The Fluctuating Phenotype of the Lymphohematopoietic Stem Cell with Cell Cycle Transit

BALB/c male and female mice (Taconic Farms Inc., Germantown, NY) and C57BL6 (Ly5.2) and B6SJL (Ly5.1/Ly5.2) mice (The Jackson Laboratory, Bar Harbor, ME) were maintained under virus-free conditions. Mice were purchased at either 16–18 or 18–22 g (age 6–8 wk), were given food and acidified water ad libitum, and were housed for a minimum of 1 wk before experimental use. All experiments were approved by the University of Massachusetts Institutional Animal Care and Use Committee under docket numbers A707 and A729.

Bone marrow was flushed from bones of 6–8-wk-old mice in HBSS from two tibiae and two femurs. Bone marrow cells were cultured in nonadherent Teflon flasks at densities ranging from 1 to 5 × 10⁶ nucleated cells/ml (see Table 1) in DMEM low glucose, supplemented with 15% heat-inactivated FCS with cytokines as follows: recombinant mouse (rm) steel factor at 50 ng/ml, rmIL-3 at 50 U/ml, rmIL-6 at 50 U/ml, and recombinant human IL-11 at 50 ng/ml at 37°C in a fully humidified atmosphere of 5% CO₂ in air. Cells were harvested at different times in culture by vigorous pipetting with ice-cold PBS. In one experiment, conditions were altered, and cells were cultured under serum-free conditions with IL-3, IL-6, IL-11, and steel factor.

Female mice were exposed to 650–1,000 cGy (the latter split 500 cGy 2 h apart) and then injected with 50% cytokine-cultured male BALB/c cells (corresponding to the equivalent starting cell culture number) and 50% fresh noncultured female cells. The percentage of male cells in host marrow, spleen, or thymus was determined at 8, 15, or 24 wk after transplantation using Southern blot analysis with a Y chromosome–specific probe (PY2) as described previously (5, 6). In brief, genomic DNA was digested with Dra-1, separated by electrophoresis in 0.8% agar, and transferred to nylon membranes (Zetaprobe; Bio-Rad Laboratories, Richmond, CA): ³²P-labeled PY2 was hybridized to the blots for 18–22 h, and autoradiography was performed using x-ray film (XRP; Eastman Kodak Co., Rochester, NY). Quantitation was determined on a PhosphorImager (Bio-Rad Laboratories). Percent male chimerism in female hosts was calculated as the proportion of male signal in transplanted mice to the signal in male marrow. Signal from normal female marrow DNA was the negative background and was subtracted from the male signal. PY2 values were normalized for loading variance by reprobing with an IL-3 cDNA.

Cells were harvested and cultured as described above. B6SJL recipients were exposed to 1,000 cGy of radiation (500 cGy given 2 h apart) from a cesium source irradiator (Gammacell 40; Nordion International, Kanata, Ontario, Canada). After 2 h, the mice were injected intravenously with a mixture of 50% cytokine-cultured B6SJL cells (equivalent to 5 × 10⁶ starting cells harvested after the appropriate hour of incubation with cytokines) and 50% noncultured C57BL/6J cells (5 × 10⁶ cells). Volume for injection was 0.5 ml per recipient. The percentage of bone marrow cells expressing Ly5.1 was determined 18 wk after transplant by flow cytometric analysis. 10⁶ suspended cells were incubated for 25 min with anti-CD45.1 FITC-conjugated antibody (PharMingen, San Diego, CA) at a total volume per tube of 0.2 ml (0.12 μg/ml in HBSS⁺ 1% FCS) at 4°C. The supernatant was aspirated, and cells were washed twice with cold HBSS and preserved with 2% paraformaldehyde. The cells were subsequently analyzed with a fluorescence-activated cell sorter (MoFlo; Cytomation, Inc., Fort Collins, CO).

