Spontaneous circulation of myeloid-lymphoid–initiating cells and SCID-repopulating cells in sickle cell crisis (original) (raw)
Increased numbers of primitive progenitors are present in the PB during acute crisis of SCD. We initially evaluated the number of CD34+CD38–Lin– cells in PB of SCD patients and normal (NL) donors (Figures 1 and 2). PB from NL donors contained 6.7 (range 1–11) CD34+CD38–Lin– cells per 105 PBMCs, and PB from patients with SCD in steady state contained 13.7 (range 1–23) CD34+CD38–Lin– cells per 105 PBMCs. PB from patients with SCD in acute crisis (AC-SCD patients) had an average of 250 (range 38–880) CD34+CD38–Lin– cells per 105 PBMCs, significantly more than in SS-SCD patients or normal controls (P < 0.01).
FACS analysis for CD34+CD38–Lin– cells in PB of NL donors and SS-SCD and AC-SCD patients. CD34+ cells were selected as described in Methods, stained with CD34 APC, CD38 FITC, and lineage phycoerythrin (Lin-PE) antibodies, and selected by FACS. Live cells were gated in R1, and then CD34 cells were gated against lineage cells. R2 represents the CD34+Lin– gate. Cells in R2 were gated as CD34 versus CD38. The R3 gate, CD34+CD38– (not shown here), was used to sort for single-cell deposition for assays.
Significantly more CD34+CD38–Lin– cells are present per 105 PBMCs from AC-SCD patients compared with NL donors and SS-SCD patients. Mononuclear cells were selected from blood from normal donors, SS-SCD patients, and AC-SCD patients. Following enrichment for CD34+ cells and Lin– cells, CD34+CD38–Lin– cells were selected by FACS. Significantly more CD34+CD38–Lin– cells were detected among mononuclear cells from AC-SCD blood compared with SS-SCD and NL-donor blood. Differences between the groups were evaluated by Student’s t test.
We next examined the proportion of CD34+CD38–Lin– cells, from the three donor groups, that were ML-ICs, LTC-ICs, and NK-ICs. For these studies, CD34+CD38–Lin– cells were selected from 50–200 ml PB. Because the number of CD34+CD38–Lin– cells was significantly lower in blood of NL donors and SS-SCD patients, fewer CD34+CD38–Lin– cells were available for evaluation from these individuals (11–66 from NL donors and 28–88 from SS-SCD patients) than from AC-SCD patients (66–132 cells). We could not detect ML-ICs among PB CD34+CD38–Lin– cells from NL donors or SS-SCD patients. CD34+CD38–Lin– cells in PB from normal donors included no cells that gave rise to NK-ICs, and 0.29% cells that gave rise to LTC-ICs only (range 0–0.5%; only three of ten PB samples contained enough CD34+CD38–Lin– cells to evaluate LTC-IC frequency). CD34+CD38–Lin– cells from SS-SCD patients included 1.51% (range 0–3.03%) cells that gave rise to LTC-ICs only and 0.5% (range 0–1.51%) cells that gave rise to NK-ICs only upon replating. In contrast, CD34+CD38–Lin– cells from AC-SCD patients included 4.55% (range 3.03–9.09%) and 0.82% (range 0–3.03%) cells that gave rise to either LTC-ICs only or NK-ICs only upon replating (Figure 3, a and b). However, because 4.35% of CD34+CD38–Lin– cells gave rise to at least one LTC-IC and one NK-IC, they were considered ML-ICs (Figure 3c). No CD34+CD38–Lin– cells from SS-SCD patients or from NL donors gave rise to LTC-ICs and NK-ICs; therefore, no ML-ICs could be detected (P < 0.05 for LTC-IC frequencies, P < 0.01 for NK-IC frequencies, and P < 0.001 for ML-IC frequencies vs. NL or SS-SCD). A strong correlation was seen between the presence of LTC-ICs (_r_2 = 0.74) or NK-ICs (_r_2 = 0.96) and the presence of ML-ICs in the blood of AC-SCD patients.
