Materials and methods

Mice

Female BALB/c mice (6-8 weeks old) (in vitro CAFC and in vivo short-term repopulation studies), male B6.SJL-Ptprc^aPep3^b/BoyJ (6-8 weeks old) (CD45.1/Ly5.1) (donor) and male (donor) and female (recipient) C57BL/6J (CD45.2/Ly5.2) mice (6-8 weeks old) (in vitro CAFC and competitive repopulation studies) mice were purchased from Jackson Laboratories (Bar Harbor, ME, USA). Mice were maintained under strict microisolator procedures and provided with acidified water (pH 2.2) and irradiated food ad libitum. Care of animals was in accordance with the University of Nebraska Medical Center Institutional Animal Care and Use Committee guidelines.

Bone marrow preparation

Mice were euthanized by anesthesia overdose and femurs and tibias removed. Femoral and tibial marrow was collected into cold Iscove's modified Dulbecco's medium (IMDM) containing 2% (v/v) fetal bovine serum (FBS) (Summit Biotechnology Inc., Fort Collins, CO, USA) (Hoechst-staining medium) and a single-cell suspension prepared by repeated, gentle aspiration through a 22-G needle. Samples were stored at +4°C until further processed.

Hoechst-33342 staining

To reveal the SP profile, BM cells were stained with Hoechst-33342 as detailed previously.^{2, 27} Cellularity was determined using a hemacytometer and adjusted to produce 10⁶ cell/ml in prewarmed (37°C) Hoechst-staining medium. Hoechst-33342 (Sigma, St Louis, MO, USA) was added to a final concentration of 5 g/ml cell suspension and incubated for 90 min with frequent mixing by inversion. After incubation, cells were rapidly cooled on ice and concentrated to 5–8 10⁶ cell/ml by centrifugation, filtered through a 40 m nylon mesh and aliquoted for flow cytometric analysis and fluorescence-activated cell sorting (FACS).

Fluorescence-activated cell sorting

FACS analysis was performed on a dual laser FACSVantage SE (Becton-Dickinson, San Jose, CA, USA). Hoechst-33342 was excited at 350–360 nm (UV) and emission fluorescence measured at two wavelengths: 670 nm (red) (670LP filter, Omega Optical Inc., Battleboro, VT, USA) and 485 nm (blue) (485DF22 filter, Omega Optical Inc.). Emissions were separated using a 510 nm long pass dichroic mirror (510 DCLP) and data acquired on linear axes. Erythrocytes and debris were excluded by gating on forward and side scatter. Four sort regions were identified (see Figure 1): (a) whole SP (WSP); (b) a region defined as non-SP (NSP); (c) upper SP^* (USP) and (d) lower SP^* (LSP) (^*the WSP profile was divided into three approximately equal fractions (upper, middle and lower) by a proportion of gated events). Cells were sorted into FBS for subsequent culture or Hoechst-staining medium for transplantation.

Figure 1.

Hoechst-33342-staining of BM cells. Flow cytometry reveals the Hoechst-33342-staining profile of BM cells. The figure illustrates the profile obtained when BALB/c BM cells were stained at 10⁶ cells/ml for 90 min at 37°C in IMDM medium containing 2% FBS and 5 g/ml Hoechst-33342. Four cell populations were sorted from Hoechst-33342-stained WBM: (1) WSP; (2) a region defined as NSP; (3) USP; and (4) LSP. Sorted cells were assayed in the in vitro CAFC or in vivo hematologic short-term (BALB/c) and competitive repopulation (C57BL/6) assays.

Full figure and legend (65K)

In vitro CAFC assay

The frequency of CAFC in samples of (a) WBM; (b) WSP; (c) NSP; (d) USP and (e) LSP cells was determined by limiting dilution analysis (LDA), as originally described.^{24, 28} In all, 20 wells were prepared per dilution, nine dilutions per sample (samples a–e) and the experiment repeated three times. Cultures were refed weekly and wells scored 'positive' or 'negative' for the presence of cobblestone areas (CA) using an inverted microscope. The proportion of negative wells at each concentration at each week was used in a Poisson-based LDA calculation to determine CAFC frequency.²⁴

Short-term hematologic reconstitution study

Balb/c mice (8–12 weeks old) received lethal (8.5 Gy) whole-body -irradiation (Picker V90 Cobalt-60 source, Chicago, IL, USA. Dose rate=0.61 Gy/min). Approximately 24 h post-irradiation, mice (five mice/group) were anesthetized (Isoflurane, Abbot Laboratories, North Chicago, IL, USA) and received (a) 10⁶ WBM cells; (b) 3300 WSP cells; (c) 40300 NSP cells; (d) 3200 USP cells or (e) 3200 LSP cells intravenously by retro-orbital injection in 200 l Hoechst-staining medium diluted in sterile saline. Mice were monitored daily for survival. After day 10 post transplant, 20 l of blood was drawn from the retro-orbital plexus of anesthetized mice once, or twice weekly for white blood cell (WBC), red blood cell (RBC) and platelet enumeration. Data from nonirradiated, nontransplanted BALB/c mice provided control data. WBC, RBC and platelet numbers were determined using a Serono-Baker System 9000 Hematology Series Cell Counter (Serono-Baker Diagnostics Inc., Allentown, PA, USA).

Competitive repopulation studies

C57BL/6J (CD45.2/Ly5.2) female mice (8–12 weeks old) received a lethal (9.5 Gy) whole-body -irradiation dose (Picker V90 cobalt-60 source, Chicago, IL, USA at a dose rate of 0.61 Gy/min). Sorted USP and LSP fractions from CD45.1/Ly5.1 and CD45.2/Ly5.2 marrow were combined such that recipient female mice received an intravenous injection containing a known proportion of either (1) CD45.1/Ly5.1 USP male cells and CD45.2/Ly5.2 LSP male cells or (2) CD45.2/Ly5.2 USP male cells and CD45.1/Ly5.1 LSP male cells. In one experiment, irradiated female mice received a known number of CD45.1/Ly5.1 USP or LSP male cells combined with a known number of CD45.2/Ly5.2 WBM male cells. The proportion of SP cells in the CD45.2/Ly5.2 WBM was determined by FACS, such that the recipient mice received an equivalent number of CD45.1/Ly5.1 USP or LSP cells and CD45.2/Ly5.2 WSP cells in the WBM transplanted. In another group, irradiated recipient female mice received 100% CD45.1/Ly5.1, male WBM cells to measure levels of endogenous (CD45.2/Ly5.2) reconstitution in the recipients.

