| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Medicine, School of Medicine, University of California at San Diego, La Jolla, CA 92093; and
Department of Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, PA
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- or 
-L chain positive. They did not express CD3, CD4, CD5, CD10,
CD11b, CD13, CD25, CD56, CD14, CD33, or CD64. Isolated 5G9+
cells were potent APCs in allogeneic MLR, compared with
5G9- PBMC, 5G9- B cells, monocytes, and
monocytes cultured in IL-4 and GM-CSF for 24 h. mAb 5G9 defines a
novel peripheral blood cell with B cell phenotype and DC morphology and
function: DC-like B cells. The significance of this cell in immune
responses requires further study. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most current studies of human DC have focused on myeloid DC. However,
Galy et al. (3) recently found that a subpopulation of
CD34+
lin-CD45RA+CD10+
cells gives rise only to T and B lymphocytes, NK cells, and DC.
Limiting dilution analysis demonstrated the existence of multipotential
B/NK/DC progenitor clones in the CD34high
lin-CD10+ adult bone
marrow cell population, indicating that nonprimitive progenitors for
lymphoid cells and for DC can be distinct from those of myeloid,
implying that the DC lineage is developmentally more closely related to
the lymphoid than the myeloid lineage. Serreze et al. (4)
studied the APCs that contribute to the development and activation of
diabetogenic T cells in the nonobese diabetic (NOD) mouse model of
insulin-dependent diabetes mellitus (IDDM). They utilized a
functionally inactivated Ig µ allele to generate a "speed
congenic" stock of B cell-deficient NOD mice. They found that these
mice were free of overt IDDM and insulin resistance, although with
normal numbers of T cells, implying that B cells play an unrecognized
role (i.e., functioning as potent APCs), which is essential for the
initial development and/or activation of T cells autoreactive with
pancreatic 
cells in NOD mice. Bjorck’s recent study demonstrated
that murine CD19+ pro-B cells develop into DC
with T cell stimulatory properties when cultured with murine IL-1,
IL-3, IL-7, TNF-
, stem cell factor (SCF), and human Flt-3 ligand
(5). The above studies suggested a close relationship
between DC and B cells and B cell precursors having B/DC bipotential
lineage capacity.
For decades, immunologists have reported the conversion of B
lymphocytic tumors to cells with properties of macrophages. Borrello
and Phipps recently discussed the identification of normal B/macrophage
(B/M
) cells, indicating that B/M cells are the normal counterpart
of the well-described endpoint of CD5+ B cell
lymphoma lineage switch to macrophages. These B/M cells are
morphologically distinct from classical macrophages, having unique
surface characteristics that suggest their B cell origin
(6). The identification of B/M imply that switching can
occur between subsets of distantly related cells that derived from
separate lineages with specialized functions.
In an attempt to generate novel anti-B cell mAbs, we discovered a hybridoma that reacted with B lymphoma cells and a subpopulation of tonsillar cells. Further work revealed the mAb to define what appears to be a novel Ag on a subset of B cells. In this paper, we describe the production of this unique mAb reactive with a 220 kDa Ag expressed on a B cell subpopulation that has the properties of a DC and functions as a potent stimulator of allogeneic T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Daudi cells (American Type Culture Collection (ATCC), Manassas, VA) were washed in serum-free RPMI 1640, incubated at 37°C for 2 h, then washed 2x in PBS. Immunizations were conducted by injecting 1 x 106 Daudi cells i.p. into a 7-wk-old female BALB/c mouse (Charles River Laboratory, Wilmington, MA) once a week for 4 wk. Four days after the last injection, the mouse splenocytes were dissected and fused with P3X63Ag8.653 myeloma cells (ATCC) according to the procedure described by Kohler and Milstein (7). Supernatants were tested for the presence of Abs against Daudi cells and the absence of reactivity with HL-60 cells by an ELISA. The hybridoma clone producing mAb 5G9 was subcloned twice by the limiting dilution method. The subtype of mAb 5G9 was determined by the mAb Check Kit for isotyping of mouse mAb in culture supernatant (Sterogene Bioseparations, Arcadia, CA) following the manufacturer’s instructions.
Ab purification
Hybridoma cells were injected i.p. into BALB/c mice. Ascites was collected and centrifuged at 8000 x g for 15 min. Supernatant was removed and filtered through a 0.45-µm filter. The protein was precipitated by adding a saturated ammonium sulfate solution to a final concentration of 45% saturation. The precipitate was dialyzed in PBS (pH 8.0) at 4°C for 24 h. Ab was purified by a protein A-Sepharose column (Pharmacia, Piscataway, NJ), and the purity was determined by SDS-PAGE.
