Cell, Vol. 118, 757–766, September 17, 2004, Copyright 2004 by Cell Press
Identification of a Human Heme Exporter
that Is Essential for Erythropoiesis
through a balance among heme biosynthesis, utilization
for hemoproteins, and catabolism by the microsomal
enzyme, heme oxygenase (HO) (Ponka, 1997), to gener-
ate carbon monoxide, iron, and biliverdin.
Red cells have a unique requirement for heme as the
prosthetic group of hemoglobin, which in the mature
erythrocyte comprises over 90% of cellular protein con-
tent. Thus, coordinate regulation of heme and globin
synthesis isrequired to avoidheme toxicityyet preserve
genitors, burst-forming units-erythroid (BFU-E), mature
to colony-forming units-erythroid (CFU-E) and then ery-
iron levels (via transferrin-iron uptake), and heme bio-
synthesis initiates (Ponka, 1997; Wickrema et al., 1992).
Previous studies have demonstrated that heme induces
the erythroid differentiation of erythroid cell lines (e.g.,
it has been shown that intracellular heme regulates both
the transcription and translation of globin mRNAs
through interference with the DNA binding of Bach1, a
transcriptional repressor of the globin genes, and inhibi-
tion of an erythroid-specific eIF-2? kinase (Ogawa et al.,
2001; Rafie-Kolpin et al., 2000). Thus, it appears that
increasing intracellular hemelevels during early erythro-
poiesis triggers the onset of globin protein synthesis,
which ultimately reduces the levels of uncommitted
heme. The response of nonerythroid cells to any excess
in intracellular heme includes the rapid induction of HO.
development of erythroid progenitors, induction of this
enzyme in erythroid cells may be inappropriate given its
prolonged half-life (20 hr; Ibrahim et al., 1982), sug-
Previously, we and others cloned FLVCR (Quigley et
al., 2000; Tailor et al., 1999), a cell surface protein that
serves as the receptor for feline leukemia virus, sub-
group C (FeLV-C). By homology, FLVCR is a member of
the major facilitator superfamily (MFS; Pao et al., 1994)
of secondary permeases, which transport small solutes
(e.g., sugars, amino acids) across membranes in re-
sponse to chemico-osmotic gradients. However, the
transport function of this viral receptor was unknown.
Cats viremic with FeLV-C, a simple (nononcogene-
containing) retrovirus, develop profound anemia (Ab-
1988; Wardrop et al., 1986); while BFU-E are present
at normal frequency, there is a paucity of CFU-E and
erythroid precursors, suggesting that erythropoiesis is
arrested at the CFU-E/proerythroblast stage (Abkowitz
et al., 1987). Studies indicate that in infected cells, the
FeLV-C envelope surface unit protein acts as a domi-
tion of its cell surface receptor (feFLVCR) and cause this
phenotype (Weiss and Tailor, 1995; Riedel et al., 1988;
Rigby et al., 1992). Further investigations using viral en-
virus whichalso infectsall felinebone marrow(BM) cells
(Dean et al., 1992), demonstrate that replacement of as
John G. Quigley,1,6,7Zhantao Yang,1,6
Mark T. Worthington,3John D. Phillips,4
Kathleen M. Sabo,1Daniel E. Sabath,2
Carl L. Berg,3Shigeru Sassa,5Brent L. Wood,2
and Janis L. Abkowitz1,*
1Department of Medicine/Hematology
2Department of Laboratory Medicine
University of Washington
Seattle, Washington 98195
3Department of Medicine/Gastroenterology
University of Virginia
Charlottesville, Virginia 22908
4Department of Medicine
University of Utah
Salt Lake City, Utah 84132
5Laboratory of Biochemical Hematology
New York, New York 10021
FLVCR, a member of the major facilitator superfamily
of transporter proteins, is the cell surface receptor for
feline leukemia virus, subgroup C. Retroviral interfer-
ence with FLVCR display results in a loss of erythroid
progenitors (colony-forming units-erythroid, CFU-E)
and severe anemia in cats. In this report, we demon-
strate that human FLVCR exports cytoplasmic heme
and hypothesize that human FLVCR is required on
developing erythroid cells to protect them from heme
toxicity. Inhibition of FLVCR in K562 cells decreases
heme export, impairs their erythroid maturation and
leads to apoptosis. FLVCR is upregulated on CFU-E,
indicating that heme export is important in primary
cells at this stage. Studies of FLVCR expression in
cell lines suggest this exporter also impacts heme
trafficking in intestine and liver. To our knowledge,
Heme, a complex of iron and protoporphyrin IX, is an
important component of a diverse group of hemopro-
teins, including those involved in oxygen transport and
storage (hemoglobin, myoglobin), electron transfer and
drug metabolism (cytochromes), and signal transduc-
tion (nitric oxide synthases) (Ponka, 1997). Heme, how-
ever, is toxic, promoting oxidative cell membrane dam-
age through lipid peroxidation, which necessitates tight
regulation of heme’s intracellular concentration (Ryter
and Tyrrell, 2000). This regulation is thought to occur
6These authors contributed equally to this work.
7Present address: Section of Hematology/Oncology (MC 734), De-
partment of Medicine, University of Illinois at Chicago, Chicago,
few as 11 amino acids of FeLV-A envelope variable re-
gion 1 (VR1) with amino acids contained in the VR1 of
FeLV-C alters both the host range and receptor use of
cats (Rigby et al., 1992). Thus, the genetic determinants
of viral binding to feFLVCR and the impairment of ery-
throid differentiation map to VR1 of the FeLV-C enve-
lope. Since feFLVCR expression is downregulated by
intracellular envelope production in all infected BM cells
(including BFU-E and CFU-GM; Abkowitz et al., 1987),
the specific loss of CFU-E implies that this transporter
serves a cellular function uniquely required for CFU-E
differentiation or survival (a conclusion that is further
supported by studies in which normal feline progenitors
infected in vitro with FeLV-C/Sarma have deficient ery-
throid, but preserved myeloid, differentiation [K.M.S.
and J.L.A., unpublished data]).
Here, we demonstrate that FLVCR functions as an
exporter of cytoplasmic heme and provide evidence of
FLVCR’s importance for human erythroid cell differenti-
FLVCR Expression in Cells and Cell Lines
FLVCR is widely expressed as demonstrated by North-
ern blot analysis (Tailor et al., 1999; Z.Y., J.G.Q., and
J.L.A., unpublished data) and by the broad range of
tissues easily infected by FeLV-C in vivo and in vitro
(Dean et al., 1992 and references therein). Thus, as an
initial study to gain insight into FLVCR function, we
quantitated human FLVCR mRNA and cell surface pro-
tein expression in cell lines and primary cells (Figure
1) using RT-PCR and a polyclonal antibody specific to
FLVCR (?-FLVCR, Figure 2). These experiments demon-
high in Caco-2 (small intestinal phenotype) and HepG2
(hepatic phenotype) cells—cell lines previously utilized
ton et al., 2001). In addition, as predicted from Unigene
and FLVCR EST expression information (enter FLVCR
as a search term at http://source.stanford.edu), FLVCR
may have increased heme utilization due to rapid cell
turnover. Notably, mRNA and protein levels are high in
mobilized peripheral blood (PB) CD34?stem/progenitor
cells and in hematopoietic cell lines with erythroid fea-
tures (e.g., K562 and HEL-DR), but cell surface protein
expressionis absentin amore matureerythroid cellline,
HEL-R, with spontaneous hemoglobinization, which is
derived from HEL-DR (Papayannopoulou et al., 1987).