Mice were monitored daily for survival. At weekly intervals (after day 10 post transplantation), 20 l of blood was drawn from the retro-orbital plexus of anesthetized mice and WBC, RBC and platelet numbers were determined using a Serono-Baker System 9000 Hematology Series Cell Counter (Serono-Baker Diagnostics Inc., Allentown, PA, USA). Once mice achieved hematologic reconstitution (levels of WBC, RBC and platelets similar to those observed in nonirradiated, nontransplanted mice), they were rested until 3 months post transplantation at which time short-term competitive repopulation was measured and at 5 and 9 months post transplantation at which time longer-term competitive repopulation was measured. At 3, 5 and 9 months post transplantation, approximately 200 l of blood was drawn from the retro-orbital plexus of each mouse for the measurement of CD45.1 (Ly5.1) and CD45.2 (Ly5.2) chimerism within the WBC compartment by flow cytometry (see below). At the 9 month time point, mice were euthanized and blood, femoral BM and spleen tissue removed from each animal. Single-cell suspensions of femoral BM and spleen were prepared in cold IMDM by gentle repeated aspiration through a syringe fitted with a 22 G needle for the measurement of CD45.1 (Ly5.1) and CD45.2 (Ly5.2) chimerism within each tissue by flow cytometry (see below).

Secondary BM transplantation

Three million BM cells from different groups of the chimeric mice established >9 months after receipt of various mixtures of USP and LSP BM cells (see above) were transplanted intravenously into 8 to 12-week-old, female, C57BL/6J (CD45.2/Ly5.2) (n=4) secondary recipients approximately 24 h after 9.5 Gy whole-body -irradiation (Picker V90 cobalt-60 source, Chicago, IL, USA at a dose rate of 0.61 Gy/min). After 7.5 months, WBC and RBC cellularities and platelet numbers and BM and spleen cellularities of these secondary recipients were measured using an H-2000 Hematology Analyzer (Careside, Culver City, CA, USA) and CD45.1/CD45.2 chimerism in the blood, BM and spleen was measured by flow cytometry (see below).

Flow cytometric analysis of CD45.1 (Ly5.1)/CD45.2 (Ly5.2) chimerism

Hematopoietic reconstitution was considered to have occurred once the recipient mice achieved stable hematologic reconstitution (WBC, RBC and platelet numbers similar to those in control animals). Short- and long-term competitive repopulation was investigated by measuring the CD45.1/CD45.2 chimerism in the WBC compartment of the blood (and, at necropsy, in WBC, BM and spleen tissue) of the recipient mice. In the blood, the proportion of CD45.1/CD45.2 cells was measured by drawing 200 l of blood from the retro-orbital plexus of anesthetized mice. Blood was drawn into heparinized tubes and RBC removed by hypotonic shock. WBC (or at terminal assay, WBC, BM and spleen cells) were washed in Dulbecco's phosphate-buffered saline (DPBS, Gibco) containing 2% (v/v) FBS. Cells were pelleted by centrifugation and incubated with rat anti-mouse CD16/CD32 monoclonal antibody (Fc Block, BD Biosciences Pharmingen) according to the manufacturer's instructions. The relative proportion of CD45.1/Ly5.1 and CD45.2/Ly5.2 WBC (or at terminal assay, WBC, BM and spleen cells) was determined by the use of a phycoerythrin (PE)-conjugated mouse anti-mouse CD45.1 monoclonal antibody (BD Biosciences Pharmingen) and a fluorescein isothiocyanate (FITC)-conjugated mouse anti-mouse CD45.2 monoclonal antibody (BD Biosciences Pharmingen) used according to the manufacturer's instructions. To analyze specifically chimerism in the granulocyte and lymphocyte compartments, samples were also stained with allophycocyanin (APC)-conjugated rat anti-mouse Ly-6C (Gr-1) monoclonal antibody (BD Biosciences Pharmingen) and PE-CY5-conjugated hamster anti-mouse CD3e monoclonal antibody (BD Biosciences Pharmingen).

Cells were stained in subdued light at +4°C for 30 min. Cells were washed twice in DPBS–2%FBS to remove unbound antibodies and fixed in a 1% paraformaldehyde solution. Samples were stored at +4°C in darkness until analyzed. Data were acquired on a Becton Dickinson FACSCalibur using a 40-mW argon laser tuned to 488 nm and a 100-mW krypton laser tuned to 647 nm. Data acquisition and analysis was performed using Becton Dickinson CELLQuest software. Forward and side scatter were collected on a linear scale, while PE, FITC, APC and PE-CY5 signals were collected on a four-decade log scale. Overlaps of emission spectra were electronically compensated. Using the threshold on forward scatter to eliminate debris, >50 000 events were collected from each sample. During analysis, a gate was set on the dot plot of forward and side scatter to include all nucleated cells and to exclude erythrocytes and cell debris. PE, FITC, APC and PE-CY5 fluorescence revealed the proportion of cells staining positively for each phenotype. A quadrant was applied to acquired events to determine the proportion of WBC (and BM and spleen cells at terminal assay), granulocytes and lymphocytes that were CD45.1 (Ly5.1) or CD45.2 (Ly5.2) positive (Figure 2).

Figure 2.

Example of FACS analysis. Blood cells (and at terminal assay, BM and spleen cells) were stained with a cocktail of PE-antiCD45.1, FITC-antiCD45.2, APC-antiGr1 and PE-CY5-antiCD3 to reveal chimerism. An example of FACS analysis (CD45.1 (y-axis) vs CD45.2 (x-axis)) of BM from a CD45.2 recipient mouse 9 months after 9.5 Gy whole-body -irradiation and transplantation with 2400 CD45.1 LSP cells and 3 10⁶ CD45.2 WBM cells is shown.

Full figure and legend (44K)

Controls included the analysis of WBC (and latterly BM and spleen) from (1) nonirradiated, nontransplanted CD45.2 (Ly5.2) female mice, to determine levels of CD45.1 (Ly-5.1)/CD45.2 (Ly5.2) antibody crossreactivity and (2) lethally irradiated CD45.2 (Ly5.2) female mice transplanted with 100% CD45.1 (Ly5.1) male marrow, to determine the levels of endogenous CD45.2 (Ly5.2) female repopulation.

