Granulocyte colony-stimulating factor preferentially stimulates proliferation of monosomy 7 cells bearing the isoform IV receptor.

Abstract

Granulocyte colony-stimulating factor (GCSF) administration has been linked to the development of monosomy 7 in severe congenital neutropenia and aplastic anemia. We assessed the effect of pharmacologic doses of GCSF on monosomy 7 cells to determine whether this chromosomal abnormality developed de novo or arose as a result of favored expansion of a preexisting clone. Fluorescence in situ hybridization (FISH) of chromosome 7 was used to identify small populations of aneuploid cells. When bone marrow mononuclear cells from patients with monosomy 7 were cultured with 400 ng/ml GCSF, all samples showed significant increases in the proportion of monosomy 7 cells. In contrast, bone marrow from karyotypically normal aplastic anemia, myelodysplastic syndrome, or healthy individuals did not show an increase in monosomy 7 cells in culture. In bone marrow CD34 cells of patients with myelodysplastic syndrome and monosomy 7, GCSF receptor (GCSFR) protein was increased. Although no mutation was found in genomic GCSFR DNA, CD34 cells showed increased expression of the GCSFR class IV mRNA isoform, which is defective in signaling cellular differentiation. GCSFR signal transduction via the Jak/Stat system was abnormal in monosomy 7 CD34 cells, with increased phosphorylated signal transducer and activation of transcription protein, STAT1-P, and increased STAT5-P relative to STAT3-P. Our results suggest that pharmacologic doses of GCSF increase the proportion of preexisting monosomy 7 cells. The abnormal response of monosomy 7 cells to GCSF would be explained by the expansion of undifferentiated monosomy 7 clones expressing the class IV GCSFR, which is defective in signaling cell maturation.

Full text

Page 1

Granulocyte colony-stimulating factor preferentially
stimulates proliferation of monosomy 7 cells
bearing the isoform IV receptor
Elaine M. Sloand*, Agnes S. M. Yong, Shakti Ramkissoon, Elena Solomou, Tullia C. Bruno, Sonnie Kim, Monika Fuhrer,
Sachiko Kajigaya, A. John Barrett, and Neal S. Young
Hematology Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD 20892
Edited by Harvey J. Alter, National Institutes of Health, Bethesda, MD, and approved July 28, 2006 (received for review June 23, 2006)
Granulocyte colony-stimulating factor (GCSF) administration has
been linked to the development of monosomy 7 in severe con-
genital neutropenia and aplastic anemia. We assessed the effect of
pharmacologic doses of GCSF on monosomy 7 cells to determine
whether this chromosomal abnormality developed de novo or
arose as a result of favored expansion of a preexisting clone.
Fluorescenceinsituhybridization(FISH)ofchromosome7wasused
to identify small populations of aneuploid cells. When bone mar-
row mononuclear cells from patients with monosomy 7 were
cultured with 400 ng?ml GCSF, all samples showed significant
increases in the proportion of monosomy 7 cells. In contrast, bone
marrow from karyotypically normal aplastic anemia, myelodys-
plastic syndrome, or healthy individuals did not show an increase
in monosomy 7 cells in culture. In bone marrow CD34 cells of
patients with myelodysplastic syndrome and monosomy 7, GCSF
receptor (GCSFR) protein was increased. Although no mutation
was found in genomic GCSFR DNA, CD34 cells showed increased
expression of the GCSFR class IV mRNA isoform, which is defective
in signaling cellular differentiation. GCSFR signal transduction via
the Jak?Stat system was abnormal in monosomy 7 CD34 cells, with
increased phosphorylated signal transducer and activation of tran-
scription protein, STAT1-P, and increased STAT5-P relative to
STAT3-P. Our results suggest that pharmacologic doses of GCSF
increase the proportion of preexisting monosomy 7 cells. The
abnormal response of monosomy 7 cells to GCSF would be ex-
plained by the expansion of undifferentiated monosomy 7 clones
expressing the class IV GCSFR, which is defective in signaling cell
maturation.