In some cells, there were disparities between protein
and mRNA expression levels (Figure 1). These dispari-
ties may reflect posttranscriptional regulation of protein
levels, for example, regulation of the rate of cycling of
FLVCR protein to and from the cell surface, a control
mechanism important in the cell surface expression of
another MFS permease, the Glut-4 transporter (Bryant
et al., 2002).
Figure 1. FLVCR Expression in Human Cells
Small intestinal (Caco-2) and hepatic phenotype (HepG2) cell lines
quantitative RT-PCR and flow cytometry with a polyclonal antibody
specific to FLVCR, ?-FLVCR, respectively. Studies of cancer cell
lines of breast (MCF7), neuroblastoma (SH-SY5Y), lung (H-460 and
are also shown. Mobilized PB CD34?cells and cell lines with an
protein on their cell surface, while the more mature erythroid line,
HEL-R, fails to express FLVCR protein.
tent of a rat renal epithelial cell line (NRK) engineered
to express the feline ortholog of FLVCR, NRK/feFLVCR,
was compared with that of control NRK cells. NRK cells
are resistant to FeLV-C infection, implying that the ro-
dent ortholog of FLVCR is poorly expressed or does not
efficiently bind FeLV-C; however, once transduced with
feFLVCR, the cells can be easily infected by FeLV-C
cellswas significantly(though minimally)lower thanthat
of control cells (10.71 vs. 11.93 pmol heme/106cells,
p ? 0.04 by two-tailed Student’s t test). We reasoned
that compensatory events during the in vitro selec-
tion and expansion of the cell lines might account for
the smallness of the difference in steady-state heme
Wethen examinedthe hemecontent ofcell linesupon
viral interference with FLVCR cell surface expression.
Feline embryonic fibroblasts (FEA) are readily infected
feFLVCR (Quigley et al., 2000; Rudra-Ganguly et al.,
Cellular Heme Content Is Dependent on FLVCR
Cell Surface Expression
As an initial screen, to determine if FLVCR could have
a role in heme trafficking or metabolism, the heme con-
FLVCR Exports Cytoplasmic Heme
at baseline ? 0, after the 30 min load ? 74.9% ? 10.5%,
and after 90 min of washout [120 min total time] ?
36.5% ? 5.2%; p ? 0.001), and thus 51.3% ? 7.1% of
ZnMP was exported. However, the MFI of control cells
did not change significantly and export was 5.3% ?
0.6% (p ? NS) (Figure 3B). In addition, there was no
when washout experiments were performed at 4?C,
demonstrating that export is a temperature-dependent
process (Figures 3A and 3B).
The uptake and washout studies were repeated using
55Fe-hemin so that heme export could be directly as-
sessed. These studies required preincubation with
ZnMP to prevent the breakdown of
cytosolic enzyme HO and thus confirm that both ZnMP
and55Fe-hemin were internalized. Cells were incubated
with55Fe-hemin then placed in washout buffer, and the
decrease in the cells’ radioactivity was measured over
time. As shown in Figure 3C, the results were compara-
ble to those seen when the export of the mesoporphyrin
content ofcontrol NRK/ev cellsand NRK/14qcells, cells
engineered to overexpress an FLVCR paralog on chro-
mosome 14q (Lipovich et al., 2002), did not decrease
significantly during the washout period.
Quantification of heme in the supernatant confirmed
cells were preincubated with ZnMP (to inhibit HO),
loaded with heme over 30 min, then washed extensively
and placed in washout buffer as described above. After
tion of heme and porphyrins from the media, and their
raphy (HPLC; Lu ¨bben and Morand, 1994). NRK/FLVCR
cells exported 3292 ? 424 pmol of heme/107cells into
the washout buffer, while NRK/ev cells exported 1762 ?
106 pmol (p ? 0.004). Consistent with these results,
NRK/FLVCR cells also exported 1504 ? 10 pmol of
ZnMP into the washout buffer, while NRK/ev cells ex-
reduced porphyrin loading of NRK/FLVCR cells due to
continued export during the loading phase (see Figure
3B), there is an approximately 2-fold increase in both
onstrating that FLVCR exports heme (and ZnMP).
Lastly,to demonstratethatFLVCRexports hemefrom
human hematopoietic cells and is required for erythroid
differentiation, we studied its cell surface expres-
sion and function in K562 cells. As shown in Figure 1
and Table 1, FLVCR is highly expressed in undifferenti-
ated K562 cells. Since human FLVCR efficiently binds
FeLV-C, these cells can be infected with FeLV-C (K562/
FeLV-C). As expected, the cell surface expression of
FLVCR decreases through viral interference (despite a
compensatory increase in FLVCR mRNA levels). This
results in a dramatic reduction in
55Fe-heme, while naive K562 cells export 54% ? 8%,
and cells infected with FeLV-B, which decreases the
cell surface expression of human Pit-1 (Takeuchi et al.,
1992), export 54% ? 10%. Notably, there is a similar
55Fe-heme by the
Figure 2. ?-FLVCR Is Specific for Human FLVCR
Western blot analysis demonstrates the specificity of the rabbit
polyclonal antibody, ?-FLVCR. A single band is detected in this
study of HepG2 cell lysates. Its location is in agreement with the
predicted mass of FLVCR.
the surface expression of its specific cell surface recep-
tor through binding of the viral envelope protein to re-
ceptor in the ER or Golgi apparatus. This viral interfer-
of thesame subgroup (Weissand Tailor,1995). Infection
of FEA cells with FeLV-B (FEA/FeLV-B) has no effect
on cell heme content. However, infection with FeLV-C,
which impairs cell surface expression of FLVCR, results
in asignificant increase in theintracellular heme content
of FEA cells (FEA, 5.91 pmol heme/106cells; FEA/FeLV-B,
6.30 pmol heme/106cells; FEA/FeLV-C, 10.96 pmol
heme/106cells; p ? 0.05 FEA/FeLV-C versus FEA/FeLV-B
cellular heme content).
Heme Is Exported by FLVCR
We next evaluated heme transport using zinc meso-
porphyrin (ZnMP), a fluorescent heme analog previously
validated in heme transport studies (Worthington et al.,
2001). NRK cells transduced with human FLVCR (NRK/
FLVCR) or empty vector alone (NRK/ev) were incubated
in ZnMP for 30 min at 37?C, and quantitative fluores-
cence microscopy was performed. The mean fluores-
cent intensity (MFI) of NRK/FLVCR cells was 74.5% ?