Statistical analyses

The mean and standard error (s.e.m.) for the frequency of CAFC in each sample (WBM, WSP, NSP, USP or LSP) was calculated from the three replicate experiments and data compared using the Student's t-test with the significance set at P0.05 (Microsoft Excel, Microsoft Corp.). The means.e.m. for WBC, RBC and platelet numbers for each group (receiving WBM, WSP, NSP, USP or LSP cells) following irradiation was calculated. Data from FACS analyses of control samples (measuring levels of CD45.1 (Ly5.1)/CD45.2 (Ly5.2)) antibody crossreactivity and levels of endogenous CD45.2 (Ly5.2) repopulation in lethally irradiated CD45.2 (Ly5.2) mice receiving 100% CD45.1 marrow were subtracted from experimental data. Means were generated for each group of mice (%CD45.1/%CD45.2 WBC, granulocytes (Gr-1⁺) and lymphocytes (CD3⁺)) and data compared using the Student's t-test assuming significance at P0.05 (Microsoft Excel, Microsoft Corp.)

Results

Frequency of WSP in WBM samples

The proportion of WSP in the BALB/c WBM samples was 0.320.12% (mean s.e.m., n=5 individual experiments with each data point from a pool of BM generated from three to four mice). The proportion of WSP in WBM from female C57BL/6J mice was 0.120.03% (meanSEM, n=3) and from male C57BL/6J mice was 0.080.01% (means.e.m., n=4). Combining the C57BL/6J data (female and male), the proportion of WSP in WBM was 0.100.01% (means.e.m., n=7). The proportion of WSP in WBM from male B6.SJL-Ptprc^aPep3^b/BoyJ (CD45.1/Ly5.1) mice (used in the in vivo competitive repopulation studies) was 0.060.01% (means.e.m., n=3). The C57BL/6 data are consistent with those reported by other investigators.^{2, 6, 20, 21, 29}

CAFC frequency in WBM and WSP, NSP, USP and LSP fractions of WBM from BALB/c mice

The change in CAFC frequencies (per 10⁶ cells plated) with time in the WBM, WSP and NSP fractions is shown in Figure 3a (each data point is the means.e.m. of three individual experiments). Maximum CAFC frequencies were observed at 3 weeks of culture for WBM, WSP and NSP. The frequency of week 3 CAFC (CAFC wk3) in WBM was approximately 12 10³/10⁶ WBM cells plated. By comparison, the CAFC wk3 frequency was significantly greater (>18-fold, P=0.0002) in the WSP fraction (approximately 221 10³/10⁶ WSP cells plated) and markedly lower (approximately 1% of WBM) in the NSP fraction (approximately 10²/10⁶ NSP cells plated). The frequency of CAFC decreased with time in culture and by week 6, the CAFC frequency in WBM was approximately 1.1 10³/10⁶ WBM cells plated. This was significantly greater (>32-fold, P=0.041) in the WSP fraction (approximately 37 10³/10⁶ WSP cells plated) and markedly reduced (to <2% of WBM) in the NSP fraction (approximately 18/10⁶ NSP cells plated).

Figure 3.

CAFC frequency in WBM, WSP and NSP cells. The frequency of CAFC was measured weekly over a 10–12 week period in samples of WBM (), WSP () and NSP () in BALB/c (Figure 3 upper panel) and C57BL6 (Figure 3 lower panel) mice. The figure demonstrates that the CAFC activity in samples of WBM is concentrated in the WSP and not in the NSP fraction. (Data are presented as the mean of three separate experimentss.e.m. for each strain. ^*significantly different (P0.05) from WBM.)

Full figure and legend (36K)

The change in CAFC frequency (per 10⁶ cells plated) in the USP and LSP fractions is shown in Figure 4a (each data point is the means.e.m. of three individual experiments). Maximum CAFC frequencies were observed at week 3 for the USP fraction (CAFC wk3: approximately 127 10³/10⁶ USP cells plated) and week 4 for the LSP fractions (CAFC wk4: approximately 272 10³/10⁶ LSP cells plated), respectively. As observed with the WBM, WSP and NSP fractions, the frequency of CAFC reduced with time in culture. The CAFC wk6 frequency in the USP fraction was approximately 1.8 10³/10⁶ USP cells plated and was significantly greater (>60-fold) in the LSP fraction (approximately 113 10³/10⁶ LSP cells plated, P=0.00001).

Figure 4.

CAFC frequency in USP and LSP cells. The frequency of CAFC was measured weekly over a 10–12 week period in samples of USP () and LSP () in BALB/c (Figure 4, upper panel) and C57BL6 (Figure 4, lower panel) mice. The figure demonstrates that CAFC present in the LSP fraction persist significantly longer in culture than do CAFC present in the USP fraction. The persistence of CAFC in culture correlates with their primitiveness, suggesting that CAFC derived from the LSP fraction are of a higher quality than those derived from the USP fraction. (Data are presented as the mean of three separate experimentss.e.m. for each strain). ^*significantly different (P0.05) from USP.)

Full figure and legend (29K)

CAFC frequency in WBM and WSP, NSP, USP and LSP fractions of WBM from C57BL/6 mice

The change in CAFC frequencies (per 10⁶ cells plated) with time in the WBM, WSP and NSP fractions is shown in Figure 3b (each data point is the means.e.m. of three individual experiments). The frequency of CAFC in WBM shows a steady reduction with time from a maximum of approximately 2 10³ at week 1 (CAFC wk1). This pattern is consistent with that reported by Parmar et al,²¹ who similarly observed a maximum CAFC frequency in male C57BL/6J WBM at week 1 (approximately 10³/10⁶ WBM cells plated). The frequency of CAFC in the NSP fraction was also maximal at week 1, although at a frequency of approximately 160/10⁶ NSP cells plated, it was only 8% of the CAFC wk1 frequency observed in WBM. The frequency of CAFC in the WSP fraction of WBM reached a maximum at week 3 (approximately 95 10³/10⁶ WSP cells plated) and was approximately 170-fold that observed in WBM and approximately 1700-fold that observed in the NSP fraction. These data indicate that the CAFC activity present in WBM is concentrated in the WSP fraction, rather than the NSP fraction. Further, since CAFC present at week 3 are considered to represent more primitive hematopoietic progenitors than those at weeks 1 or 2, these data suggest that more of the more primitive CAFC are concentrated within the WSP fraction, as compared to the WBM or NSP fractions. By week 10 of culture, while CAFC frequencies in the WBM and the NSP fraction were <5 CAFC/10⁶ WBM or NSP cells plated, the frequency of CAFC in the WSP fraction was approximately 10³ CAFC/10⁶ WSP cells plated.