hematopoiesis ? myelodysplasia ? clonal evolution ? leukemia ?
marrow failure
T
improvesbloodcounts,butaminorityofpatientsdeveloprecurrent
or refractory cytopenias, or frank leukemia, associated with the
expansioninthemarrowofdysplasticandcytogeneticallyabnormal
clones, a pathophysiology termed ‘‘clonal evolution’’ (1, 2). Of the
cytogenetic abnormalities, monosomy 7 is the most common and
has the worst prognosis (3, 4). Some studies of AA in children and
adults, and observations in children with congenital neutropenia
(Kostmann’s syndrome), have suggested a correlation between
development of monosomy 7 and long-term use of high doses of
granulocytecolony-stimulatingfactor(GCSF)(5–7).Itisuncertain
whether GCSF is directly responsible for cytogenetic abnormalities
because other linked factors such as unresponsiveness to immuno-
suppressive therapy, persistent neutropenia, and concurrent infec-
tions might also be implicated (8). Some investigators report a
comparable incidence of monosomy 7 in patients who have not
been treated with GCSF (2, 9).
Using the sensitive quantitative method fluorescence in situ
hybridization(FISH),wepreviouslyfoundsmallnumbersofmono-
somy 7 cells (among other chromosomal abnormalities) in AA
reatment of aplastic anemia (AA) with immunosuppressive
drugs such as antithymocyte globulin and cyclosporine usually
patients sampled at the time of presentation of their disease and
before treatment (10). Routine cytogenetic analysis of AA bone
marrow samples is often unsuccessful because many patients have
too few dividing bone marrow cells to permit analysis of 20
metaphasecells.However,FISHmakespossibletheexaminationof
400ormoreinterphasecellsbecausethetechniquedoesnotrequire
cells to be in metaphase. In one study of marrow-failure patients,
only 28% of chromosomal abnormalities detected by FISH were
observed by conventional cytogenetics (10). GCSF levels are gen-
erally an order of magnitude higher in severe AA than in neutro-
penia of other origins (11, 12). An increased prevalence of mono-
somy 7 among untreated cases could, therefore, be explained by
high endogenous levels of GCSF in AA.
Monosomy 7 in AA could be a consequence of GCSF-induced
errorsinthecellcycle,DNAreplicationandrepair,orchromosome
segregation errors provoked by increasing cell replication. Alter-
natively, GCSF might favor the expansion of a preexisting clone. In
support of the latter hypothesis, dominant proliferative signaling
from a mutant GCSF receptor (GCSFR) has been reported in
children with severe congenital neutropenia (SCN) who require
prolonged GCSF treatment for survival (13). To distinguish be-
tween these two possibilities, we used FISH for chromosome 7 to
examine cells from patients with AA, myelodysplastic syndrome
(MDS) with normal cytogenetics, MDS with cytogenetically doc-
umented monosomy 7, and normal controls, in order to assess the
effect of GCSF on bone marrow cells grown in short-term culture.
We also cultured cytogenetically normal samples obtained at
presentation of pancytopenia from patients who later developed
monosomy 7 with GCSF. Lastly, we sought evidence of abnormal
GCSF signal transduction in monosomy 7 cells.
Results
Study Population. Twenty-eight de novo MDS patients with mono-
somy 7 (24 positive by FISH and conventional cytogenetics and 4
MDS patients positive by FISH alone), 4 additional patients with a
history of AA who developed monosomy 7, and 26 individuals with
normal karyotype by cytogenetics and FISH (6 MDS patients, 10
AA patients, and 10 healthy controls) were studied. Four AA
patients who initially had normal cytogenetics, but who later
developed monosomy 7 were also studied. Data for these subjects
are provided in Table 6, which is published as supporting informa-
tion on the PNAS web site.
