G-CSF)-mobilized Peripheral Blood Mononuclear Cells Research Articles

Pharmacological Immunosuppression Reduces But Does Not Eliminate the Need for Total-Body Irradiation in Nonmyeloablative Conditioning Regimens for Hematopoietic Cell Transplantation

In the dog leukocyte antigen (DLA)-identical hematopoietic cell transplantation (HCT) model, stable marrow engraftment can be achieved with total-body irradiation (TBI) of 200 cGy when used in combination with postgrafting immunosuppression. The TBI dose can be reduced to 100 cGy without compromising engraftment rates if granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells (G-PBMC) are infused with the marrow. T cell-depleting the G-PBMC product abrogates this effect. These results were interpreted to suggest that the additional T cells provided with G-PBMC facilitated engraftment by overcoming host resistance. We therefore hypothesized that the TBI dose may be further reduced to 50 cGy by augmenting immunosupression either by (1) tolerizing or killing recipient T cells, or (2) enhancing the graft-versus-host (GVH) activity of donor T cells. To test the first hypothesis, recipient T cells were activated before HCT by repetitive donor-specific PBMC infusions followed by administration of methotrexate (MTX) (n = 5), CTLA4-Ig (n = 4), denileukin diftitox (Ontak; n = 4), CTLA4-Ig + MTX (n = 8), or 5c8 antibody (anti-CD154) + MTX (n = 3). To test the second hypothesis, recipient dendritic cells were expanded in vivo by infusion of Flt3 ligand given either pre-HCT (n = 4) or pre- and post-HCT (n = 5) to augment GVH reactions. Although all dogs showed initial allogeneic engraftment, sustained engraftment was seen in only 6of 42 dogs (14% of all dogs treated in 9 experimental groups). Hence, unless more innovative pharmacotherapy can be developed that more forcefully shifts the immunologic balance in favor of the donor, noncytotoxic immunosuppressive drug therapy as the sole component of HCT preparative regimens may not suffice to ensure sustained engraftment.

Common Myeloid Progenitor Content in Allogeneic Graft Products Correlates with Time to Platelet Engraftment.