Immunoprecipitation
Daudi cells (2 x 107) were cultured for 1 h in 5 ml of methionine- and cysteine-free medium. Labeling was initiated by adding 300 µci of [35S]methionine-cysteine, and cells were collected after 3 h of culture. Labeled Daudi cells were lysed by suspension in 500 µl ice-cold lysis buffer (50 mM Tris-HCl, 1% Nonidet P-40, 50 mg/ml BSA, with a mixture of protease inhibitors (5 mM EDTA, 100 µg/ml PMSF, and 5 µg/ml aprotinin)). Lysates were mixed with 100 µl of normal mouse (NM) serum and precleared with killed fixed Staphylococcus aureus cells that had been washed in lysis buffer. The precleared lysate was divided into 100-µl aliquots. MAb 5G9, SCCL-1 (a mouse IgG2a mAb reactive with the transferrin receptor used for positive control; Ref. 8), and NM serum (negative control) were added to the aliquots of lysate and mixed for 1 h at 4°C. Rabbit anti-mouse IgG Ab was then added and mixed for another hour. The aliquots of lysate were adsorbed to 50 µl of packed, washed protein A-Sepharose, and each sample was divided into two parts, one kept under nonreducing conditions, the other exposed to DTT. The samples were analyzed by SDS-PAGE.
Cell isolation procedure
Leukocyte-enriched leukopaks (obtained from Central Blood Bank,
Pittsburgh, PA) were treated with Gey’s solution to lyse erythrocytes
and washed 4x at 210 x g with PBS to remove
platelets. PBMC were cultured overnight in RPMI 1640 medium containing
10% heat-inactivated FCS, 10 mM glutamine, and
penicillin/streptomycin. A DC-enriched population of cells was obtained
as previously described with some modifications (9, 10).
Briefly, the mononuclear cells were suspended in IMDM (Life
Technologies, Grand Island, NY) medium containing 0.2% BSA (Sigma,
Louis, MO) at 1 x 107/ml, and incubated
with 2–3 subsequent 45-min periods at 37°C in 150 x 15-mm
petri dishes (Becton Dickinson Labware, Lincoln Park, NJ) to remove
adherent monocytes. Monocyte-derived DC (MoDC) were obtained by culture
of the adherent monocytes in RPMI 1640 complete medium containing 1000
U/ml IL-4, 100 ng/ml GM-CSF, and 10 ng/ml TNF-. The remaining
nonadherent cells were resuspended at 2 x
107 cells/ml in RPMI 1640 containing 1% FCS and
incubated with mAb OKT-3 for 30 min at 4°C, and washed 3x and
incubated with sheep anti-mouse IgG Dynabeads M-450 (N-0212; Dynal,
Oslo, Norway) for 30 min at 4°C. The cell suspension was washed 3x
against Dynal MPC to remove T cells. The T cells and monocyte-depleted
mononuclear cells were resuspended at 5 x
106/ml and layered onto hypertonic 14.5%
metrizamide (Sigma) in 15-ml conical centrifuge tubes and sedimented at
650 x g for 15 min at room temperature. DC-enriched
cell populations were harvested from the interface of the metrizamide
gradient.
To isolate 5G9-positive population of cells, monocytes were depleted from mononuclear cells by 2–3 rounds of adherent to petri dishes. Nonadherent cells were either isolated immediately or after culturing 1–5 days in RPMI 1640 complete medium with 1000 U/ml IL-4. Cells were resuspended in 5G9 culture supernatant at 2–4 x 107/ml, incubated for 30 min at 4°C, washed 3x, resuspended in RPMI 1640, 2% FCS and mixed with sheep anti-mouse IgG Dynabeads M-450 continuously for 60 min at 4°C. 5G9+ cells were obtained by washing 3x the cell suspension against Dynal MPC. The cells attached to magnetic beads were suspended in RPMI 1640 10% FCS and incubated at 37°C overnight. About 30–50% of isolated 5G9+ cells were detached from the Dynabeads and washed three times against Dynal MPC.