The change in CAFC frequencies (per 10⁶ cells plated) with time in the USP and LSP fractions is shown in Figure 4b. Maximum CAFC frequencies in the USP fraction are observed at week 2 (approximately 31 10³/10⁶ USP cells plated) and decrease rapidly, such that at week 6 of culture, <1% (approximately 160/10⁶ USP cells plated) of the week 2 maximum CAFC frequency remains. In contrast, maximum CAFC frequencies in the LSP fraction are observed at weeks 3–4 of culture (approximately 78 10³ CAFC wk 3/10⁶ LSP cells plated and approximately 81 10³ CAFC wk 4/10⁶ LSP cells plated, respectively) and reduce less rapidly than was observed in the USP fraction, with approximately 30% of the week 4 maximum, persisting in culture at week 6. Indeed, while it was not possible to measure CAFC frequencies in the USP fraction beyond week 7 (<80 CAFC wk 7/10⁶ USP cells plated), it was possible to measure CAFC frequencies in the LSP fraction up to week 10 (>3 10³ CAFC wk 10/10⁶ LSP cells plated). These data demonstrate that the CAFC present in the LSP fraction of the WSP persist longer in culture than do CAFC present in the USP fraction of the WSP and, since persistence in culture is a measure of primitiveness, these data suggest that more primitive CAFC are concentrated in the LSP fraction of the WSP than are concentrated in the USP fraction.

Survival of lethally irradiated BALB/c mice following transplantation

By day 17 post transplant, all mice (n=5) receiving 40 300 NSP cells died due to hematopoietic failure (2 at day 10, 2 at day 13 and 1 at day 17). Two mice receiving 3200 USP cells (days 13 and 56) and one mouse receiving 3200 LSP cells (day 8) also died due to hematopoietic failure. Surviving mice from the USP (n=3) and LSP groups (n=4) and all mice (n=5) in the groups receiving 3300 WSP and 10⁶ WBM cells survived >120 days post transplantation (data not shown).

Hematologic reconstitution of lethally irradiated BALB/c mice following transplantation

Hematologic reconstitution of lethally irradiated (8.5 Gy) recipient BALB/c mice following transplantation of 10⁶ WBM; 3300 WSP cells; 40300 NSP cells; 3200 USP cells or 3200 LSP cells is shown in Figure 5 (WBC – upper panel, RBC – middle panel and platelet numbers – lower panel). For comparison, data were also accrued from control (nonirradiated, nontransplanted) mice. Hematopoietic failure in mice receiving 40300 NSP cells was marked by a significant reduction in WBC, RBC and platelet numbers with time and all mice died by day 17. WBC reconstitution occurred at approximately day 21 for mice receiving 10⁶ WBM cells, day 24 for mice receiving 3300 WSP cells, day 30 for mice receiving 3200 USP cells and day 16 for mice receiving 3200 LSP cells. RBC reconstitution occurred at approximately day 25 for mice receiving 3300 WSP cells, or 3200 USP cells, and day 15 for mice receiving 3200 LSP cells. RBC cellularity remained within control values following transplantation of 10⁶ WBM cells. Platelet reconstitution occurred at approximately day 29 for mice receiving 10⁶ WBM cells, day 27 for mice receiving 3300 WSP cells, day 37 for mice receiving 3200 USP cells and day 16 for mice receiving 3200 LSP cells.

Figure 5.

Hematologic reconstitution in lethally irradiated and transplanted mice. Hematologic reconstitution (upper panel – WBC, middle panel – RBC and lower panel – platelet) was followed in lethally irradiated (8.5 Gy) BALB/c mice transplanted with (1) 10⁶ WBM cells; (2) 3300 WSP cells; (3) 40 300 NSP cells; (4) 3200 USP cells or (5) 3200 LSP cells. Hematologic data (means.e.m.) from nontransplanted, nonirradiated, control mice are indicated (shaded box). Mice receiving (1) 10⁶ WBM cells () achieved WBC and platelet reconstitution at days 21 and 29, respectively; (2) 3300 WSP cells () achieved WBC, RBC and platelet reconstitution at days 24, 25 and 27, respectively; (3) 40 300 NSP cells () died within 17 days post transplant due to hematopoietic failure; (4) 3200 USP cells () achieved WBC, RBC and platelet reconstitution at days 30, 25 and 37, respectively, and (5) 3200 LSP cells () achieved WBC, RBC and platelet reconstitution at days 16, 15 and 16, respectively.

Full figure and legend (60K)

Survival of lethally irradiated female C57BL/6J mice following transplantation

With the following exceptions, all animals in the study achieved levels of WBC, RBC and platelets comparable to those found in nonirradiated, nontransplanted controls by 30 days post transplant (data not shown) and subsequently survived >120 days post transplant. A single mouse (excipient control) that received no cells in transplant died 10 days after irradiation due to hematopoietic failure, demonstrating the lethality of the 9.5 Gy irradiation dose. One animal in Group 2 (transplanted with 10⁶ WBM cells) died 14 days after transplantation (cause of death not determined).

Competitive repopulation (USP vs LSP) as measured in the blood at 12 weeks (3 months) post transplantation

The proportion of WBC, Gr-1⁺ cells (granulocytes) and CD3⁺ cells (lymphocytes) revealed as CD45.1⁺ or CD45.2⁺ is shown in Table 1. In control (nonirradiated, nontransplanted) female CD45.2 (Ly5.2) mice (Group 1), 99.50.03% (means.e.m., n=2) WBC, 99.70.1% Gr-1⁺ cells and 98.30.1% CD3⁺ cells were CD45.2⁺. A background of 0.50.03% (WBC), 0.30.1% Gr-1⁺ and 1.70.1% CD3⁺/CD45.1⁺ staining was observed and was subtracted from subsequent analyses. In these mice, WBC were composed of 11.62.9% Gr-1⁺ cells and 28.05.6% CD3⁺ cells. In Group 2 (n=4), CD45.2 female mice were lethally irradiated and transplanted with 100% male CD45.1 marrow cells. These data provide a measure of the levels of endogenous (CD45.2) hematopoietic repopulation following a lethal 9.5 Gy total body irradiation dose 12 weeks following transplantation. Assay of CD45.1/CD45.2 chimerism in the blood at this time revealed that although 100% CD45.1 marrow was transplanted, 5.80.4% WBC, 3.90.6% Gr-1⁺ cells and 12.70.5% CD3⁺ cells expressed the CD45.2 antigen and were therefore of recipient, not donor, origin. These data, indicating endogenous hematopoietic reconstitution in the irradiated recipients, were subtracted from subsequent analyses. In these mice, WBC were composed of 18.01.8% Gr-1⁺ cells and 29.11.4% CD3⁺ cells. These differential data were not significantly different from those of control (nonirradiated, nontransplanted) mice. With the above controls included in analyses, CD45.1 vs CD45.2 expression was subsequently used to track the hematopoietic reconstitution of the lethally irradiated mice by transplanted BM composed of known proportions of USP, LSP or WSP.