Author contributions: E.M.S. and N.S.Y. designed research; E.M.S., A.S.M.Y., S.R., E.S.,
T.C.B., S. Kim, M.F., and S. Kajigaya performed research; E.M.S., S. Kajigaya, N.S.Y., and
A.J.B. analyzed data; and E.M.S., A.J.B., and N.S.Y. wrote the paper.
The authors declare no conflict of interest.
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: AA, aplastic anemia; BMMNC, bone marrow mononuclear cells; GCSF,
granulocyte colony-stimulating factor; GCSFR, GCSF receptor; MDS, myelodysplastic syn-
drome; SCN, severe congenital neutropenia.
*To whom correspondence should be addressed. E-mail: sloande@nih.gov.
www.pnas.org?cgi?doi?10.1073?pnas.0605245103PNAS ?
September 26, 2006 ?
vol. 103 ?
no. 39 ?
14483–14488
MEDICAL SCIENCES

Page 2

GCSF Promotes Expansion of Preexisting Monosomy 7 Clones. When
bone marrow mononuclear cells (BMMNC) from eight patients
with monosomy 7 (as determined by cytogenetics) were cultured in
liquid media for 7 days in the presence or absence of pharmaco-
logical doses of GCSF (400 ng?ml), all cytokine-treated samples
showed increased proportions of cells with monosomy 7 (Fig. 1A)
(P ? 0.01). Similarly, samples from four patients with MDS and
monosomy 7 identified only by FISH (presumably below the
thresholdofdetectionbycytogenetics)showedexpansionofmono-
somy 7 cell numbers after culture (Fig. 1B). When MDS BMMNC
fromfiveadditionalpatientswithmonosomy7bycytogeneticswere
cultured in semisolid media for 2 weeks and individual colonies
were picked for FISH analysis, a similar increase was seen in the
number and proportion of hematopoietic colonies with monosomy
7, primarily among myeloid progeny (Table 1). Colonies were
homogeneous(i.e.,theywerecomposedofeitheralldiploidcellsor
all monosomy 7 cells). GCSF-stimulated cultures of monosomy 7
cells showed large numbers of undifferentiated cells by microscopy
(data not shown). In contrast, BMMNC from 10 AA and 6 MDS
patients with normal chromosome 7 numbers by FISH had no
increaseinmonosomy7cellsafterculturewithGCSF(Fig.1B;P?
0.01). Four patients with AA had BMMNC stored at the time of
first clinical presentation; FISH had disclosed a detectable mono-
somy 7 clone of cells in all of the samples, despite normal conven-
tional cytogenetics, and we were able to expand this clone in vitro
with GCSF (Table 2). The proportion and number of monosomy 7
cells were unaltered in vitro by incubation with other cytokines,
IL-3,orgranulocyte-macrophagecolony-stimulatingfactor,evenat
high concentrations (data not shown).
Decreased numbers of monosomy 7 cells relative to diploid cells
were evident in cultures after 4 days at low GCSF concentrations,
but significantly more monosomy 7 cells were seen growing at very
high GCSF concentrations, in the range of those present in the
circulation of patients with marrow failure (11, 14) (Fig. 1C).
Inferior survival of monosomy 7 cells after 14 days of culture with
GCSF was evident at low concentrations of GCSF (Fig. 1D).
GCSFR Expression Is Increased on CD34 Cells from Monosomy 7
Patients and Preferentially in Monosomy 7 Cells. We next examined
GCSFR expression on cells from monosomy 7 patients by using
Fig. 1.
a preexisting monosomy 7 clone. BMMNC were placed in culture with and
without pharmacologic doses (400 ng?ml) of GCSF for 7 days, as described in
MaterialsandMethods.(A)All24patientswhodemonstratedmonosomy7by
conventionalcytogeneticsshowedincreasedproportionsandnumbersofcells
with monosomy 7. Results are shown for eight patients. (B) None of the 10
normal controls or 6 AA patients without clinical monosomy 7 showed a
significant increase in the proportion of cells with monosomy 7 in tissue
cultureasaresultofexposuretohighGCSFdoses,whereasMDSpatientswith
monosomy 7 detectable by FISH all showed expansion of this clone after GCSF
exposure.(C)Adose–responsecurveforBMMNCperformedonthreepatients
with monosomy 7 showing decreased sensitivity to low doses of GCSF
GCSF increases the proportion of monosomy 7 cells in patients with
when cultured short-term (4 days) compared with diploid cells. (D) When the
cells described in C were subjected to longer culture times (14 days), mono-
somy 7 cells declined at lower GCSF concentrations.