Cells and cell lines
Peripheral blood samples from adult patients with chronic
lymphocytic leukemia (CLL) were obtained with informed consent under
the auspices of the Institutional Review Board of the respective
institution. Mononuclear cells were isolated by Ficoll-Hypaque
centrifugation. The cell lines are listed in Table I
. Cell lines were cultured in RPMI 1640
(Life Technologies) containing 10% FBS (HyClone, Logan, UT) in a
humidified atmosphere of 5% CO2 and air. Cells
in log phase were freshly harvested from culture, washed three times,
and resuspended in RPMI 1640 medium before use.
|
PE-labeled mouse anti-human mAbs CD3, CD4, CD5, CD10, CD11b,
CD13, CD14, CD19, CD20, CD21, CD22, CD33, CD38, CD40, and CD56 HLA-Dr
were purchased from Becton Dickinson (San Jose, CA); CD83-PE and
CD19-energy-coupled dye (ECD) from Immunotech (Westbrook, ME); and
CD80-FITC and CD86-PE from PharMingen (San Diego, CA). The samples were
washed twice with IFA medium (PBS supplemented with 4% FCS and 0.1%
sodium azide), and the cells were resuspended in IFA medium containing
2% human AB serum and 2% goat serum (to block Fc
R). The lack of
reaction of control tubes that were stained with mouse IgG1-FITC/mouse
IgG1-PE or goat anti-mouse IgG-FITC demonstrated that nonspecific
binding was well blocked. Indirect immunofluorescence staining was
conducted after washing the cells twice with PBS containing 1% BSA and
0.1% sodium azide. Cells were incubated with mAb 5G9 for 30 min at
4°C, washed twice, incubated with FITC-labeled goat
F(ab')2 anti-mouse IgG (HL) (Caltag
Laboratories, Burlingame, CA), and washed before analysis on FACScan
system (Becton Dickinson Immunocytometry System) or a Coulter XL
(4-color studies; Beckman Coulter, Miami, FL). Double staining involved
labeling with mAb and goat F(ab')2 anti-mouse
IgG followed by incubation 10 min with 10% mouse serum at 4°C before
adding a PE-conjugated mAb.
Light microscopy
The cytospin slides of peripheral blood, cultured moDC, and purified 5G9+ cells were stained with Wright-Giemsa by using an Ames HEMA-TEK Slide stainer (Curtis Matheson Scientific, Houston, TX). Tissue was fixed in 10% neutral buffered formalin and embedded in paraffin; and 4-micron sections were cut and stained with hematoxylin-eosin (H-E) for microscopic examination.
MLR
Stimulators for the MLR were total PBMC, freshly isolated 5G9+ cells, 5G9- cells, 5G9- B cells, and monocytes derived from the separation procedures, cultured in IL-4/GM-CSF for 24 h. 5G9- B cells were obtained by negative selection of 5G9- PBMC population with a mixture of mAbs CD3, 3G8 (CD16), 251 (CD33) and sheep anti-mouse IgG Dynabeads M-450. These cells were irradiated (3000 rad) and added in graded doses to 2 x 105 allogeneic or syngeneic (autologous) mononuclear cells in 96-well U-bottom tissue culture plates in final volume of 0.2 ml. The medium used for MLR was AIM-V medium (Life Technologies) containing 10% inactivated human AB serum. Proliferation was measured by the uptake of [3H]thymidine (1 µCi/well) added 8 h before the end of culture (37°C, 5% CO2) on the 5th day and counted in 1205 Betaplate counter (Wallac, Gaithersburg, MD).
In alternative experiments, 5G9+ cells isolated by beads were cultured in RPMI 1640 with 10% FCS supplemented with IL-4 (1000 U/ml), or soluble CD40 ligand (CD40L) from a stable transfected cell supernatant (three parts RPMI:one part CD40L supernatant) (kindly provided by Dr. Thomas Kipps, University of California San Diego), or IL-4 plus CD40L supernatant (IL-4/CD40L 5G9+ cells) for 4 days. The cultured 5G9+ cells were then used as MLR stimulators to compare with MoDC cultured in IL-4/GM-CSF medium for 5 days and 5G9+ cells freshly isolated from PBMC that were cultured for 5 days in medium supplemented with IL-4 (IL-4/PBMC 5G9+ cells).
Immunohistochemistry
Vectastain ABC kit (Vector Laboratories, Burlingame, CA) was used for immunohistochemical analysis (biotin-avidin-peroxidase-AEC) of 5G9, CD19, or CD21 expression on cytospin slides for PBMC, DC, MoDC, 5G9+ cells, or tissue section slides, following the manufacture’s instruction. The slides or sections were counterstained with hematoxylin. For double staining, cytospin slides were first incubated with mAb B4 (anti-CD19), washed, and incubated with alkaline phosphatase-conjugated goat F(ab')2 anti-mouse IgG (Caltag), then developed with Vector blue alkaline phosphatase substrate kit III following the manufacture’s instruction. After blocking with goat serum and blocking serum of ABC kit, the slides were incubated with mAb 5G9, and Vectastain ABC kit was used for developing second color (red). Control slides were stained with double staining procedure without mAb CD19 or 5G9 or both.
