model of these forms of anemia. Additionally, the role of oxidant in the progression of
this anemia may be susceptible to manipulation by antioxidants such as N-acetylcysteine or vitamin E. Viable motheaten mice may be an excellent model for assessment of the
therapeutic effect of antioxidants.
14.
15.
16.
17.
18.
Acknowledgments
This work was supported in part by National Institutes of Health
Grants CA20408 (L.D.S.), T32RR07068 (B.L.L.), and ES07408
(M.A.L.). The authors thank Dr. Eugene Goldwasser (University
of Chicago, Chicago, IL) for measurement of erythropoietin levels
and Drs. Jane Barker and David Serreze (The Jackson Laboratory,
Bar Harbor, ME) for critical review of this manuscript.
References
1. Cotran R. Red cells and bleeding disorders. In: Cotran RS, Kumar V,
Collins T, eds. Robbins Pathologic Basis of Disease, 6th ed. Philadelphia: W.B. Saunders; 1999, p. 601–643.
2. Baxter AG, Mandel TE. Hemolytic anemia in non-obese diabetic
mice. Eur J Immunol. 1991;21:2051–2055.
3. Sen G, Mukhopadhyay R, Ghosal J, Biswas T. Oxidative damage of
erythrocytes: a possible mechanism for premature hemolysis in experimental visceral leishmaniasis in hamsters. Ann Hematol. 2001;80:32–37.
4. Shultz LD, Schweitzer PA, Rajan TV, et al. Mutations at the murine
motheaten locus are within the hematopoietic cell protein-tyrosine
phosphatase (Hcph) gene. Cell. 1993;73:1445–1454.
5. McCulloch J, Siminovitch KA. Involvement of the protein tyrosine
phosphatase PTP1C in cellular physiology, autoimmunity and oncogenesis. Adv Exp Med Biol. 1994;365:245–254.
6. Zhang J, Somani AK, Siminovitch KA. Roles of the SHP-1 tyrosine
phosphatase in the negative regulation of cell signalling. Semin Immunol. 2000;12:361–378.
7. Shultz LD, Coman DR, Bailey CL, Beamer WG, Sidman CL. “Viable
motheaten,” a new allele at the motheaten locus. I. Pathology. Am J
Pathol. 1984;116:179–192.
8. Shultz LD, Rajan TV, Greiner DL. Severe defects in immunity and hematopoiesis caused by SHP-1 protein-tyrosine-phosphatase deficiency. Trends Biotechnol. 1997;15:302–307.
9. Wickrema A, Chen F, Namin F, et al. Defective expression of the SHP-1
phosphatase in polycythemia vera. Exp Hematol. 1999;27:1124–
1132.
10. Van Zant G, Shultz L. Hematologic abnormalities of the immunodeficient mouse mutant, viable motheaten (mev). Exp Hematol. 1989;
17:81–87.
11. Arvieux J, Roussel B, Ponard D, Colomb MG. Reactivity patterns of
anti-phospholipid antibodies in systemic lupus erythematosus sera in
relation to erythrocyte binding and complement activation. Clin Exp
Immunol. 1991;84:466–471.
12. Arvieux J, Schweizer B, Roussel B, Colomb MG. Autoimmune
haemolytic anaemia due to anti-phospholipid antibodies. Vox Sang. 1991;
61:190–195.
13. Bailey NC, Bona A, Dikman S, Bonilla F, Alt FW, Bona C. Correla-
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
tion between the occurrence of lupus nephritis, anti-erythrocyte autoantibodies and V kappa haplotype in NZB x 129/J and NZB x SM/J
recombinant inbred murine strains. Eur J Immunol. 1991;21:959–965.
Pani G, Siminovitch KA, Paige CJ. The motheaten mutation rescues B
cell signaling and development in CD45-deficient mice. J Exp Med.
1997;186:581–588.
Kitamura D, Roes J, Kuhn R, Rajewsky K. A B-cell-deficient mouse
by targeted disruption of the membrane exon of the immunoglobulin
mu chain gene. Nature. 1991;350:423–426.
Palmiter RD, Sandgren EP, Koeller DM, Brinster RL. Distal regulatory elements from the mouse metallothionein locus stimulate gene expression in transgenic mice. Mol Cell Biol. 1993;13:5266–5275.
Masters BA, Kelly EJ, Quaife CJ, Brinster RL, Palmiter RD. Targeted
disruption of metallothionein I and II genes increases sensitivity to
cadmium. Proc Natl Acad Sci U S A. 1994;91:584–588.
Emmendorffer A, Hecht M, Lohmann-Matthes ML, Roesler J. A fast
and easy method to determine the production of reactive oxygen intermediates by human and murine phagocytes using dihydrorhodamine
123. J Immunol Methods. 1990;131:269–275.
Jain SK. Hyperglycemia can cause membrane lipid peroxidation and osmotic fragility in human red blood cells. J Biol Chem. 1989;264:21340–
21345.