Table 1. Expansion of monosomy 7 CD34 cells exposed to
in vitro pharmacologic concentrations of GCSF
Patient
no.
% monosomy
7 CD34
% CFU-GM
monosomy 7
% CFU-E
monosomy 7
?GCSF
?GCSF
?GCSF
?GCSF
24
21
7
22
13
25
9
3
6
8
35
12
0
2
10
75
35
8
8
25
5
0
2
0
2
13
7
12
5
12
CD34 cells were obtained by flow cytometry, and FISH was performed by
using a centrometric probe for chromosome 7. CD34 cells were plated in
semisolid media with and without pharmacologic concentrations of GCSF.
Colonies derived from granulocyte-macrophage hematopoietic progenitors
(CFU-GM) and erythroid hematopoietic progenitors (CFU-E) were isolated,
andFISHwasperformedonindividualcolonies.Colonieswerehomogeneous,
containing only monosomy 7 cells or diploid cells. Results are shown as a
percentage of colonies in which FISH showed monosomy 7. The monosomy 7
cloneexpandedinsamplescontainingpharmacologicconcentrationsofGCSF,
mostly in myeloid colonies.
14484 ?
www.pnas.org?cgi?doi?10.1073?pnas.0605245103Sloand et al.

Page 3

flowcytometry.Becausehematopoieticcellsincreaseexpressionof
GCSFR as they mature, we specifically studied flow-cytometrically
purifiedbonemarrowCD34cells.WhenCD34cellsweresortedon
the basis of GCSFR expression, the percentage of cells with
monosomy 7 assessed by FISH was significantly increased in the
fraction of cells with higher expression of the receptor (Fig. 2).
CD34 Cells Expressing GCSFR Isoform IV Are Increased in Monosomy
7 Patients. Mutations in the genomic DNA encoding the GCSFR,
as have been described in SCN (15), or alternative splicing of the
mRNA, as occurs in leukemia (16), could account for altered
sensitivity to GCSF, especially if the genetic alteration were to
remove the distal cytoplasmic domain responsible for myeloid
lineage differentiation. The nucleotide sequences of the GCSFR of
two patients and of a monosomy 7 cell line showed several
polymorphisms but no mutations (see Table 7, which is published
as supporting information on the PNAS web site). We then
examined GCSFR mRNA, assaying specifically for the differenti-
ation-defective isoform IV, which lacks 87 aa at the C terminus as
a result of alternative splicing. Overexpression of this isoform has
been described in some patients with leukemia (17) and has been
implicated in disordered myeloid maturation. Monosomy 7 CD34
cell mRNA was extracted and amplified. PCR products were
electrophoresed, and band intensities were quantified and com-
pared with those of CD34 cells from 10 normal donors and 6
patients with MDS with normal cytogenetics. Analysis of relative
amountsofGCSFRisoformsinsixmonosomy7patientswithMDS
showed an increased ratio of isoform IV relative to isoform I in the
monosomy7cellpopulation(seeexampleinFig.3A)butnotinfour
patients with MDS and normal cytogenetics or in six healthy
controls. To confirm these findings, real-time PCR was performed
on lymphocyte-depleted BMMNC from an additional seven pa-
tients with monosomy 7 MDS patients, eight healthy controls, and
sevenMDSpatientswithothercytogeneticabnormalitiesornormal
cytogenetics. The numbers of monosomy 7 cells correlated with
increased relative expression of isoform IV (Fig. 3B).
