Handling of CB cells
The isolation of pure mononuclear cell (MNC) fractions from CB, and subpopulations thereof, brings about a special challenge. This appears to be due to the large number of thrombocytes and erythroid progenitors in CB. In the Ficoll-Paque density gradient, all erythroid cells do not necessarily sediment to the bottom layer, but are retained in the interphase of plasma and Ficoll-Paque. The erythroid cells remaining in the interphase are nucleated progenitors that may hamper the subsequent immunomagnetic selection of HSC populations. Treatment with ammonium chloride or diethylene glycol may be used to deplete red blood cells, but depletion was not performed in this study as the nucleated erythroid progenitors are not easily lyzed and purities up to 97% were reached without any additional treatment. The unusually slow sedimentation of erythroid cells is not seen when working with PB.
When handling cryopreserved CB cells, aggregation was observed. Aggregation was reduced by replacing ethylenediamine tetraacetic acid (2 mM EDTA, Merck, Darmstadt, Germany) with anticoagulant citrate dextrose solution, formula A (0.6% ACD/A, Baxter Healthcare, Lessines, Belgium) in the sample buffer. In some cases aggregation was so substantial that the cells needed to be resuspended in DNaseI containing buffer. DNaseI digests the DNA released from dead cells and prevents aggregation. The DNase treatment did not affect the viability or colony-forming potential of selected CB cells. No cell aggregation was seen when handling fresh CB cells.
In cryopreserved CB, the mean MNC concentration was 5.14 × 109/l (range 0.96–10.00, SD = 3.43) (Figure ), and the mean platelet concentration was 8.30 × 109/l (range 0.00–17.00, SD = 5.46). Cryopreserved CB contained a mean of 0.14 × 1012/l erythrocytes (range 0.02–0.67, SD = 0.20), and the mean hematocrit was 2%.
Figure 1 Mean MNC count in cryopreserved and fresh CB. The mean MNC concentration for cryopreserved and fresh CB was 5.14 × 109/l (range 0.96–10.00, median 4.30, SD = 3.43) and 2.68 × 109/l (range 1.24–3.62, median 2.77, SD = 0.64), (more ...)
Fresh CB contained a mean of 2.68 × 109
/l MNCs (range 1.24–3.62, SD = 0.64) (Figure ). The difference in MNC concentration between cryopreserved and fresh samples was not statistically significant (P = 0.06). The remarkably high disparity in the standard deviation of MNC concentration between cryopreserved and fresh CB may be due to the processing and freezing of cells performed to bank the CB units [19
]. The mean concentration was 205.89 × 109
/l (range 84.00–505.00, SD = 130.58) for platelets and 0.12 × 1012
/l (range 0.03–0.50, SD = 0.14) for erythrocytes. The mean hematocrit was 1%.
Immunomagnetic separation of HSC populations
When using the Direct CD34 Progenitor Cell Isolation Kit with single column separation and the labeling protocol recommended by the manufacturer (Miltenyi Biotec, Bergisch Gladbach, Germany), a purity of less than 50% was reached for the CD34+ cells (Figure ). To obtain highly pure CD34+ cells, the immunomagnetic selection method was optimized. Several washing steps (3–10) were tested for single column separation. A purity of 80% was achieved with extensive washing, but the yield was poor (less than 50% of the expected yield). Two successive column separations resulted in 77% purity, but a great number of CD34+ cells were still lost during the process indicating a further need to optimize the protocol. An additional labeling step between the two column separations increased the purity to >90% (results for a representative sample shown in figure ) and resulted in an acceptable yield as well. The optimized two-column method with additional labeling proved reliable and was applied to the separation of both CD34+ and CD133+ cells. The average yield of CD34+ and CD133+ cells from one cord blood unit was 106 and 105, respectively. The purity of positively selected CD34+/CD133+ cells was reproducibly over 90% and their negative counterparts were nearly 100% pure.
Figure 2 Purity assessment of the CD34+ cell fraction by flow cytometry. The initial purity of CD34+ cells after separation through single column was 47.5%. The CD34- fraction was 99.4% pure. CD34+ and CD34- cell populations were defined by first gating on forward (more ...)
Figure 3 Purity assessment of CD34+ cell fraction after one or two column separations. A) The CD34+ cell fraction was 78% pure after the first column separation. B) A 92% pure CD34+ cell faction was obtained by an additional labeling step in connection with a (more ...)
Generally, the recovery of CD34+ and CD133+ cells was 0.86% (range 0.56–1.45, SD = 0.36) and 0.21% (range 0.04–0.41, SD = 0.12), respectively. The recovery of CD34+ cells was higher from fresh CB (0.97%) when compared to cryopreserved CB (0.78%), although the difference was not statistically significant (P= 0.54). The results are consistent with the study by Almici et al. showing no significant difference in yield or in purity for fresh CB CD34+ cells in comparison to crypreserved cells [20
]. This was the case with CD133+ cells as well, the recovery being 0.29% for fresh CB and 0.12% for cryopreserved CB (P = 0.11). The purities were not affected by the initial percentage of HSC populations in CB. The results of the purity assessment for representative samples of CD34+/-, CD133+/- and Lin-/+ cells are shown in figure .
Figure 4 Purity assessment of CD34+/-, CD133+/- and Lin-/+ cell fractions. A) Purities for CD34+ and CD34- cell factions were 97.1% and 99.1%, respectively. B) Purities for CD133+ and CD133- fractions were 93.6% and 99.1%, respectively. C) Purities for Lin- and (more ...)
