SEARCH

SEARCH BY CITATION

Keywords:

  • fetal;
  • migration;
  • endothelium;
  • adhesion

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Summary. Fetal haemopoietic cells continually circulate and migrate into tissues, and thus may have specialized homing capabilities. In this study we investigated the in vitro features of haemopoietic cells in fetal blood and liver which are relevant to homing and engraftment. Fetal cells were examined for long-term culture-initiating cell (LTC-IC) and progenitor content, adhesion molecule expression, cell cycle behaviour and transendothelial migratory activity. The LTC-IC content of fetal CD34+ cells is similar to that of CD34+ cells from cord and adult mobilized blood. In contrast to adult and cord blood CD34+ cells, fetal CD34+ cells were actively cycling (11·0 ± 1·7% and 28 ± 1·1% of fetal blood and liver CD34+ cells, respectively, in S+G2M, P < 0·001, compared with cord and adult cells). The striking finding was that fetal haemopoietic cells (both LTC-ICs and committed progenitors) displayed significantly higher levels of migration across endothelium (P < 0·05 compared with cord, P < 0·01 compared with adult blood and bone marrow CD34+ cells), which were further increased by chemokines and growth factors. The superior migratory activity of fetal haemopoietic cells may underlie a more efficient homing ability, in keeping with their physiological role.

The use of primitive haemopoietic cells for stem cell transplantation is now an established treatment modality for many malignant and inherited disorders; however, our understanding of the molecular basis for stem cell engraftment remains fragmentary. Infused stem cells leave the circulation by migrating across vascular endothelium; this is an early but critical step in the homing of these cells into extravascular haemopoietic tissue. Tracking of labelled stem cells suggests that this occurs within a few hours of transplantation (Hendrikx et al, 1996); thus, cells which fail to transmigrate successfully are eliminated, probably through uptake by the reticuloendothelial system. We have previously shown that the transendothelial migration of human mobilized adult blood progenitor cells is mediated by adhesion molecules such as CD18 and platelet endothelial cell adhesion molecule (PECAM)-1, and is regulated differently from the transmigration of mature inflammatory cells (Yong et al, 1998). Our studies also indicate that, while adult progenitor cells require growth factor stimulation in order to transmigrate, migration occurs preferentially in cells which are in G0/G1 phase of the cell cycle (Yong et al, 1999a). Thus, cell cycle behaviour can impose constraints on the migratory, and hence homing, ability of haemopoietic cells, with optimal homing occurring in cells residing in G0/G1. This may account for the observation made by Gothot et al (1998) that CD34+ cells engraft best when in G0/G1 phase of the cell cycle, and similar observations that prior exposure to cytokines in vitro reduces the engraftment capability of haemopoietic cells (Peters et al, 1996).

The homing of transplanted cells to haemopoietic tissue and their subsequent engraftment recapitulates the physiological process in fetal life, whereby haemopoiesis is established sequentially and simultaneously in different tissues by the seeding of circulating haemopoietic stem cells. Intriguingly, the cell cycle constraints on engraftment by adult cells may not apply to fetal repopulating cells. Recently, two groups reported preliminary results indicating that fetal cells in S+G2/M are able to engraft, albeit to a lesser degree than their counterparts in G0/G1 (Bhatia et al, 1999; Wilpshaar et al, 1999). Hence, fetal haemopoietic cells may be intrinsically different from their adult counterparts, particularly with respect to engraftment behaviour. Fetal haemopoietic cells, by virtue of their ontogeny and function, represent a potentially valuable source of primitive haemopoietic cells, not only for use in stem cell transplantation and gene therapy, but also for studying in detail the biological processes involved in stem cell homing and engraftment.

The aim of this study was to conduct a detailed analysis of the features and properties of fetal cells which are directly relevant to homing and engraftment. Thus, we have examined the stem cell and progenitor content, adhesion molecule expression, cell cycle profile and transendothelial migratory ability of first trimester fetal blood and liver cells, and compared these with our observations on cord and adult mobilized CD34+ cells.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Haemopoietic cells.  Cord blood was obtained from discarded placentas within minutes of delivery, and centrifuged over Ficoll (Nycomed, Coventry, UK) to obtain low-density mononuclear cells (MNCs). Bone marrow samples (adult bone marrow, ABM) were obtained from normal donors during harvesting of bone marrow. CD34+ cells were selected using anti-CD34 Multisort beads and the VS MACS system (MACS, Miltenyi Biotec, Bisley, Surrey, UK). Mobilized peripheral blood (MPB) cells were obtained from patients with relapsed or resistant lymphoma, or myeloma, undergoing peripheral blood stem cell transplantation (PBSCT), or from normal donors. Single apheresis collections on the ‘Cobe Spectra’ were processed for clinical scale CD34+ cell purification using a Ceprate SC immunoaffinity column or the Miltenyi CliniMACS separation device (Jones et al, 1994). The cell preparations had a median purity of 87% (range 65–98%), as determined by dual immunofluorescence staining with anti-CD45-fluoroscein isothiocyanate (FITC) and anti-CD34-phycoerythrin (PE), and flow cytometry.

Fetal blood samples (gestation ages 7–14 weeks) were collected using ultrasound-guided transabdominal fetal cardiocentesis from anatomically normal fetuses during elective surgical terminations performed for psychosocial reasons under general anaesthesia. Fetal liver samples (gestation ages 7–14 weeks) were collected using dissecting forceps after the operation. Low-density MNCs were obtained from heparinized fetal blood and homogenized liver samples by centrifugation over Ficoll. Cell viability, as assessed by trypan blue exclusion, was consistently > 95%.