Cultured male cells were competed with noncultured female marrow cells (harvested separately for each time point) in irradiated female hosts, and engraftment of the male cells in marrow was evaluated at 8 wk or 6 mo after marrow infusion (Fig. 1, A and B, and Tables 1 and 2). These studies showed a loss of engraftment over the 28– 36-h interval, a later restoration of engraftment, and a subsequent loss of engraftment again at 44–48 h. A total of five experiments was carried out evaluating cultured cells and analyzing engraftment at 8 wk after infusion and two separate experiments for the analysis at 6 mo. Tables 1 and 2 present the actual engraftment percentages seen in individual experiments, and Fig. 2 illustrates the mean time of first nadir and time of first recovery superimposed on the previously ascertained cell cycle status of purified cells from these populations and analyzed at 8 wk. The first nadir occurred on average with cells cultured for 33.2 h, with a range within 1 SD of 30.2–36.2 h. The first nadir fell within the late S/G₂ phase as determined from previous experiments (Fig. 2). The first nadir compared with 0 h showed a significant decrease (Wilcoxon's rank sum test, P <0.001). Similarly, the first recovery from nadir, which occurred on average at 39.6 h, was significant (P <0.001) compared with the first nadir value. There was a secondary nadir seen at the 44–48-h range (P <0.001). In one experiment, there was a second late recovery seen at 60 h. In four of five experiments at 8 wk (Table 1), recovery values after the first nadir exceeded the time 0 value, whereas in the fifth experiment, the recovery value was >90% of the 0-h value.

Similar results were obtained in the 6-mo experiments (Fig. 1,B, and Table 2): the mean first nadir occurred at 32 h of culture (P = 0.001) compared with 0 h and the first recovery at a mean of 42 h (P = 0.001 compared with first nadir). In one experiment using the Ly5.1/Ly5.2 transplant model and assessing engraftment at 15 wk, there was a nadir, recovery, nadir, recovery pattern, albeit with somewhat different kinetics (Table 2).

Spleen engraftment was analyzed from BALB/c marrow cells cultured for 32 or 40 h and compared with input: nadir (32 h) and recovery (40 h) values paralleled those seen with marrow, both at 8 and 24 wk after transplantation (data not shown), indicating the multilineage nature of engraftment from cytokine-cultured cells.

A striking aspect of these studies was the dramatic shifts in engraftment occurring over 2–4-h time intervals. An alteration in engraftment was felt to be biologically significant if it represented a relative difference more than two times the coefficient of variation (SD divided by the mean). Considering the experiments presented in Tables 1 and 2 (8, 15, and 24 wk after cell infusion), a total of 16 and 8 significant alterations in engraftment were observed at 2-and 4-h intervals, respectively.

Cells cultured under different conditions also evidenced these engraftment shifts during cytokine-stimulated culture. When cells were cultured serum-free with IL-3, IL-6, IL-11, and steel factor, there was a drop in engraftment at 24 h, but thereafter the same type of fluctuations was seen with time in culture (Fig. 3), along with a late recovery at 56 and 80 h.

Previous data have indicated an expansion of progenitor cells in this cytokine combination (5, 6), and other studies have shown progressive expansions of surrogate progenitor stem cells in cytokine-stimulated cultures (8, 9). However, the present data indicate that at least during the first several cell cycles, there are dramatic fluctuations in engraftment phenotype, which are reversible and thus presumably not tied to irreversible differentiation steps.

The mechanism underlying this phenotypic engraftment plasticity may relate to alterations in stem cell “stickiness” and “trafficking.” Separate studies have shown marked fluctuations of adhesion protein and mRNA levels at different points in cell cycle (10). The integrin pathways are complicated, and homing via these pathways probably represents a multistep process. Other possible mechanisms could relate to differential sensitivity to transplantation- induced apoptosis at certain points in the cell cycle, or to altered responsiveness of transplanted cells at certain points in cycle to host proliferation and differentiation signaling.

These data demonstrate dramatically different outcomes with stem cell engraftment based on the precise time in cell cycle at which the cells are infused. Several gene therapy approaches with retroviral vectors use in vitro cytokine stimulation to induce cycling of hematopoietic stem cells to facilitate incorporation of the vectors into the host genome (11–13). Clearly, the end results of such approaches with infusion of gene-transfected stem cells will depend on whether the hematopoietic stem cells are in an appropriate state to facilitate engraftment, i.e., presumably G₁, or in a state where engraftment is markedly diminished, i.e., late S/G₂. Similarly a relatively large number of investigators have developed approaches to expand both human (14–17) and murine (8, 9) hematopoietic stem cells. These approaches usually use in vitro cytokine stimulation of hematopoietic stem cell populations with subsequent engraftment. Our studies show that a knowledge of the cell cycle status of these cells will be critical to ensure adequate clinical results with expanded stem cells.

We thank our laboratory colleagues for help with mouse work, especially Mark Dooner, Brian Benoit, Lizhen Pang, Allen Stering, Bernice Fraioli, and Caron Engstrom.