|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Prepublished online as a Blood First Edition Paper on November 27, 2002; DOI 10.1182/blood-2002-09-2898.
HEMATOPOIESIS
From the Department of Hematology of the University of
Crete School of Medicine and Institute of Molecular Biology and
Biotechnology, Heraklion, Crete, Greece; Cancer Research
United Kingdom Institute for Cancer Studies of the University of
Birmingham Medical School, United Kingdom; and Institute
of Clinical Immunology and Transfusion Medicine of the Justus-Liebig
University, Giessen, Germany.
To probe the pathophysiologic mechanisms underlying neutropenia in
patients with chronic idiopathic neutropenia (CIN) with hypoplastic and
left-shifted granulocytic series in the bone marrow (BM), we have
studied granulocytopoiesis in 32 adults with CIN by evaluating the
number and survival characteristics of cells in several stages of
granulocyte differentiation using flow cytometry and BM culture assays.
We found that patients with CIN displayed a low percentage of
CD34+/CD33+ cells, defective granulocyte
colony-forming unit (CFU-G) growth potential of BM mononuclear or
purified CD34+ cells, and low CFU-G recovery in long-term
BM cultures (LTBMCs), compared with controls (n = 46). A low
percentage of CD34+/CD33+ cells in patients was
associated with accelerated apoptosis and Fas overexpression within
this cell compartment compared with controls. No significant difference
was documented in the percentage of apoptotic cells or the
Fas+ cells within the fractionated
CD34+/CD33 The term chronic idiopathic neutropenia (CIN)
is addressed to any persistent unexplained reduction in the number of
circulating neutrophils below the lower limit of the normal
distribution for a given ethnic population.1-4 The
disorder can be identified among the other types of chronic
neutropenias by its acquired character, the absence of phasic
variations in neutrophil count, lack of clinical and laboratory
evidence for any underlying systemic disease to which neutropenia might
be ascribed, and absence of any drug relationship. Typically, CIN is
characterized by low incidence of infections and usually benign
outcome.5,6 The disorder displays a female
predominance2,5,7 and an HLA class II genetic
predisposition.8
Disparate mechanisms have been implicated in the pathogenesis of CIN,
including decreased neutrophil production in the bone marrow
(BM),1,2,4,9 excessive margination or enhanced neutrophil
extravasation into the tissues,10,11 and immune-mediated destruction of mature blood neutrophils or their marrow
progenitors.12,13 BM morphology varies according to the
presumed underlying pathophysiology. Neutropenia due to increased
peripheral sequestration or destruction of mature neutrophils is
usually associated with a compensatory hyperplasia of the BM myeloid
precursor cells, whereas severe myeloid hypoplasia might imply rare
cases of immune-mediated destruction of BM immature myeloid progenitor
cells by humoral or cellular cytotoxic mechanisms.14 In
the majority of CIN cases, however, BM morphology does not clearly
explain the degree of neutropenia and a mild myeloid hypoplasia
affecting mainly the postmitotic, maturating pool of the granulocytic
series is typically recognized.14-16
In vitro BM growth studies using culture techniques have long been used
to assess patterns of granulocytopoiesis and determine possible
pathogenetic mechanisms underlying CIN.17,18 No conclusive evidence, however, has been reported regarding the number and functional characteristics of the BM myeloid colony-forming cells since
increased,19 normal,20 or decreased myeloid
progenitor cell growth has been described in the affected
subjects.9,21 These inconsistencies might be explained by
the diversity of the patients studied in regard to the underlying
pathophysiology, given that cases with immune- and nonimmune-mediated
neutropenia have been recorded.
In the present study, we probed more deeply into the mechanisms of
granulocytopoiesis in patients with CIN displaying myeloid hypoplasia
and negative tests for antineutrophil antibodies. This type of the
disorder is the most commonly seen in adults.2,15 We
specifically examined the BM myeloid cell reserve and function in
several stages of granulocyte differentiation, from the early progenitors to the mature neutrophils. We also explored the influence of patients' marrow microenvironment on myelopoiesis by investigating the capacity of BM stromal cells to induce and support the growth of
the myeloid progenitor cells.