10.9% SD of that of NRK/ev cells (n ? 5 experiments,
p ? 0.001, two-tailed Student’s t test). As ZnMP blocks
its own catabolism within cells by inhibition of micro-
somal HO (Worthington et al., 2001), the reduced net
intracellular accumulation (or loading) of ZnMP ob-
served in NRK/FLVCR cells after 30 min must result
from either increased export or impaired import of the
the washout of ZnMP. Cells were loaded with ZnMP,
washed extensively, then placed in buffer alone for 90
min, and the percentage of ZnMP exported was deter-
mined by calculating the ratio of the cellular MFI after
cells markedly decreases during the washout period (MFI
decrease in heme export when the function of FLVCR
is inhibited by incubation with ?-FLVCR (0.25 mg/ml).
In addition, the inhibition of FLVCR cell surface ex-
pression (by FeLV-C) or FLVCR function (by ?-FLVCR)
significantly impairs the ability of these cells to undergo
erythroid differentiation, as assessed by the frequency
of benzidine-positive cells present after 3 days of expo-
sure to hemin (50 ?M) or imatinib (0.25 ?M; Table 1; or
butyrate [0.5 mM] or TGF-? [5 ng/ml]; K.M.S. and J.L.A.,
unpublished data]). As importantly, when FLVCR is in-
hibited and erythroid differentiation is induced, apopto-
sis initiates, as shown by an increase in annexin V bind-
ing (Table 2).
The Expression of FLVCR on the Surface
of CFU-E Is Higher Than on Other
Hematopoietic Progenitor Cells
The specific loss of CFU-E in cats infected with FeLV-C
suggests that the function of FLVCR as a heme exporter
is uniquely important at this stage of erythropoiesis. We
therefore examined the expression of FLVCR on CFU-E
tor cells. CD34?cells were incubated in stem cell factor
(SCF), interleukin 3 (IL3), and IL6 for 48 hr, and then in
an erythroid differentiation cocktail containing SCF, IL3,
and erythropoietin for 72 hr. Next, cells were sorted by
FACS (using ?-FLVCR) into two populations, FLVCRhi,
comprising the brightest 13%, and FLVCRlo, the bottom
77%, and the prevalence of erythroid and myeloid pro-
genitors (BFU-E, CFU-E, and CFU-GM) was quantitated
using methylcellulose assays. The FLVCRhipopulation
contained 48.7% ? 9.2% of all CFU-E present in the
initial samples, an enrichment of CFU-E content of 3.8-
fold, but as expected, BFU-E and CFU-GM were distrib-
erage enrichments of 1.1-fold (n ? 4 experiments). A
representative study is shown in Figure 4.
FLVCR Cell Surface Expression Decreases
as Erythropoiesis Proceeds
zation proceeds. To investigate this possibility, the
FLVCR cell surface expression of normal human BM
cells was examined by flow cytometry. Maturing ery-
throid forms (CD71hi, CD34neg) had consistently less
FLVCR expression than did CD34?cells (1.98 ? 0.87
versus 4.71 ? 2.61 relative fluorescent units, p ? 0.01,
n ? 10).
In additional experiments, we observed the FLVCR
protein and mRNA expression of mobilized PB CD34?
cells undergoingerythroid differentiation during10 days
of culture (Figure 5). At day 0, the FLVCR cell surface
binized cells predominated in the culture, protein ex-
ulation of expression with erythroid maturation is also
in agreement with the results from studies of the mature
and immature human erythroid cell lines, HEL-R and
HEL-DR (Figure 1).
Figure 3. Overexpression of FLVCR Increases Export of the Heme
Analog ZnMP and55Fe-Heme
Cells engineered to overexpress FLVCR (NRK/FLVCR) and control
cells (NRK/ev) were incubated with ZnMP (5 ?M) for 30 min at 37?C,
then washed and incubated for 90 min in buffer alone at 37?C or
4?C. The relative fluorescence of NRK/FLVCR cells is reduced at
37?C (A, right panels). No effects are seen at 4?C (A, left panels).
The analysis of ZnMP washout after 30 and 90 min (B) demonstrates
a significant time-dependent reduction in NRK/FLVCR cell fluores-
cence at 37?C, but not at 4?C (n ? 4 independent studies).
Similar studies of NRK/FLVCR and control cells were performed
with55Fe-heme. After a brief incubation in ZnMP to inhibit HO activity,
55Fe-hemin (0.9 ?M) was added to the cell media for 30 min. Next,
cells were washed and the cellular55Fe-heme content was quantified.
Cells were then incubated in buffer alone for 90 min to observe55Fe-
heme washout (n ? 3). There is a significant decrease in the55Fe-heme
content of NRK/FLVCR, but not NRK/ev cells or NRK/14q cells (cells
engineered to overexpress a FLVCR paralog on chromosome 14q)
(C). When the ZnMP pre-incubation step was omitted,
disappeared quickly and equivalently from NRK/FLVCR and control
cells (data not shown).
FLVCR Exports Cytoplasmic Heme
Table 1. Impairment of FLVCR Expression or Function in K562 Cells Inhibits Heme Export and Erythroid Differentiation
Cell Lines FLVCR Expression Heme ExportAbility to Differentiate
Benzidine-Positive After 3
days of Incubation in (%):
mRNA (copies ?
1000/50 ng RNA)
Exported (%)Protein (MFI)a) Hemin b) Imatinib
K562 ? control IgG
K562 ? ?-FLVCR
34.3 ? 10.4
26.8 ? 3.2
18.4 ? 3.3
51.4 ? 2.2
61.4 ? 11.7
88.4 ? 6.2
54 ? 8
54 ? 10
3 ? 6
44 ? 4
6 ? 4
51.2 ? 4.1
51.5 ? 2.0
20.7 ? 5.5
50.6 ? 5.2
28.2 ? 1.2
60.7 ? 0.9
61.8 ? 0.6
31.4 ? 0.7
59.5 ? 1.7
24.4 ? 1.5
K562/FeLV-C versus K562
K562 ? ?-FLVCR versus
K562 ? control IgG
p ? 0.05p ? 0.05p ? 0.01p ? 0.05p ? 0.01
p ? 0.05
p ? 0.01
p ? 0.01
p ? 0.01
p ? 0.05
p ? 0.05
p ? 0.01
p ? 0.01
K562 cells were differentiated utilizing hemin (50 ?M) or imatinib (0.25 ?M). Results represent the mean ? SD of 4–6 independent studies.
p values were derived with two-tailed Student’s t-tests. As a further control, ?-FLVCR IgG was incubated with NRK/FLVCR cells (to adsorb the
?-FLVCR activity) prior to incubation with K562 cells. Adsorption abrogated the ?-FLVCR-induced impairment of K562 erythroid differentiation.
feFLVCR have a significant decrease in cellular heme
content, while specific (FeLV-C) interference with
feFLVCR expression or function significantly increases
cellular heme content.