Table 1 - Chimerism in blood as measured by FACS at 3, 5 and 9 months post transplantation (data presented as CD45.1/CD45.2%).

Full table

At 12 weeks post transplant, lethally irradiated mice receiving BM composed of the following: Group 3 – 88% USP and 12% LSP cells (USP/LSP; 88/12) (n=4, approximately 4 10³ total cells transplanted per mouse) reconstituted approximately (USP/LSP) 21/79 WBC, 6/94 Gr-1⁺ and 31/69 CD3⁺ cells; Group 4 – 43% USP and 57% LSP cells (USP/LSP; 43/57) (n=3, approximately 2x10³ total cells transplanted per mouse) reconstituted approximately (USP/LSP) 13/87 WBC, 4/96 Gr-1⁺ and 20/80 CD3⁺ cells; Group 7 – 52% USP and 48% LSP cells (USP/LSP; 52/48) (n=2, approximately 3.3x10³ total cells transplanted per mouse) reconstituted approximately (USP/LSP) 7/93 WBC, 1/99 Gr-1⁺ and 7/93 CD3⁺ cells; Group 8 – 76% USP and 24% LSP cells (USP/LSP; 76/24) (n=4, approximately 4.7 10³ total cells transplanted per mouse) reconstituted approximately (USP/LSP) 55/45 WBC, 47/53 Gr-1⁺ and 58/42 CD3⁺ cells.

In one series of experiments, 2400 USP (Group 5) or 2400 LSP cells (Group 6) sorted from CD45.1 WBM were competed against 3 10⁶ CD45.2 BM cells containing approximately 2400 WSP cells (WSP frequency in the sample of WBM of approximately 0.08%) for the reconstitution of lethally irradiated CD45.2 female recipients. At 3 months post transplant, lethally irradiated mice receiving BM composed of approximately 2400 USP cells and approximately 2400 WSP cells (contained in 3 10⁶ WBM cells, n=4 – Group 5) reconstituted approximately (USP/WSP) 3/97 WBC, 1/99 Gr-1⁺ and 3/97 CD3⁺ cells. By comparison, 12 weeks post transplant, lethally irradiated mice receiving BM composed of approximately 2400 LSP cells and approximately 2400 WSP cells (contained in 3 10⁶ WBM cells, n=2 – Group 6) reconstituted approximately (LSP/WSP) 43/57 WBC, 35/65 Gr-1⁺ and 51/49 CD3⁺ cells.

Competitive repopulation (USP vs LSP) as measured in the blood at 20 weeks (5 months) post transplantation (Table 1)

In control (nonirradiated, nontransplanted) female CD45.2 (Ly5.2) mice (Group 1), 99.90.04% (means.e.m., n=2) WBC, 98.60.2% Gr-1⁺ cells and 98.00.3% CD3⁺ cells were CD45.2⁺. A background of 0.10.04% (WBC), 1.40.2% Gr-1⁺ and 2.00.3% CD3⁺/CD45.1⁺ staining was observed and was subtracted from subsequent analyses. In these mice, WBC were composed of 9.83.4% Gr-1⁺ cells and 26.23.3% CD3⁺ cells. In Group 2 (n=3) (CD45.2 female mice lethally irradiated and transplanted with 100% male CD45.1 marrow cells), levels of endogenous (female CD45.2) hematopoietic repopulation were followed. The assay of CD45.1/CD45.2 chimerism in the blood at this time revealed that, although 100% CD45.1 marrow was transplanted, 2.20.4% WBC, 2.80.4% Gr-1⁺ cells and 8.50.6% CD3⁺ cells expressed the CD45.2 antigen and were therefore of recipient, not donor, origin and subtracted from subsequent analyses. In Group 2, WBC were composed of 14.60.9% Gr-1⁺ cells and 26.52.0% CD3⁺ cells. These differential data were not significantly different from those from control (nonirradiated, nontransplanted) mice.

At 5 months post transplant, lethally irradiated mice received BM composed of the following: Group 3 – 88% USP and 12% LSP cells reconstituted approximately (USP/LSP) 12/88 WBC, 9/91 Gr-1⁺ and 21/79 CD3⁺ cells; Group 4 – 43% USP and 57% LSP cells reconstituted approximately (USP/LSP) 11/89 WBC, 11/89 Gr-1⁺ and 15/85 CD3⁺ cells; Group 7 – 52% USP and 48% LSP cells reconstituted approximately (USP/LSP) 4/96 WBC, 0/100 Gr-1⁺ and 3/97 CD3⁺ cells; Group 8 – 76% USP and 24% LSP cells reconstituted approximately (USP/LSP) 62/38 WBC, 43/57 Gr-1⁺ and 61/39 CD3⁺ cells.

In Group 5, 2400 USP and Group 6, 2400 LSP cells were sorted from CD45.1 WBM and competed against 3x10⁶ CD45.2 BM cells containing approximately 2400 WSP cells for the reconstitution of lethally irradiated CD45.2 female recipients. At 20 weeks post transplant, lethally irradiated mice receiving 2400 USP and 2400 WSP cells (Group 5) reconstituted (USP/WSP) 2/98 WBC, 0/100 Gr-1⁺ and 2/98 CD3⁺ cells. By comparison, 20 weeks post transplant, lethally irradiated mice receiving 2400 LSP cells and 2400 WSP cells (Group 6) reconstituted (LSP/WSP) 42/58 WBC, 22/78 Gr-1⁺ and 51/49 CD3⁺ cells.