STAT1-P and STAT5-P Are Up-Regulated in CD34 Cells of Patients with
Monosomy 7. In experimental mouse models and in children with
SCN, abnormal GCSFRs have been associated with defects in
internalization, resulting in membrane GCSFR overexpression and
ahyperproliferativecellularresponse(18).Transfectedmurinecells
overexpressing class IV and SCN cells expressing the truncated
GCSFR show constitutive STAT1 phosphorylation and an in-
creased ratio of STAT5-P?STAT3-P. To determine whether STAT
expression was similarly disordered in monosomy 7 MDS, we
measured STAT1-P expression in CD34 cells from patients with
significant proportions of monosomy 7 cells in the bone marrow, as
Table 2. Serial analysis of BMMNC from AA patients with small monosomy 7 clones exposed
in vitro to GCSF
Patient
no.
?18 mo
NT, %
?18 mo
? GCSF, %
?12 mo
NT, %
?12 mo
? GCSF, %
?6 mo
NT, %
?6 mo ?
GCSF, %
0 mo
NT, %
29
30
31
32
5
6
74
7
7 10
12
18
24
30
45
35
41
60
46
38
53
10
23
ND
14
16
ND
12
ND
10
ND
Frozen samples from AA patients were obtained before the development of any clinical cytogenetic abnor-
mality BMMNC cultured with GCSF. Results are shown as percentage of monosomy 7 cells. Time 0 is when
monosomy 7 was clinically first evident on routine cytogenetics. NT, not treated with GCSF; ND, not done.
Fig. 2.
cells from monosomy 7 patients were sorted into GCSFR-negative and -posi-
tive fractions by flow cytometry and FISH was performed, the monosomy 7
cells were predominantly in the GCSFR-positive population.
GCSFR protein is increased on monosomy 7 CD34 cells. When CD34
Fig. 3.
abnormalamountsoftheshortenedisoformIVofGCSFR.RNAfromCD34cells
from monosomy 7 patients was subjected to DNase treatment (Invitrogen,
Carlsbad,CA)followedbyoligo(dT)-primedRT-PCRusingtheSuperScriptFirst
Strand synthesis system for RT-PCR (Invitrogen). GCSFR isoform amplification
was performed as described in ref. 17. (A) Samples from three healthy volun-
teers, three patients, and cell lines KGi1, HL60, and THP-1, showing increased
expressionofshortenedisoformIV.(B)Whenreal-timePCRwasperformedon
BMMNC, as described in Materials and Methods, an increased ratio of GCSFR-
IV?I was observed that was proportional to the number of monosomy 7 cells
in the bone marrow sample.
Bone marrow CD34 cells from patients with monosomy 7 express
Sloand et al.
PNAS ?
September 26, 2006 ?
vol. 103 ?
no. 39 ?
14485
MEDICAL SCIENCES

Page 4

determined by flow cytometry (Fig. 4A). Immunoblot verified the
overexpression of STAT1-P in monosomy 7 patients (Fig. 4B).
Similarly, when BMMNC were sorted by flow cytometry into
CD34-positive fractions containing STAT1-P-positive and -nega-
tive fractions, the STAT1-P-positive fraction contained most of the
monosomy 7 cells (Fig. 4C). Next, we stimulated BMMNC from
monosomy 7 patients and compared levels of STAT1-P, STAT3-P,
and STAT5-P with those of normal controls. GCSF-stimulated
BMMNC demonstrated increased STAT1-P, decreased STAT3-P,
and increased STAT5-P compared with normal by immunoblot
(Fig. 4D), and confirmed on flow cytometry studies (Table 3).
