Download Structure-Function Relationships of Stem Cell Factor

From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
Structure-Function Relationships of Stem Cell Factor: An Analysis Based on
a Series of Human-Murine Stem Cell Factor Chimera and the Mapping of a
Neutralizing Monoclonal Antibody
By Jeffrey V. Matous, Keith Langley, and Kenneth Kaushansky
Although much is now known about the biological properties of the c-kit receptor and its ligand, stem cell factor (SCF),
little is known of the structural basis for the binding and
function of this hematopoietic cytokine. By analyzing the
activities of chimeric interspecies and homologue muteins
and epitope mapping of a monoclonal antibody (MoAb) to
the human protein, we havefound that three distinct regions
of SCF are essential for full biological function. Homologue
and interspecies swapping of polypeptide sequences between the amino terminus and 635, between L79 and N97,
and between R121 and D128 reduced or eliminated the ability of the chimera to act in synergy with murine granulocyte-
macrophagecolony-stimulatingfactor (GM-CSF)to promote
hematopoieticcolony formation. Moreover, a nonconformation-dependent MoAb that neutralizes human, but not murine SCF, was found to bind to residues within the L79-N97
segment of the human homologue. As these three regions
localize to the putative first, third, and fourth helices of the
protein, findings remarkably similar to previous studies of
cytokines as diverse as growth hormone, GM-CSF, and interleukin (IL)-4, our resultssuggest that cytokinesof multiple
classes share a common functional organization.
0 1996 by The American Society of Hematology.
T
scaffold upon which sit critical side chains of proper charge
density, hydrogen bonding potential, hydrophobicity, and
shape to allow docking with its cell surface receptor. A
thorough understanding of the mechanisms involved in receptor-mediated signal transduction is dependent on the detailed study of those regions of ligand required for docking
to and triggering of the receptor. In addition, an understanding of the structure-function relationships of any growth factor is imperative when considering the rational design of
therapeutic derivatives. Towards this end, we have begun to
map regions of SCF critical for receptor engagement and
biological activity. Using interspecies chimeric analysis and
mapping of a neutralizing monoclonal antibody (MoAb),
techniques that have been extremely informative in our and
other laboratorie.~,3~-~*
we have identified two nonlinear regions of SCF necessary for full species-specific activity. This
strategy was particularly informative for SCF, as the murine
growth factor stimulates both human and murine c-kit receptor bearing cells, but human SCF fails to engage the murine
receptor. Therefore, should an interspecies chimera lose murine SCF activity, its tertiary folding and quaternary associations can be assessed by measuring human receptor binding
capacity. In addition to the information gained from interspecies chimeras, site directed mutagenesis studies of residues
predicted to be in the fourth helix of the folded polypeptide
point to an important role for this region in receptor engagement. Based on these findings, we propose that SCF demonstrates a marked conservation of receptor binding topology
HE TYROSINE KINASE receptor, c-kit, is critically
involved in the proliferation and differentiation of early
hematopoietic progenitor cells. The cloning of this re~eptor’‘~
and its ligand (here termed stem cell factor [SCF], but also
known as c-kit ligand, mast cell growth factor and steel
factor4-”); culminated decades of work begun by developmental biologists who first noted the pleiotropic effects
(macrocytic anemia, mast cell deficiency, sterility, and pigmentation defects) of the Steel (Sl) and White Spotting (W)
mutations of mice.’* The gene products encoded by the W
and S1 loci were subsequently found to be c-kit and SCF,
respectively. SCF exists in both soluble and transmembrane
forms. Alternative splicing of exon 6 is responsible for the
generation of the critical proteolytic signal and cleavage sites
responsible for the release of the soluble form of the protein
from the cell surface.13The full-length transcript includes a
25-amino acid secretory leader, a 189-amino acid extracellular domain, a 23-amino acid transmembrane region, and a
36-amino acid cytoplasmic tail. The first 164 or 165 residues
of the extracellular domain comprise soluble SCF, which
exists in solution as a noncovalently linked h~modimer.’~
The severe hematologic defects of the S1 and W mutants,
and the dramatic in vitro and in vivo effects of SCF support
its role in hematopoie~is.~.’~.’’~’~~’~
Alone, SCF is a relatively
weak proliferative stimulus of marrow hematopoietic colony-forming cells. In contrast, together with erythropoietin
(Epo), granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage (GM)-CSF, interleukin (1L)- 1, IL-3, IL6, IL-7, IL- 11, or thrombopoietin, the cytokine exerts a profound synergistic response in both the number and size of
the resultant hematopoietic c o l ~ n i e s . ’ ~ *As
”*~
a ~result
- ~ ~ of
these activities, studies have begun to investigate the potential clinical role of SCF both alone and in combination with
other growth factors to augment hematopoiesis in states of
marrow f a i I ~ r e . ~ ~ - ~ *
Despite a growing body of data characterizing the biological effects of SCF, very little is known of the structural
features of the protein responsible for receptor binding and
signal transduction. Several strategies can and have been
employed to elucidate the structure-function relationships of
the hemopoietins. The methods and results of these studies
have been summarized in recent report^.'^-^^ Briefly, a hematopoietic growth factor can be thought of as a structural
Blood, VO; 88,N O 2 (July 15). 1996: pp 437-444
From the Division of Hematology, University of Washington Seattle, WA; and AMGen, Inc, Thousand Oaks, CA.
Submitted December 13, 1995; accepted March 13, 1996.
Supported by National Institutes of Health Grant No. ROI
CA31615 to K.K.