Competitive repopulation (USP vs LSP) as measured in the blood (Table 1), BM and spleen (Table 2) at 9 months post transplantation

Blood

In control (nonirradiated, nontransplanted) female CD45.2 (Ly5.2) mice (Group 1), 98.70.39% (means.e.m., n=2) WBC, 98.30.6% Gr-1⁺ cells and 95.21.5% CD3⁺ cells were CD45.2⁺. A background of 1.30.39% (WBC), 1.70.6% Gr-1⁺ and 4.81.5% CD3⁺/CD45.1⁺ staining was observed and was subtracted from subsequent analyses. In these mice, WBC were composed of 24.50.3% Gr-1⁺ cells and 24.10.6% CD3⁺ cells. In Group 2, providing a measure of the levels of endogenous (CD45.2) hematopoietic repopulation following a lethal 9.5 Gy total body irradiation dose, 3.10.7% WBC, 5.11.0% Gr-1⁺ cells and 8.51.0% CD3⁺ cells expressed the CD45.2 antigen and were therefore of recipient, not donor, origin. In these mice, WBC were composed of 33.12.5% Gr-1⁺ cells and 23.51.2% CD3⁺ cells. These differential data were not significantly different from those from control (nonirradiated, nontransplanted) mice.

Chimerism 9 months post transplantation in each group was as follows: Group 3 – receiving 88% USP and 12% LSP cells (USP/LSP) 10/90 WBC, 6/94 Gr-1⁺ and 13/87 CD3⁺ cells; Group 4 – 43% USP and 57% LSP cells (USP/LSP) 15/85 WBC, 20/80 Gr-1⁺ and 15/85 CD3⁺ cells; Group 7 – 52% USP and 48% LSP cells (USP/LSP) 2/98 WBC, 1/99 Gr-1⁺ and 4/96 CD3⁺ cells; Group 8 – 76% USP and 24% LSP cells (USP/LSP) 60/40 WBC, 52/48 Gr-1⁺ and 57/43 CD3⁺ cells.

In Group 5, 2400 USP or in Group 6, 2400 LSP cells were sorted from CD45.1 WBM and competed against 3 10⁶ CD45.2 BM cells containing approximately 2400 WSP cells. At 9 months post transplant, Group 5 reconstituted (USP/WSP) 0/100 WBC, 0/100 Gr-1⁺ and 0/100 CD3⁺ cells and Group 6) reconstituted (LSP/WSP) 36/64 WBC, 26/74 Gr-1⁺ and 44/56 CD3⁺ cells.

Bone marrow

In control (nonirradiated, nontransplanted) female CD45.2 (Ly5.2) mice (Group 1), 99.30.37% (means.e.m., n=2) BM cells, 99.50.2% Gr-1⁺ cells and 92.12.3% CD3⁺ cells were CD45.2⁺. A background of 0.70.37% BM cells, 0.50.2% Gr-1⁺ and 7.92.3% CD3⁺/CD45.1⁺ staining was observed. In these mice, BM cells were composed of 67.31.1% Gr-1⁺ cells and 8.41.7% CD3⁺ cells. Femur cellularity was 18.94.7x10⁶ cells (means.e.m., n=2). In Group 2 (providing a measure of the levels of endogenous (CD45.2), hematopoietic repopulation), 2.60.6% BM cells, 1.80.3% Gr-1⁺ cells and 22.91.6% CD3⁺ cells expressed the CD45.2 antigen and were therefore of recipient, not donor, origin. In Group 2, BM cells were composed of 66.40.6% Gr-1⁺ cells and 8.40.5% CD3⁺ cells. Femur cellularity was 19.61.1 10⁶ cells (means.e.m., n=3). These differential and cellularity data were not significantly different from those from control (nonirradiated, nontransplanted) mice.

Chimerism 9 months post-transplant in lethally irradiated mice receiving BM composed of the following: Group 3 – 88% USP and 12% LSP cells (USP/LSP) was 6/94 BM cells, 5/95 Gr-1⁺ and 16/84 CD3⁺ cells. Femur cellularity was 21.10.8 10⁶ cells (means.e.m., n=3); Group 4 – 43% USP and 57% LSP cells (USP/LSP) was 18/82 BM cells, 20/80 Gr-1⁺ and 7/93 CD3⁺ cells. Femur cellularity was 29.52.4 10⁶ cells (means.e.m., n=3); Group 7 – 52% USP and 48% LSP cells (USP/LSP) was 0/100 BM cells, 0/100 Gr-1⁺ and 0/100 CD3⁺ cells. Femur cellularity was 31.16.7 10⁶ cells (means.e.m., n=2); Group 8 – 76% USP and 24% LSP cells (USP/LSP) was 60/40 BM cells, 59/41 Gr-1⁺ and 52/48 CD3⁺ cells. Femur cellularity was 23.21.7 10⁶ cells (means.e.m., n=4).

In Group 5, 2400 USP or in Group 6, 2400 LSP cells were sorted from CD45.1 WBM and competed against 2400 WSP cells. At 9 months post transplant, Group 5 reconstituted (USP/WSP) 1/99 BM cells, 1/99 Gr-1⁺ and 0/100 CD3⁺ cells. Femur cellularity was 23.11.1 10⁶ cells (means.e.m., n=4). By comparison, Group 6 reconstituted (LSP/WSP) 17/83 BM cells, 10/90 Gr-1⁺ and 39/61 CD3⁺ cells. Femur cellularity was 19.20.0 10⁶ cells (means.e.m., n=2).

Spleen

In control (nonirradiated, nontransplanted) female CD45.2 (Ly5.2) mice (Group 1), 98.90.08% (means.e.m., n=2) spleen cells, 96.90.5% Gr-1⁺ cells and 97.00.2% CD3⁺ cells were CD45.2⁺. A background of 1.10.08% spleen cells, 3.10.5% Gr-1⁺ and 3.00.2% CD3⁺/CD45.1⁺ staining was observed and was subtracted from subsequent analyses. In these mice, spleen cells were composed of 9.20.2% Gr-1⁺ cells and 29.52.4% CD3⁺ cells. In Group 2 (n=3), providing a measure of the levels of endogenous (CD45.2) hematopoietic repopulation following transplantation assay of CD45.1/CD45.2 chimerism in the spleen revealed that 4.71.6% spleen cells, 14.62.4% Gr-1⁺ cells and 9.20.7% CD3⁺ cells expressed the CD45.2 antigen and were therefore of recipient, not donor, origin. These data, indicating endogenous hematopoietic reconstitution in the irradiated recipients, were subtracted from subsequent analyses. In these mice, spleen cells were composed of 11.60.2% Gr-1⁺ cells and 35.61.4% CD3⁺ cells. The percentage of Gr-1⁺ cells was significantly greater (P0.05) than that from control (nonirradiated, nontransplanted) mice. With the above controls included in analyses, CD45.1 and CD45.2 expression were subsequently used to track reconstitution of lethally irradiated mice by BM composed of known proportions of USP, LSP or WSP.