In order to directly demonstrate the role of FLVCR
in heme export, we performed quantitative fluorescent
microscopy studies on cell lines overexpressing human
FLVCR, utilizing a fluorescent heme analog, zinc meso-
porphyrin (ZnMP). ZnMP has been used to examine
heme transport in intestinal and hepatic cell lines (Wor-
thington et al., 2001) and yeast (M.T.W., unpublished
data). Due to its potent inhibition of HO, changes in
cellular fluorescence reflect mesoporphyrin trafficking
and notmodification ofheme orZnMP catabolism(Wor-
thington et al., 2001). Our studies demonstrate a sig-
nificant reduction in ZnMP accumulation in cells over-
expressing FLVCR as compared to control cells under
steady-state conditions. More importantly, washout
studies show a dramatic time-dependent decrease in
the fluorescence of NRK/FLVCR but not control cells
Numerous reports, beginning in the 1950s, allude to the
presence of a heme binding protein on the surface of
mammalian cells (enterocytes, hepatocytes, and hema-
topoietic cell lines) (Galbraith, 1990; Worthington et al.,
2001; reviewed in Uzel and Conrad, 1998). However,
these putative mammalian heme importers were not
identified. Here, we have performed a series of experi-
ments which indicate that the human ortholog of
feFLVCR, FLVCR, exports cytoplasmic heme and is im-
portant for erythroid differentiation.
Identification of FLVCR as a Mammalian
FLVCR mRNA and protein expression levels are in-
creased in cell lines derived from tissues that either
transport heme (e.g., intestinal and hepatic cell lines) or
have increased heme synthesis (e.g., erythroid cell
lines). In addition, cell lines engineered to overexpress
Table 2. Impairment of FLVCR Expression or Function in K562 Cells Is Associated with Apoptosis
Cell Lines Induction by Hemin Induction by Imatinib
Binding (%) PI Positive (%)PI Positive (%)
K562 ? control IgG
K562 ? ?-FLVCR
9.6 ? 0.1
8.7 ? 1.2
23.6 ? 0.8
10.6 ? 1.3
25.3 ? 0.6
0.01 ? 0.01
0.01 ? 0.01
0.02 ? 0.02
0.02 ? 0.01
0.02 ? 0.02
12.9 ? 0.4
11.3 ? 0.5
21.4 ? 3.8
10.9 ? 0.4
21.9 ? 0.9
0.04 ? 0.05
0.02 ? 0.03
0.32 ? 0.14
0.01 ? 0.01
0.02 ? 0.01
K562/FeLV-C versus K562
K562 ? ?-FLVCR versus K562 ?
p ? 0.005p ? NSp ? 0.07p ? NS
p ? 0.002
p ? 0.005
p ? NS
p ? NS
p ? 0.09
p ? 0.03
p ? NS
p ? NS
K562 cells were differentiated utilizing hemin (50 ?M) or imatinib (0.25 ?M) for 24 hr. Early and late apoptosis was assessed by binding of
annexin V and staining with propidium iodide (PI), respectively. Results represent the mean ? SD of two independent studies. p values were
derived with two-tailed Student’s t-tests.
Figure 4. FLVCR Expression Is High on
CFU-E Progenitor Cells
Mobilized PB CD34?cells were exposed to
an erythroid differentiation cocktail for 72 hr.
FLVCR?cells were then sorted into FLVCRhi
(13% with highest fluorescence) and FLVCRlo
(bottom 77%) subpopulations. The number
of progenitors derived from sorting 106cells
is shown. In this representative study, the
FLVCRhipopulation contains 54% of all
FLVCR?CFU-E, when 13% is expected, a
4.2-fold enrichment, while the early erythroid
progenitors, BFU-E, and granulocyte-macro-
phage progenitors, CFU-GM, are distributed
normally. Similar results were observed in
three independent experiments.
indicating that FLVCR exports heme. Similar fluores-
cence methodologies and experimental design have
been utilized to demonstrate therole of other secondary
transporters (e.g., Bmr, Neyfakh et al., 1991 and QacA,
Mitchell et al., 1999) in multidrug export from bacteria.
In confirmatory studies, the
NRK/FLVCR cells and NRK/ev cells was quantified after
55Fe-hemin exposure and washout conditions—studies
performed in the presence of ZnMP to competitively
inhibit HO-mediated catabolism of55Fe-heme. Identical
experiments with NRK cells overexpressing a paralog of
Lipovich et al., 2002) demonstrate that heme export is
a specific function of FLVCR (Figure 3C). Furthermore,
of NRK/FLVCR cells (with HPLC) proves that heme (and
ZnMP) are exported as intact molecules across the cell
human erythroid cell line, demonstrate that specific in-
terference with either the cell surface expression of
FLVCR (by FeLV-C) or its membrane transport function
(by ?-FLVCR) significantly reduces the export of heme.
taining cell membranes, experiments where FLVCR is
selectively modulated are critical. The ?-FLVCR experi-
ments are especially informative as they demonstrate
that targeted (and nonretroviral-mediated) cell surface
inhibition of FLVCR function impairs heme export.
We have no evidence to suggest that FLVCR is a
bidirectional heme transporter. The observations dem-
onstrating increased accumulation of ZnMP in NRK/ev
compared with NRK/FLVCR cells after a 30 min loading
phase speak against an import function (Figures 3A and
3B). Also, studies following the overnight incubation of
thesecells withthehemesynthesis inhibitorsuccinylac-
etone (Ponka et al., 1982; Worthington et al., 2001) sug-
gest that FLVCR does not reverse function even when
intracellular heme is depleted (M.T.W., unpublished
Prokaryotic heme import systems consisting of an
outer membrane TonB-dependent hemoreceptor (e.g.,
HemR), a periplasmic heme binding protein, and an ATP
binding permease that delivers heme to the cytoplasm
have been described in gram-negative bacteria (see
Wandersman and Stojiljkovic, 2000 for review). Detailed
receptor domain containing invariant histidine residues,
FRAP and NPNL amino acid boxes, suggesting their
importance in prokaryotic heme binding and/or trans-
port. In addition, a number of mammalian intracellular
and more general tetrapyrrole binding proteins (e.g.,
p22HBP; Jacob Blackmon et al., 2002; Taketani et al.,
1998) have been identified. Sequence comparisons,
however, show no similarity between FLVCR and these
proteins or HemR.
55Fe-heme content of
Heme Export Is of Critical Importance
Heme is essential to oxidation-reduction reactions in
all aerobic cells, thus it is not surprising that FLVCR
orthologs are present in bacterial, plant, and animal ge-
nomes (Lipovich et al., 2002; see also http://www.ncbi.
nlm.nih.gov/sutils/blink.cgi?pid?7661708). However, in
mammals, heme has a unique role in oxygen binding
and transport as the prosthetic group of hemoglobin in
erythrocytes. At, or just after, the CFU-E stage of ery-
umented by59Fe incorporation into heme (Wickrema et
al., 1992). We propose that this results in a unique re-
quirement for FLVCR to control intracellular heme con-
tent. In agreement with this hypothesis, these progeni-
tors have increased levels ofFLVCR on their cell surface
compared to levels on the less-mature progenitor,
BFU-E (Figure 4). Also, a region of the FLVCR promoter,
between ?370 and ?1030 nucleotides relative to the
translation initiation site,contains four potential STAT5a
binding sites, as well as consensus GATA-1, GATA-2,
c-myb and NF-E2 binding sites, providing potential
during early erythroid commitment and differentiation.