Address reprint requests to Kenneth Kaushansb, MD, Division
of Hematology, Box 357710, University of Washington, Seattle, WA
98195.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1996 by The American Society of Hematology.
0006-4971/96/8802-OO19$3.OO/0
431
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
438
MATOUS, LANGLEY. AND KAUSHANSKY
with cytokines seemingly as diverse as growth hormone and
GM-CSF.
MATERIALS AND METHODS
Generution of chimeric cDNA expression vectors and proteins.
Full-length cDNA clones for the soluble form of human SCF and
murine SCF were subjected to site-directed mutagenesis using a
polymerase chain reaction (PCR)-based technique3' to introduce useful restriction sites for subcloning and to generate interspecies chimeric cDNA. The resultant cDNA were subcloned into the mammalian
cell expression vector pD540and the nucleotide sequences confirmed
by the method of Sanger.,' Cell lines expressing high levels of
recombinant SCF and muteins were established by cotransfecting
BHK cells with the various expression vectors and one encoding
dihydrofolate reductase (DHFR), as previously de~cribed.~','~
Cell
lines expressing high levels of SCF-specific mRNA by slot blot
analysis and/or expressing high levels of soluble SCF protein by
immunoblotting (see below) were identified for further analysis. Cell
lines expressing specific muteins were washed in phosphate-buffered
saline and cultured in Dulbecco's modified Eagle's medium
(DMEM) supplemented with 2% dialyzed fetal calf serum (dFCS)
and antibiotics. Conditioned culture media containing soluble recombinant SCF protein was collected after 3 to 4 days. Sham-conditioned
media was made by transfecting BHK cells with DHFR alone. Recombinant '%-SCF proteins were generated by metabolically labeling established cell lines in DMEM devoid of cysteine and methionine (-CysNet), supplemented with 2% dFCS, antibiotics, and SO
pCi/mL "S CysNet (Express; New England Nuclear, Boston, MA).
The conditioned culture medium was then harvested after 2 to 3
days.
Western blot analysis. The presence and relative level of specific
recombinant SCF proteins were determined by Western blotting,
as described previo~sly.'~Aliquots of conditioned medium from
appropriate cell lines were denatured, size-fractionated by NaDodS04/10% polyacrylamide gel electrophoresis, transferred to nitrocellulose and probed with a 1:lOO dilution of a polyclonal rabbit
antiserum raised against a synthetic peptide representing residues 331 of human SCF. This antiserum recognizes human SCF and murine
SCF equally well, making quantitation of SCF chimeras reliable. An
Iz5Igoat antirabbit IgG (Amersham, Arlington Heights, IL) was used
to detect immune complexes on the blot. After autoradiography,
phosphorimager analysis (Molecular Dynamics, Sunnyvale, CA) was
used to quantitate the I2'I detecting reagent present, and thereby the
relative amounts of specific SCF protein produced by each cell line.
Radioimmunoprecipitation of metabolically-labeled SCF proteins.
"S Met/Cys-labeled conditioned culture media (2.0 to 2.5 mL) from
each of the native and mutein SCF-producing cell lines was incubated with 10 pg of an affinity purified murine antihuman SCF
MoAb, 7H6, for 16 to 24 hours at 4°C. Twenty-five micrograms of
antimurine Ig (Cappel Laboratories, Chochrancille, PA) were next
added for 6 to 8 hours at 4"C, and the immune complexes precipitated
with an excess of Staph A (Pansorbin; Calbiochem Corp, La Jolla,
CA). The immunoprecipitates were size-fractionated by 10% NaDodSO, polyacrylamide gel electrophoresis. Gels were fixed, soaked
in Amplify solution (Amersham Corp), dried, and exposed to film.
Bone marrowprogenitor cell assays. A human marrow erythroid
progenitor cell assay for burst-forming unit-erythroid (BFU-E) was
performed as previously described42 using marrow cells obtained
from normal adult volunteers giving informed consent under a University of Washington approved protocol. Low density, nonadherent
cells (1 X 105/mL) were plated in semisolid media, 15% FCS, 5%
to 10% bovine serum albumin (BSA), antibiotics, 50 U/mL gibbon
IL-3, 2 U/mL Epo, and the various muteins under study. Cultures
were maintained in 5% C 0 2 at 37°C and BFL-E enumerated by
inverted microscopy on day 14. The capacity of murine SCF or SCF
muteins to synergize with murine GM-CSF was tested in a murine
marrow colony-forming assay (CFU-GM). Marrow cells from (CS7/
B16 X DBN2) F, hybrid (BDF,) mice were obtained as previously
described3?and plated as above except that a suboptimal concentration of murine GM-CSF (50 UlmL) was used instead of IL-3 and
Epo. CFU-GM-derived colonies were enumerated on day 6 of culture.
c-kit receptor binding assays. Cells from the human erythroleukemia cell line M07e. which express high numbers of the human
c-kit receptor, were maintained in RPMI 1640 medium supplemented
with 10% FCS and antibiotics. A total of 150 pg of I2'I human SCF
was mixed with S X lo4 cells, 1 to 2 ng of recombinant SCF or
SCF mutein, and binding buffer (RPMI 1640, 50 mmol/L Hepes,
1% BSA, 10 pg/mL cytochalasin B, and 0.1% sodium azide) in a
volume of 100 pL. The mixture was incubated at 37°C for 1 hour,
after which the cell bound "'I was separated from unbound '''1
protein by centrifugation through phthalate oil." Results are expressed as the percentage competition in the presence of the SCF
recombinant proteins. The maximal competition achievable with an
excess of native human SCF is defined as 100%. Triplicate measurements were performed for each experiment.