Significantly more LTC-ICs (a), NK-ICs (b), and ML-ICs (c) are present in CD34+CD38–Lin– cells from AC-SCD patients than in those from NL donors or SS-SCD patients. Single CD34+CD38–Lin– cells, selected by FACS from normal donors (44–66 wells plated), SS-SCD patients (66–88 wells plated), and AC-SCD patients (66–132 wells plated), were plated in ML-IC assays as previously described (24). After 2 weeks, single CD34+CD38–Lin– cell progeny were replated in four individual wells, two of which were maintained under LTC-IC and two under NK-IC conditions. For LTC-IC cultures, wells were overlaid after 5 weeks with clonogenic medium, and the presence of CFCs was determined 2 weeks later. NK-IC cells were harvested after 5 weeks and evaluated by FACS for the presence of CD56+ NK cells or CD19+ B cells. An LTC-IC was determined as a well in LTC-IC cultures where CFCs were present without NK and/or B cells being present in the companion NK-IC cultures. An NK-IC was determined as a well in NK-IC cultures where NK and/or B cells were present without CFCs being present in the companion LTC-IC cultures. An ML-IC was identified when progeny of the initial CD34+CD38–Lin– cell gave rise to at least one LTC-IC and at least one NK-IC. Significantly more LTC-ICs, NK-ICs, and ML-ICs were present among CD34+CD38–Lin– cells in AC-SCD blood compared with blood from SS-SCD patients and NL donors. Values shown as 0 represent frequencies below the detection level of the assays. As we plated between 44 and 132 wells, this indicates that fewer than one ML-IC was present among 44–132 CD34+CD38–Lin– cells. Differences between the groups were evaluated by t test.
We next tested whether PB CD34+ cells from SCD patients included SRCs. Fifty milliliters of blood was obtained from patients with SCD in acute crisis or steady state, and 200 milliliters was obtained from normal donors. CD34+ cells were selected, and all CD34+ cells were injected in sublethally irradiated NOD-SCID mice. Because of the differences in the number of CD34+ cells obtained from a given volume of PB among AC-SCD patients, SS-SCD patients, and NL donors, the total number of CD34+ cells injected varied significantly between the groups: 2.68 × 105 ± 0.62 × 105 for AC-SCD patients, 2.92 × 104 ± 1.32 × 104 for SS-CSD patients, and 2.88 × 104 ± 0.6 × 104 for normal donors (P < 0.01). Engraftment levels are shown in Table 2. All animals that were injected with CD34+ cells from AC-SCD patients showed multilineage human cell engraftment by FACS, at levels ranging from 0.6% to 1.9% (Table 2; Figure 4). These results were confirmed by quantitative PCR for human DNA, which showed engraftment at levels ranging from 0.6% to 1.4% (Table 2; Figure 5). PCR for the β-globin gene also showed that engrafted cells contained the valine-to-glutamic-acid substitution at codon 6 (not shown). In contrast, animals that received CD34+ cells from SS-SCD patients did not show any human engraftment. Likewise, animals that received CD34+ cells from normal donors did not have human CD45+ cells. Control animals that received either irradiated normal cells alone or CD34– cells also did not show engraftment, as assessed by FACS or PCR (Table 2).
Multilineage engraftment of AC-SCD CD34+ cells in NOD/SCID mice. Bone marrow from mouse no. 3 was harvested as described and then stained with anti–human CD45 and anti–human lineage antibodies to demonstrate multilineage engraftment. Live cells were gated in R1, and then human CD45 was gated against the antibody of interest. Percentages of human lineage cells are shown on the FACS plots.
PCR confirmation of human hematopoietic cell engraftment in NOD/SCID mice transplanted with AC-SCD CD34+ cells. PCR for human DNA is shown. Lanes 1–3, spleens from mice nos. 1–3. Lanes 4–9, bone marrow from mice nos. 1–6. Lanes 10–12, PB from mice nos. 1–3. Lane 13, K562 cells. Lane 14, cord blood CD34+ cells. Normal mouse bone marrow was used as a control.
Engraftment of CD34+ cells from PB of normal donors, SS-SCD patients, and AC-SCD patients
Bone marrow from mice with engraftment levels of about 2% human CD45+ cells (n = 2) was transplanted into secondary irradiated NOD/SCID recipients, and their marrow was again analyzed at 6 weeks for human engraftment. Both recipients showed 0.2% engraftment by FACS and PCR.