Scanning electron microscopy
Cytospin glass discs were fixed in 1% glutaraldehyde and dehydrated in a concentration gradient of ethanol. Cells were then critical-point dried through CO2 to maintain cellular ultrastructure and attached to carbon planchettes with colloidal graphite, coated with a 250 A layer of gold in a sputter-coater, and studied in a scanning electron microscope.
Statistical analysis
The significance level was determined by the student t test when applicable.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
mAb 5G9 was obtained by fusing murine myeloma cells with BALB/c
mouse spleen lymphocytes immunized with Daudi cells. The subtype of mAb
5G9 was determined to be a murine IgG1. By Western blot and
immunoprecipitation, it reacted with a 220-kDa Ag on Daudi cells, which
reduced to four subunits with apparent m.w. of 55, 65, 80, and 85 kDa
(Fig. 1).
|
). mAb 5G9 did not react
with peripheral blood erythrocytes, platelets (data not shown), T cells
(CD3+), and NK (CD56+)
cells, and reacted weakly with monocytes and granulocytes
(n = 8). Double-staining 5G9+
cells in resting PBMC with CD83 and CD86 showed no double-positive
cells (Fig. 2). This mAb reacted with lymphoma cell lines and cells
from CLL patients (four of four), but did not react with AML and ALL
cell lines or AML and ALL patient-derived cells (Table I
).
|
Immunohistochemical study of 5G9 using Vectastain ABC demonstrated
that the mAb stained scattered cells on human tonsil sections, which
appeared morphologically to be DCs (Fig. 3, A and B). On
cytospin slides of MoDC cultured with IL-4, GM-CSF, and TNF- for
4–10 days, mAb 5G9 strongly reacted with cells with typical dendritic
morphology that demonstrated exceptionally long, fine, and often beaded
membrane processes, morphologically different from those of other MoDC
(Fig. 3, C and D). These cells comprised about
2.5 ± 1.7% of MoDC cells (n = 6, range
1.25–6.1%), and were double stained by mAb CD19 and CD20 by flow
cytometry (data not shown).
|
5G9+ cells were successfully isolated by
positive selection with mAb 5G9 and goat anti-mouse IgG magnetic
beads as described in Materials and Methods above. From five
experiments, where the average starting PBMC was 5.7 x
108 (range = 2.8–8.2 x
108), 4.3 x 106
(range = 3–5.6 x 106)
5G9+ cells were obtained after overnight culture.
Immunohistochemical staining of cytospin slides of
5G9+ cells with mAb 5G9 (Fig. 3E) or
double staining (Fig. 3F) with mAb 5G9 (red) and CD19 (B4)
(blue) immediately following isolation showed that the
5G9+ DC-like cells coexpressed CD19. Immediately
following isolation, the 5G9+ cells showed a
fraction (4.7 ± 1.9%, n = 7, range 2–7%) of
DC-like cells with long and thin processes; the remaining were small
lymphocytes (90%). After 1–4 days in culture, most small lymphocytes
died, and the percentage of DC-like cells increased to 50–80%.
Morphological changes occurred in DC-like cells, with larger cell size
and shorter processes. When human PBMC were cultured in IL-4 for 4 days
before bead isolation, the majority of isolated cells showed DC
morphology (64.5 ± 8.6%, range 54–75%, n = 4).
Fig. 3G shows the immunohistochemical staining of cytospin
slides of PBMC cultured in medium supplemented with 1000 U/ml IL-4 for
4 days. There were scattered 5G9+ DC-like cells
in the T cell clumps, suggesting that these cells may be related to T
cell proliferation. When 5G9+ cells were
solitary, they tended to have long and beaded processes. The cells also
displayed terminal branches or synaptic terminal-like structures
characteristic of the terminal portion of a neuron axon, suggesting the
possibility that 5G9+ cells are capable of
specialized cell:cell communication.