At 9 months post transplant, Group 3 (88% USP and 12% LSP cells) repopulated at (USP/LSP) 18/82 spleen cells, 29/71 Gr-1⁺ and 23/77 CD3⁺ cells; Group 4 (43% USP and 57% LSP cells) (USP/LSP) at 12/88 BM cells, 16/84 Gr-1⁺ and 11/89 CD3⁺ cells; Group 7 (52% USP and 48% LSP cells) (USP/LSP) at 5/95 spleen cells, 6/94 Gr-1⁺ and 7/93 CD3⁺ cells and Group 8 (76% USP and 24% LSP cells) (USP/LSP) at 60/40 spleen cells, 48/52 Gr-1⁺ and 56/44 CD3⁺ cells.

In Group 5, 2400 USP or in Group 6, 2400 LSP cells were sorted from CD45.1 WBM were competed against 2400 WSP cells for the reconstitution of lethally irradiated CD45.2 female recipients. Group 5 reconstituted approximately (USP/WSP) 2/98 BM cells, 4/96 Gr-1⁺ and 5/95 CD3⁺ cells. By comparison, Group 6 reconstituted approximately (LSP/WSP) 41/59 spleen cells, 44/56 Gr-1⁺ and 50/50 CD3⁺ cells.

Cellularities (Table 3) and chimerism (Table 4) measured in BM marrow and spleen of secondary BM recipients, 7.5 months post transplantation

Lethally irradiated secondary recipients (n=4 mice per group) received 3 10⁶ BM cells from different groups of primary recipients (Groups 3, 4, 6–8) 9 months after primary transplantation of known mixtures of LSP and USP. Chimerism (CD45.1/CD45.2) of WBC, granulocytes and lymphocytes in the BM of the primary mice at this 9-month time point is shown in Table 2. After 7.5 months, WBC, RBC, platelet, BM and spleen cellularities of the secondary recipients were determined and compared to control (nonirradiated, nontransplanted) mice (Table 3) and WBC, granulocyte and lymphocyte chimerism (CD45.1/CD45.2) measured in the blood, BM and spleen (Table 4). Generally, the chimerism observed in the spleen and blood closely reflected that observed in the BM. Further, chimerism observed in the granulocyte compartment of the BM, spleen and blood was generally reflected in the lymphocyte compartment.

Table 3 - Blood, BM and spleen cellularities 7.5 months after serial transplantation (means.e.m.).

Full table

Table 4 - Chimerism in blood, BM and spleen as measured by FACS at 7.5 months post serial transplantation (data presented as CD45.1/CD45.2%).

Full table

RBC and platelet cellularities for all groups were not significantly different from control. Significant differences from control cellularities were observed for the WBC cellularity of Group 4 (reduced to 56% of control, P0.05), the femur cellularity of Group 7 (reduced to 74% of control, P0.05) and spleen cellularities of Group 3, 6–8 (reduced to 49, 62, 75 and 74% of control, respectively P0.05).

Primary recipients in Group 3 initially received BM composed of 88% USP/12%LSP. By 9 months, this had established BM chimerism of 6/94 (USP/LSP) and was transplanted into lethally irradiated secondary recipients. After 7.5 months, a chimerism of 3/97 (USP/LSP) was measured in the BM of the secondary recipient. Primary recipients in Group 4 initially received BM composed of 57% LSP/43%USP. By 9 months, this had established a BM chimerism of 82/18 (LSP/USP) and was transplanted into lethally irradiated secondary recipients. After 7.5 months, a chimerism of 91/9 (LSP/USP) was measured in the BM of the secondary recipient. Primary recipients in Group 6 initially received BM composed of 2400 LSP/3x10⁶ WBM. By 9 months, this had established a BM chimerism of 17/83 (LSP/WBM) and was transplanted into lethally irradiated secondary recipients. After 7.5 months, a chimerism of 3/97 (LSP/WBM) was measured in the BM of the secondary recipient. Primary recipients in Group 7 initially received BM composed of 52% USP/48% LSP. By 9 months, this had established a BM chimerism of 0/100 (USP/LSP) and was transplanted into lethally irradiated secondary recipients. After 7.5 months, a chimerism of 1/99 (USP/LSP) was measured in the BM of the secondary recipient. Primary recipients in Group 8 initially received BM composed of 24% LSP/76% USP. By 9 months, this had established BM chimerism of 40/60 (LSP/USP) and was transplanted into lethally irradiated secondary recipients. After 7.5 months, a chimerism of 25/75 (USP/LSP) was measured in the BM of the secondary recipient.

Discussion

While the efflux of Hoechst-33342 dye has been proposed as a phenotypic and functional regulator of hematopoietic,^{1, 2, 3, 4, 5} nonhematopoietic^{8, 9, 10, 11, 12, 13} and tumor stem cells,¹⁴ data presented here and elsewhere provide evidence that there is a qualitative hierarchical organization within the SP isolated from murine BM.^{2, 3, 21, 23} Here, data are presented that demonstrate that SP cells that more strongly efflux the Hoechst-33342 dye (LSP) appear to be more primitive, higher quality stem cells as measured in vitro by the CAFC assay and in vivo by short-term and competitive repopulation studies, when compared to those that less effectively efflux the dye (USP). The proportions of WSP in WBM from 6- to 8- week-old BALB/c (female), C57BL/6J (male and female combined data) and B6.SJL-Ptprc^aPep3^b/BoyJ (male) were 0.32%, 0.1% and 0.06%, respectively. These C57BL/6 data are consistent with and extend those reported by other groups.^{2, 6, 21, 29}

These studies demonstrate that the primitive hematopoietic progenitor cell (CAFC) activity associated with WBM in BALB/c and C57BL6 mice is concentrated within the WSP, rather than the NSP fraction (Figure 3a and b). Further, consistent with these observations, Guo et al.⁶ report a significant increase in colony-forming units, representative of the more mature, committed hematopoietic progenitor cell compartment, in the SP rather than the NSP from C57BL6 BM. We also demonstrate that for both mouse strains assayed (BALB/c and C57BL6), within the WSP, higher quality hematopoietic activity, as defined by the persistence of CA in culture, is concentrated within the LSP rather than the USP (Figure 4a and b).