Discussion
Here, we show that GCSF does not induce the karyotypic abnor-
mality of monosomy 7 in patients with bone marrow failure; rather,
high concentrations of GCSF favor the expansion of preexisting
monosomy 7 clones. This conclusion is based on our ability to
demonstratethatthesurvivalandproliferationofmonosomy7cells
were inferior to those of diploid cells at very low GCSF concen-
trations. These results suggest that the GCSFR, which mediates
proliferation, maturation, and survival of myeloid progenitor cells,
was abnormal in monosomy 7. Although genomic DNA GCSFR
was normal in monosomy 7 cells, there was an increase in the class
IVGCSFRisoform(whichlacksthemembrane-distalregionofthe
receptor responsible for differentiation). The class IV GCSFR
isoform, like the GCSFR in SCN, lacks the same 87 C-terminal
amino acids that are present in the full-length receptor; however,
the class IV isoform differs from the GCSFR in SCN because this
segment is replaced by a shorter, novel 34-aa sequence. The class
IV isoform is the product of alternative splicing, which is respon-
siblefortissue-specificvariationinthereceptor.Myeloidblastswith
normal genomic GCSFR DNA also express an increased amount
of isoform IV, again resulting from alternative splicing of the
mRNA (17). In murine CD34 cells transfected with the isoform IV
gene, relatively small increases in the ratio of isoform IV relative to
full-lengthisoformIcausesubstantialdefectsinmaturationwithout
affecting proliferation (which is controlled by the membrane-
proximal cytoplasmic domain of the receptor) (19). Our data are
consistentwiththoseofotherinvestigatorswhoreportthattheclass
IV isoform results in reduced, rather than increased, proliferative
signaling in BAF3 and 32D cells at GCSF concentrations ?100
ng?ml (20), with decreased survival at lower concentrations of
GCSF and increased proliferation at concentrations ?200 ng?ml
(21). In monosomy 7, increased class IV GCSFR isoform expres-
sionwascomparableto,orgreaterthan,thatmeasuredinleukemia
and in the murine model (17, 21). The pattern of STAT activation
we observed in monosomy 7 was similar to that described in SCN
(22–24), consistent with functional dominance of the class IV over
theclassIisoforms.Theapparentfunctionaldominanceoftheclass
IV isoform in monosomy 7 cells in leukemic blasts and in trans-
fectedmurinecellscouldbeexplainedif,likethetruncatedformof
GCSFR seen in SCN (18), the class IV isoform failed to normally
internalize following GCSF binding. Both the leukemic class IV
isoform and the truncated SCN GCSFR lack the dileucine residue
required for normal internalization (21), and defective internaliza-
tion of GCSFR in response to GCSF binding is well described in
SCN (18).
Monosomy 7 cells with up-regulated membrane GCSFR would
be expected to have a survival advantage compared with diploid
cells because forced cell surface expression of the mutated GCSFR
of SCN (25) or of isoform IV (26) confers resistance to apoptosis
by up-regulation of Akt (27, 28). If GCSF only facilitates prolifer-
ation of a stable, preexisting monosomy 7 clone, a decline in
monosomy 7 cells should occur at lower GCSF levels. In support of
this hypothesis, two of our patients showed substantial decreases in
the number of monosomy 7 bone marrow cells (one to a fully
normal karyotype) following clinical responses to immunosuppres-
sive therapy (unpublished data); others have reported similar
declines in monosomy 7 after successful treatment (29) or on
withdrawal of GCSF administration (3, 30). Previous studies dem-
onstrated a dosage effect in gene transcription in trisomy 8 and
monosomy 7 (31, 32). Multiple genes on chromosome 7 regulate
myeloid proliferation. CDP?cut or CCAAT displacement is a
transcription factor, located on the long arm of chromosome 7,
which indirectly down-regulates membrane GCSFR expression by
negatively regulating C?EBP-? (33, 34). Down-regulation of CDP?
cut would be expected to be accompanied by up-regulation of
C?EBP-?, which in turn increases GCSFR expression (35).
In conclusion, monosomy 7 cells are abnormally sensitive to high
concentrations of GCSF. Our findings link a GCSF-mediated
advantage for monosomy 7 cells to an overexpressed isoform of the
GCSFR. The abnormalities of proliferation and differentiation
closely mimic those of SCN, which has defects in the membrane
Table 3. Monosomy 7 BMMNC constitutively express STAT1-P
and have an increased STAT5?STAT3 ratio
Patient no.