Isolation of peripheral blood DC
Based on these findings, human blood mononuclear cells were
enriched for DC by metrizamide density gradient centrifugation as
described in Materials and Methods to determine whether this
DC population expressed detectable levels of 5G9 Ag. mAb 5G9 reacted
with a fraction of metrizamide low density mononuclear cells
representing 
17 ± 4% (n = 4) of the
DC-enriched fraction of cells. This fraction was not the
lineage-negative, CD83-positive DC described previously
(10). 5G9+ cells were CD19, CD20,
CD40, HLA-Dr positive, and CD83 dim (Fig. 4).
|
Flow cytometry analysis of magnetic bead-isolated
5G9+ cells showed a population of small cells in
the low light scatter gate and a population of larger cells in the high
light scatter gate. Both populations showed the same phenotype (CD19,
CD20, CD22, CD40, HLA-Dr, CD83, CD86, IgD-positive, and either - or
-L chain positive). They were CD38 dim and did not express CD3, CD4,
CD5, CD10, CD13 CD14, CD16, CD25, CD33, CD56, or CD64. The difference
was that the cells with high light scatter had higher percentages of
positive cells and higher mean fluorescence intensity (MFI) for HLA-Dr,
CD83 and CD86, CD19, and CD40, suggesting that a higher DC composition
existed in the cells with high light scatter (Fig. 5, A and B; Table II). The expression of DC morphology and
of CD83 and CD86 on these cells of enlarged cell size may have occurred
during the separation procedure or after the binding of 5G9 to these
cells. Previous studies proved that, regardless of the culture period
(0–72 h), B cells did not express detectable levels of CD83 as
determined by flow cytometry analysis (10). Therefore, the
expression of CD83 on 5G9+ cells after isolation
by mAb 5G9 and magnetic beads (and following 24-h culture) may have
been induced by the binding of mAb 5G9 to B cells. Four color flow
cytometry analysis of mAb 5G9 and magnetic bead-purified cells showed
that these cells expressed CD83-PE, CD19-ECD, HLA-Dr-PC5, and 5G9-FITC
simultaneously (Fig. 6). Three
populations (R4, R5, and R6) were defined based on CD19 and 5G9
expression of normal PBMC control (Fig. 6B) in the same
experiment (R4: 5G92+,
CD19+; R5: 5G92+,
CD192+; R6: 5G9+,
CD192+). We can see from the four-color FACS
data, the major population of isolated cells fell in R4, a smaller
portion in R5 and fewer in R6. The separation procedure selectively
obtained higher 5G9 expressing cells.
|
|
|
).
|
).
|
The ability of 5G9+ B cells to induce the
proliferation of allogeneic T cells was compared with that of other
leukocyte populations, such as total PBMC, resting monocytes, monocytes
cultured in IL-4 and GM-CSF for 24 h, 5G9-B cells, and
5G9- blood PBMC. Because of the cell isolation procedure
for obtaining 5G9+ cells, adherent monocytes were
cultured in GM-CSF/IL-4 for 24 h. To study the APC function of
freshly isolated 5G9+ cells, monocytes cultured
for 24 h from the same donor were used in MLR as stimulator to
compare with 5G9+ cells after isolation. The
first day after isolation, 5G9+ B cells were
consistently found to be the most potent allogeneic MLR stimulators
(Fig. 9A). At a
stimulator:responder ratio of 1:20, the stimulating activity of the
5G9+ B cell population was 2.4-fold higher than
the 5G9- B cells (n = 5;
p < 0.026), 1.5-fold higher than monocytes cultured in
IL-4 and GM-CSF for 24 h (n = 6;
p < 0.01), and 4.1-, 3.5-, 3.4-fold higher than
resting monocytes, 5G9-PBMC, and total PMBC, respectively.
|
B). At a
stimulator:responder ratio of 1:20, the stimulating activity of MoDC
was 1.87-fold higher than the IL-4/CD40L-5G9+
cells (n = 3; p < 0.048), 1.79 fold
higher than IL-4/PBMC-5G9+ cells
(n = 3; p < 0.043). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- or -L chain positive. CD83 and CD86 were expressed only
after positive selection with 5G9. They did not express CD3, CD4, CD5,
CD10, CD25, CD13, CD14, CD15, CD16, CD33, CD56, or CD64. The
5G9+ cells were either or positive. This
provides the definitive proof that the cells detected are B cells and
not another cell lineage capable of binding cytophilic Abs. Functional studies showed the freshly isolated 5G9+ cells to be the most potent APCs in allogeneic MLR compared with 5G9- PBMC cells, 5G9- B cells, resting monocytes, and monocytes cultured in IL-4/GM-CSF for 24 h. However, after culture for 4 days, the APC function of 5G9+ cells was reduced significantly. Thus, in comparison with MoDC cultured for 5 days, 5G9+ cells cultured for 5 days were weak stimulators. Notably, the T cell stimulation induced by freshly isolated 5G9+ cells was similar to that of MoDC cultured for 5 days. Three possibilities can be put forward to explain the reduction of APC function of 5G9+ B-DC after culture. First, the distinct population of DC in various tissues or organs may have different functions in the immune system. It has already been suggested that the myeloid-related DC play a major role in triggering immune response, while the lymphoid-related DC participate in the regulation of these responses (11, 12). It is possible that B-DC as an immune regulator may function differentially in different circumstances. Second, DC differ in their maturation stage and are heterogeneous in turnover. The freshly isolated B-DC may be fully mature DC. Some DC have a rapid turnover rate with a t1/2 < 1 wk, such as airway epithelial DC with t1/2 < 2 days (13, 14, 15). 5G9+ B-DC may be terminally differentiated DC and rapidly lose their APC function. Third, other growth factor or factors besides IL-4 or CD40L may play important roles in maintaining the APC function of B-DC.