Abstract 3215Poster Board III-152In clinical hematopoietic cell transplantation (HCT), CD34 is used as a surrogate marker to determine the content of hematopoietic stem and progenitor cells in the graft. Increasing numbers of CD34+ cells in the graft were associated with significantly accelerated neutrophil and platelet engraftment after myeloablative allogeneic HCT with granulocyte colony-stimulating factor (G-CSF) mobilized peripheral blood mononuclear cells (G-PBMC). However, the CD34+ population is very heterogeneous and contains several subpopulations, such as hematopoietic stem cells (HSC), common myeloid progenitors (CMP), granulocyte- macrophage progenitors (GMP), megakaryocyte erythroid progenitors (MEP), common lymphoid progenitors (CLP), T/ NK cell progenitors, and pro B cells. So far, these populations have not been quantified and functionally correlated with time to engraftment in patients after allogeneic HCT. This study was designed to (1) assess the distribution of the CD34+ subpopulations in G-PBMC, bone marrow (BM) and cord blood (CB) products and (2) to correlate the content of CD34+ subpopulations with time to engraftment in patients after myeloblative allogeneic HCT. The following populations within the Lin-CD34+ fraction were quantified by flow cytometry: CD38- HSC, CD38+IL3RaloCD45RA- CMP, CD38+IL3RaloCD45RA+ GMP, CD38+IL3Ra-CD45RA- MEP, CD10+CD19- CLP, CD10+CD19+ Pro B cells and CD10-CD7+CD45RA+ T/ NK cell progenitors. First, the relative distribution of these populations in % was determined in G-PBMC, BM and CB. In G-PBMC the percentage of CMP was highest whereas HSC, GMP and T/ NK progenitor percentage was lowest, when compared to BM or CB (Table 1). MEP and CLP percentages were not significantly different. Second, the functional correlation of absolute progenitor counts per kilogram (kg) recipient body weight with regards to neutrophil and platelet engraftment was analyzed in 24 patients with myeloablative allogeneic HCT with G-PBMC (n=21) or BM (n=3) from matched related (n=15) or unrelated donors (n=9). The underlying diseases were acute leukemia in 14, myelodysplastic syndrome in 2, chronic myeloid leukemia in 3, lymphoproliferative disease in 3, severe aplastic anemia in 1 or non-malignant disease in 1. Median age was 51 years (range 12- 65). Conditioning regimens were cyclophosphamide (Cy)/ busulfan in 15, Cy/ total body irradiation (TBI) in 3, BEAM/ fludarabine/ TBI in 3, other in 3. A Cox model was used to determine the correlation between progenitor counts and time to engraftment of neutrophils >0.5×10e9/L, platelets >20×10e9/L (Plt20) or >50×10e9/L (Plt50). In this cohort, only graft source had a significant impact on neutrophil recovery (p=0.01). However, platelet engraftment also correlated with the total CD34+ cell count/ kg (Plt20 p=0.018, Plt50 p=0.020). This correlation was attributed to the content of CMP/ kg contained in the graft (Plt20 p=0.006, Plt50 p=0.005). In conclusion: (1) G-PBMC, BM and CB not only differ in absolute CD34+ cell numbers but also in the relative distribution of CD34+ subpopulations showing that clinical hematopoietic grafts are more complex than simple CD34 counts. (2) In this patient cohort, accelerated platelet engraftment after myeloablative allogeneic HCT is correlated with the content of CMPs. These findings underline that there is a need for better understanding of hematopoietic graft compositions as this could have further implications for graft manipulation in the future.Table 1:Distribution of Lin-CD34+ populations in % in G-PBMC, BM, or CB. Comparisons by Mann Whitney test, NS= not significant.Populations in % of Lin-CD34+ cellsG-PBMC N=30BM N=6CB N=7p value, (G-PBMC vs BM)p value, (G-PBMC vs CB)p value, (BMvsCB)HSCMedian1.764.156.570.003<0.001NS95% CI (low/high)1.54/2.962.51/6.474.51/14.65CMPMedian73.436.63.19<0.001<0.0010.00195% CI (low/high)66.12/76.3522.71/42.881.06/10.73GMPMedian15.8021.5572.400.027<0.0010.00195% CI (low/high)13.52/22.7617.73/27.2761.89/77.02MEPMedian0.210.910.12NSNSNS95% CI (low/high)0.20/0.47-0.15/2.45-1.17/4.14CLPMedian0.760.421.09NSNSNS95% CI (low/high)0.58/1.050.11/1.070.33/3.43Pro BMedian0.4928.253.19<0.001<0.0010.00195% CI (low/high)0.37/0.7315.51/42.092.13/5.01T NK progenitorMedian5.107.2316.90NS<0.0010.00195% CI (low/high)4.66/7.205.08/9.3112.35/21.22 DisclosuresNo relevant conflicts of interest to declare.

Administrations of Peripheral Blood CD34-Positive Cells Contribute to Medial Collateral Ligament Healing via Vasculogenesis