Increased levels of mobilizing cytokines are present in the PB during acute crisis of SCD. Levels of SCF (Figure 6a), GCSF (Figure 6b), GM-CSF (Figure 6c), and IL-8 (Figure 6d) were measured in serum of NL donors and SCD patients. SCF was undetectable in the serum of normal donors. Levels of SCF in serum of SS-SCD patients were 11.67 pg/ml (range 1–22 pg/ml), whereas levels in serum of AC-SCD patients were 4,150 pg/ml (range 88–7,000 pg/ml), representing a 200-fold increase compared with SS-SCD patients (P < 0.01). Serum GCSF levels in normal controls averaged 859 pg/ml (range 594–966 pg/ml), similar to those for SS-SCD patients (759 pg/ml; range 654–871 pg/ml). In contrast, serum levels of GCSF in AC-SCD patients were 2,000 pg/ml (range 1,820–2,180 pg/ml) (P < 0.05 vs. SS-SCD group). GM-CSF was undetectable by ELISA in the serum of normal donors; SS-SCD patients had an average serum level of 1.5 pg/ml (range 0–12 pg/ml), whereas levels in serum of AC-SCD patients averaged 1,049 pg/ml (range 431–1,914 pg/ml) (P < 0.01 for AC-SCD patients vs. SS-SCD patients and NL donors). Levels of IL-8 were also significantly higher in AC-SCD patients (4,820 pg/ml; range 466–7,015 pg/ml) than in SS-SCD patients (86 pg/ml; range 41–465 pg/ml) and NL donors (58 pg/ml; range 46–98 pg/ml) (P < 0.01 for AC-SCD vs. SS-SCD patients). ML-IC frequency in blood of AC-SCD patients correlated with SCF levels at _r_2 = 0.78, GCSF levels at _r_2 = 0.50, IL-8 levels at _r_2 = 0.72, and GM-CSF levels at _r_2 = 0.58.
Elevated serum levels of SCF (a), GCSF (b), GM-CSF (c), and IL-8 (d) in serum of AC-SCD patients compared with NL donors and SS-SCD patients. Serum was collected from ten normal donors, seven SS-SCD patients, and 13 AC-SCD patients, and then frozen at –80°C. Levels of SCF, GCSF, GM-CSF, and IL-8 were measured using ELISA. Differences between groups were analyzed by unpaired t test.
Addition of SCF to GCSF results in increased mobilization of primitive progenitors in the PB of lymphoma patients. To further determine whether the presence of high levels of multiple cytokines in serum of AC-SCD patients might be responsible for the mobilization of ML-ICs in the blood, we evaluated the number of ML-ICs, LTC-ICs, and NK-ICs in the PB of patients with Hodgkin disease or non-Hodgkin lymphoma who were treated with either GCSF alone or GCSF plus SCF to mobilize PB progenitors for transplantation. Increased numbers of ML-ICs were present in lymphoma patients whose blood was mobilized with GCSF plus SCF (6.25% ML-ICs [range 3.03–8.33%] per CD34+CD38–Lin– cell) compared with patients whose blood was mobilized with GCSF alone (1.7% ML-ICs [range 1.41–2.27%] per CD34+CD38–Lin– cell). As we observed for AC-SCD and SS-SCD patients, the number of LTC-ICs among PB CD34+CD38–Lin– cells was only marginally different (GCSF plus SCF, 18.22%, range 15.9–31.82%; GCSF alone, 13.45%, range 7.58–18.94%), whereas fewer NK-ICs were detected among PB CD34+CD38–Lin– cells from lymphoma patients treated with GCSF alone (1.52%, range 0–3.13%) compared with patients treated with GCSF plus SCF (8.91%, range 4.55–15.17%) (Figure 7, a–c).
More LTC-ICs, NK-ICs, and ML-ICs are mobilized in PB following GCSF+SCF than following GCSF-only mobilization. Single CD34+CD38–Lin– cells, selected by FACS from lymphoma patients who received GCSF+SCF mobilization (66–132 wells plated) and lymphoma patients who received GSCF-only mobilization (66–132 wells plated), were plated in the ML-IC assay as previously described (24). The frequency of LTC-ICs was enumerated by overlaying of cultures with clonogenic medium after 5 weeks. Cells were maintained for 2 weeks in expansion medium, and progeny were replated in two LTC-IC and two NK-IC cultures. The frequency of NK-ICs was enumerated by harvesting of plates after 5 weeks and evaluation of the cells for presence of CD56+ NK cells or CD19+ B cells by FACS as previously described (24). An ML-IC was identified when progeny of the initial CD34+CD38–Lin– cell gave rise to at least one LTC-IC and at least one NK-IC. Significantly more NK-ICs and ML-ICs were present in CD34+CD38–Lin– cells from blood mobilized with GCSF+SCF than in those from blood mobilized with GCSF only. Differences in LTC-IC frequency between the two groups were not significant. Differences between the groups were evaluated by t test.