sign in sign up
<<
Home , Hematopoietic stem cell

Marrow Transplantation, (1999) 23, 63–69  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Comparison of -dependent inhibitory activity in GM- CSF vs G-CSF mobilized PSC products

AG Ageitos1, ML Varney1, PJ Bierman2, JM Vose2, PI Warkentin1 and JE Talmadge1

1Department of Pathology and Microbiology, and 2Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE, USA

Summary: following transplant compared to normal individuals. PSCT patients, however, have a more rapid immune recovery than This study compares the immune properties of periph- do BMT patients.5–7,10 eral (PSC) products mobilized with dif- Stem cells are mobilized from the (BM) ferent hematopoietic growth factors (HGFs) as well as into the peripheral blood (PB) after myelosuppressive apheresis products and peripheral blood leukocytes chemotherapy,12 chemotherapy combined with HGFs13–17 (PBL) from normal individuals. We found that mono- or during administration of HGFs.18–24 Accelerated hemato- cytes in mobilized PSC products appear to inhibit T cell poietic recovery is observed following PSCT as compared function independent of whether colony- to transplantation with steady-state BM and is only stimulating factor (G-CSF) or granulocyte– observed when peripheral blood stem cell (PSC) products colony-stimulating factor (GM-CSF) was used for mobi- are collected after mobilization.21,24,25 The choice of HGF lization. In addition, the GF used to mobilize the stem used for mobilization may also be important as granulo- cell product may be less important to the CD4:CD8 cyte–macrophage colony-stimulating factor (GM-CSF) ratio than the extent of prior chemotherapy, as we interacts with earlier progenitors than does granulocyte found an inverse correlation between chemotherapy and colony-stimulating factor (G-CSF)26 and increases the fre- the CD4:CD8 ratio. In other observations, all apheresis quency of , , , mega- products, whether mobilized or unmobilized, contained karyocytes, erythroid precursors, and dendritic cells from significantly more compared to normal PBL. hematopoietic progenitors.27 In addition, the administration The mononuclear cells (MNC) from G-CSF or GM-CSF of GM-CSF28 or G-CSF29 after autologous BMT or PSCT mobilized PSC products had a similar T cell phytohem- also can enhance the endogenous cytotoxic and antimicro- agglutinin (PHA) mitogenic response that was signifi- bial30 activity of immune cells. cantly lower (P = 0.001 and P = 0.005, respectively) than G-CSF is used most frequently for mobilization and has non-mobilized apheresis products. We also examined been shown to enhance the production and maturation of neu- the T cell inhibitor (TI) activity of the MNC from the trophils in vitro and in vivo.27,31,32 G-CSF is unique among PSC products for allogeneic proliferation the regulators of in that it not only stimulates and found that PSC products significantly reduced the the proliferation, but also induces the terminal maturation of proliferation of allogeneic PBL to PHA. A significant myeloid progenitor cells.33 Like other , G-CSF correlation (P = 0.001, r = 0.517) between the frequency exerts a pleiotropic effect on different cell populations. The of monocytes and TI activity also was observed. G-CSF receptor can be found on myeloid progenitor cells, Keywords: monocytes; stem cell mobilization; T cells; mature , , monocytes, and some T and B GM-CSF; G-CSF lymphoid cell lines.33 In addition, the PSC pool can be expanded by both G-CSF and GM-CSF.34–36 While it is well documented that the injection of HGFs mobilizes stem cells with different hematopoietic properties dependent on the The therapeutic efficacy of hematopoietic growth factor HGF, no studies to date have compared the immunologic (HGF) administration has been shown in a variety of dis- characteristics of PSC products mobilized with different 1,2 eases. One such utility includes rapid myelorestoration HGFs. following high-dose chemotherapy (HDT) and stem cell Analysis of G-CSF mobilized PSC products has shown transplantation. Although peripheral blood stem cell trans- that they contain at least three times more T cells than plantation (PSCT) has largely replaced bone marrow trans- steady-state BM products. However, the frequency of natu- plantation (BMT) due to a lower incidence of tumor-cell ral killer (NK) and T cells in these PSC products is in the 3 4–6 contamination and more rapid hematopoietic recovery, range expected for normal PB.37 We and others have 5–11 both BMT and PSCT patients are immunosuppressed reported9,38 that GM-CSF mobilized PSC products have a high frequency of monocytes. Furthermore, the phenotypic analysis of mobilized PSC products revealed normal levels Correspondence: Dr JE Talmadge, Department of Pathology and Micro- + + biology, University of Nebraska Medical Center, 600 South 42nd Street, of CD4 cells and an increased frequency of CD8 cells, Omaha, NE 68198-5660, USA which results in a decreased CD4:CD8 cell ratio as com- Received 29 May 