Patients
BM samples
Purification of BM myeloid progenitor and precursor cells BMMCs from CIN patients and healthy controls were fractionated into CD34+ early progenitor, CD34 /CD33+ myeloid progenitor, and
CD34/CD33/CD15+ granulocyte
precursor cells using sequential immunomagnetic labeling and sorting
according to the manufacturer's protocol (Miltenyi Biotec, Bergisch
Gladbach, Germany). In all experiments, purity of each subpopulation
was more than 96% as estimated by flow cytometry.
Flow cytometric analysis of CD34+ cells An indirect immunofluorescence technique was used to quantitate the CD34+ cells and their subpopulations in the BMMC fraction. In brief, 1 × 106 BMMCs were stained with phycoerythrin (PE)-conjugated mouse antihuman CD34 monoclonal antibody (mAb) (QBEND-10; Immunotech, Marseille, France) and fluorescein isothiocyanate (FITC)-conjugated mouse antihuman CD33 mAb (D3HL60-251; Immunotech) for 30 minutes on ice. PE- and FITC-conjugated mouse IgG isotype-matched controls were used as negative controls. Cells were washed twice in phosphate-buffered saline (PBS)/1% fetal bovine serum (FBS; Gibco)/0.05% sodium azide and fixed in 500 µL 2% paraformaldeyde solution (PFA; Sigma). Data were acquired and processed on 500 000 events using an Epics Elite model flow cytometer (Coulter, Miami, FL). The estimation of CD34+/CD33+ cells was performed in the gate of cells with low forward (FSC) and low right-angle side scatter (SSC) properties.7-AAD staining for the study of apoptosis Aliquots of 1 × 106 BMMCs were stained with PE-conjugated anti-CD34 and FITC-conjugated mouse antihuman Fas (CD95; LOB 3/17; Serotec, Oxford, United Kingdom) mAbs as described (see "Flow cytometric analysis of CD34+ cells"). In some experiments, aliquots of 1 × 106 purified CD34+ or CD34/CD33+ and
CD34/CD33/CD15+ cells were also
stained with FITC-conjugated anti-Fas and PE-conjugated anti-CD33 mAbs
or FITC-conjugated anti-Fas and PE-conjugated mouse antihuman CD15
(80H5; Immunotech) mAbs, respectively. The cells were further stained
prior to fixation with 100 µL 7-amino-actinomycin D solution (200 µg/mL; 7-AAD; Calbiochem-Novabiochem, La Jolla, CA) as previously
described24 and analyzed on 500 000 events using 5 parameters: FSC, SSC, and triple-color immunofluorescence from FITC,
PE, and 7-AAD. For the BMMCs, a scattergram was created by combining
SSC with CD34 fluorescence in the gate of cells with low FSC and SSC
properties and a second scattergram by combining CD34 and Fas
fluorescence in the gate of CD34+ cells. Finally, a
scattergram was created by combining FSC with 7-AAD fluorescence to
quantitate 7-AAD (live), 7-AADdim (early
apoptotic), and 7-AADbright (late apoptotic) cells in the
gate of CD34+ BMMCs. For the purified CD34+
cells, after creating a scattergram combining SSC with CD33
fluorescence, a second scattergram was created by combining CD33 with
Fas fluorescence gated on the CD33+ purified
CD34+ cells (Figure 1).
Similarly, for the purified CD34/CD33+ and
CD34/CD33/CD15+ cells, a
scattergram was created by combining SSC with CD15 fluorescence gated
on each purified population and a second scattergram by combining CD15
with Fas fluorescence gated on the CD15+ cells of each
purified population (Figure 2). In each
case, a scattergram of FSC versus 7-AAD fluorescence was generated for quantification of live, early, and late apoptotic cells in the gates of
CD34+/CD33,
CD34+/CD33+,
CD33+/CD15,
CD33+/CD15+, and
CD33/CD15+ cells, representing the sequential
stages of granulocyte differentiation. In each cell population, a
subset analysis in the Fas+ and Fas cells was
also performed. A representative example of this analysis is shown in
Figures 1 and 2. Results of apoptosis were expressed as total
7-AAD+ (7-AADbright plus
7-AADdim) cells.