An extension of our hypothesis is that FLVCR expres-
sion may be downregulated when globin protein levels
of the erythroid cell lines HEL-DR and HEL-R, data from
55Fe-heme washout studies in K562 cells, a
FLVCR Exports Cytoplasmic Heme
Figure 5. FLVCR mRNA and Cell Surface Protein Expression Decrease as CD34?Cells Differentiate
FLVCR mRNA and protein expression were observed during 10 days of erythroid differentiation. The number of CD34?cells decreases (A),
CD71?cells increase (not shown), and hemoglobinization occurs (as measured by benzidine staining [B]) during this time. Initially there are
high frequencies of CFU-GM (not shown) and BFU-E (C) but few CFU-E. The number of CFU-E increases throughout the study; however, their
frequency remains constant due to an increase in the number of maturing erythroid cells (D). The FLVCR expression in this heterogeneous
population of cells is initially high but decreases with erythroid differentiation (E–F). The results reflect the mean (?SD) of three independent
experiments plus two additional studies observing days 0–3.
normal BM CD34?versus maturing (bright CD71?) ery-
human peripheral blood CD34?cells (and K562 cells;
Z.Y. and J.L.A., unpublished data) all support this
The studies of K562 cells are also in accordance with
the concept that FLVCR is required for erythroid differ-
or specific antibody-mediated inhibition of FLVCR func-
tion impairs both heme export and the subsequent he-
moglobinization of these cells (Table 1). Also, there is
increased apoptosis (Table 2). It is possible that heme
excess (e.g., through oxidative cell membrane damage)
induces apoptosis and, in turn, the failure of erythroid
ulate erythropoiesis (e.g., by prematurely initiating glo-
bin translation) and cause apoptosis.
where there is normal development of BFU-E but severe
anemia due to a block of CFU-E and further erythroid
maturation, attest to the importance of FLVCR function
in vivo. As the viral envelope interferes with the expres-
sion/function of its receptor (FLVCR) in all infected cells
(including BFU-E and CFU-GM progenitors; Abkowitz
et al., 1987), the specific loss of cells at the CFU-E stage
of differentiation argues for a unique requirement for
heme export at this time.
K562 cells and improves the differentiation of erythroid
progenitors (BFU-E) in vitro (Holden et al., 1983; Ruther-
ford et al., 1979). Heme upregulates the erythroid heme
biosynthetic pathway, possibly through inhibition of the
transcriptional repressor, Bach1, and also contributes
Furthermore, erythroid differentiation is arrested upon
disruption of the murine erythroid-specific 5-aminolevu-
linate synthase gene (ALAS-2), the critical initial enzyme
in erythroid intracellular heme synthesis (Nakajima et
al., 1999). Thus, the results presented here, demon-
strating that FLVCR-mediated heme export is required
during early erythroid development, may appear coun-
terintuitive. Heme, however, is also toxic, necessitating
lar concentration. Thus, we hypothesize that FLVCR func-
tions as a cell membrane export channel (overflow
valve), providing a safety mechanism that is uniquely
important at the CFU-E stage of differentiation.
Regulation of intracellular heme levels differs mark-
edly between erythroid and nonerythroid cells. In hepa-
a combination of synthetic and degradative mecha-
nisms (through heme-mediated negative feedback con-
trol of ALAS-1 and heme induction of HO). In contrast,
is not regulated by heme, but rather by iron supply (re-
viewed in Ponka, 1997). In addition, there is no evidence
that HO is induced by endogenous heme; studies of the
ation is in fact associated with a reduction in HO mRNA
levels (Fujita and Sassa, 1989). Thus, apart from the
FLVCR Functions as an Overflow Valve
Previous studies demonstrate that exogenous heme in-
duces the erythroid differentiation of cell lines such as
Heme Content of NRK Cell Lines
Heme content (reported as pmol heme/106cells) was determined
as described previously (Sassa, 1976).
ability to upregulate globin protein synthesis, there is a
paucity of identified heme control mechanisms in ery-
throid cells, presumably related to their requirement for
maximal heme synthesis. We conclude that expression
of a heme exporter on the cell membrane is therefore
required during early erythropoiesis to allow for a rapid
cellular response to fluctuating cytosolic heme levels,
thus complementing the more static role of induction of
globin in preventing heme toxicity.
Although FLVCR is downregulated with terminal ery-
throid differentiation, we suspect that FLVCR does not
significantly interfere with heme binding to globin. Over-
expression of FLVCR in K562 cells via retroviral gene
transfer does not impede erythroid differentiation (Z.Y.,
J.G.Q., and J.L.A., unpublished data). These observa-
ZnMP Uptake and Washout Studies
NRK/FLVCR and NRK/ev cell lines were grown on chamber slides
in standard media, with parallel 30 min uptake studies performed
in cell culture-grade DMSO to a concentration of 4 mM, followed
by dilution in washout buffer (25 mM HEPES, [pH 7.4], 130 mM NaCl,
described (Worthington et al., 2001). For washout experiments, the
cells were washed 3 times and then incubated in working buffer
and 5% BSA at 37?C or 4?C for 90 min.
Fluorescence microscopic imaging and quantitative digital micros-
copy measurement of cellular ZnMP uptake were performed as
cate images for each data point. Digital images were analyzed as
described previously (Worthington et al., 2001). The MFI of cellular
ZnMP in thecontrol NRK/ev cells after 30 minwas considered 100%
in each study.
The Function of FLVCR in Other Tissues
FLVCR is highly expressed in cell lines with a small
intestinal or hepatic phenotype (Figure 1). As two-thirds
of body iron derives from dietary heme iron (Carpenter
and Mahoney, 1992), there must be significant heme
trafficking into the small intestine. The liver has the sec-
ond highest rate of heme biosynthesis (for cytochrome
P450) and serves as a destination for heme linked to
hemopexin or albumin. Thus, FLVCR might facilitate
heme transport or, more likely, complement the role of
HO induction in the control of cellular heme content at
these sites. FLVCR is also expressed in hematopoietic
stem cells (Z.Y. and J.L.A., unpublished data; and see
Figures 1 and 5), which presumably require stringent
Although these cells are relatively quiescent, heme syn-
thesis is needed for oxidative metabolism. FLVCR may
function in their protection from perturbations in endog-
enous heme levels or from extracellular heme in the
BM microenvironment (e.g., hemolysis in the BM is a
consequence of ineffective erythropoiesis). Notably, re-
cent studies demonstrate that the ABC transporter,
ABCG2, which protects stem cells from environmental
brane export pump for protoporphyrin IX, the direct pre-
cursor of heme in the heme synthesis pathway (Jonker
et al., 2002; Krishnamurthy et al., 2004). These data lend
further support for the concept that stem cells require
redundant mechanisms for regulation of heme content.