1998; accepted 11 August 1998 pared to the PB of normal individuals. Monocytes in PSC products AG Ageitos et al 64 While the effects of GM-CSF and G-CSF on myeloid Isolated mononuclear cells (MNC) cells have been studied, the effect of HGF mobilization on lymphoid cells is not well understood. GM-CSF has been MNC were obtained from apheresis products and the hep- reported to amplify (IL)-2 supporting T cell pro- arinized venous blood of 49 normal donors by Ficoll– liferation,39 and in patients with non-Hodgkin’s lymphoma Hypaque gradient centrifugation. The interphase cells were (NHL), there is an increase in the absolute lymphocyte harvested, washed three times in Hanks’ balanced salt sol- count, particularly activated after GM-CSF ution (HBSS; Life Technologies; Gaithersburg, MD, USA), 40 41 and resuspended in RPMI-1640 medium supplemented with administration. Other reports have examined the ability ␮ of GM-CSF to induce T cell proliferation and found that 1mm HEPES, glutamine, 40 g/ml gentamycin (Life it upregulated the expression of IL-2 receptors on T lym- Technologies), and 10% fetal bovine serum (FBS). phocytes. We report here a comparison of the effect of G- CSF and GM-CSF mobilization on lymphocyte and mono- Proliferation cyte function and phenotype in PSC products. Phytohemagglutinin (PHA)-induced T cell proliferation was measured by determining 3H-thymidine incorporation into DNA. Cells (100 ␮l/well) were plated in quadruplicate Materials and methods in flat-bottomed wells of microtiter plates at 1 × 106 cells/ml. The cells were cultured in PHA (100 ␮l/well) at Patients and control group a final concentration of 0.5 ␮g/ml of media. On day 3, the ␮ 3 A total of 54 apheresis products, including HGF mobilized wells were pulsed for 18 h with 1 Ci of H-thymidine leukapheresis products from 36 patients and 18 non-mobil- (Amersham Life Sciences; Arlington Heights, IL, USA) ized apheresis products from healthy donors, were exam- and the cells harvested on day 4 using an automatic multi- ined. The samples were divided into three groups, accord- well harvester (Packard Instruments, Downers Grove, IL, ing to whether HGF was used for mobilization and whether USA). The filters were allowed to air dry, scintillation G-CSF or GM-CSF was used for mobilization. PSC apher- cocktail was added, and the samples were counted in a esis products were collected from 21 patients following Packard multi-well beta-plate counter. In all studies, a con- GM-CSF and from 15 patients following G-CSF mobiliz- trol of normal peripheral blood leukocytes (PBL) was ation. NHL patients who were eligible for PSCT at the Uni- included and the data are reported as the percent of the versity of Nebraska Medical Center (UNMC) were entered control sample. into these studies. The samples were obtained using proto- cols approved by the Institutional Review Board of UNMC and following written informed consent. The clinical characteristics of both groups are shown in Table 1. Using Flow cytometric analyses were used for the phenotyping the same protocol as for the patients, non-mobilized, studies. The cell populations were studied prior to Ficoll– apheresis products were obtained from normal donors and Hypaque separation and erythrocytes removed by lysis. Ali- served as a control group for the mobilized PSC apheresis. quots of 5 × 104 cells were stained with biotin-conjugated monoclonal antibodies (MoAb) to CD8 and CD14, a fluor- escein isothiocyanate (FITC)-labeled MoAb to CD3, and a phycoerythrin (PE)-labeled MoAb to CD4. After incubation Table 1 Characteristics of study patients and washing, streptavidin allophycocyanin (APC) was added as the third fluorochrome. The data were acquired Clinical features G-CSF GM-CSF on a FACStarPlus using a 40 mw, 488 nm argon laser and mobilized mobilized a 100 mW, 647 nm HeNe laser for excitation. Detailed data analysis was performed using Paint-A-Gate Plus software Number 15 21 Sex from Becton Dickinson (San Jose, CA, USA). Male 7 9 Female 8 12 Age (years) T cell inhibition assay Median 48 45 Range 27–64 14–63 The coculture assay used to measure the cellular inhibition Prior therapy of T cell function has been described previously.9,38,42,43 Chemotherapy 15 21 Briefly, Ficoll–Hypaque purified, normal PBL (1 × 105)as Radiation 6 responder cells were cocultured with varying numbers of Months of chemotherapy median 2 9a irradiated (500 cGy) putative inhibitor cells (MNC from PB range 1–5 2–25 or PSC product) starting at an inhibitor to responder ratio Tumor type (I:R) of 4:1. Four, two-fold dilutions of inhibitor cells (50 Non-Hodgkin’s lymphoma 15 16 ␮l) were made in 96-well flat bottom plates. The responder HD 1 cells were then added (50 ␮l) together with the PHA (100 Breast cancer 2 ␮ ␮ Sarcoma Ewing 1 l) to achieve 0.5 g/ml. ° Cells were cultured at 37 C (5% CO2) for 3 days in aSignificant difference with prior cycles number of chemotherapy in G- RPMI-1640 medium plus 10% FBS and the mitogenic CSF mobilized cancer patients (P = 0.0001). response by the