For the study of apoptosis of peripheral blood neutrophils, aliquots of 100 µL EDTA (ethylenediaminetetraacetic acid)-anticoagulated, fresh or stored overnight at room temperature, blood samples were washed twice and stained with 7-AAD as described (see "7-AAD staining for the study of apoptosis"). Following centrifugation and removal of the supernatant, contaminating red cells were lysed with 0.12% formic acid and samples were fixed in 0.2% PFA using the Q-prep reagent system (Coulter). The samples were analyzed on 10 000 events by creating a scattergram of FSC versus 7-AAD fluorescence gated on the cells with high forward and high side scatter properties where neutrophils are included. Clonogenic progenitor cell assays BMMCs or purified CD34+ BM cells were cultured in 35-mm Petri dishes in 1 mL IMDM supplemented with 30% FBS, 1% bovine serum albumin (Gibco), 104 M mercaptoethanol (Sigma),
0.075% sodium bicarbonate (Gibco), 2 mM L-glutamine
(Sigma), 0.9% methylcellulose (Stem Cell Technologies, Vancouver, BC,
Canada), in the presence of 5 ng granulocyte-macrophage colony-stimulating factor (R & D Systems, Minneapolis, MN), 50 ng
interleukin 3 (R & D Systems), and 2 IU erythropoietin (Janssen-Ciliag, Athens, Greece) at a concentration of 105 BMMCs or
3 × 103 CD34+ cells/mL culture medium.
Following 14 days of culture in a 37°C-5%CO2 fully
humidified atmosphere, myeloid colonies were scored and classified as
granulocyte colony-forming units (CFU-Gs), macrophage colony-forming
units (CFU-Ms), and granulocyte-macrophage colony-forming units
(CFU-GMs), according to established criteria.25 Results were finally expressed as CFU-Gs, total CFU-Ms (CFU-Ms plus CFU-GMs), and total CFU-GMs (CFU-Gs plus total CFU-Ms).
Limiting dilution assay for quantification of LTC-ICs Seven dilutions of a single suspension of CD34+ cells were overlaid on preformed murine MS-5 stromal layers26 at concentrations ranging from 10 to 1000 cells/well in 96-well culture plates as previously detailed.27 Cultures were fed weekly by demi-depopulation, and after 5 weeks were overlaid with methylcellulose culture medium with growth factors as described (see "Clonogenic progenitor cell assays"). The frequency of long-term culture-initiating cells (LTC-ICs) was calculated by determining the CD34+ cell dilution that resulted in 37% or fewer wells negative for colonies28 using a Fig.P Biosoft PC program (Biosoft, Cambridge, United Kingdom). Using this culture system, one can also calculate the proliferative potential of LTC-ICs for each subject by dividing the sum total of colonies at week 5 by the number of LTC-ICs among the CD34+ cells plated.29Assessment of BM stromal cell function Standard long-term BM cultures.
Long-term bone marrow cultures (LTBMCs) from 107 BMMCs
were grown according to the standard technique25 in 10 mL
IMDM supplemented with 10% FBS, 10% horse serum (Gibco), 100 IU/mL
PS, 2 mmol L-glutamine and 10 Cytokine measurement in LTBMC supernatants.