Computer analysis. Predictions of secondary structure for SCF
were done with the software program PC/Gene (IntelliGenetics,
Mountain View, CA) according to the algorithm of Garnie~.~'
Amphipathic helixes were predicted using the program H e l ~ h e e lThe
.~~
computer coordinates for the crystalline structure of human M-CSF
were kindly provided by Dr Sung-Hou Kim at the University of
California, Berkeley, CA.45Computer modeling to align the structure
of SCF on that of M-CSF was performed by Paul 0. Sheppard,
ZymoGenetics, Inc, Seattle, WA, using software from Biosym Technologies, Inc, San Diego, CA.
RESULTS
Generation of interspecies chimeric SCF molecules. The
constructs illustrated in Fig 1 were designed to independently
test each of the four predicted alpha helices of SCF, based
on the SCFA4-CSF sequence alignment of Bazan>6and were
generated by PCR-induced mutagenesis of cDNA clones for
human and murine SCF. On sequence verification and subcloning into a mammalian cell expression vector, cell lines
were generated expressing each construct by methotrexate
selection of BHK cells cotransfected with an SCF expression
vector and one encoding DHFR. Isolated, methotrexate-resistant colonies were picked, expanded, and culture supematants then subjected to protein blotting to determine the relative amounts of SCF proteins (Fig 2). As the Western blots
used a polyvalent antiserum raised against an amino-terminal
peptide of SCF identical in the murine and human proteins,
the intensity of antibody staining of the size-fractionated
spent tissue culture medium was proportional to the levels
of specific protein present. We have successfully used this
strategy to quantitate cytokines while studying the structurefunction relationships of both GM-CSF and IL-3.'2,3'
Biological activity of SCF muteins. To determine if the
chimeric proteins were properly folded and could dimerize,
we tested whether the muteins could support the growth of
human erythroid progenitor cells. As both human and murine
SCF support the growth of human progenitor cells equally
well,9 each of the muteins, unless improperly folded, should
be active on human cells. This was the case. In the presence
of suboptimal levels of IL-3 and optimal levels of Epo, all
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
- -
STRUCTURE-FUNCTION OF SCF
439
n
t
I
I
El
A165
K1
A165
hSCF
mSCF
G35 W44
hml
hm2
hm3
L79
L59
P95
hm5
G35 W44
mhl
Fig 1. Humanlmurine chimeric cDNA constructs.
Constructs are numbered from the first residue of
the mature, soluble form of SCF. Human specific sequence is indicated by cross-hatched bars, murine
sequence by open bars, and regions of sequence
identity at junction points by black bars. The single
letter amino acid code is used. The predicted alpha
helical regions based on computer modeling is
shown above, and the residues mutated t o M-CSF
specific sequence in construct pD-helix are also
shown by vertical hatched lines. The precise mutations were R121N, D124N. K127D. and D128K.
mh2
. .,-
PQI;
mh3
L59
mh5
L79
I
-
N97
I
R121 D128
P Dhelix
0
murine
human
common
"3
hM-CSF
human leukemic cell line. As shown in Table 1 , all of the
interspecies chimera competed for binding to M07e cells in
the presence of '"1 human SCF.
Next, based on the findings that rodent SCF supports the
growth of murine hematopoietic progenitor cells, but the
of the chimeric SCF proteins acted in synergy to enhance
the growth of BFU-E, thereby assuring the tertiary and quaternary structural integrity of the muteins. These results were
confirmed by assessing the capacity of the muteins to compete for binding to the human SCF receptor, c-kit, on a
X
.r(
r(
ti
68
ti
s
ln
N
rr)
L;
s s
ti
ti
ti
2 2 2 4n
ti ti ti 3.
+
43-
29
+
Fig 2. Western blot analysis. Equal volumes of conditioned culture media from BHK cell lines expressing human or murine SCF, or the
chimera illustrated in Fig 1 were size-fractionated on a 10% polyacrylamide gel, transferred t o nitrocellulose, and probed with a polyclonal
anti-SCF-peptide antibody that recognizes both human SCF and murine SCF equally well. After incubation with '*51-labeled goat antirabbit
lg, autoradiography and phosphorimaging were performed. Relative amounts of specific protein were thus determined. Molecular mass
markers (kD) are indicated at the left. The four prominent bands of M, -32 t o 45 kD represent differentially glycosylated forms of SCF. A
similar analysis was performed on at least three separate preparations of protein.
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
MATOUS, LANGLEY, ANDKAUSHANSKY
440
Table 1. Human c-kit Receptor Competition Assay
Protein
Human SCF
Murine SCF
hml
hm2
hm3
hm5
mhl
mh2
mh3
mh5
@helix*
Sham
glL-3
Cornpetltion
% Specific
-
(100)
100
76
80
90
93
100
100
20
6
0
Conditioned media from BHK-derived cell lines expressing soluble
SCF chimeric proteins was tested for the ability to compete with
labeledhuman SCF in an M07e c-kit receptorcompetition assay.
Equivalent amounts of protein, as determined by quantitative Western blotting (Fig21, were used. The maximal percent competition was
defined as 100% specific competition. The data represent the mean
of triplicate determinations.