responder population assayed by pulsing Monocytes in PSC products AG Ageitos et al 65 with 1 ␮Ci/well 3H-thymidine over the final 18 h of culture. pared to normal PB (4.1 Ϯ 0.7%), suggesting that apheresis Cells were harvested with a multi-well harvester onto a enriches for monocytes. Although the frequency of mono- fiberglass strip and the radioactivity determined in a Pack- cytes from G-CSF mobilized PSC products was slightly ard beta scintillation counter as described above. The data higher than that found in PSC products mobilized with GM- are presented as the mean Ϯ the standard error of the mean CSF, it did not achieve significance (P = 0.061). Thus, it (s.e.m.). Results were calculated as: appears that the frequency of monocytes within mobilized PSC products was independent of the type of HGF used [(c.p.m. (counts per minute) control Ϫ c.p.m. in mobilization. experimental)/c.p.m. control] × 100. The frequency of T cells within mobilized and non-mobi- This was then normalized to the activity of normal PB lized products was similar to that found in normal PB. analyzed in parallel experiments and reported as a percent Thus, there were 16.2 Ϯ 2.2% CD3+ cells in G-CSF mobil- of control cells. ized PSC products as compared to 14.8 Ϯ 7.9% CD3+ cells in GM-CSF mobilized products (P = 0.629). The frequency of CD3+ cells in normal PB, which has not been Ficolled, Statistical analysis (17.2 Ϯ 2.3%) also was statistically identical to that SPSS for Windows was used to apply Pearson, Kendall’s observed in the G-CSF products (P = 0.771) and GM-CSF tau-b, and Spearman correlation analyses (two-tail) to the mobilized PSC products (P = 0.421). In contrast to the data. The independent sample’s t-test was used to compare similar frequencies of CD3+ cells found in mobilized PSC the difference between two independent means. Mean dif- products and normal PB, the frequencies of CD4+ and CD8+ ferences were considered significant when P р 0.05. cells differ significantly between the two mobilized PSC products (P = 0.016 and 0.043, respectively), ie a signifi- cant increase in CD4+ cells and a significant decrease in Results CD8+ cells were found in G-CSF mobilized products rela- tive to GM-CSF mobilized products. The frequencies of Monocyte and T cell frequency within PSC products CD4+ cells (lymphocyte gate) within G-CSF and GM-CSF mobilized PSC products and normal PB were The frequency of monocytes (mean Ϯ s.e.m.) in mobilized 35.2 Ϯ 2.14%, 22.17 Ϯ 5.06% and 31.1 Ϯ 3.75%, respect- PSC products, 30.3 Ϯ 3.2% and 41.3 Ϯ 4.6% for GM-CSF + ively, while CD8 cells were 24.7 Ϯ 3.68%, 35.5 Ϯ 5.25% and G-CSF mobilized products, respectively, were signifi- Ϯ р and 18.6 2.84%, respectively. This resulted in a signifi- cantly higher (P 0.0001 for both) than for non-mobilized = Ϯ cant difference (P 0.034) in the CD4:CD8 cell ratio apheresis products (11.7 1.1%) (Figure 1). In addition to between GM-CSF (0.898 Ϯ 0.028) and G-CSF mobilized the significant increase in the frequency of monocytes in PSC products (1.99 Ϯ 0.47). The CD4:CD8 ratio of the the mobilized PSC products, there was also a significant GM-CSF mobilized products, but not the G-CSF mobilized increase in the frequency of monocytes in the non-mobil- products, also differed significantly (P = 0.014) from nor- ized apheresis products from normal individuals as com- mal PB cells (1.87 Ϯ 0.2).

PSC mobilized with G-CSF PSC mobilized with GM-CSF T cell function PSC non-mobilized Normal PB & $ * T cell proliferation by the MNC from PSC products was always compared to the PHA response of MNC from nor- *&+ mal PBL examined at the same time. The proliferation of lymphocytes from PSC products (c.p.m. experimental) is $+ reported as the percent proliferation of lymphocytes from normal PB (c.p.m. control) to normalize for day-to-day PSC products $+ variability. Although the response to PHA mitogenesis by G-CSF was lower than that of GM-CSF mobilized apher- esis products (16.8 Ϯ 5.4% vs 32.3 Ϯ 12% of control, respectively), it was not a significant difference (P = 0.257) 0 10 20 30 40 50 (Figure 2). This suggests that the low PHA mitogenic Monocyte frequency (%) response was independent of G-CSF or GM-CSF mobiliz- ation and did not correlate with the extent of prior chemo- Figure 1 Frequency of monocytes from PSC products mobilized or not therapy (analysis not shown). Further, MNC from PSC pro- mobilized with HGF and normal PB. Cells were stained with antibodies to CD14 and analyzed using flow cytometry. Monocytes were defined as ducts mobilized with G-CSF or GM-CSF had a decreased cells with high CD14 expression and intermediate side scatter. (&) Sig- PHA mitogenic response, which was significantly lower (P nificantly different from mobilized PSC products with GM-CSF (n = 21); = 0.0001 and P = 0.005, respectively) than that of non- (#) significantly different from PSC products mobilized with G-CSF (n = mobilized apheresis products (94.6 Ϯ 16.3%). Thus, HGF 15); ($) significantly different from non-mobilized PSC products from nor- mal donors (n = 9); and (+) significantly different from normal PB (n = mobilization contributes to the decreased PHA proliferation 18). Values represent the mean percent positive cells Ϯ standard error of by MNC from apheresis products and is independent of the the mean (s.e.m.). HGF used (G-CSF vs GM-CSF). Monocytes in PSC products AG Ageitos et al 66 PSC non-mobilized PSC Non-mobilized $ PSC mobilized with G-CSF PSC Mobilized with GM-CSF 1:1 $ PSC mobilized with GM-CSF $ PSC Mobilized with G-CSF &*