Cell-free supernatants of confluent LTBMCs (week 3-4) were stored at
RT-PCR for the detection of IFN- Recharged LTBMCs. A 2-stage culture procedure was used to test the capacity of patient BM stromal layers to support normal hematopoiesis. Confluent stromal layers from patients and healthy controls grown in standard LTBMCs were irradiated (10 Gy) and recharged with 5 × 104 normal allogeneic CD34+ BM cells as previously described.27,32,33 In each experiment, flasks were recharged in triplicate and CD34+ cells from the same healthy control were used to test cultures from patients and healthy controls. Cultures were monitored weekly by determining the number of nonadherent cells and the frequency of clonogenic progenitor cells. Statistical analysis Data were analyzed in the Graphpad Prism statistical PC program (Graphpad Software, San Diego, CA) by means of the Student t test and the Pearson coefficient of correlation test. Standard 2-way analysis of variance was applied to define differences in the number of nonadherent cells and the number of clonogenic progenitor cells in LTBMCs and the 2 test to define differences in IFN-
and FasL expression in LTBMC stromal layers between patients and
controls. Grouped data are expressed as mean ± 1 SD.
BM progenitor cells Flow cytometric analysis of CD34+ cells in the BMMC fraction of CIN patients and healthy controls is presented in Table 2. The patients displayed significantly lower percentage of CD34+ cells compared with the controls (P < .0001) due to the lower proportion of both the committed myeloid CD34+/CD33+ (P < .0001) and the more primitive CD34+/CD33 (P = .0039) cells.
However, the percentage of CD33+ cells detected within the
CD34+ cell fraction was significantly lower in the patients
(12.04% ± 4.42%) compared with the controls (24.60% ± 15.82%;
P = .0002) suggesting that the reduced proportion of the
CD34+/CD33+ cells in CIN patients does not
simply reflect the lower total CD34+ cell numbers but
concerns specifically the myeloid progenitors. It has been shown that
the CD34+/CD33 BM cell fraction contains a
significant proportion of the early progenitors cells.34
However, despite the low percentage of CD34+/CD33 cells in our CIN patients, the
frequency of LTC-ICs, representing the best available approximation of
primitive stem cells,35 did not differ significantly
between patients (3.53 ± 1.54/103 CD34+
cells; n = 15) and healthy controls (3.20 ± 1.86/103
CD34+ cells; n = 20; P = .582). Similarly,
the proliferative potential of patient LTC-ICs (0.88 ± 0.34) did not
differ statistically from the respective control subjects
(0.97 ± 0.28; P = .513). Taken together, these data
suggest that CIN patients display normal number and proliferative
characteristics of primitive stem cells but display low frequency of
committed myeloid progenitors.
Survival characteristics of BM myeloid cells and peripheral blood neutrophils To explore whether the decrease of CD34+ cell percentages in CIN patients is due to increased apoptosis, we first evaluated the proportion of apoptotic cells within the CD34+ fraction of BMMCs (Table 2). We found that patient (n = 31) CD34+ BMMCs contained a significantly higher number of apoptotic cells compared with the controls (n = 29; P = .0189). In contrast, no statistically significant difference was found between patients and controls in the percentage of apoptotic cells detected in the non-CD34+ BMMC fraction (P = .162).Because Fas antigen expression has been associated with apoptosis of BM
hematopoietic progenitor cells,36 we next evaluated the
expression of this molecule on patient CD34+ cells (Table
2). The proportion of Fas+ cells detected in the
CD34+ BMMC fraction was significantly higher in the
patients (n = 31) compared with the controls (n = 28;
P = .0043) and individual Fas+ cell
proportions correlated positively with the percentages of apoptotic
CD34+ cells (r = .469, P = .049), suggesting
that Fas expression is possibly actively involved in the apoptotic
depletion of patient progenitor cells. The highly significant
difference in the percentage of apoptotic cells detected between the
Fas+ (38.84% ± 24.71%) and Fas Having demonstrated that increased apoptosis is probably involved in
the reduction of patient CD34+ cells, we next evaluated the
survival characteristic of the committed myeloid progenitor cells in
CIN using immunomagnetically sorted highly purified CD34+
cells (Table 2). We found that the percentage of apoptotic cells within
the CD34+/CD33+ cell fraction was significantly
higher in the patients (n = 18) compared with the controls (n = 10;
P = .0054). In contrast, no statistically significant
difference was found between patients and controls in the percentage of
apoptotic cells detected within the