* A fivefold excess of this construct was used in the assay.
human homologue doesso only very poorly (800-fold reduction in specific activity; Martin eta19), we tested the proteins
for their ability to synergize with suboptimal
levels of murine
GM-CSF to enhance the growth of murine granulocyte and
macrophage colonies. Synergisticassays of SCF activity
were used in this study, as by itself, SCF is a poor stimulus
of any type of hematopoietic cell.I5 As shown in Fig 3, only
mutein mh3 was as active as equal amounts of murine SCF
in themurineGM-CSF
synergyassay. Thus,thehuman
SCF-specific regions downstream of P95 were not required
for species-specific activity. This result was independently
confirmed by analysisof construct hm3. It showed
absolutely
no capacity to support the growth of murine CFU-GM derived colonies, indicating that alone, sequences downstream
of N97 are insufficient for either murine SCF receptor engagement or signal transduction.
The results illustrated in Fig 3 also indicate that constructs
hml and hm2 were equally, but only partially, active in this
assay of synergistic activity. The partialactivity of hml,
compared with murineSCF,indicates that the regionbetween theaminoterminusandG35
of SCF is necessary
for full activity. However, as construct mhl failed to fully
stimulate GM colony growth, this region is not sufficient for
full biological responsiveness.
Additional inferences can be drawn from the data in Fig
3. Theequal activity of hml andhm2 indicatesthat the
region between W44 and L59 does not contributetothe
synergistic activity of murine SCF. Otherwise, construct hm2
should have been less active than h m l . This conclusion is
also supported by comparing constructs mhl and mh2. The
addition of murine specific sequence between W44 and L59
(mh2 compared withmhl) failed to favorably affect murinespecific biological function of these partially active proteins.
Moreover, as hm2 is partially active, despite having human
specific sequence in the amino terminal region, it appears
that the murine specific residues beyond position L59 also
contribute to murine SCF
receptor-specific activity. Comparison of the partial activity of hm2 and the inactivity of hm3
suggest that these additional interactive residues liebetween
positions L79 and N97. This conclusion was supported by
analysis of the biological activities of constructs hm5 and
mh5. Thepartial activity of hm5, a construct in which murine
SCF-specific residues from L79 to N97 were substituted into
a human SCF background, indicates the importance of this
region for murine SCF function. Using a similarstrategy,
substitutionof human specific residues from L79 to N97
into a murine SCF backgroundlead to substantial reduction
in the biological function of murine SCF (construct mh5).
Anulysis of U neutmlizing MoAb of human SCF. The
binding epitope of the murine antihuman SCF MoAb 7H6
was next mapped using the chimeric constructs illustrated
in Fig 1, in an attempt to glean furtherinformation from our
analysis. Previous work has shown that both intact antibody
and an Fab fragmentof 7H6 neutralize the activity of human
SCF, but fail tobind the murine homologue (data
not shown).
Moreover, the antibody is highly reactive in a Western blotting format, even when the target protein is denatured and
carboxymethylated on theblot. Such characteristicsstrongly
suggest that the antibody binds to a simple, linear epitope
on human SCF. Thus, identifying its binding epitope might
shed light on at least one of the sites necessary for receptor
binding. Using metabolically-labeled human SCF and SCF
muteins, 7H6 wasfoundtoimmunoprecipitateconstruct
hm3, but not hml or hm2, and to bind to mhl, mh2, and
mh5, but not mh3 (Fig 4). From this data it can be deduced
that 7H6 binds to residues between L79 and N97, precisely
the same residues found to be responsible for some
of the
synergistic activity in the colony-forming assays.
Analysis ofputative D-helix residues. The results of chimeric analysis are highly dependent on the presence of critical surface residue differences between two homologues of
a protein. Often, the sites critical for binding of a cytokine to
its receptor are madeup of both species-specific and species250
200
E:
5
5
l
150
100
8
50
0
l
KL Chimeric Protein
Fig 3. Murine bone marrow colony forming assay. The data represent the mean number of CFU-GM-derived colonies 2 SEM produced
in the presence of equivalent amounts of each of the chimeric SCF
proteins and a suboptimal amount of murine GM-CSF, from at least
three separate experiments. The relative amount of protein assayed
was determined so that CFU-GM colony growth for the native murine
SCF would be submaximal,thereby insuring accuraterepresentation
of the relative activities of the various SCF proteins.
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
STRUCTURE-FUNCTION OF SCF
441
Fig 4. Radioimmunoprecipitation of metabolically labeled
SCF. Aliquots of 35S metabolically-labeled conditioned-culture media from cell lines that
express chimeric SCF proteins
were immunoprecipitated with
the antihuman MoAb 7H6 and
size-fractionated on a 10% polyacrylamide gel. An autoradiogram exposed for 14 days is
shown. Molecular mass markers
(kDl are indicated to the left.
conserved residues. As human and murine SCF share 79%
sequence identity, this event is likely indeed. Structure-function analysis of a number of cytokines have shown that
the fourth alpha helix provides a critical site of receptor
interaction.33.35-3K
However, based on our computer modeling, human and murine SCF are almost identical in the putative fourth helix. Thus, chimeric analysis of this region (as
exemplified by mh3 v murine SCF) may not reliably identify
critical fourth helix residues.
There are striking similarities between the c-kitlSCF receptor-ligand pair and that of another tyrosine kinase receptor, c-fms, and its ligand, M-CSF. Both are involved in developmental processes, occur in soluble and membrane-bound
forms, assemble into homodimers, and share significant sequence and probable structural h o m o l ~ g y . ~To
" ~further
~
explore the putative fourth helical region of SCF, we constructed a chimeric SCFM-CSF mutant, based on their
structural similarities. Although only four highly polar residues of murine SCF were substituted by the homologous
murine M-CSF residues (Fig I), all shown to be on the
surface of M-CSF and predicted to be surface-oriented on
murine SCF, the pD-helix construct was devoid of biological
activity and bound very poorly to the c-kit receptor (Table
1). Thus, using homologue chimeric analysis, the putative
fourth helix of SCF appears to be essential for receptor engagement.