$ $

2:1 $ &* # & S:R ratio PSC products

$ 4:1 &* $

0 10 20 30 40 50 60 70 80 90 100 110 120 -20 -10 0 10 20 30 40 50 60 70 80 Percent of control (PBL-%) Percentage of inhibition of T cell blastogenesis (Relative Inhibition Index %) Figure 2 Response to PHA by irradiated MNC from PSC products that were non-mobilized or mobilized with G-CSF or GM-CSF. The results Figure 3 Inhibition of T cell proliferation by mobilized PSC products are normalized compared to that of MNC from normal donors compared expressed as a percent inhibition of PHA responses by allogeneic MNC ϫ 6 the same day. Cells from apheresis products were cultured at 1 10 /ml in the presence of graded numbers of irradiated MNC from mobilized PSC ␮ 3 in PHA at a final concentration of 0.5 g/ml for 72 h followed by H- products. MNC from normal donor PBL (1 ϫ 105) as responder cells were TdR incorporation (c.p.m.). (&) Significantly different from PSC products cocultured with varying numbers of irradiated (500 cGy) MNC from PSC = # mobilized with GM-CSF (n 21); ( ) significantly different from PSC products starting at an inhibition to responder cell ratio (I:R) of 4:1 and = products mobilized with G-CSF (n 15); and ($) significantly different an optimal concentration (0.5 ␮g/ml) of PHA in 96-well flat bottom plates. = from non-mobilized PSC products from normal donors (n 18). Data are Cells were cultured for 4 days and pulsed with 1 ␮Ci/well 3H-thymidine Ϯ the mean s.e.m. over the final 18 h of culture. The relative inhibition index was calculated using the formula 1 Ϫ c.p.m. experimental/c.p.m. control × 100. (&) Sig- = Cellular T cell inhibitor (TI) activity nificantly different from GM-CSF mobilized PSC products (n 21); ($) significantly different from non-mobilized PSC products from normal We previously reported that monocytes from mobilized donors (n = 18); and (*) significantly different from normal MNC from normal PB (paired studies and tests). The data are presented as the mean PSC products have the ability to kill activated T cells by Ϯ s.e.m. of results obtained with different patients. apoptosis.43 To determine whether the HGF used for the mobilization of stem cells affected TI activity, irradiated MNC from G-CSF and GM-CSF mobilized PSC products man’s rho P = 0.013, r =−0.479 analyses), suggesting an were added in graded numbers to cultures of allogeneic association between these two parameters. As shown in MNC with PHA. Figure 3 shows the ability of MNC from Figure 4, the frequencies of monocytes in the mobilized or PSC products to inhibit the mitogenic response to PHA. non-mobilized products significantly correlated (r = 0.517 The data are reported as the percent inhibition of T cell and P = 0.001) with TI activity. In addition, the inhibition blastogenesis at I:R ratios of 4:1, 2:1 and 1:1 as compared of T cell blastogenesis, and level of TI activity, was asso- to irradiated MNC from normal PB. The proliferative ciated with HGF mobilization and independent of the HGF response of the allogeneic T cells to PHA at each I:R ratio in the presence of irradiated MNC from G-CSF or GM-CSF mobilized PSC products was similar (4:1, 38.3 Ϯ 12.3% vs 80 Ϯ Ϯ Ϯ r = 0.517 51.3 15.5%; 2:1, 30.6 8.8% vs 40 9.8%; and 1:1, P = 0.001 18.6 Ϯ 9.6% vs 21.3 Ϯ 7.6%, respectively). Thus, the cells 70 from both HGF mobilized PSC products were potent inhibi- 60 tors of normal MNC proliferation in response to PHA, while cells from non-mobilized apheresis products showed 50 a slight or negative inhibition (Ϫ14.5 Ϯ 12%, Ϫ5.5 Ϯ 7% and Ϫ7.7 Ϯ 6%, at I:R ratios of 4:1, 2:1 and 1:1, 40

respectively). 30

Clinical and immunologic correlations 20 Monocyte frequency (%) As shown in Table 1, patients mobilized with GM-CSF 10 received significantly (P = 0.0001) more chemotherapy 0 compared to patients mobilized with G-CSF (8.6 Ϯ 1.1 vs -200 -150 -100 -50 0 50 100 150 Ϯ 1.8 0.3 months, respectively). There was no correlation Percent of inhibition of T cell blastogenesis between the number of months of chemotherapy prior to (Relative inhibition index-RII%) transplant and PHA or TI functional activity. Interestingly, Figure 4 Correlation of T cell inhibitor (TI) activity and monocyte fre- an analysis of the CD4:CD8 cell ratio with the extent of quency within MNC from non-mobilized and mobilized PSC products. A prior chemotherapy (months) revealed an inverse corre- positive and significant correlation (correlation coefficient, r = 0.517 and lation (Kendall’s tau-b P = 0.021, r = Ϫ0.337 and Spear- P = 0.001) was found between the frequency