55Fe-Hemin Uptake and Washout Studies
The procedures and buffers were the same as for the ZnMP studies.
After a brief rinse with working buffer, the cells (105cells/well in 12-
well dishes) were incubated with 5 ?M of ZnMP in washout buffer
and 2.5 ?M BSA for 15 min, and then
LLC, Hudson, New Hampshire) was added (to a final concentration
of 0.9 ?M) for 30 additional minutes. The cells were then washed
three times, and placed in washout conditions (washout buffer and
5% BSA) for 90 min at 37?C. To measure cellular radioactivity, the
cells were rinsed twice with washout buffer, detached from the
plates, mixed with liquid scintillation cocktail (ICN), and placed in a
liquid scintillation counter. Counts were normalized by total cell
protein content. The maximum scintillation count of control NRK/
ev cells after the 30 min uptake of55Fe-heme was considered 100%
in each study.
55Fe-hemin (RI Consultants
Measurement of Heme and ZnMP Export
The procedures and buffers were the same as described for the
ZnMP and55Fe-hemin uptake and washout studies, except that cell
numberwasincreased to107cellsinT-75 flasks.Afterpreincubation
with 5 ?M ZnMP in working buffer, hemin was added to a final
concentration of 50 ?M for 30 min (loading). The cells were then
washed and placed in washout conditions as above. After 90 min,
the washout supernatant (30 ml) was collected and kept at ?80?C
until analysis. The extraction of porphyrins was adapted from pre-
extracted in parallel three times with 10 ml of ethyl acetate/acetic
acid 2:1, then washed twice with 0.5 volumes of saturated sodium
acetate. The sodium acetate washings were also extracted with
fresh ethyl acetate and combined with the original extract, which
was then washed once with 0.1 volumes of 3% sodium acetate and
dried down with argon. Samples were dissolved in buffer and the
tier Scientific) were used for peak identification and quantification.
cellswereobtainedfromM.-C. King,RD-4was agiftof M.Hentze,and
all other cell lines were obtained from the American Type Culture
Collection. NRK/FLVCR and NRK/ev cells were generated by trans-
duction of NRK-52E cells with the retroviral vector MSCVneo (BD
Biosciences Clontech) containing the human FLVCR cDNA (ampli-
fied from a human kidney cDNA library [BD Biosciences Clontech],
using primers specific to the human FLVCR cDNA [GenBank acces-
sion number: AF118637]) or no additional cDNA, respectively. Cells
were selected in G418. The generation of NRK/feFLVCR has been
described previously (Quigley et al., 2000). The cDNA encoding
FLVCRL14q (I.M.A.G.E. clone 4866427) was inserted into MSCVneo,
and NRK/14q cells generated in a similar fashion.
Western Blot Analysis
Western blotsof HepG2lysates wereprobed withrabbit anti-FLVCR
(1 ?g/ml), or preimmune serum (data not shown), followed by a
1:2000 dilution of HRP-labeled goat anti-rabbit IgG (PharMingen).
Detection used a chemiluminescence kit (Supersignal West Pico
Chemiluminescent kit, Pierce).
RNA Isolation, Quantitative RT-PCR, and Northern
Cells were pelleted, and total RNA isolated using Trizol (Invitrogen),
as suggested by the manufacturer. Total RNA was reverse tran-
scribed using the SuperScript first-strand synthesis system (In-
FLVCR Exports Cytoplasmic Heme
istry and the ABI PRISM 7700 sequence detection system (Applied
Biosystems). The results are expressed as copies ? 1000/50 ng
total RNA. GADPH and ?2-microglobulin were used for the internal
normalization of CD34?cell differentiation and cellline experiments,
respectively. PCR primers and probes (available on request) were
designed using the Primer Express program (Applied Biosystems)
and synthesized at Integrated DNA Technologies, Inc. (Coralville,
our RT-PCR results, Northern blot analyses were performed ac-
cording to standard protocols, using a 360 bp midsection fragment
of human FLVCR cDNA that does not crossreact with FLVCR para-
logs in the human genome (Lipovich et al., 2002; Z.Y., J.G.Q., and
J.L.A., unpublished data).
Sorting of Mobilized Human PB CD34?Cells
For sorting experiments, cells were thawed and differentiated as
above. At day ?3, cells were pelleted and resuspended at 5 ?
serum, then labeled either with ?-FLVCR or with rabbit IgG, followed
by R-PE-conjugated donkey anti-rabbit IgG. Cells were washed be-
tween incubations with X-VIVO15 and 1% BSA. Labeled cells were
resuspended at 5 ? 106/ml in X-VIVO15 and 1% BSA and sorted
using FACSVantage SE (BD Biosciences). From a dot plot of live
cells, the highest expressing 13% were sorted as FLVCRhiand a
distinct population as FLVCRlo. The sorted subpopulations were
then assayed for the prevalence of BFU-E, CFU-E, and CFU-GM
progenitors using the methylcellulose culture assay described
Generation and Characterization of ?-FLVCR
Polyclonal rabbit antisera were raised against NRK/FLVCR cell sur-
face membranes isolated as described (Kinne-Saffran and Kinne,
1989) (performed at R&R Rabbitry, Stanwood, Washington). Prepa-
ration of acetone precipitates of NRK/ev cells was performed as
to normal rat proteins, 1 ml of anti-FLVCR rabbit serum was incu-
bated with 20 mg of NRK/ev cell acetone precipitate overnight at
4?C, centrifuged, and the supernatant applied to a protein A car-
tridge connected to a desalting cartridge (protein A antibody purifi-
cation kit, Sigma) to purify the anti-FLVCR rabbit IgG, termed
Flow Cytometric Analysis of Normal Human
Bone Marrow Cells
BM was aspirated from ten volunteers, after informed consent, in
accordance with institutional guidelines. Cells were labeled with
anti-CD71 FITC (clone YDJ1.2.2, Beckman Coulter), anti-CD34 PE-
Cy5 (Beckman Coulter, clone 581), either polyclonal rabbit IgG or
?-FLVCR, and an R-PE-conjugated anti-rabbit IgG as a secondary
antibody prior to flow cytometric analysis using a Coulter-XL (Beck-
man Coulter). CD34?cells were identified by high CD34 expression
and low sidescatter while maturing erythroid forms were identified
by high CD71 expression without CD34 expression. For these stud-
ies, FLVCR expression was defined as the ratio of the mean fluores-
cence of cells labeled with FLVCR to the mean fluorescence of cells
labeled with the IgG control.