CD34+/CD33 To investigate whether accelerated apoptosis characterizes specifically
the BM myeloid progenitor cells or is a feature of all stages of
granulocyte differentiation in CIN, we next evaluated the survival
characteristics of BM myeloid precursor cells. The percentage of
apoptotic cells and the proportion of Fas+ cells detected
in the CD34 Finally, we examined the apoptotic rate of patient peripheral blood
neutrophils using fresh or overnight stored EDTA-anticoagulated blood
samples as previously suggested.37,38 The percentage of
apoptotic neutrophils detected in freshly drawn blood samples from CIN
patients (2.36% ± 1.62%, n = 23) was not statistically different
from the controls (1.73% ± 0.70%, n = 13;
P = .150), nor were the proportions of Fas+
neutrophils significantly different between patients
(51.54% ± 26.01%) and control subjects (61.95% ± 23.55%;
P = .209). Following overnight storage, a highly
significant increase was observed in the percentage of apoptotic
neutrophils in both CIN patients (31.99% ± 16.17%) and healthy
controls (30.91% ± 17.07%) compared to baseline
(P < .0001 and P < .0001, respectively),
which was associated with a significant reduction in the proportion of
Fas+ neutrophils after storage (20.70% ± 24.77%
in the patients and 14.61% ± 11.31% in the controls) in
comparison to baseline (P < .001 in patients and
P < .0001 in the controls). However, no significant
difference was documented between patients and controls in the
percentage of apoptotic or the proportion of Fas+ stored
neutrophils (P = .833 and P = .351,
respectively), suggesting that peripheral blood neutrophils display
normal survival characteristics in CIN. Notably, no significant
difference was found in apoptosis between the Fas+ and
Fas Colony-forming cells The frequency of clonogenic progenitor cells in the BMMC fraction of CIN patients (n = 31) and healthy controls (n = 34) is depicted in Figure 3. The mean number of total CFU-GMs obtained by 107 BMMCs was significantly lower in the patients (4516 ± 2073) compared with the controls (7650 ± 2136; P < .0001). This decrease reflected probably the lower number of CFU-Gs in CIN patients compared with controls (2290 ± 1289 versus 4556 ± 1919; P < .0001) because no statistically significant difference was found between patients and control subjects in the number of total CFU-Ms (2226 ± 1559 versus 3094 ± 2365; P = .0886).
To investigate whether the decreased CFU-GM formation in CIN is due to the low progenitor cell number in patients' BMMC fraction or is the consequence of a progenitor cell defect, we tested the clonogenic potential of purified CD34+ BM cells (Figure 3). We found that the total CFU-GM number obtained by 5 × 104 CD34+ cells was significantly lower in the patients (668 ± 499, n = 23) compared with the controls (993 ± 362, n = 15; P = .037) and this reduction was essentially due to the low number of CFU-Gs in the patients (331 ± 186 versus 537 ± 180 in the controls; P = .0018) because no significant difference was found between patients and controls in the number of CFU-Ms (337 ± 327 and 456 ± 294, respectively; P = .262). These findings are in agreement with the flow cytometric data suggesting low frequency of committed myeloid progenitor cells in CIN and provide strong evidence for selective decrease of the granulocyte progenitor cells in these patients. Standard LTBMCs Typical confluent stromal layers were formed over the first 3 to 4 weeks in both patient (n = 31) and normal (n = 34) LTBMCs. However, the average nonadherent cell recovery was significantly lower in patient LTBMCs compared with controls (F = 5.415 > F  
Cytokine production by LTBMC stromal layers It has been reported that the constitutive production of inhibitory cytokines in the BM microenvironment may mediate hematopoietic progenitor cell inhibition.32,39-41 To examine the possible inhibitory effect of BM stromal cells on hematopoiesis in CIN, we first evaluated the levels of TNF- ,
IFN-, and FasL, molecules known to have proapoptotic
properties,42 in the supernatants of confluent LTBMCs from
patients and controls. We found that TNF- levels were significantly
higher in the patients (7.41 ± 6.19 pg/mL; median, 6.05 pg/mL;
range, 1.26-32.3 pg/mL; n = 26) than the controls (3.24 ± 1.76
pg/mL; median, 3.52 pg/mL; range, 0.35-6.06 pg/mL; n = 12;
P = .029; Figure 5) and
individual TNF- values inversely correlated with the numbers of
CFU-GMs (r =  0.367, P = .030) and positively with the
proportions of CD34+/Fas+ cells (r = .709,
P < .0001) and the percentages of
CD34+/7-AAD+ cells (r = .783,
P < .0001; Figure 6). These
data suggest that increased TNF- production by patient stromal cells
may exert a negative effect on the myeloid progenitor cell growth by
inducing apoptosis in the CD34+ cell population.