of monocytes and TI activity. Monocytes in PSC products AG Ageitos et al 67 used for mobilization. In other studies, we have found that T cells in the PSC product. This is consistent with prior monocytes at a high frequency result in TI activity,43 and studies7 where the monocytes were shown to actively their removal restores T cell function.43–45 inhibit T cell proliferation.43,44,46 In contrast to the similar frequency of CD3+ cells, there was a significant decrease in the frequency of CD4+ cells in GM-CSF mobilized pro- Discussion ducts and a significant increase in the frequency of CD8+ cells related to G-CSF products. This resulted in a signifi- Stem cell mobilization is used to obtain PSC products that cant difference in the CD4:CD8 ratio. Whilst not directly accelerate hematopoietic recovery following HDT and comparable, this differs from the results of San-Miguel et appear to contain a low frequency of contaminating tumor al58 who found that GM-CSF administration resulted in a cells. However, the influence of the mobilization procedure higher frequency of CD4+ cells in stem cell products. How- and the HGF used on lymphocyte function within the mobi- ever, because our studies examined patients at different dis- lized PSC products has not been studied extensively. ease stages, there were significant differences between the Despite the loss of T cell function in stem cell products, a two cohorts in the extent of prior chemotherapy. In agree- similar frequency of lymphocytes is found in PSC products ment with our prior studies,6 which demonstrated a corre- mobilized with HGF alone (G-CSF or GM-CSF) as com- lation between PHA response and prior chemotherapy, we pared to normal PB.46 However, little data exist comparing found an inverse relationship between chemotherapy and the effect of G-CSF to GM-CSF47–50 on the immune pheno- the CD4:CD8 ratio. Thus, in these studies the extent of type and function of the PSC products. Because we have prior chemotherapy is a critical determinant in the found that monocytes in GM-CSF mobilized PSC products CD4:CD8 ratio rather than the HGF used in mobilization. can inhibit T cell proliferation via the induction of This does not preclude a differential effect by HGFs on apoptosis,43,44 the purpose of this study was to compare the lymphocyte populations and function, but does demonstrate immune-mediating effect of mobilization with G-CSF vs that additional factors can compound the interpretation of GM-CSF. these data. Indeed, preclinical studies with G-CSF have Monocytes from GM-CSF mobilized PSC products were revealed a shift toward type 2 production.59 reported by Triozzi et al51 to have antitumor and accessory Recently, it was reported that an increase in monocytes immune activities. In contrast, we previously have predicts an optimal mobilization of CD34+ cells after reported44,52 that phagocytic CD14+ cells from GM-CSF chemotherapy followed by G-CSF administration24 and was mobilized PSC products can inhibit T and NK cell func- a useful parameter to follow in the timing of PSC collec- tions.43–45 These monocytes express a high level of HLA- tions. Further, it has been shown that the efficacy of CD34+ DR and low or non-existent levels of B7–1 (CD80). In stem cell purification from mobilized PSC is inversely pro- addition, these ‘activated monocytes’ have increased levels portional to the number of monocytes in the initial cell of IL-10 mRNA,53 which might relate to the low expression population. Nakamura et al60 also reported an increased of CD80 on the mobilized monocytes.54–56 Furthermore, the purification of CD34+ cells from mobilized PSC products GM-CSF mobilized monocytes induced the apoptosis of followed depletion of adherent cells. Similarly, Yamaguchi activated lymphocytes.43 Recently, Mielcarek et al57 also et al61 reported an inverse correlation between the fre- reported TI activity within G-CSF mobilized PSC products quency of CD14+ cells in the initial cell population and the used in allotransplantation, which may result in lower graft- efficacy of CD34+ cell purification using immunomagnetic versus-host disease (GVHD). In other studies, when mono- microspheres. Overall, we suggest that the depletion of cytes were removed from PSC products, T cell function HGF mobilized monocytes from within PSC products returned to normal values43–45 and TI activity was lost.43 might provide a strategy with therapeutic implications for We report here that HGF mobilized apheresis products hematologic and immunologic reconstitution post-trans- also have a significant increase in monocyte frequency plantation. We also speculate that the HGF mobilized mon- compared to normal PB and non-mobilized apheresis pro- ocytes are at least partially responsible for the immune dys- ducts. Non-mobilized apheresis products also have a sig- function found post-transplantation, a hypothesis currently nificantly higher monocyte frequency than normal PB, under investigation. which suggests that apheresis enriches monocytes as well Schneider et