Neoangiogenesis is a key process in the initial phase of ligament healing. Adult human circulating CD34+ cells, an endothelial/hematopoietic progenitor-enriched cell population, have been reported to contribute to neoangiogenesis; however, the therapeutic potential of CD34+ cells for ligament healing is still unclear. Therefore, we performed a series of experiments to test our hypothesis that ligament healing is supported by CD34+ cells via vasculogenesis. Granulocyte colony-stimulating factor-mobilized peripheral blood (GM-PB) CD34+ cells with atelocollagen (CD34+ group), GM-PB mononuclear cells (MNCs) with atelocollagen (MNC group), or atelocollagen alone (control group) was locally transplanted after the creation of medial collateral ligament injury in immunodeficient rats. Reverse transcriptase-polymerase chain reaction (RT-PCR) and immunohistochemical staining at the injury site demonstrated that molecular and histological expression of human-specific markers for endothelial cells was higher in the CD34+ group compared with the other groups at week 1. Endogenous effect, assessed by capillary density and mRNA expression of vascular endothelial growth factor, was significantly higher in CD34+ cell group than the other groups. In addition to the observation that, as assessed by real-time RT-PCR, gene expression of ligament-specific marker was significantly higher in the CD34+ group than in the other groups, ligament healing assessed by macroscopic, histological, and biomechanical examination was significantly enhanced by CD34+ cell transplantation compared with the other groups. Our data strongly suggest that local transplantation of circulating human CD34+ cells may augment the ligament healing process by promoting a favorable environment through neovascularization.

Duration of Immunosuppressive Therapy for Chronic Graft-vs.-Host Disease (cGVHD) Following Non-Myeloablative Allogeneic Hematopoietic Cell Transplantation (HCT).

BACKGROUND: Chronic GVHD is a major determinant of quality of life after allogeneic HCT. Systemic immunosuppressants (IS) are the mainstay of treatment for cGVHD. Patients (pts) with cGVHD after myeloablative HCT are able to discontinue IS at a median of 23 months after HCT (Stewart et al., Blood 2004 Dec 1;104[12]:3501–6). We examined the duration of treatment with IS in patients with cGVHD after non-myeloablative HCT.METHODS: All pts who underwent non-myeloablative HCT at our institutions through 31 Dec 2003 were assessed for inclusion. Two hundred sixty-nine pts were alive and free of their underlying disease at day +84 after HCT; 138 of these (51%) began systemic IS for cGVHD, at a median of 112 days after HCT. We analyzed these 138 pts with cGVHD to determine duration of treatment with IS. Diagnoses are listed in Table 1. Sixty-eight pts received related allografts and 70 received unrelated allografts; median age at HCT was 53.3 years. The allograft source was granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cells (G-PBMC) in 137 pts and bone marrow in 1 pt. Relapsed malignancy and death were considered competing risk events in the estimation of cumulative incidence of discontinuation of IS; follow-up was censored at the time of death, relapsed malignancy, or discontinuation of all systemic IS.RESULTS: Thirty-seven pts (27%) had recurrent malignancy and 23/138 (16%) died without recurrent malignancy while on IS; 41/138 (30%) discontinued all systemic IS after resolution of cGVHD; and 37/138 (27%) were alive without malignancy at last follow-up (median 3.7 years after cGVHD diagnosis), but remained on IS for cGVHD. Five-year cumulative incidence estimates by outcome are listed in Table 2. Of pts alive and free of malignancy at 5 years after cGVHD diagnosis, 58% had discontinued IS. In pts who discontinued IS, the median time to discontinuation was 2.0 (range, 0.3 – 5.7) years (see Figure).CONCLUSION: The median time to resolution of cGVHD and discontinuation of IS in this group is comparable to that reported after myeloablative HCT (24 months vs. 23 months), despite the older age of these pts (median age 53.3 years, vs. 39.5 years in myeloablative cohort) and the prevalence of G-PBMC allografts (>99%, vs. 20% in myeloablative cohort). In this older population with cGVHD after non-myeloablative allogeneic HCT, the majority of pts who remain alive and free of relapsed malignancy have resolution of cGVHD over time.Table 1Underlying diagnoses among pts with cGVHDDiagnosisNo. of pts (%)AML19 (14%)ALL6 (4%)CML6 (4%)CLL11 (8%)MDS14 (10%)HD9 (6%)NHL31 (22%)MM32 (23%)MPD1 (<1%)Waldenstrom's1 (<1%)Renal cell carcinoma1 (<1%)Wilms' tumor1 (<1%)Immunodeficiency syndrome6 (4%)Total138Table 2Five-year cumulative incidencesOutcome5-year cumulative incidence % (95% CI)Relapsed malignancy/death45 (36–54)Alive, malignancy-free, on IS23 (15–33)Alive, malignancy-free, discontinued IS32 (23–41) [Display omitted]