At the beginning of this study, 5G9+ DC-like
cells were first detected, on the cytospin slides of MoDC cultured in
medium supplemented with IL-4 and GM-CSF, by immunohistochemical
staining. We considered that these scattered 5G9+
cells with long and beaded dendrites on MoDC cytospin were DC.
Therefore, MoDC-related growth factors originally used in the culture,
such as IL-4, GM-CSF, and TNF-, were used to try to stimulate these
cells in PBMC. Growth factors, such as IL-10, IL-2, and different
combinations of these growth factors, were also compared in cultures of
PBMC in this study. Only IL-4 enhanced 5G9 expression on B cells and
gave DC morphology of these cells in PBMC culture. Thus, IL-4 was
chosen for later experiments to enhance the isolation of DC-like B
cells from PBMC. However, the rapid loss of viability of purified
5G9+ cells in culture with IL-4 suggested there
was some other growth factor, or factors, needed for the growth of
DC-like B cells. CD40L also significantly enhanced the expression of
5G9 Ag, and both IL-4 and CD40L were needed to keep
5G9+ cells in good viability in culture for up to
5 days. The morphology of 5G9+ cells in such
cultures demonstrated short dendrites resembling hairy cells. In PBMC
or MoDC culture supplemented with IL-4 or IL-4/GM-CSF,
5G9+ cells grew with good viability and expressed
long dendrites for up to 15 days. Purified 5G9+
cells could also be cultured in MoDC culture supernatant for up to 10
days. The increased B-DC in isolated 5G9+ cells
from cultured total PBMC suggests that the growth factors secreted by
monocytes, MoDC, or other cells in the culture may induce the
differentiation of DC-like B cells. We plan to screen a spectrum of
growth factors to find the best growth factor or combination of growth
factors for B-DC growth.
The percentage of 5G9+ DC-like cells in human
peripheral blood was studied by flow cytometry analysis of
5G9+ cells in the high light scatter gate and low
light scatter gate. The higher HLA-Dr, CD83, and CD86 expression of
5G9+ cells in the high light scatter gate and the
relatively larger cell size of DC-like cells shown in
immunohistochemical and Wright-Giemsa of cytospin slides of
5G9+ cells suggested that cells with DC-like
morphology were confined to the high light scatter gate. Typically,
5G9+, CD19+,
CD3-, CD14-, and
CD56- cells in the high light scatter gate
represented 0.29 ± 0.06 (n = 6) of total
mononuclear cells and about 3% of a B cell subpopulation. Therefore,
5G9+ DC-like B cells represented a trace
population of PBMCs. Total PBMC were cultured for 4 days in RPMI 1640,
10% FCS, and 1000 U/ml IL-4. Flow cytometry analysis of isolated
5G9+ cells from these cultured cells revealed the
overlapping of the percentage of CD19+ cells
(90%) and percentage of cells with DC morphology studied by
immunohistochemical and Wright-Giemsa stain of cytospin slides (64%),
thus indicating the B cell phenotype of the DC-like cells. We also
confirmed the presence of CD19 on 5G9+ cells of
DC-like morphology by double immunohistochemistry staining (Fig. 3F). The marked increase in percentage of DC-like B cells in
the 5G9+ fraction of cultured PBMC compared with
uncultured cells (64% vs 4.7%) implies the existence of both B-DC and
progenitors of B-DC in PBMC. Upon stimulation by IL-4 or growth factors
secreted by other cells during the culture period, the progenitors
developed into DC-like cells.