al48 have suggested that both the mobiliz- as CD34+ cells. In addition, the non-mobilized apheresis ation regimen and pre-treatment patient characteristics are products have a significantly higher PHA response than critical to the successful mobilization of PSC autografts. In PSC products mobilized with either G-CSF or GM-CSF. these studies, the mobilization of PSC with G-CSF or GM- Further, the MNC from mobilized PSC products have sig- CSF in patients who had received prior myelosuppressive nificantly higher TI activity for allogeneic lymphocyte pro- therapy produced autografts that were deficient in commit- liferation compared to MNC from non-mobilized products. ted progenitors and absolutely deficient in primitive pro- The monocyte content of the leukapheresis products genitors. In our studies, there was a significant difference directly correlated with TI activity. However, the TI in the extent of prior chemotherapy between the two groups activity was independent of the type of HGF used in mobil- of mobilized patients. Nonetheless, the T cell proliferative ization. Further, the depressed PHA mitogenic response response to PHA from the G-CSF mobilized PSC products was not due to a decreased CD3+ cell frequency, as the was insignificantly (P = 0.257) lower than that of GM-CSF frequency of CD3+ cells within normal PB and G-CSF and mobilized apheresis products (16.8 Ϯ 5.4% of control vs GM-CSF mobilized PSC products was statistically similar. 32.3 Ϯ 12% of control, respectively). Further, we found Thus, the depressed T cell function is not due to a loss of that the negative effect of monocytes on T cell proliferation Monocytes in PSC products AG Ageitos et al 68 was independent of whether G-CSF or GM-CSF was used 11 Foot ABM, Potter MN, Donaldson C et al. Immune reconsti- for mobilization and the frequency of CD3+ cells in the tution after BMT in children. Bone Marrow Transplant 1993; products. Thus, the low PHA proliferative response of 11: 7–13. MNC from mobilized PSC products is due to the presence 12 To LB, Shepperd KM, Haylock DN et al. Single high-doses of a high frequency of monocytes, while prior treatment of enable the collection of high numbers of hemopoietic stem cells from the peripheral blood. Exp with chemotherapy affects the CD4:CD8 cell ratio. In keep- 46 Hematol 1990; 18: 442–447. ing with this conclusion, data reported by Mielcarek et al 13 Copelan EA, Ceselski SK, Ezzone SA et al. Mobilization of found similar TI activity in leukapheresis products from peripheral-blood progenitor cells with high-dose etoposide and normal donors after G-CSF mobilization. In conclusion, we granulocyte colony-stimulating factor in patients with breast suggest that the extent of prior chemotherapy and the mobi- cancer, non-Hodgkin’s lymphoma and Hodgkin’s disease. J lizing HGF are independent of TI activity for allogeneic Clin Oncol 1997; 15: 759–765. lymphocyte proliferation. Furthermore, an inverse corre- 14 To LB, Haylock DN, Dowse T et al. A comparative study of lation is found between the frequency of monocytes in aph- the phenotype and proliferative capacity of peripheral blood (PB) CD34+ cells mobilized by four different protocols and eresis products and TI activity in agreement with our pre- + vious reports of monocytes as mediators of TI activity.43–45 those of steady-phase PB and bone marrow CD34 cells. Blood 1994; 84: 2930–2939. 15 Siena S, Bregni M, Brando B et al. Circulation of CD34+ hematopoietic stem cells in the peripheral blood of high-dose Acknowledgements cyclophosphamide-treated patients: enhancement by intra- venous recombinant granulocyte–macrophage colony-stimul- This research was supported in part by grant No. RO1-CA61593 ating factor. Blood 1989; 74: 1905–1914. from the National Institutes of Health and Nebraska Cancer and 16 Pettengell R, Demuynck H, Testa NG, Dexter TM. The Smoking Disease Research Program No. 97-71. We also would engraftment capacity of peripheral blood progenitor cells like to thank Lisa Chudomelka for her editorial assistance. (PBPC) mobilized with chemotherapy+G-CSF. Int J Cell Cloning 1992; 10: 59. 17 Tarella C, Ferrero D, Bregni M et al. Peripheral blood expan- References sion of early progenitor cells after high dose cyclophospham- ide and rH GM-CSF. Eur J Cancer 1991; 27: 22–27. 1 Brugger W, Bross K, Frisch J et al. Mobilization of peripheral 18 Nemunaitis J. Cytokine-mobilized peripheral blood progenitor blood progenitor cells by sequential administration of interleu- cells. Semin Oncol 1996; 23: 9–14. kin-3 and granulocyte–macrophage colony-stimulating factor 19 Sheridan WP, Begley CG, Juttner CA et al. Effect of periph- following polychemotherapy with etoposide, ifosfamide and eral blood progenitor cells mobilized by filgrastim (G-CSF) cisplatin. Blood 1992; 79: 1193–1200. on recovery after high-dose chemotherapy. Lancet 2 Lieschke GJ, Burgess AW. Granulocyte colony-stimulating 1992; 339: 640–644. factor and granulocyte–macrophage colony-stimulating factor. 