Sherwood JB, Goldwasser E. A radioimmunoassay for erythropoietin.
Blood. 1979;54:885–893.
Ballas SK, Tabbara KF, Murphy DL, Mohandas N, Clark MR, Shohet
SB. Erythrocyte deformability changes in autoimmune hemolytic anemia during development of NZB mice and their (NZB/NZW)F1 hybrid. J Clin Lab Immunol. 1985;16:217–222.
Macpherson AJ, Lamarre A, McCoy K, et al. IgA production without
mu or delta chain expression in developing B cells. Nat Immunol.
2001;2:625–631.
Connor J, Pak CC, Schroit AJ. Exposure of phosphatidylserine in the
outer leaflet of human red blood cells. Relationship to cell density, cell
age, and clearance by mononuclear cells. J Biol Chem. 1994;269:2399–
2404.
Boas FE, Forman L, Beutler E. Phosphatidylserine exposure and red
cell viability in red cell aging and in hemolytic anemia. Proc Natl Acad
Sci U S A. 1998;95:3077–3081.
Lok CN, Ponka P. Identification of an erythroid active element in the
transferrin receptor gene. J Biol Chem. 2000;275:24185–24190.
Oldenborg PA, Zheleznyak A, Fang YF, Lagenaur CF, Gresham HD,
Lindberg FP. Role of CD47 as a marker of self on red blood cells. Science. 2000;288:2051–2054.
Kruger J, Butler JR, Cherapanov V, et al. Deficiency of Src homology 2containing phosphatase 1 results in abnormalities in murine neutrophil
function: studies in motheaten mice. J Immunol. 2000;165:5847–5859.
Zhang J, Somani AK, Yuen D, Yang Y, Love PE, Siminovitch KA. Involvement of the SHP-1 tyrosine phosphatase in regulation of T cell
selection. J Immunol. 1999;163:3012–3021.
Sharlow ER, Pacifici R, Crouse J, Batac J, Todokoro K, Wojchowski
DM. Hematopoietic cell phosphatase negatively regulates erythropoietin-induced hemoglobinization in erythroleukemic SKT6 cells. Blood.
1997;90:2175–2187.
Asimakopoulos FA, Hinshelwood S, Gilbert JG, et al. The gene encoding hematopoietic cell phosphatase (SHP-1) is structurally and transcriptionally intact in polycythemia vera. Oncogene. 1997;14:1215–1222.
Klingmuller U, Lorenz U, Cantley LC, Neel BG, Lodish HF. Specific
recruitment of SH-PTP1 to the erythropoietin receptor causes inactivation of JAK2 and termination of proliferative signals. Cell. 1995;80:
729–738.
Guyton AC, Hall JE. Transport of oxygen and carbon dioxide in the
blood and body fluids. In: Guyton AC, Hall JE, eds. Textbook of
Medical Physiology, 10 ed. Philadelphia: W.B. Saunders; 2000 p.
463–473.
Johnson RM, Ravindranath Y, el-Alfy M, Goyette G Jr. Oxidant damage to erythrocyte membrane in glucose-6-phosphate dehydrogenase
B.L. Lyons et al./Experimental Hematology 31 (2003) 234–243
deficiency: correlation with in vivo reduced glutathione concentration
and membrane protein oxidation. Blood. 1994;83:1117–1123.
34. Rahman MT, De Ley M. Metallothionein isogene transcription in red
blood cell precursors from human cord blood. Eur J Biochem. 2001;
268:849–856.
35. Chukhlovin AB. Apoptosis and red blood cell echinocytosis: common
features. Scanning Microsc. 1996;10:795–803.
36. Oldenborg PA, Gresham HD, Lindberg FP. CD47-signal regulatory
protein alpha (SIRPalpha) regulates Fcgamma and complement receptor-mediated phagocytosis. J Exp Med. 2001;193:855–862.
243
37. Fossati-Jimack L, Azeredo da Silveira S, Moll T, et al. Selective increase of autoimmune epitope expression on aged erythrocytes in mice:
implications in anti-erythrocyte autoimmune responses. J Autoimmun.
2002;18:17–25.
38. Balasubramanian K, Bevers EM, Willems GM, Schroit AJ. Binding of
annexin V to membrane products of lipid peroxidation. Biochemistry.
2001;40:8672–8676.
39. Liu J, Liu Y, Habeebu SS, Klaassen CD. Metallothionein-null mice are
highly susceptible to the hematotoxic and immunotoxic effects of chronic
CdCl2 exposure. Toxicol Appl Pharmacol. 1999;159:98–108.