Although evidence (in vitro and in vivo) is presented that hematopoietic activity contained within the WBM is concentrated within the WSP (rather than the NSP), there is also evidence (from the CAFC assay) for the existence of accessory cells, or rare stromal elements, that are not contained within the SP profile. Data presented here enable a measure of the frequency of CAFC in WBM and the frequency of WSP in WBM to be made. These calculations allow an estimate of the frequency of CAFC in the WSP fraction to be made. However, the measured frequency of CAFC in the WSP fraction was found to be markedly less (greater than 10-fold) than estimated. While this difference may, in part, be a consequence of cell trauma associated with FACS and SP isolation, it may also be a consequence of the exclusion of accessory cells, or rare stromal elements that are required for optimal performance of the CAFC assay. Evidence for such accessory cells, or rare stromal elements, has been reported elsewhere.^{30, 31, 32, 33} Interestingly, Uchida et al²⁹ also report a disparity between calculated long-term culture-initiating cell (LTC-IC) frequencies and estimates of primitive hematopoietic progenitor cell numbers in isolated SP subsets. One possible interpretation of this observation may also be the exclusion of rare accessory cells required for optimal performance of the LTC-IC assay by the SP selection process.

Comparison of the survival and short-term multilineage (erythroid, WBC and platelet) repopulation ability of WBM, WSP and NSP, and USP and LSP cells in lethally irradiated BALB/c recipients provides evidence that LSP cells are of higher quality than WBM, WSP and USP cells (Figure 5a–c). While 100% of mice receiving approximately 40 000 NSP cells died of hematopoietic failure within 17 days, 100% of mice receiving approximately 3000 WSP, or 10⁶ WBM cells (containing approximately 3000 SP) survived >120 days post transplantation, as did 60% of mice receiving approximately 3000 USP and 80% of mice receiving approximately 3000 LSP. Of these recipients, mice receiving approximately 3000 LSP cells demonstrated more rapid hematologic reconstitution (5–8 days earlier for WBC, 13–14 days earlier for platelets and 10 days earlier for RBC) than mice receiving 3000 WSP, or 10⁶ WBM cells (containing approximately 3000 SP) and significantly more rapid hematologic reconstitution (14 days earlier for WBC, 21 days earlier for platelets and 10 days earlier for RBC) than mice receiving approximately 3000 USP cells. The more rapid, multilineage hematologic reconstitution observed following transplantation with LSP cells provides evidence that these cells are functionally of a higher quality than the WSP, WBM or USP cells, reinforcing the observations made in the in vitro CAFC assay. These conclusions are consistent with a previous report²³ detailing that significant donor engraftment was observed in mice receiving 10 'LSP' – comparable ('Tip'-SP) cells, while 1000 'USP' – comparable cells were required before significant donor engraftment was observed. These data may provide an explanation of the more rapid engraftment observed when LSP cells were transplanted into the lethally irradiated recipients by suggesting that LSP cells may have a markedly greater marrow-seeding efficiency than do USP cells.

Competitive repopulation studies provide further evidence of heterogeneity within the SP profile with the observation that LSP cells consistently out-compete USP cells for the repopulation of lethally irradiated recipients as measured in the blood (3, 5 and 9 months) (Table 1) and BM and spleen (9 month) (Table 2). Analysis of the blood demonstrated that, generally, chimerism measured in total WBC, granulocyte and lymphocyte populations at the 3-month time point was stable through 5 and 9 months, with the exception of Group 3 (88% USP and 12% LSP cells) and a blood lymphocyte chimerism of 31/69 at 3 months, 21/79 at 5 months and 13/87 at 9 months. Also, trends in chimerism observed in the blood (WBC, granulocyte and lymphocyte populations) were mirrored in the BM and spleen when assayed at the 9-month time point and, generally, chimerism observed in the granulocyte compartment of the three tissues (blood, BM and spleen) was mirrored in the lymphocyte compartment.

Chimeric BM from the primary recipients 9 months after transplantation was transplanted into lethally irradiated secondary recipients and chimerism measured in the BM, blood and spleen after 7.5 months. The chimerism observed in the blood, BM and spleen of the secondary recipients 7.5 months after transplantation (Table 4) was similar to that measured in the donor BM (from primary recipients 9 months after receipt of initial LSP/USP BM mixture) (Table 2). This suggests that the composition of the stem cell compartment established in the primary recipient is stable, which potentially implicates microenvironmental regulation. Furthermore, since this composition is also stably transferred to secondary recipients, and considering that it was established with as few as 2400 LSP cells in the primary recipients, these observations strongly suggest that LSP both self-renew as well as produce differentiated progeny, consistent with a minimal definition of stem cells.³⁴ These conclusions are supported by the findings of Matsuzaki et al,²³ who also report that 'LSP' equivalent 'Tip'-SP cells have a high marrow-seeding efficiency in addition to long-term, multilineage repopulation activity.

In conclusion, consistent with and extending previous reports,^{2, 3, 21, 23} these data (in vitro CAFC and in vivo hematopoietic reconstitution assays) provide evidence that the Hoechst-33342 SP is heterogeneous and that a qualitative hierarchy exists within the SP profile, with more primitive, potentially higher quality hematopoietic progenitors, concentrated in the population of cells with the greatest Hoechst dye efflux. This qualitative measure of primitive hematopoietic progenitors may permit characterization of genetic variations in the quality of stem cells as well as help to define their stem cell-specific gene expression.^{35, 36} Further, this information should permit the assessment of changes in the hematopoietic system during, for example, therapeutic or transplantation strategies including cord blood, or during aging, and possibly allow the characterization, isolation and targeting of primitive hematopoietic progenitors for gene-marking studies or gene therapy.

References

References

Acknowledgements

We gratefully acknowledge Dr John D Jackson, Department of Pathology and Microbiology, UNMC, for his advice with the preparation of this paper; Linda Wilkie, Cell Processing Facility, UNMC, for her assistance and advice with FACS analysis, Craig Boyer, Department of Genetics, Cell Biology and Anatomy, UNMC, for his technical assistance and Dr Elizabeth J Shpall, Department of Blood and Bone Marrow Transplantation, The University of Texas MD Anderson Cancer Center, Houston, Texas, for her encouragement and support. Research supported by the Dean's Office of the College of Medicine, UNMC and a Nebraska Research Initiative award.