20 Vose JM, Bierman PJ, Kessinger A et al. The use of recombi- New Engl J Med 1992; 327: 99–106. nant human granulocyte–macrophage colony stimulating fac- 3 Sharp JG. Micrometastases and transplantation. J Hematother tor for the treatment of delayed engraftment following high 1996; 5: 519–524. dose therapy and autologous trans- 4 To LB, Roberts M, Haylock D. Comparison of haematological plantation for lymphoid malignancies. Bone Marrow Trans- recovery times and supportive care requirements of autologous plant 1991; 7: 139–143. recovery phase peripheral blood stem cell transplants, autolog- 21 Kessinger A. Circulating stem cells: waxing hematopoietic. ous bone marrow transplants and allogeneic bone marrow New Engl J Med 1995; 333: 315–316. transplants. Bone Marrow Transplant 1992; 9: 277–284. 22 Haas R, Ho D, Bredthauer U et al. Successful autologous 5 Henon PR, Liang H, Beck-Wirth G et al. Comparison of hem- transplantation of blood stem cells mobilized with recombi- atopoietic and immune recovery after autologous bone marrow nant human granulocyte–macrophage colony-stimulating fac- or blood stem cell transplants. Bone Marrow Transplant 1992; tor. Exp Hematol 1990; 18: 94–98. 9: 285–291. 23 Pettengell R, Testa NG, Swindell R et al. Transplantation 6 Talmadge JE, Reed E, Ino K et al. Rapid immunologic recon- potential of hematopoietic cells released into the circulation stitution following transplantation with mobilized peripheral during routine chemotherapy for non-Hodgkin’s lymphoma. blood stem cells as compared to bone marrow. Bone Marrow Blood 1993; 82: 2239–2248. Transplant 1997; 19: 161–172. 24 Hansson M, Svensson A, Engervall P et al. Increase of mono- 7 Roberts MM, To LB, Gillis D et al. Immune reconstitution cytes predicts mobilization of peripheral stem and progenitor following peripheral blood stem cell transplantation, autolog- cells after chemotherapy followed by G-CSF administration. ous bone marrow transplantation and allogeneic bone marrow Eur J Haematol 1995; 54: 321–328. transplantation. Bone Marrow Transplant 1993; 12: 469–475. 25 Lobo F, Kessinger A, Landmark JD et al. Addition of periph- 8 Ashihara E, Shimazaki C, Yamagata N et al. Reconstitution eral blood stem cells collected without mobilization tech- of lymphocyte subsets after peripheral blood stem cell trans- niques to transplanted autologous bone marrow did not hasten plantation: two color cytometric analysis. Bone Marrow marrow recovery following myeloablative therapy. Bone Mar- Transplant 1994; 13: 377–381. row Transplant 1991; 8: 389–392. 9 Talmadge JE, Reed EC, Kessinger A et al. Immunologic attri- 26 Weisbart RH, Golde DW, Clark SC. Human granulocyte– butes of cytokine mobilized peripheral blood stem cells and macrophage colony-stimulating factor is activator. recovery following transplantation. Bone Marrow Transplant Nature 1985; 314: 361–363. 1996; 17: 101–109. 27 Schwarzmeier JD. The role of cytokines in . 10 Atkinson K. Reconstitution of the hematopoietic and immune Eur J Haematol 1996; 57: 69–74. recovery systems after human marrow transplantation. Bone 28 Richard C, Baro J, Bello-Fernandez C et al. Recombinant Marrow Transplant 1990; 5: 209–226. human granulocyte–macrophage colony stimulating factor (rh Monocytes in PSC products AG Ageitos et al 69 GM-CSF) administration after autologous bone marrow trans- cell products: monocyte inhibition. J Immunother 1999 (in plantation for acute myeloblastic enhances activated press). killer cell function and may diminish leukemic relapse. Bone 46 Mielcarek M, Roecklein BA, Torok-Storb B. CD14+ cells in Marrow Transplant 1995; 15: 721–726. granulocyte colony-stimulating factor (G-CSF) mobilized per- 29 Dreger P, Suttorp M. Influence of G-CSF on NK cell recovery ipheral blood mononuclear cells induce secretion of interleukin- after autologous BMT. Bone Marrow Transplant 1996; 13: 6andG-CSFbymarrowstroma.Blood 1996; 87: 574–580. 667–668. 47 Bolwell BJ, Goormastic M, Yanssens T et al. Comparison of 30 Fabian I, Kletter Y, Bleiberg I et al. Effect of exogenous G-CSF with GM-CSF for mobilizing peripheral blood pro- recombinant human granulocyte and granulocyte–macrophage genitor cells and for enhancing marrow recovery after autolog- colony-stimulating factor on neutrophil function following ous bone marrow transplant. Bone Marrow Transplant 1994; allogeneic bone marrow transplantation. Exp Hematol 1991; 14: 913–918. 19: 868–873. 48 Schneider JG, Crown JP, Wasserheite CJ et al. Factors affect- 31 Bennett CL, Smith TJ, Weeks JC et al. Use of hematopoietic ing the mobilization of primitive and committed hematopoietic colony-stimulating factors: The American Society of Clinical progenitors into the peripheral blood of cancer patients. Bone Oncology Survey. J Clin Oncol 1996; 14: 2511–2520. Marrow Transplant 1994; 14: 877–884. 