Compared with freshly isolated CD19+ and
CD21+ cells, only cells isolated with mAb 5G9
were double positive for CD83, further suggesting the correlation of
5G9 Ag and DC-like cells. The high percentage of cells reactive with
mouse IgG1 and goat anti-mouse IgG-FITC in the CD19-isolated
fraction (Fig. 7) implies that the CD19 mAb remained on the cell
surface after the isolation procedure. The low percentage of CD21
expression on isolated CD21+ cell fraction
suggests that, after isolation, CD21 Ag was lost from the B cell
surface. The negative reaction of isolated 5G9+
cells with goat anti-mouse IgG-FITC suggests that 5G9 Ag/Ab
complexes were shed or completely internalized after overnight culture.
The positive reaction with goat anti-mouse IgG-FITC after
incubating with freshly added mAb 5G9 suggests that 5G9 Ag may recycle
to the cell surface or that new Ag may be
synthesized.
Morphologically, 5G9+ B-DC are most similar to human tonsil follicular dendritic cells (FDC) isolated by Hart et al. (11). Both 5G9+ B-DC and FDC have long, fine, and often beaded membrane processes, which suggests the close relationship between the two cell types. The beaded structures on the processes of FDC were described by Hart et al. as "consisting of a series of interconnected immune complex coated bodies, termed ‘iccosomes.’" The function of "iccosomes" of FDC was demonstrated to mediate the delivery of Ag to germinal center B cells and macrophages (12, 13). The similar "iccosomes" structure was observed on the processes of 5G9+ B-DC, implying a common function of these cells. However, FDC were lineage negative. The lineage relationship of DCs with other hemopoietic cells and within the broader class of DC is not well understood. DC in different tissue sites and having slightly different characteristics all play a specialized role in maintaining self tolerance by the endocytosis and presentation of Ags within their environment. Recent evidence now suggests a possible lineage relationship between T cells and lymphoid DCs. Like B-DC, the discovery of T cell-derived DC also appears to conflict with the view that DCs have an exclusive common origin with myeloid cells. One possibility is that DCs mature in different tissue sites from bone marrow-derived precursors and develop region-specific characteristics that could reflect lineage differences.
Although pioneering studies of DC by Steinman and Cohn identified mouse spleen DC in 1973 and established lymphoid tissue-derived DC as potent stimulators of primary immune response (2, 14, 15, 16, 17), most current studies of DC are concentrated on myeloid lineage-derived DC. One generally recognized DC feature is its lineage negativity. A recent working definition of blood DC by Hart (18) emphasized a cell-surface Ag phenotype distinguishing it from other leukocytes, notably monocytes/macrophages and B lymphocytes . The most current separation procedures of blood DC are also based on the lineage-negative feature of DC. The finding in this study of DC-like cells with B cell surface Ag phenotype appears to conflict with current definitions of DC. However, the DC ontogeny has remained controversial since the discovery of DC. The DC-like B cells described in the current study possess consensus DC properties and characteristics. The distinctive morphology, the potent ability to stimulate allogeneic T cell proliferation, marked cell motility, and active membrane extension/retraction, and the expression of certain DC-associated Ags such as CD83 and HLA-Drhigh lead to the conclusion that B-DC are previously unrecognized members of the DC family.
Galy et al. recently found that a subpopulation of
CD34+
lin-CD45RA+CD10+
cells gives rise only to lymphoid T, B, NK, and DC. Limiting dilution
analysis demonstrates the existence of multipotential B/NK/DC
progenitor clones in the CD34high
lin-CD10+ adult bone
marrow cell population, indicating that nonprimitive progenitors for
lymphoid cells and for DC can be distinct from those of myeloid DC. The
implication is that at least one DC subset is developmentally more
closely related to lymphoid lineage than to the myeloid lineage
(3). A subset of B cells capable of assuming dendritic
morphology after stimulation with phorbol esters and attachment to a
surface was reported by Corradi and Lipsky in 1987 (19).
Serreze (4) studied the APCs that contribute to the
development and activation of diabetogenic T cells in the NOD mouse
model of IDDM. They utilized a functionally inactivated Ig µ allele
to generate a "speed congenic" stock of B cell-deficient NOD mice.