DISCUSSION
SCF is an early-acting hematopoietic cytokine produced
by the marrow stroma. Its physiologic importance is certain,
owing to its role in a congenital macrocytic anemia in mice.
In addition, potential therapeutic roles are under active investigation. Little is known, however, of the structural basis
through which SCF interacts with its receptor, c-kit, to produce biological effects. To date, only two reports detailing
truncation mutagenesis have been reported, demonstrating
that at least 141 of the 165 residues of soluble SCF are
required for activity>"." In the present studies, we have explored the structure-function relationships of the soluble
form of SCF by assaying the species-specific activity of a
series of human-murine chimeric polypeptides. We report
that at least two nonlinear regions of the glycoprotein are
critical for species-specific activity. In addition, based on the
activity of a site-directed mutant of a region close to the
carboxyl terminus of the protein, an eight amino acid region
was also identified as critical for the biological activity of
SCF. Finally, we mapped the binding epitope of a neutralizing antihuman MoAb to one of these regions, corroborating
some of the findings derived from biological activity assays.
The ability of chimeric structure-function analysis to identify important ligand-receptor interactive regions and residues is well documented for hematopoietic cytokines, including GM-CSF, IL-4, IL-5, and leukemia inhibitory factor
In contrast to the
(LIF), and for other important
results from simple deletional mutagenesis, in which regions
critical for both structural integrity and receptor binding will
result in ablation of functional activity, and hence be indistinguishable, the risk that loss-of-function interspecies chimeric
mutants are due to gross structural abnormalities are substantially reduced, because proteins that share as little as 25%
primary sequence identity display conserved tertiary folding
patterns." As human and murine SCF share 79% sequence
identity, they are almost certain to retain highly similar tertiary structures. Therefore, substituting the structural backbone regions of one homologue (eg, human SCF) for those
of another (eg. murine SCF) should not affect biological
activity through gross changes in structural folding. Consistent with this assumption is our finding that all of the chimeric SCF muteins bound to and activated the human c-kit
receptor.
In contrast to human c-kit bearing cells, in which both
human and murine SCF are equally active, only the murine
homologue can synergize with GM-CSF and stimulate hematopoietic colony formation from murine marrow cells. Thus,
substitution of human SCF-specific sequences involved in
receptor interaction, for the homologous regions of murine
SCF, would be expected to alter the capacity of the resultant
interspecies chimeric protein to activate murine cells. Based
on this strategy, two regions of the protein, between residues
1-35 and residues between positions 79-97, were identified
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
442
as contributory to the synergistic biological activity of murine SCF. At least part of these sequences reside in the
predicted first and third helices of murine SCF.
In related studies reported herein, we mapped the binding
epitope of a neutralizing MoAb of human SCF to many of
these same residues, located between L79 and N97. As this
MoAb blocks the biological activity of human SCF, both
as an intact Ig and as an Fab fragment, it is unlikely that
neutralization is accomplished by steric hindrance. Also, as
7H6 recognizes denatured human SCF in a Western blotting
format, even when the target protein is denatured and acetylated under harsh conditions on the blot, it is unlikely that
the MoAb binds to a complex epitope so that neutralization
of SCF activity is accomplished by MoAb binding to a site
distant from where our chimeric protein analysis mapped it.
Determining the binding epitopes of more neutralizing and
nonneutralizing antibodies would yield additional important
information about the structure-function relationships of
SCF. However, it is of considerable interest that the binding
epitope of MoAb 7H6 was assigned to residues between
positions 79-97, further emphasizing the critical nature of
this region.
Chimeric analysis is not without its deficiencies.'" The
biological activity of a growth factor may not be due entirely
to species-specific residues. Although not required for species-specific activity, other surface residues that are conserved between species may also be essential for receptor
engagement or for triggering signal transduction. Such residues are not identified by chimeric analysis. One striking
example of the failure of chimeric analysis to identify critical
receptor interactive residues is E21 of hGM-CSF. Despite
the fact that our previous chimeric analysis of GM-CSF
suggested that residues E14 through E21 were necessary
for full a~tivity,~'
the importance of position 21 was not
appreciated as both the murine and human homologues contain glutamic acid at this site. It was not until this residue
was specifically mutated to Arg that its critical role was
identified.5' Thus, it remained possible that amino acid residues outside of regions 1-35 and 79-97 contribute to the
functional activity of SCF.
The D or fourth helix of the hematopoietic cytokine family
assumes great importance in receptor-ligand interactions.29.33..'7.38A s the residues of the predicted fourth helix of
SCF are nearly identical in the human and murine homologues (only R117 [to SI and V130 [to MI vary in the two
SCFs), the usefulness of interspecies chimera in this region
was limited. Instead, we chose to substitute some of the
predicted surface residues of murine SCF for the corresponding residues in M-CSF.46 The justification for this strategy
is sound. First, helical wheel analysis in this region of SCF
clearly delineates hydrophilic and hydrophobic faces, and
the crystal-derived structure of M-CSF establishes the onentation of these sites.45 Second, we mutated only four of the
putative surface residues in SCF to the corresponding MCSF surface residues, all highly polar amino acids (thus
unlikely to represent internal, structure-determining core residues). Third, we easily obtained mammalian cell lines that
secreted significant quantities of mutant protein. As it is clear
from the literature and our own personal experience that
MATOUS, LANGLEY, AND KAUSHANSKY
abnormally folded proteins are rarely secreted from mammalian cells, the ease of obtaining high level pD-helix expression argues against its loss of activity based on gross structural alteration. Fourth, this region is not involved in the
formation of M-CSF hom~dimerization,~'
again making it
unlikely that the crippling effect of the yD-helix mutation
is mediated by mechanisms other than alteration of a critical
site of receptor interaction. And fifth, similar types of helix
swapping experiments have been reported for IL-5 and GMCSF, in which a critical yet conserved helical region of the
former was substituted into the latter, without any adverse
effect on biological activity.54Thus, the loss of activity and
receptor competition of the yD-helix construct clearly points
to the importance of the putative fourth helix of SCF in its
biological function.