Quantification of Cell Surface Expression of FLVCR
taker), blocked by 1% donkey serum and 1% BSA, and labeled with
?-FLVCR (20 ?g/ml) followed by R-phycoerythrin (PE)-conjugated
donkey anti-rabbit IgG (5 ?g/ml, Jackson ImmunoResearch Labora-
tories). The cells were washed with X-VIVO15 between incubations,
fixed, and kept at 4?C until analysis. MFI unit measurements were
calculated by subtraction of the fluorescence of cells labeled con-
currently with preimmune rabbit IgG. The mean fluorescence of
cells labeled with control rabbit IgG was set between 5 and 10
We thank R.T. Doty and H.A. Dailey for critical reading of the manu-
script. This work was supported by grants from the NIH to J.L.A.
(HL31823), S.S. (DK32890), and J.D.P. (DK20503) and the Center for
Clinical Research at the University of Washington (RR00037).
Received: March 2, 2004
Revised: July 22, 2004
Accepted: July 27, 2004
Published: September 16, 2004
Studies of FLVCR Function in K562 Cells
Undifferentiated K562 cells (2 ? 105/ml) were pre-incubated with
?-FLVCR (0.25 mg/ml) or control IgG for 16–18 hr prior to study of
55Fe-Heme washout. For differentiation studies, this preincubation
period was followed by exposure to hemin (50 ?M; Benz et al.,
1980), or imatinib (0.25 ?M; Druker et al., 1996; or Butyrate, 0.5 mM;
or TGF-?, 5 ng/ml [K.M.S. and J.L.A., unpublished data]) for 3 days
in the continued presence of antibody. Differentiation was assessed
by the frequency of benzidine-positive cells (Borsook et al., 1969)
present after 3 days of exposure to the inducing agent.
Abbott, B.L. (2003). ABCG2 (BCRP) expression in normal and malig-
nant hematopoietic cells. Hematol. Oncol. 21, 115–130.
Abkowitz, J.L., Holly, R.D., and Grant, C.K. (1987). Retrovirus-
induced feline pure red cell aplasia: Hematopoietic progenitors are
infected with feline leukemia virus and erythroid burst-forming cells
are uniquely sensitive to heterologous complement. J. Clin. Invest.
Benz, E.J., Jr., Murnane, M.J., Tonkonow, B.L., Berman, B.W., Ma-
zur, E.M., Cavallesco, C., Jenko, T., Snyder, E.L., Forget, B.G., and
Hoffman, R. (1980). Embryonic-fetal erythroid characteristics of a
human leukemic cell line. Proc. Natl. Acad. Sci. USA 77, 3509–3513.
esis. II. A method of segregating immature from mature adult rabbit
erythroblasts. Blood 34, 32–41.
of the glucose transporter GLUT4. Nat. Rev. Mol. Cell. Biol. 3,
Carpenter, C.E., and Mahoney, A.W. (1992). Contributions of heme
and non-heme iron to human nutrition. Crit. Rev. Food Sci. Nutr.
Charles, R.J., Sabo, K.M., Kidd, P.G., and Abkowitz, J.L. (1996). The
pathophysiology of pure red cell aplasia: Implications for therapy.
Blood 87, 4831–4838.
Dean, G.A., Groshek, P.M., Mullins, J.I., and Hoover, E.A. (1992).
Hematopoietic target cells of anemogenic subgroup C versus non-
anemogenic subgroup A feline leukemia virus. J. Virol. 66, 5561–
Druker, B.J., Tamura, S., Buchdunger, E., Ohno, S., Segal, G.M.,
Fanning, S., Zimmermann, J., and Lydon, N.B. (1996). Effects of a
In Vitro Differentiation of Mobilized PB CD34?Cells
Mobilized human PB CD34?cells (from the Program in Excellence
for GeneTherapy, HematopoieticCell Processing Core,Fred Hutch-
inson Cancer Research Center) were thawed, then incubated in
PeproTech), and rhIL6 (50 ng/ml, PeproTech) for 48 hr (day ?2 to
0), andthen resuspended at5 ? 105/mlin anerythroid differentiation
cocktail consisting of X-VIVO15, containing rhSCF (100 ng/ml), rhIL3
(50 u/ml), and rhEp (2 u/ml) for 10 days (termed days 0–10). Fresh
mediumwasadded at3–4dayintervalstomaintain cellsatdensities
less than 4 ? 106/ml during differentiation. At the time points indi-
cated, aliquots of cells were removed from the culture. These ali-
quots were analyzed for (1) CD34, CD71, and FLVCR cell surface
expression; (2) the degree of hemoglobinization (using benzidine
staining); (3) FLVCR mRNA expression using quantitative RT-PCR;
and (4), the frequency of BFU-E, CFU-E, and CFU-GM progenitors
using the methylcellulose culture assay (a modification of previous
methods; Charles et al., 1996). Erythroid colonies (cluster of 8–50
hemoglobin-containing cells from CFU-E) were quantitated at day
7 and erythroid bursts (?250 hemoglobin-containing cells from
BFU-E) and GM colonies (?50 cells, from CFU-GM) at day 14.
Cell Download full-text
selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-
Abl positive cells. Nat. Med. 2, 561–566.
Fujita, H., and Sassa, S. (1989). The rapid and decremental change
in haem oxygenase mRNA during erythroid differentiation of murine
erythroleukaemia cells. Br. J. Haematol. 73, 557–560.
Galbraith, R.A. (1990). Heme binding to Hep G2 human hepatoma
cells. J. Hepatol. 10, 305–310.
Harlow, E., and Lane, D. (1988). Antibodies, A Laboratory Manual
(Cold Springs Harbor, NY: Cold Spring Harbor Laboratory Press).
Holden, S.A., Steinberg, H.N., Matzinger, E.A., and Monette, F.C.
(1983). Further characterization of the hemin-induced enhancement
of primitive erythroid progenitor cell growth in vitro. Exp. Hematol.
Ibrahim, N.G., Lutton, J.D., and Levere, R.D. (1982). The role of
haem biosynthetic and degradative enzymes in erythroid colony
development: The effect of haemin. Br. J. Haematol. 50, 17–28.
Iwahara, S., Satoh, H., Song, D.X., Webb, J., Burlingame, A.L., Na-
gae, Y., and Muller-Eberhard, U. (1995). Purification, characteriza-
tion, and cloning of a heme-binding protein (23 kDa) in rat liver
cytosol. Biochemistry 34, 13398–13406.
Jacob Blackmon, B., Dailey, T.A., Lianchun, X., and Dailey, H.A.
(2002). Characterization of a human and mouse tetrapyrrole-binding
protein. Arch. Biochem. Biophys. 407, 196–201.
Jonker, J.W., Buitelaar,M., Wagenaar, E., Van DerValk, M.A., Schef-
fer,G.L.,Scheper, R.J.,Plosch,T.,Kuipers,F., Elferink,R.P.,Rosing,
H., et al. (2002). The breast cancer resistance protein protects
against a major chlorophyll-derived dietary phototoxin and proto-
porphyria. Proc. Natl. Acad. Sci. USA 99, 15649–15654.
Kinne-Saffran, E., and Kinne, R.K.H. (1989). Membrane isolation:
Strategy, techniques, markers. Methods Enzymol. 172, 3–17.