%
monosomy 7
% STAT1-P
expression
STAT5-P?STAT3-P
ratio
4
2
9 14
30
24
4
6.6
20
26
2.6
65
68
7
15
13
MDS* mean (n ? 5)
Normal (n ? 10)
2
0
2
1
0.8
0.12
CD34 cells from monosomy 7 patients were stained for STAT1-P and
STAT5-Pandexaminedbyflowcytometry.Thecellsdemonstratedconstitutive
expression of STAT1-P and an increased STAT3?STAT5 ratio consistent with
abnormalities in GCSFR signal transduction. Patients with small numbers of
monosomy 7 cells showed STAT-P and isoform values similar to those for
normal controls.
*Denotes MDS with normal cytogenetics.
Fig.4.
werestainedwithCD34-PEandanti-STAT1-PmAbs.(A)WhenCD34cellsfrom
monosomy7cellsnotsubjectedtoGCSFstimulationwerestainedwithSTAT1-
P-FITC mAb, there was increased STAT1-P expression in patients compared
with controls. (B) Constitutive expression of STAT1-P in monosomy 7 patients
is shown on immunoblot. (C) When CD34 cells were sorted by flow cytometry
and FISH was performed, there was a preponderance of monosomy 7 cells in
the STAT1-P?group, indicating constitutive expression of this protein in
monosomy 7 cells. (D) When BMMNC from two healthy donors, four MDS
patients without monosomy 7, and three patients with monosomy 7 were
stimulated with GCSF (400 ng?ml) for 30 min, immunoblotting showed
increased expression of STAT1-P and decreased STAT3-P (example shown).
STATPexpressionisabnormalinmonosomy7.UnstimulatedBMMNC
14486 ?
www.pnas.org?cgi?doi?10.1073?pnas.0605245103 Sloand et al.

Page 5

distal portion of the GCSFR. Although this study reveals similar-
ities between SCN and monosomy 7, the mechanism of GCSFR
classIVoverrepresentationinmonosomy7remainsunknown.Our
data strongly suggest a mechanism in which GCSF expands an
existing monosomy 7 clone; however, we cannot exclude the
possibility that GCSF facilitates genomic instability with longer
exposure. Nevertheless, our findings should be reassuring to phy-
sicians who administer this agent to healthy donors of blood
products and to patients not experiencing bone marrow failure.
Materials and Methods
Patients.Bonemarrowwasobtainedfrom35patientswithMDS(as
defined in ref. 36): 28 with monosomy 7, 10 with AA with normal
cytogenetics, and 10 healthy adult volunteers. All individuals gave
written informed consent, in accordance with protocols approved
bytheInstitutionalReviewBoardoftheNationalHeart,Lung,and
Blood Institute.
Cell Preparation. BMMNC were aspirated from the posterior iliac
crest into syringes containing media supplemented 1:10 with hep-
arin(O’NeillandFeldman,St.Louis,MO)andpreparedbydensity
gradient centrifugation using lymphocyte separation medium (Or-
ganon, Durham, NC). Duplicate aliquots of freshly isolated
BMMNC were placed in Myelocult (StemCell Technologies, Van-
couver, Canada) in the presence of growth factors (37). In some
instances, cells were plated in methylcellulose with similar growth
factor support. At 4, 7, and 14 days later, colonies were counted,
individual colonies were picked, cells were lysed, and each colony’s
nuclei were subjected to FISH for chromosome 7.
FISH. FISH was performed as described in ref. 37, using probes for
chromosome 7 (Vysis, Downers Grove, IL). Percent positive stain-
ing was based on counting of 400 cells assessed. Three different
observers,whowereblindedwithrespecttosampleidentity,scored
three different sets of slides. Scores were averaged, and the mean
of the three was recorded.