The results of the present study support a model in which
the first, third, and fourth helices of SCF are necessary for
engagement of c-kit and signal transduction. This conclusion
is remarkably similar to the functional anatomy of growth
hormone, GM-CSF, and IL-4, whose receptors are members
of the hematopoietic cytokine receptor family, a group of
transmembrane proteins distinct from the receptor tyrosine
kinases for SCF and M-CSF. However, the data for the
model are clear, interspecies chimera and site-directed mutants of the three distinct regions of SCF clearly affect biological activity. And a neutralizing, conformation independent MoAb binds to one of these regions. Additional
experiments using single alanine substitution mutants will
be required to further localize the precise binding sites involved in the SCFlc-kit interaction, but the present work
allows us to extend the common functional organization of
the hematopoietic cytokine receptor family to those of the
receptor tyrosine kinase family of ligands.
ACKNOWLEDGMENT
The authors would like to thank A. Turner for assistance with the
receptor binding assay, J. McCarty for graphic presentation analysis,
S-H. Kim for coordinates of the crystalline structure of M-CSF
before its publication, and P. Sheppard for performing the computer
modeling.
REFERENCES
1. Yarden Y, Kuang W-J, Yang-Feng T, Coussens L, Munemitsu
S, Dull TJ, Chen E, Schlessinger J, Francke U, Ullrich A: Human
proto-oncogene c-kit: A new cell surface receptor tyrosine kinase
for an unidentified ligand. EMBO J 6:3341, 1987
2. Qiu F, Ray P, Brown K, Barker PE, Jhanwar S, Ruddle FH,
Besmer P: Primary structure of c-kit: Relationship with the CSF-I/
PDGF receptor kinase family-oncogenic activation of v-kit involves deletion of extracellular domain and C terminus. EMBO J
7:1003, 1988
3. Geissler EN, Ryan MA, Housman DE: The dominant-white
spotting (W) locus of the mouse encodes the c-kit proto-oncogene.
Cell 55:185, 1988
4. Zsebo KM, Wypuch J, McNiece IK, Hsieng SL, Smith KA,
Karkare SB, Sachdev RK, Huschenkoff VN, Birkett NC, Williams
LR, Satyagal VN,Tung W, Bossleman RA, Mendiaz EA, Langley
KE: Identification, purification, and biological characterization of
the hematopoietic stem cell factor from Buffalo rat liver-conditioned
medium. Cell 63:195, 1990
5. Williams DE, Eisenmann J, Baird A, Rauch C, Van Ness K,
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
STRUCTURE-FUNCTION OF SCF
March CJ, Park LS, Martin U, Mochizuki DY, Boswell HS, Burgess
GS, Cosman D, Lyman SD: Identification of a ligand for the c-kit
proto-oncogene. Cell 63:1667, 1990
6. Huang E, Nocka K, Beier DR, Chu TY, Buck J, Lahm HW,
Wellner D, Leder P, Besmer P: The hematopoietic growth factor
KL is encoded by the SI locus and is the ligand of the c-kit receptor,
the gene product of the W locus. Cell 63:225, 1990
7. Anderson DM, Lyman SD, Baird A, Wignall JM, Eisenman J,
Rauch C, March CJ, Boswell HS, Gimpel SD, Cosman D, Williams
DE: Molecular cloning of mast cell growth factor, a hemopoietin
that is active in both membrane bound and soluble forms. Cell
63:235, 1990
8. Flanagan JG, Leder P: The kit ligand A cell surface molecule
altered in steel mutant fibroblasts. Cell 63:185, 1990
9. Martin FH, Suggs SV, Langley KE, Lu HS, Ting J, Okino KH,
Moms CF, McNiece IK, Jacobsen FW, Mendiaz EA, Birkett NC,
Smith KA, Johnson MJ, Parker VP, Flores JC, Patel AC, Fisher EF,
Erjavec HO, Herrera CJ, Wypych J, Sachdev K, Pope JA, Leslie I,
Wen D, Lin C-H, Cupples RL, Zsebo KM: Primary structure and
functional expression of rat and human stem cell factor DNAs. Cell
63:203, 1990
10. Witte ON: Steel locus defines new multipotent growth factor.
Cell 635, 1990
11. Zsebo KM, Williams DA, Geissler EN, Broudy VC, Martin
FH, Atkins HL, Hsu R-Y, Birkett NC, Okino KH, Murdock DC,
Jacobsen FW,Langley KE, Smith KA, Takeishi T, Cattanach BM,
Galli SJ, Suggs SV: Stem cell factor is encoded at the SI locus of
the mouse and is the ligand for the c-kit tyrosine kinase receptor.