Krishnamurthy, P., Ross, D.D., Nakanishi, T., Bailey-Dell, K., Zhou,
S., Mercer, K.E., Sarkadi, B., Sorrentino, B.P., and Schuetz, J.D.
(2004). The stem cell marker Bcrp/ABCG2 enhances hypoxic cell
survival through interactions with heme. J. Biol. Chem. 279, 24218–
Lipovich, L., Hughes, A.L., King, M.C., Abkowitz, J.L., and Quigley,
J.G. (2002). Genomic structure and evolutionary context of the hu-
man feline leukemia virus subgroup C receptor (hFLVCR) gene: Evi-
dence for block duplications and de novo gene formation within
duplicons of the hFLVCR locus. Gene 286, 203–213.
Lu ¨bben, M., and Morand, K. (1994). Novel prenylated hemes as
cofactors of cytochrome oxidases: Archea have modified hemes A
and O. J. Biol. Chem. 269, 21473–21479.
Mitchell, B.A., Paulsen, I.T., Brown, M.H., and Skurray, R.A. (1999).
Bioenergetics of the Staphylococcal multidrug export protein QacA;
identification of distinct binding sites for monovalent and divalent
cations. J. Biol. Chem. 274, 3541–3548.
Nakajima, O., Takahashi, S., Harigae, H., Furuyama, K., Hayashi, N.,
Sassa, S., and Yamamoto, M. (1999). Heme deficiency in erythroid
lineage causes differentiation arrest and cytoplasmic iron overload.
EMBO J. 18, 6282–6289.
Neyfakh, A.A., Bidnenko, V.E., and Chen, L.B. (1991). Efflux-medi-
ated multi-drug resistance in Bacillus subtilis: Similarities and dis-
similarities with the mammalian system. Proc. Natl. Acad. Sci. USA
Ogawa, K., Sun, J., Taketani, S., Nakajima, O., Nishitani, C., Sassa,
K. (2001). Heme mediates derepression of Maf recognition element
through direct binding to transcription repressor Bach1. EMBO J.
Onions, D., Jarrett, O., Testa, N., Frassoni, F., and Toth, S. (1982).
Selective effect of feline leukaemia virus on early erythroid precur-
sors. Nature 296, 156–158.
Pao, S.S., Paulsen, I.T., and Saier, M.H., Jr. (1994). Major facilitator
superfamily. Microbiol. Mol. Biol. Rev. 62, 1–34.
Papayannopoulou, T., Nakamoto, B., Kurachi, S., and Nelson, R.
(1987). Analysis of the erythroid phenotype of HEL cells: Clonal
variation and the effect of inducers. Blood 70, 1764–1772.
Ponka, P. (1997). Tissue-specific regulation of iron metabolism and
heme synthesis: Distinct control mechanisms in erythroid cells.
Blood 89, 1–25.
in rabbit reticulocytes. A study using succinylacetone as an inhibitor
of heme synthesis. Biochim. Biophys. Acta 720, 96–105.
Quigley, J.G., Burns, C.C., Anderson, M.M., Lynch, E.D., Sabo, K.M.,
Overbaugh, J., and Abkowitz, J.L. (2000). Cloning of the cellular
receptor for feline leukemia virus subgroup C, a retrovirus that in-
duces red cell aplasia. Blood 95, 1093–1099.
Rafie-Kolpin,M., Chefalo,P.J., Hussain,Z.,Hahn, J.,Uma, S.,Matts,
R.L., and Chen, J.J. (2000). Two heme-binding domains of heme-
regulated eukaryotic initiation factor-2?kinase. J. Biol. Chem. 275,
Riedel, N., Hoover, E.A., Gasper, P.W., Nicolson, M.O., and Mullins,
anemia retrovirus, feline leukemia virus C-Sarma. J. Virol. 60,
Riedel, N., Hoover, E.A., Dornsife, R.E., and Mullins, J.I. (1988).
Pathogenic and host range determinants of the feline aplastic ane-
mia retrovirus. Proc. Natl. Acad. Sci. USA 85, 2758–2762.
Rigby, M.A., Rojko, J.L., Stewart, M.A., Kociba, G.J., Cheney, C.M.,
Rezanka, L.J., Mathes, L.E., Hartke, J.R., Jarrett, O., and Neil, J.C.
iour in feline leukaemia viruses with chimeric envelope genes. J.
Gen. Virol. 73, 2839–2847.
Rudra-Ganguly, N., Ghosh, A.K., and Roy-Burman, P. (1998). Retro-
virus receptor PiT-1 of the Felis catus. Biochim. Biophys. Acta
Rutherford, T.R., Clegg, J.B., and Weatherall, D.J. (1979). K562 hu-
to haemin. Nature 280, 164–165.
Ryter, S.W., and Tyrrell, R.M. (2000). The heme synthesis and degra-
dation pathways: Role in oxidant sensitivity. Free Radic. Biol. Med.
Sassa, S. (1976). Sequential induction of heme pathway enzymes
during erythroid differentiation of mouse Friend leukemia virus-
infected cells. J. Exp. Med. 143, 305–315.
Smith, K.M. (1975). Porphyrins and Metalloporphyrins (Amsterdam:
Tailor, C.S., Willett,B.J., and Kabat, D. (1999). Aputative cell surface
receptor for anemia-inducing feline leukemia virus subgroup C is a
member of a transporter superfamily. J. Virol. 73, 6500–6505.
Taketani, S., Adachi, Y., Kohno, H., Ikehara, S., Tokunaga, R., and
Ishii, T. (1998). Molecular characterization of a new heme-binding
protein induced during differentiation of urine erythroleukemia cells.
J. Biol. Chem. 273, 31388–31394.
Takeuchi, Y., Vile, R.G., Simpson, G., O’Hara, B., Collins, M.K., and
surface receptor as gibbon ape leukemiavirus. J.Virol. 66, 1219–1222.
Uzel, C., and Conrad, M.E. (1998). Absorption of heme iron. Semin.
Hematol. 35, 27–34.
Wandersman, C., and Stojiljkovic, I. (2000). Bacterial heme sources:
The role of heme, hemoprotein receptors and hemophores. Curr.
Opin. Microbiol. 3, 215–220.
son, J.W. (1986). Quantitative studies of erythropoiesis in the clini-
cally normal, phlebotomized, and feline leukemia virus-infected cat.
Am. J. Vet. Res. 47, 2274–2277.
Weiss, R.A., and Tailor, C.S. (1995). Retrovirus receptors. Cell 82,
Wickrema, A., Krantz, S.B., Winkelmann, J.C., and Bondurant, M.C.
(1992). Differentiation and erythropoietin receptor gene expression
in human erythroid progenitor cells. Blood 80, 1940–1949.
Worthington, M.T., Cohn, S.M., Miller, S.K., Luo, R.Q., and Berg,
C.L. (2001). Characterization of a human plasma membrane heme
transporter in intestinal and hepatocyte cell lines. Am. J. Physiol.
Gastrointest. Liver Physiol. 280, G1172–G1177.