Sequence Analysis of the GCSFR Coding Regions. Cryopreserved
BMMNC samples from patients confirmed by FISH to contain
Table 4. Primers for real-time PCR
GeneDesignation Sequence (5? 3 3?)
ABL
Forward primer
Reverse primer
TaqMan probe
GATACGAAGGGAGGGTGTACCA
CTCGGCCAGGGTGTTGAA
TGCTTCTGATGGCAAGCTCTACGTCTCCT
GCSFR1
Forward primer
Reverse primer
TaqMan probe
GGATCCGGGTCCATGGG
TTAAGAGGCAGGCCCAAGAAG
AACCCCAGGAAGCCCTA
GCSFR4
Forward primer
Reverse primer
TaqMan probe
CAGCCCCAATCCCAGTCT
GGAGCATGATCTGGTCCTTAAAGT
TCAGGCTGGGCCTCC
Table 5. Primer sequences for GCSFR sequencing
RegionsPCR primersSequence primers
I
F1-1 (8512): CTT GAA GGG TGA GAC AGG
AAG GTG
R1-1 (10991): CTT CTG TGC CTC TGT CTC
TAG GTC
FS1-A (8552): GGA GTA CAA AGG CCT
AGA TTA TAG C
FS1-B (10295): CAG ACA AGG TGA CAA
CAG AGG AAC
8529–
10970
F1-2 (8529): GAA GGT GAG GCC TCT GAG
TCT AGG
R1-2 (10970): G TCT TCC AGT GTC TTC
TTG TCT GTC
RS1-A (9024): GAA TAC AGG CGT GAG
CCA ACG TGC
RS1-B (10948): GT CGC TAG TTC TGT
GTT TCC CTC TCC
II
F2-1 (11477): CAT AGA GAT CGA ACC ACA
GAA ACG
R2-1 (10034): GAA CTC CTG ACC TCG GAT
GAT CTG
FS2-A (11527): GAA CCA CAG ACA GGG
AGA CAA GAG
FS2-B (12560): CTC CGT TTG TAT GAT
CCA TCC ACC
11499–
13000
F2-2 (11499): CGT GTC ATT GGC AAC GTG
GAC AGA TG
R2-2 (13000): CCT CCC AAA GTG CTG GGA
TTA CAG
RS2-A (12332): CTC CCA GAC CTG TTG
GAG TCC TAA G
RS2-B (12950): GGA CTA GAT TT AAC
CCA GGC AGT C
III
F3-1 (14307): CTA CTT TAC AGA TGA GGA
AAC TGA G
R3-1 (15308): CTG TAG GGA TCC AGT GTA
AGC CAG
FS3-A (14381): CAG GAT TTG AAC CCA
GGC TTC TGG
RS3-A (15234): GTT CCC TAT ACT TCT
GAT TGC TGG
14332–
15263
F3-2 (14332): GCT AAG ATG TTG ACC TAC
CAA AGG
R3-2 (15263): CAG GGC TGG AAG TAT GGT
AGG AAG
IV
F4-1 (15917): GAT GGA ATC TGG ATT AAA
TCC CTG G
R4-1 (18330): GAA GAA AGG ACC AAC TCT
GAG AAG
FS4-A (15997): GTG TCT GGG AAG CCA
CAA GAA GTC
FS4-B (16561): GGA CTG ACT TTG AAT
CCC CTG GTC
FS4-C (17362): CAC CAG TAA CAA CAG
TTG AAT GTC
15962–
18314
F4-2 (15962): GAG TCC CCT TGG ACT CTT
TTC TCA C
R4-2 (18314): CTG AG AAG TTT CCC CCT
GAC TGC TG
RS4-A (16553): CTC TCA AAA TCA GCA
TCC TTT GGG
RS4-B (17146): GCT TCA GAA GGT GTC
CCT TCA CTG
RS4-C (18276): AAA TAA CAG AGA CTC
AGG TAC ACC
Sloand et al.
PNAS ?
September 26, 2006 ?
vol. 103 ?
no. 39 ?
14487
MEDICAL SCIENCES