Cell 63:213, 1990
12. Russell ES: Hereditary anemias of the mouse: A review for
geneticists. Adv Genet 20:357, 1979
13. Flanagan JG, Chan DC, Leder P: Transmembrane form of
the kit ligand growth factor is determined by alternative splicing
and is missing in the Sld mutant. Cell 64:1025, 1991
14. Arakawa T, Yphantis DA, Lary JW, Narhi LO, Lu HS,
Prestrelski SJ, Clogston CL, Zsebo KM, Mendiaz EA, Wypych J,
Langley KE: Glycosylated and unglycosylated recombinant-derived
human stem cell factors are dimeric and have extensive regular
secondary structure. J Biol Chem 266:18942, 1991
15. McNiece IK, Langley KE, Zsebo KM: Recombinant human
stem cell factor synergises with GM-CSF, C-CSF, IL-3 and Epo to
stimulate human progenitor cells of the myeloid and erythroid lineages. Exp Hematol 19:226, 1991
16. Andrews RG, Bartelmez SH, Appelbaum FR, Bemstein ID,
Knitter GH, Langley KE, Farrer D, Hendren W, Zsebo KM: Recombinant human SCF, a e-kit ligand, stimulates hematopoiesis in primates. Blood 78:1975, 1991
17. Molineux G , Migdalska A, Szmitkowski M, Zsebo K, Dexter
TM: The effects of hematopoiesis of recombinant stem cell factor
(ligand for c-kif) administered in vivo to mice either alone or in
combination with granulocyte colony-stimulating factor. Blood
78:961, 1991
18. Ulich TR, del Castillo J, Yi ES, McNiece I, Yung YP, Zsebo
KM: Hematologic effects of stem cell factor in vivo and in virro in
rodents. Blood 78:645, 1991
19. Zsebo KM, Smith KA, Hartley CA, Greenblatt M, Cooke K,
Rich W, McNiece I K Radioprotection of mice by recombinant rat
stem cell factor. Proc Natl Acad Sci USA 89:9464, 19912
20. Lowry PA, Deacon D, Whitefield P, McGrath HE, Quesenberry PJ: Stem cell factor induction of an in vitro murine hematopoietic colony formation by “subliminal” cytokine combinations: The
role of “anchor factors”. Blood 80:663, 1992
21. Kaushansky K, Lok S, Holly RD, Broudy VC, Lin N, Bailey
MC, Forstrom JW, Buddle M, Oort PJ, Hagen FS, Roth GJ, Papayannopoulou T, Foster DC: Promotion of megakaryocyte progenitor
443
expansion and differentiation by the c-Mpl ligand thrombopoietin.
Nature 369568, 1994
22. Lemoli RM, Fogli M, Fortuna A, Motta MR, Rizzi S, Benini
C, Tura S: Interleukin-11 stimulates the proliferation of human hematopoietic CD34+ and CE34+CD33-DR-cells and synergizes
with stem cell factor, interleukin-3, and granulocyte-macrophage
colony-stimulating factor. Exp Hematol 21: 1668, 1993
23. Backx B, Broeders L, Lowenberg B: Kit ligand improves in
virro erythropoiesis in myelodysplastic syndrome. Blood 80: 1213,
1992
24. Bagnara GP, Strippoli P, Bonsi L, Brizzi MF, Avanzi GC,
Timeus F, Ramenghi U, Piaggio G, Tong J, Pedesta M, Paolucci G,
Pegoraro L, Bacigalupo A: Effect of stem cell factor in colony
growth from acquired and constitutional (fanconi) aplastic anemia.
Blood 80382, 1992
2.5. Woodnar-Filipowicz A, Tichelli A, Zsebo KM, Speck B, Nissen C: Stem cell factor stimulates the in vitro growth of bone marrow
cells from aplastic anemia patients. Blood 79:3196, 1992
26. Tong J, Gordon MS, Srour EF, Cooper RJ, Orazi CA,
McNiece I, Hoffman R: In vivo administration of recombinant methionyl human stem cell factor expands the number of human marrow hematopoietic stem cells. Blood 82:784, 1993
27. Andrews RG, Briddell RA, Knitter GH, Rowley SD, Appelbaum FR, McNiece I K Rapid engraftment by peripheral blood progenitor cells mobilized by recombinant human stem cell factor and
recombinant human granulocyte colony-stimulating factor in nonhuman primates. Blood 85:15, 1995
28. Andrews RG, Briddell RA, Knitter GH, Opie T, Bronsdon M,
Myerson D, Appelbaum FR, McNiece IK. In vivo synergy between
recombinant human stem cell factor and recombinant granulocyte
colony-stimulating factor in baboons enhanced circulation of progenitor cells. Blood 84:800, 1994
29. Wells JA: Systematic mutational analyses of protein-protein
interfaces. Methods Enzymol 202:390, 1991
30. Kaushansky K, Karplus PA: Hematopoietic growth factors:
Understanding functional diversity in structural terms. Blood
82:3229, 1993
31. Mott HR, Campbell ID: Four-helix bundle growth factors and
their receptors: Protein-protein interactions. Curr Opin Struct Biol
5:114, 1995
32. Kaushansky K, Shoemaker SG, Alfaro S, Brown C: Hematopoietic activity of granulocytehnacrophage colony-stimulating factor
is dependent upon two distinct regions of the molecule: Functional
analysis based upon the activities of interspecies hybrid growth factors. Proc Natl Acad Sci USA 861213, 1989
33. Kaushansky K, Shoemaker SG, Broudy VC, Lin NL, Matous
JV, Alderman EM, Aghajanian JD, Szklut PJ, VanDyke RE, Pearce
MK, Abrams JS: Structure-function relationships of interleukin-3.