32 Schimitz N, Dreger P, Zander AR et al. Results of a ran- 49 Ho AD, Young D, Maruyama M et al. Pluripotent and lineage- domized, controlled, multicentre study of recombinant human committed CD34+ subsets in leukapheresis products mobil- granulocyte colony-stimulating factor (filgrastim) in patients ized by G-CSF, GM-CSF vs a combination of both. Exp Hem- with Hodgkin’s disease and non-Hodgkin’s lymphoma atol 1996; 24: 1460–1468. undergoing autologous bone marrow transplantation. Bone 50 San Miguel JF, Hernandez MD, Gonzalez M et al. A ran- Marrow Transplant 1995; 15: 261–266. domized study comparing the effect of GM-CSF and G-CSF 33 Avalos BR. Molecular analysis of the granulocyte colony- on immune reconstitution after autologous bone marrow trans- stimulating factor receptor. Blood 1996; 88: 761–777. plantation. Br J Haematol 1996; 94: 140–147. 34 Gasson JC. Molecular physiology of granulocyte–macrophage 51 Triozzi PL, Tucker F, Benzies T, Balcerzak SP. Antitumor colony-stimulating factor. Blood 1991; 77: 1131–1145. and accessory immune activities of peripheral blood stem cells 35 DeLuca E, Sheridan WP, Watson D et al. Prior chemotherapy mobilized with granulocyte–macrophage colony-stimulating does not prevent effective mobilization by G-CSF of periph- factor. Bone Marrow Transplant 1996; 18: 47–52. eral blood progenitor cells. Br J Cancer 1992; 66: 893–899. 52 Tanaka J, Mielcarek M, Torok SB. Impaired induction of the 36 Socinski MA, Elias A, Schnipper L et al. Granulocyte–macro- CD28-responsive complex in granulocyte colony-stimulating phage colony stimulating factor expands the circulating factor mobilized CD4 T cells. Blood 1998; 91: 347–352. haemopoietic compartment in man. Lancet 53 Singh RK, Varney ML, Ageitos AG et al. Differential cyto- 1988; 1: 1194–1198. kine expression in growth factor mobilized peripheral stem 37 Dreger P, Marquardt P, Haferlach T et al. Effective mobiliz- cell and bone marrow products and their immunoregulatory ation of peripheral blood progenitor cells with ‘Dexa-BEAM’ significance. Proc Amer Assoc Cancer Res 1997; 58: 33. and G-CSF: timing of harvesting and composition of the leu- 54 de Wall MR, Haanen J, Spits H et al. Interleukin-10 (IL-10) kapheresis product. Br J Cancer 1993; 68: 950–957. and viral IL-10 strongly reduce antigen-specific human T cell 38 Mills KC, Gross TG, Varney ML et al. Immunologic pheno- proliferation by diminishing the antigen-presenting capacity of type and function in human bone marrow, blood stem cells monocytes via downregulation of class II histocompatibility and umbilical . Bone Marrow Transplant 1996; 18: complex expression. J Exp Med 1991; 174: 915–924. 53–61. 55 Willems F, Marchant A, Delville JP et al. Interleukin-10 39 Santoli D, Clark SC, Kreider BL et al. Amplification of IL-2 inhibits B7 and intercellular adhesion molecule-1 expression driven T cell proliferation by recombinant human IL-3 and on human monocytes. Eur J Immunol 1994; 24: 1007–1009. granulocyte–macrophage colony stimulating factor. J Immunol 56 Bromberg JS. IL-10 immunosuppression in transplantation. 1988; 141: 519–526. Curr Opin Immunol 1995; 7: 639–643. 40 Ho AD, Haas R, Wulf G et al. Activation of lymphocytes 57 Mielcarek M, Martin PJ, Torok-Storb B. Suppression of induced by recombinant human granulocyte–macrophage col- alloantigen-induced T-cell proliferation by CD14+ cells ony-stimulating factor in patients with malignant lymphoma. derived from granulocyte colony-stimulating factor-mobilized Blood 1990; 75: 203–212. peripheral blood mononuclear cells. Blood 1997; 89: 1629– 41 Al-Aoukaty A, Giaid A, Sinoff C et al. Priming effects of 1634. GM-CSF are coupled to cholera toxin-sensitive guanine nucle- 58 San-Miguel MJ, Hernandez MD, Gonzalez M et al. A ran- otide binding protein in human T lymphocytes. Blood 1994; domized study comparing the effect of GM-CSF and G-CSF 83: 1299–1309. on immune reconstitution after autologous bone marrow trans- 42 Badger AM, King AG, Talmadge JE et al. Induction of non- plantation. Br J Haematol 1996; 94: 140–147. specific suppressor cells by a novel azaspirane, SK&F 105685, 59 Pan L, Delmonte J Jr, Jalonen CK, Ferrara JLM. Pretreatment an immunomodulatory compound with therapeutic activity in of donor mice with granulocyte colony-stimulating factor pol- animal models of . J Autoimmun 1990; 3: arizes donor T lymphocytes toward type-2 cytokine pro- 485–500. duction and reduces severity of experimental graft-versus-host 43 Ino K, Singh RK, Talmadge JE. Monocytes from mobilized disease. Blood 1996; 86: 4422–4429. stem cells inhibit T cell function. J Leuk Biol 1997; 61: 60 Nakamura Y, Noma M, Kidokoro M et al. Isolation of human 583–591. CD34+ peripheral blood stem cells using an immunomagnetic 44 Ageitos AG, Ino K, Ozerol I et al. Restoration of immune nylon-wool column. J Hematother 1994; 3: 128–133. function in mobilized stem cell product by suppressor cell 61 Yamaguchi M, Sawada K, Sato N et al. A rapid nylon-fiber removal. Bone Marrow Transplant 1997; 20: 117–123. syringe system to deplete CD14+ cells for positive selection 45 Ageitos AG, Singh RK, Ino K et al. IL-2 Expansion of T and of human blood CD34+ cells. Use of immunomagnetic NK cells from growth factor-mobilized peripheral blood stem microspheres. Bone Marrow Transplant 1997; 19: 373–379.