23 INVITED T regs in cancer

In the dog leukocyte antigen (DLA)-identical hematopoietic cell transplantation (HCT) model, stable marrow engraftment can be achieved with total-body irradiation (TBI) of 200 cGy when used in combination with postgrafting immunosuppression. The TBI dose can be reduced to 100 cGy without compromising engraftment rates if granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells (G-PBMC) are infused with the marrow. T cell-depleting the G-PBMC product abrogates this effect. These results were interpreted to suggest that the additional T cells provided with G-PBMC facilitated engraftment by overcoming host resistance. We therefore hypothesized that the TBI dose may be further reduced to 50 cGy by augmenting immunosupression either by (1) tolerizing or killing recipient T cells, or (2) enhancing the graft-versus-host (GVH) activity of donor T cells. To test the first hypothesis, recipient T cells were activated before HCT by repetitive donor-specific PBMC infusions followed by administration of methotrexate (MTX) (n = 5), CTLA4-Ig (n = 4), denileukin diftitox (Ontak; n = 4), CTLA4-Ig + MTX (n = 8), or 5c8 antibody (anti-CD154) + MTX (n = 3). To test the second hypothesis, recipient dendritic cells were expanded in vivo by infusion of Flt3 ligand given either pre-HCT (n = 4) or pre- and post-HCT (n = 5) to augment GVH reactions. Although all dogs showed initial allogeneic engraftment, sustained engraftment was seen in only 6 of 42 dogs (14% of all dogs treated in 9 experimental groups). Hence, unless more innovative pharmacotherapy can be developed that more forcefully shifts the immunologic balance in favor of the donor, noncytotoxic immunosuppressive drug therapy as the sole component of HCT preparative regimens may not suffice to ensure sustained engraftment.

The Immunosuppressive Agent FK506 Enhances the Cytolytic Activity of Inhibitory Natural Killer Cell Receptor (CD94/NKG2A)-Expressing CD8 T Cells

Tacrolimus (FK506) is a potent immunosuppressive agent that inhibit transcription of cytokines such as IL-2 in T cells. The C-type lectin superfamily inhibitory NKR CD94/NKG2A-expressing cells can monitor the global status of HLA class I on tumor and leukemic cells through the recognition of HLA-E and induce cytolytic attack without an inhibitory signal against HLA class I-decreased target cells. We found that there was no effect of FK506 on the expansion of inhibitory NKR CD94/NKG2A-expressing T cells from granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells (G-PBMCs). However, cytolytic activity levels of purified CD94-expressing cells from 7-day cultures with FK506 were much higher than those from 7-day cultures without FK506. These data suggested that FK506 did not inhibit cytolytic activities of inhibitory NKR-expressing T cells and that there was a possibility of cytolytic activities being enhanced through the induction of cytotoxic molecules such as NKG2D and granzyme during a seven-day culture with FK506.

High doses of transplanted CD34+ cells are associated with rapid T-cell engraftment and lessened risk of graft rejection, but not more graft-versus-host disease after nonmyeloablative conditioning and unrelated hematopoietic cell transplantation