An analysis based on the function and binding characteristics of a
series of interspecies chimera of gibbon and murine interleukin-3. J
Clin Invest 90:1879, 1992
34. Owczarek CM, Layton MJ, Metcalf D, Lock P, Willson TA,
Gough NM, Nicola NA: Inter-species chimeras of leukaemia inhibitory factor define a major human receptor-binding determinant.
EMBO J 12:3487, 1993
35. Cunningham BC, Jhurani P, Ng P, Wells JA: Receptor and
antibody epitopes in human growth hormone identified by homologscanning mutagenesis. Science 243: 1130, 1989
36. McKenzie ANJ, Bany SC, Strath M, Sanderson CJ: Structurefunction analysis of interleukin-5 utilizing mousehuman chimeric
molecules. EMBO J 10:1193, 1991
37. Kanakura T, Cannsitra SA, Brown CB, Nakamura M, Seelig
GF, Prosise WW, Hakins JC, Kaushansky K, Griffin E:Identification of functionally distinct domains of human granulocyte-macro-
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
444
phage colony-stimulating factor using monoclonal antibodies. Blood
77:1033, 1991
38. &use N, Shen B-J, Amold S , Tony H-P, Muller T, Sebald
W: Two distinct functional sites of human interleukin-4 are identified
by variants impaired in either receptor binding or receptor activation.
EMBO J 125121, 1993
39. Horton RM, Hunt HD, Ho SN, Pullen JK, Pease IR: Engineering hybrid genes without the use of restriction enzymes: Gene
splicing by overlap extension. Gene 77:61, 1989
40. Berkner K, Busky S , Davie E, Hart C, Insley M, Kiesel W,
Kumar A, Murray M, O’Hara P, Woodburg R, Hagen F: Isolation/
expression of cDNAs encoding human factor VII. CSH Symposium
on Quantitation Biology 51531, 1986
41. Sanger F, Nicklen S, Coulson AR: DNA sequencing with
chain-terminating inhibitors. Proc Natl Acad Sci USA 745463, 1977
42. Kaushansky K, Broudy VC, Grossmann A, Humes J, Lin N,
Ren HP, Bailey MC, Papayannopoulou T, Forstrom JW, Sprugel
KH: Thrombopoietin expands erythroid progenitors, increases red
cell production, and enhances erythroid recovery after myelosuppressive therapy. J Clin Invest 96:1683, 1995
43. Gamier J: Analysis of the accuracy and implications of simple
methods for predicting the secondary structure of globular proteins.
J Mol Biol 120:97, 1988
44. Novotny J, Auffrey C: A program for prediction of protein
secondary structure from nucleotide sequence data: Application to
histocompatibility antigens. Nucleic Acids Res I2:243, 1984
45. Pandit J, Bohm A, Jancarik J, Halenbeck R, Koths K, Kim
S-H: Three-dimensional structure of dimeric human recombinant
macrophage colony-stimulating factor. Science 258: 1358, 1992
46. Bazan JF: Genetic and structural homology of stem cell factor
and macrophage colony-stimulating factor. Cell 65:9, 1991 (letter)
MATOUS, LANGLEY, AND KAUSHANSKY
47. Williams DE, de Vries P, Namen AE, Widmer MB, Lyman
SD: The steel factor. Dev Biol 151:368, 1992
48. Pollard JW, Bartocci A, Arceci R, Orlofsky A, Ladner MB,
Stanley ER: Apparent role of the macrophage growth factor, CSFI, in placental development. Nature 330:484, 1987
49. Wong GG, Temple PA, Leary AC, Witek-Giannotti JS, Yang
Y-C, Ciarletta AB, Chung M, Murtha P, Kriz R, Kaufman RJ,Ferenz
CR, Sibley BS, Turner KJ, Hewick RM, Clark SC, Yanai N, Yokota
H, Yamada M, Saito M, Motoyoshi K, Takaku F: Human CSF-I:
Molecular cloning and expression of 4-kb cDNA encoding the human urinary protein. Science 235:1504, 1987
50. Nishikawa M, Tojo A, lkebuchi K, Katayama K, Fuji N,
Ozawa K, Asano S : Deletional mutagenesis of stem cell factor defines the C-terminal sequences essential for its biological activity.
Biochem Biophys Res Commun 188:292, 1992
51. Langley KE, Mendiaz EA, Liv N, Narhi Lo, Zeni L.
Parseghian CM, Clogston CL, Leshe I, Pope JA, Lu HS, Zsebo KM,
Boone TC. Properties of variant forms of human stem cell factor
recombinantly expressed in Escherichia coli. Arch Biochem Biophys
31155, 1994
52. Chothia C, Lesk AM: The relation between the divergence
of sequence and structure in proteins. EMBO J 5:823,
1986
53. Hercus TR, Cambareri B, Dottore M, Woodcock J, Bagley
CJ, Vadas MA, Shannon MF, Lopez A F Identification of residues
in the first and fourth helices of human granulocyte-macrophage
colony-stimulating factor involved in biologic activity and in binding
to the cy- and 0-chains of its receptor. Blood 83:3500, 1994
54. Kastelein RA, Shanafelt AB: GM-CSF receptor: Interactions
and activation. Oncogene 8:231, 1993
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
1996 88: 437-444
Structure-function relationships of stem cell factor: an analysis based
on a series of human-murine stem cell factor chimera and the
mapping of a neutralizing monoclonal antibody
JV Matous, K Langley and K Kaushansky
Updated information and services can be found at:
http://www.bloodjournal.org/content/88/2/437.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.