Because ELISA failed to detect IFN-
Recharged LTBMCs To test the hematopoiesis-supporting capacity of patient stromal cells independently of the autologous progenitor cells, confluent LTBMCs from 4 CIN patients with IFN-- and FasL-expressing stromal cells and 4 healthy controls were recharged with normal
CD34+ cells. The frequency of CFU-GMs and CFU-Gs obtained
over a period of 5 weeks after the CD34+ cell inoculum were
lower in patient LTBMCs compared with controls (F
Acquired CIN in adults has long been recognized more as a syndrome than a disease generally characterized by a homogenous constellation of clinical features and a rather unifying natural history but variable underlying pathophysiology and BM morphology.2,4,5,20 Serum antineutrophil antibody screening, marrow examination, and in vitro culture studies evaluating the granulocyte progenitor growth potential have been used to determine parameters with possible pathophysiologic relevance in the clinical setting of this variable disease state.18,20 Results from such studies have revealed a common cellular defect in a subgroup of patients with negative serum antineutrophil antibody activity that concerns a selective hypoplasia of the BM granulocytic series with increased immature-mature cell ratio suggesting either a defective myeloid cell development or loss of cells during the differentiation process.15 In an attempt to probe the pathophysiologic mechanisms underlying CIN,
we have investigated cell reserves and function in several stages of
granulocyte differentiation in CIN patients using in vitro culture
assays and flow cytometry. The data of the study show significant
quantitative and functional abnormalities of the granulocyte progenitor
cells indicated by the low percentage of
CD34+/CD33+ cells, the low frequency of CFU-G
progenitors cells within the BMMCs or the purified CD34+
cell fraction, and the low CFU-G recovery in LTBMCs. To characterize further the underlying cellular defect in CIN, we conducted apoptosis studies in unfractionated and purified BM-derived granulocyte progenitor and precursor cell populations as well as in peripheral blood neutrophils. These studies demonstrated accelerated apoptosis in
the CD34+ cell fraction of the patients concerning
specifically the committed CD34+/CD33+ but not
the more primitive CD34+/CD33 Increased apoptosis of BM progenitor cells has been implicated in the
pathophysiology of marrow failure associated with myelodysplastic syndromes and aplastic anemia43 and Fas receptor
triggering on hematopoietic progenitors has been recognized as a
central pathway for the elimination of stem cells in these
conditions.42,44 We therefore evaluated Fas antigen
expression on the progenitor cells of patients with CIN and we found a
significant up-regulation of this molecule on the CD34+
cell fraction that specifically characterized the
CD34+/CD33+ committed myeloid progenitor cell
subpopulation but not the more primitive
CD34+/CD33 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
;
n = 31) and healthy controls (
; n = 34) using clonogenic assays.
The right graph represents the mean colony values (± SEM) obtained by
5 × 104 immunomagnetically sorted CD34+
cells in 23 CIN patients
(
; n = 31) and healthy controls (
; n = 34)
over a period of 8 weeks. (B) Mean frequency (± SEM) of myeloid
progenitor cells in the nonadherent cell fraction throughout 8 weeks of
culture. Comparison between patient and control cultures was performed
using the 2-way analysis of variance test.