This report examines the impact of graft composition on outcomes in 130 patients with hematological malignancies given unrelated donor granulocyte-colony-stimulating-factor-mobilized peripheral blood mononuclear cells (G-PBMC) (n = 116) or marrow (n = 14) transplantation after nonmyeloablative conditioning with 90 mg/m(2) fludarabine and 2 Gy TBI. The median number of CD34(+) cells transplanted was 6.5 x 10(6)/kg. Higher numbers of grafted CD14(+) (P = 0.0008), CD3(+) (P = 0.0007), CD4(+) (P = 0.001), CD8(+) (P = 0.004), CD3(-)CD56(+) (P = 0.003), and CD34(+) (P = 0.0001) cells were associated with higher levels of day 28 donor T-cell chimerism. Higher numbers of CD14(+) (P = 0.01) and CD34(+) (P = 0.0003) cells were associated with rapid achievement of complete donor T-cell chimerism, while high numbers of CD8(+) (P = 0.005) and CD34(+) (P = 0.01) cells were associated with low probabilities of graft rejection. When analyses were restricted to G-PBMC recipients, higher numbers of grafted CD34(+) cells were associated with higher levels of day 28 donor T-cell chimerism (P = 0.01), rapid achievement of complete donor T-cell chimerism (P = 0.02), and a trend for lower risk for graft rejection (P = 0.14). There were no associations between any cell subsets and acute or chronic GVHD nor relapse/progression. These data suggest more rapid engraftment of donor T cells and reduced rejection rates could be achieved by increasing the doses of CD34(+) cells in unrelated grafts administered after nonmyeloablative conditioning.

Cytolytic activity and regulatory functions of inhibitory NK cell receptor–expressing T cells expanded from granulocyte colony-stimulating factor–mobilized peripheral blood mononuclear cells

AbstractInhibitory natural killer cell receptor (NKR)–expressing cells may induce a graft-versus-leukemia/tumor (GVL/T) effect against leukemic cells and tumor cells that have mismatched or decreased expression of HLA class I molecules and may not cause graft-versus-host disease (GVHD) against host cells that have normal expression of HLA class I molecules. In our study, we were able to expand inhibitory NKR (CD94/NKG2A)–expressing CD8+ T cells from granulocyte colonystimulating factor (G-CSF)–mobilized peripheral blood mononuclear cells (G-PBMCs) by more than 500-fold using stimulation by an anti-CD3 monoclonal antibody with interleukin 15 (IL-15). These expanded and purified CD94-expressing cells attacked various malignant cell lines, including solid cancer cell lines, as well as the patients' leukemic cells but not autologous and allogeneic phytohemagglutinin (PHA) blasts in vitro. Also, these CD94-expressing cells prevented the growth of K562 leukemic cells and CW2 colon cancer cells in NOD/SCID mice in vivo. On the other hand, the CD94-expressing cells have low responsiveness to alloantigen in mixed lymphocyte culture (MLC) and have high transforming growth factor (TGF)–β1– but low IL-2– producing capacity. Therefore, CD94-expressing cells with cytolytic activity against the recipient's leukemic and tumor cells without enhancement of alloresponse might be able to be expanded from donor G-PBMCs.

Induction of CD94/NKG2A expression on T cells in mixed lymphocyte culture by CD14+ cells from granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cells.

We report the increased expression of inhibitory natural killer (NK) cell receptor, CD94/NKG2A, for human leucocyte antigen (HLA) class I on CD8+ T cells in granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cells (G-PBMC) after mixed lymphocyte culture (MLC). The addition of purified CD14+ cells to CD14-depleted G-PBMC-induced CD94/NKG2A expression in a dose-dependent fashion; however, this enhancing effect was inhibited by the membrane between responder cells and CD14+ cells. Therefore, CD14+ cells in G-PBMC induce CD94/NKG2A expression on CD8+ T cells, which in turn appear to downregulate alloresponses after allogeneic peripheral blood stem cell transplantation.

Effects of extending the duration of postgrafting immunosuppression and substituting granulocyte-colony-stimulating factor-mobilized peripheral blood mononuclear cells for marrow in allogeneic engraftment in a nonmyeloablative canine transplantation model.