The magnetic sorting of Lin- cells was optimized to find the optimum concentrations for antibodies and magnetic colloids. Lin- cells were separated from the MNC fraction with magnetic cell sorting and the purity of the enriched cell fraction was analyzed by flow cytometry. The average yield of Lin- cells from one cord blood unit was 106. The overall recovery of Lin- cells was 0.29% (range 0.13–0.70, SD = 0.21), being higher in fresh CB samples (0.43%) than in cryopreserved CB samples (0.16%). The difference between fresh and cryopreserved samples was not statistically significant (P = 0.19). The mean purity of Lin- cell population was 90% (range 82–98%, SD = 5.00) and it did not differ between fresh and cryopreserved CB units. To assess the effect of antibody concentration on the selection of Lin- cells, varying amounts of the antibody cocktail (50–100 μl/ml) and magnetic colloids (30–60 μl/ml) were tested. The concentration of antibodies and magnetic colloids did not affect the purity of Lin- cells based on flow cytometric analysis.
All the optimized protocols are described in detail in figure . The operation time for the selection of CD34+/- and CD133+/- cells using MiniMACS or MidiMACS is approximately 1.5 hours. The operation time for isolation of Lin-/+ cells is approximately 45 minutes. The selection can be made more effective with the AutoMACS system developed for high-speed automated cell sorting. The optimized protocols have been developed for enrichment of CB HSCs for research purposes. However, the enrichment of stem and progenitor cells is often necessary in clinical settings. The selected HSCs are increasingly used in transplantations and enrichment may be required for depletion of contaminating mature cells or tumor cells. Potentially, the methods described here could be applied to clinical-grade selection using the CliniMACS System that is CE-marked for clinical use in Europe.
Figure 5 A chart of the optimized protocols to isolate CD34+/-, CD133+/- and Lin-/+ cells from cord blood. Isolation of CD34/- and CD133+/- cells was performed using Direct CD34 Progenitor Cell Isolation Kit (#130-046-702, Miltenyi Biotec) and CD133 Cell Isolation (more ...)
With the optimized protocols, a purity of at least 90% was achieved for CD34+, CD133+ and Lin- cells. Viability was 99% for all the selected cell types. This demonstrates that the optimized protocols work well for HSC enrichment from both fresh and cryopreserved CB. Fresh CB was easier to handle and the recovery of HSCs was higher from fresh CB. Nonetheless, cryopreserved cord blood is almost exclusively used in clinical settings. Therefore, if one wishes to use selected HSCs, the cells should preferably be isolated on fresh cord blood and cryopreserved after the selection procedure to maximize the recovery of HSCs. HSCs, enriched by the protocols described here, have been used in gene expression studies with great reproducibility and consistency [21
It has been suggested that the binding of an antibody to the surface of a HSC may influence cell proliferation and differentiation by activating intracellular signaling pathways [14
]. An anti-CD34 antibody has been shown to induce tyrosine phosphorylation in BM-derived CD34+ cells [22
]. However, Lin- cells are selected through negative depletion. Thus, neither antibody binding nor activation of signaling pathways is expected. Further studies on the effect of the interaction between HSCs and the antibodies used for their selection as well as the possible impact of this contact on HSC quality are awaited.
Low number of HSCs in cord blood is a limitation for its use. Ex vivo expansion of HSCs may be used to generate the clinically relevant cell numbers needed for adult patients. However, mature cells may develop during long-term culture and result in a need for reselection of progenitor cells. The optimized protocols can be applied for enrichment of stem and progenitor cells after ex vivo expansion.
Colony forming unit assay
The CFU assay was used to measure the clonogenic capacity of CD34+, CD133+ and Lin- cells as well as MNCs. Total CFU (CFU-TOT) number was determined as the sum of granulocyte-erythroid-macrophage-megakaryocyte (CFU-GEMM), granulocyte-macrophage (CFU-GM), erythroid (CFU-E) and burst-forming erythroid (BFU-E) colonies. CFU-TOT counts were 84.5, 80.0, 57.3 and 0.5 per 1000 cells for CD34+, CD133+, Lin- and MNCs, respectively. CD34- and CD133- cell populations have shown very limited colony forming potential in our previous studies with CFU-TOT counts of 0.1 and 0.6 per 1000 cells, respectively.
CFU-GM colonies (mean 53.2%) and CFU-GEMM colonies (mean 32.6%) were the most common colony types formed by of CB-derived HSCs. The proportion of different colony types for CD34+, CD133+, Lin- and MNCs is shown in Table . BFU-Es represented a mean of 12.9% of the colony content of HSCs. However, the Lin- cell fraction formed a surprisingly large number of BFU-Es (27.3%), probably due to the inefficient removal of erythroid progenitors during the depletion procedure. Very little CFU-E colonies were observed (mean 1.3%), MNC population being the most efficient in forming them (5.5%). The high proportion of BFU-E and CFU-E colonies formed by the MNC population reflects the unusual sedimentation of erythroid progenitors in Ficoll-Paque density gradient. The results show that all the selected HSC populations have substantial clonogenic potential and are highly non-committed. Taken together, the data support the use of traditionally used markers to separate HSC populations until more specific markers are found.
Frequency of different types of CFU colonies within CD34+, CD133+, Lin- and MNC populations.