They found these mice were free of overt IDDM and were insulin
resistant, even with normal numbers of T cells. The implication is that
B cells play an unrecognized role, perhaps as potent APCs, which is
essential for initial development and/or activation of cell
autoreactive T cells in NOD mice. Cerny’s study (20)
concluded that mature B cells and/or B cell products are required for
the development of mature FDC in the mouse lymph node. A novel member
of the ubiquitin family was isolated from human DC by Bates
(26). Expression of this 777-bp mRNA was restricted to DC
and B cells, with strong expression in mature B cells. Comparative
analysis and the expression pattern of this gene suggested a function
in Ag processing and presentation. Vasef recently reported three
patients with low grade B cell lymphoproliferative disease who
developed subsequent DC tumors, and indicated the clear association
between these tumors and low grade B cell malignancies
(27). Bjorck’s recent study demonstrated that murine
CD19+ pro-B cells have the B/DC bipotential
lineage capacity (5). The above studies suggested a close
relationship between DC and B cells and the possibility that B cells
can be morphologically and functionally differentiated to DC. The
morphological and functional similarity of 5G9+
B-DC and FDC, and the fact that both 5G9+ B-DC
and FDC have beaded "iccosomes" on the long fine processes,
suggests that the study of circulating B-DC may shed new light on the
mysterious origin of FDC.
When Ag binds to the Ab molecules on the surface of a virgin or a
memory B cell, it initiates a complicated series of events, culminating
in B cell proliferation and maturation to produce either memory cells
or active Ab-producing plasma cells, apparently distinct from DC. This
classic pathway of B cell development, and the separation of
circulating DC-like cells from the blood B cell fraction by this study,
suggest that an alternative outcome may evolve during B cell
development. A fraction of B cells, which might be a fraction of memory
B cells or differentiate simultaneously with memory B cells during
primary immunization, may develop into DC. With Ag-specific Igs on
their surface and class II HLA expression, these cells may be very
efficient Ag-specific professional APCs, and function as positive
feedback cells to stimulate further T cell proliferation and subsequent
Ab-specific B cell proliferation, and thus produce a more profound
immune response. A suggested hemopoietic differentiation pathway for
DC-like B cells or B-DC is shown in Fig. 10. The exact differentiation
pathway of B-DC is unknown at present. However, the proposed pathway
suggests that B-DC might develop from B cell precursors, lymphoid DC
precursors, memory B cells, or naive B cells.
|
Almost every fraction of blood cells has been examined for biologic therapy, including T cells, NK cells, DC, monocytes/macrophages, and granulocytes. Few efforts have been conducted with B cells. As we demonstrate in this study, B cells or DC-like B cells may be the next candidates for clinical trials as APCs.
|
Footnotes |
|---|
2 Abbreviations used in this paper: DC, dendritic cells; MoDC, monocyte-derived DC; FDC, follicular DC; NOD, nonobese diabetic; IDDM, insulin-dependent diabetes mellitus; B/M, B/macrophage; CLL, chronic lymphocytic leukemia; ECD, energy-coupled dye; CD40L, CD40 ligand; HCL, hairy cell leukemia; B-DC, novel peripheral blood cell with B cell phenotype plus DC morphology and function; MFI, mean fluorescence intensity; NM, normal mouse.
Received for publication January 11, 1999. Accepted for publication May 7, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. He, X. Qiao, P. J. Klasse, A. Chiu, A. Chadburn, D. M. Knowles, J. P. Moore, and A. Cerutti HIV-1 Envelope Triggers Polyclonal Ig Class Switch Recombination through a CD40-Independent Mechanism Involving BAFF and C-Type Lectin Receptors J. Immunol., April 1, 2006; 176(7): 3931 - 3941. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ohta, E. Landis, T. Boulay, R. B. Phillips, B. Collet, C. J. Secombes, M. F. Flajnik, and J. D. Hansen Homologs of CD83 from Elasmobranch and Teleost Fish J. Immunol., October 1, 2004; 173(7): 4553 - 4560. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Burke, A. J. Stagg, P. A. Bedford, N. English, and S. C. Knight IL-10-Producing B220+CD11c- APC in Mouse Spleen J. Immunol., August 15, 2004; 173(4): 2362 - 2372. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Mohty, D. Isnardon, A. Charbonnier, M. Lafage-Pochitaloff, M. Merlin, D. Sainty, D. Olive, and B. Gaugler Generation of potent Th1 responses from patients with lymphoid malignancies after differentiation of B lymphocytes into dendritic-like cells Int. Immunol., July 1, 2002; 14(7): 741 - 750. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Lu, C. A. Bonham, X. Liang, Z. Chen, W. Li, L. Wang, S. C. Watkins, M. A. Nalesnik, M. S. Schlissel, A. J. Demestris, et al. Liver-Derived DEC205+B220+CD19- Dendritic Cells Regulate T Cell Responses J. Immunol., June 15, 2001; 166(12): 7042 - 7052. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