Stable mixed donor/host hematopoietic chimerism was uniformly achieved in dogs given 200 cGy total body irradiation (TBI) before and immunosuppression with mycophenolate mofetil (MMF) for 28 days and cyclosporine (CSP) for 35 days after transplantation of marrow from dog leukocyte antigen-identical littermates. When the TBI dose was lowered to 100 cGy, donor marrow engraftment in 6 dogs was only transient, lasting 3 to 12 weeks. In this study, we asked whether stable engraftment in this model could be achieved: (1) by substituting recombinant canine granulocyte-colony-stimulating factor-mobilized peripheral blood mononuclear cells (G-PBMCs) for marrow and (2) by extending CSP administration from 35 to 100 days. Eighteen dogs were given G-PBMC grafts and MMF for 28 days. Eight of the 18 dogs received CSP for 35 days and 10 for 100 days. We found that substituting G-PBMCs for marrow did not increase the incidence of stable allogeneic engraftment (P = .11). However, increasing the duration of posttransplantation immunosuppression with CSP from 35 to 100 days favorably influenced stable donor engraftment (P = .06).

Production of Interleukin-10 by Granulocyte Colony-Stimulating Factor–Mobilized Blood Products: A Mechanism for Monocyte-Mediated Suppression of T-Cell Proliferation

Previous reports showed that granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells (G-PBMC) are hyporesponsive to alloantigen compared with control PBMC. In the current study, neutralizing antibodies to interleukin-10 (IL-10) increased the proliferative response of G-PBMC to alloantigen by 50. 14% (+/- 12.79%; n = 8), whereas the proliferative response of control PBMC was not affected. The inhibition of OKT3-stimulated CD4 cell proliferation by G-PBMC-derived CD14(+) cells could also be abrogated by the addition of IL-10 neutralizing antibodies. Further, IL-10 levels correlated with the number of CD14 cells in these cultures. Constitutive IL-10 mRNA levels detected by quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) were 10-fold higher in G-PBMC compared with control PBMC. This translated into significantly higher IL-10 levels after 24-hour lipopolysaccharide (LPS) stimulation of G-PBMC compared with control PBMC (P = .036). IL-10 mRNA levels were also fivefold higher in isolated G-PBMC-derived CD14 cells compared with control CD14 cells. This corresponded to increased constitutive production of IL-10 by isolated G-PBMC-derived CD14 cells compared with control CD14 cells (357.2 +/- 104.5 v 51.7 +/- 30.5, P = .051). In conclusion, these data suggest that monocytes contained within G-PBMC, which, in comparison to marrow, are increased in absolute number and relative proportion to T cells, may suppress T-cell responsiveness by secretion of IL-10.

Impaired Induction of the CD28-Responsive Complex in Granulocyte Colony-Stimulating Factor Mobilized CD4 T Cells

AbstractUse of the CD28/B7 costimulatory signal for T-cell activation was analyzed in granulocyte colony-stimulating factor (G-CSF) mobilized peripheral blood mononuclear cells (G-PBMCs) and in peripheral blood mononuclear cells obtained before administration of G-CSF (preG-PBMCs). CTLA4Ig inhibition of OKT3-stimulated proliferation was significantly lower in G-PBMCs compared with preG-PBMCs (39.9% ± 5.6% and 72.2% ± 5.4%, respectively; P < .001). Furthermore, as shown in electrophoretic mobility-shift assays, the inducible level of the T-cell transcription factor CD28 responsive complex (CD28RC) was suppressed in CD4 cells derived from G-PBMC. However, depletion of CD14 cells from G-PBMCs restored CD28RC induction to normal levels. Taken together, these findings suggest that the large number of CD14 monocytes in G-PBMCs may limit T-cell responsiveness by suppressing the induction of the CD28RC.

Impaired Induction of the CD28-Responsive Complex in Granulocyte Colony-Stimulating Factor Mobilized CD4 T Cells

Use of the CD28/B7 costimulatory signal for T-cell activation was analyzed in granulocyte colony-stimulating factor (G-CSF) mobilized peripheral blood mononuclear cells (G-PBMCs) and in peripheral blood mononuclear cells obtained before administration of G-CSF (preG-PBMCs). CTLA4Ig inhibition of OKT3-stimulated proliferation was significantly lower in G-PBMCs compared with preG-PBMCs (39.9% ± 5.6% and 72.2% ± 5.4%, respectively; P &lt; .001). Furthermore, as shown in electrophoretic mobility-shift assays, the inducible level of the T-cell transcription factor CD28 responsive complex (CD28RC) was suppressed in CD4 cells derived from G-PBMC. However, depletion of CD14 cells from G-PBMCs restored CD28RC induction to normal levels. Taken together, these findings suggest that the large number of CD14 monocytes in G-PBMCs may limit T-cell responsiveness by suppressing the induction of the CD28RC.

Normal Human Peripheral Blood Mononuclear Cells Mobilized With Granulocyte Colony-Stimulating Factor Have Increased Osteoclastogenic Potential Compared To Nonmobilized Blood

Single-cell suspensions of granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells (G-PBMC) cultured in alpha minimal essential medium (alphaMEM) containing 10% fetal bovine serum formed multicellular aggregates within 24 hours. In six separate experiments, formation of aggregates appeared to be dependent on cell density per surface area, so that 5.8 +/- 1.3 aggregates formed per 1 x 10(5) cells when G-PBMC were cultured at densities greater than or equal to 1 x 10(5) cells/cm2. The frequency of aggregate formation was less than 1 per 10(5) cells when G-PBMC were cultured at densities less than 1 x 10(5) cells/cm2. Once formed, aggregates became adherent within 72 hours, and then, over the course of 21 days, released CD3/CD4/CD25-positive cells into the supernatant. This T-cell production peaked between days 7 and 14, reaching a total of 1,269 +/- 125.9 cells released per aggregate by day 21. Between days 14 and 21, the aggregates also generated macroscopic clusters of adherent mononuclear and giant multinucleated cells that stained positive for tartrate-resistant acid phosphatase (TRAP). At 4 weeks, the macroscopic foci coalesced into monolayers. Multinucleated TRAP-positive cells were distinguished from macrophage polykaryons by the absence of CD14 expression and the presence of osteoclast-specific membrane receptors for calcitonin and alphavbeta3-vitronectin. The osteoclast nature of these cells was further demonstrated by their ability to form resorption lacunae on dentine slices. Comparable osteoclast formation was not detected in cultures of normal marrow or normal nonmobilized peripheral blood.

CD14+ Cells in Granulocyte Colony-Stimulating Factor (G-CSF)-Mobilized Peripheral Blood Mononuclear Cells Induce Secretion of Interleukin-6 and G-CSF by Marrow Stroma

The ability of granulocyte colony-stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells (G-PBMCs) to induce secretion of cytokines in primary long-term marrow cultures (LTC) or in the human marrow stromal cell line HS23 was compared with that of marrow mononuclear cells. Equal numbers of G-PBMCs or marrow mononuclear cells were added to stromal cultures, supernatants were harvested at day 4 and levels of interleukin-1 alpha (IL-1 alpha), IL-1 beta, IL-2, IL-6, G-CSF, and tumor necrosis factor alpha (TNF alpha) were determined. G-PBMCs induced 21.4-fold higher levels of IL-6 and 12.5-fold higher levels of G-CSF in LTC cocultures compared with marrow mononuclear cells and induced 20.6-fold more IL-6 and 6.3-fold more G-CSF when added to HS23 cells. Experiments using sorted populations of CD20+, CD3+, and CD14+ cells showed that CD14+ cells within G-PBMCs were responsible for triggering the production of IL-6 and G-CSF. The effect did not require cell-cell contact and was inhibited when neutralizing antibodies to IL-1 alpha and IL-1 beta were used in combination. In these experiments, the greater stimulating ability of G-PBMCs is most likely attributable to the greater number of CD14+ cells in G-PBMCs (26.1+% +/- 2.3%) compared with marrow (2.5% +/- 0.8%), because equal numbers of CD14+ cells sorted from marrow and G-PBMCs showed comparable ability to induce IL-6 and G-CSF when placed directly on stromal cells.