Full informed consent was obtained from all donors, and all studies were approved by the University College London Hospitals (UCLH) committee on the Ethics of Human Research.

Endothelial cells.  Human umbilical vein endothelial cells (HUVECs) were harvested from umbilical cords by collagenase treatment [0·1% in phosphate-buffered saline (PBS)], and grown to confluence in fibronectin (2 µg/cm2)-coated Falcon tissue culture flasks in Iscove's modified Dulbecco's medium (IMDM) with 20% fetal calf serum (FCS), endothelial cell growth supplement (ECGS, 25 µg/ml) and heparin (25 U/ml). HUVECs were passaged using trypsin (0·05%) EDTA (0·01%) in PBS, and used in transmigration experiments between the second and fourth passages.

Monoclonal antibodies.  The following directly conjugated antibodies were obtained from Becton Dickinson (Kidlington, UK): mouse antibodies directed against human CD34, CD45, CD11a (LFA-1), CD62L (l-selectin), CD49d (VLA-4). Anti-CXCR4-PE was purchased from Pharmingen (San Diego, USA).

Flow cytometric analysis of fetal cells.  Fetal low-density cells obtained after Ficoll centrifugation contained a variable number of large erythroid cells. Hence, after dual staining with anti-CD45-FITC and anti-CD34-PE, samples were treated with saponin (0·25%)/paraformaldehyde (0·5%) in PBS to lyse erythrocytes. Nucleated normoblasts, however, were retained. Dual immunofluorescence was used to analyse adhesion and chemokine receptor expression on gated CD34+ cells, and negative controls were stained with isotype-matched monoclonal antibodies.

Clonogenic assays in methyl cellulose.  Granulocyte/monocyte colony-forming units (CFU-GM), erythroid burst-forming units (BFU-E) and the more primitive mixed-lineage CFU (CFU-GEMM) were simultaneously evaluated in a methylcellulose-based clonogenic assay medium which consisted of Methocult (Stem Cell Technologies, Vancouver, Canada), with 20% IMDM, and supplemented with interleukin 3 (IL-3) 30 ng/ml, granulocyte–macrophage colony-stimulating factor (GM-CSF) 25 ng/ml, G-CSF 25 ng/ml, stem cell factor (SCF) 10 ng/ml and erythropoietin 2 U/ml. Purified CD34+ cells were set up in quadruplicate at 5·0 × 102/ml, while fetal MNCs were set up at 2–4 × 103/ml. Colonies were counted after 14 d incubation at 37°C in a humidified 5% CO2 atmosphere.

Long-term culture-initiating cell (LTC-IC) assays. These were carried out using the murine stromal cell line, MS-5 (Itoh et al, 1989) (kind gift of Dr Inoue, Kirin Pharmaceutical Research Laboratory, Gunma, Japan), which is grown in [α-minimum essential (α-MEM) medium; Life Technologies, Paisley, UK] with 10% FCS. MS-5 cells were irradiated 1–3 d prior to seeding in 96-well plates. Two to 7 d later, medium was removed and plates were rinsed before the addition of limiting dilutions of fetal MNCs, or selected CD34+ cells from adult MPB cells in LTC-IC medium (75% α-MEM/12·5% horse serum/12·5% FCS and 10−4 mercaptoenthanol). Half-volume media exchanges were performed weekly, and wells scored for cobblestone area-forming cells (CAFCs) after 5 weeks. The medium was then removed from each well, and replaced with methylcellulose-based clonogenic assay medium and colonies scored after a further 2 weeks of incubation. LTC-IC frequencies were calculated using Poisson statistics, and the results obtained from CAFC analysis were similar to those derived from colony counts.

Cell cycle analysis.  Between 2 and 5 × 105 purified CD34+ cells or 5 × 105 fetal low density cells were first stained with anti-CD45-FITC and anti-CD34-PE, then washed and incubated with 0·25% saponin/0·5% paraformaldehyde at 4°C for 10 min before staining with the DNA dye, TOPRO. Analysis was carried out using an Epics-Elite flow cytometer (Coulter Electronics, High Wickham, UK) and gates were set on the CD45+CD34+ low side scatter population.

Transendothelial migration assays.  HUVECs were seeded at 1 × 105/filter onto fibronectin-coated 12 mm Transwell filters (Costar UK, High Wickham, UK) and allowed to grow to confluence over 48–96 h prior to transmigration assays. Confluency of the endothelial monolayers was confirmed by visual inspection using an inverted microscope. The integrity of the monolayers was further confirmed by testing the permeability to 125I-labelled human serum albumin (125I-hsa), as previously described (Yong et al, 1998). Equilibration of 125I-hsa was consistently less than 10% across endothelium-covered Transwell filters over 20 h, while complete equilibration occurred across non-endothelium-covered filters within 4 h.

Confluent endothelial monolayers on Transwell filters were washed three times with warm medium and placed in wells with fresh medium (IMDM/20% FCS). Cell aliquots (3 × 105 purified CD34+ cells, or 5 × 105 fetal MNCs per filter) were placed in the upper chamber. Our initial experiments using this system showed that, while neutrophils and monocytes migrated across endothelium-covered filters in 1–2 h, T cells and CD34+ cells required longer incubation times of 18–24 h for optimal levels of spontaneous migration. All transmigration experiments on haemopoietic progenitor cells were therefore conducted over a standard incubation time of 20 h. At the end of the experiment, transmigrated cells were recovered from the lower compartment and set up in clonogenic assays. Percentage migration of clonogenic cells was calculated from the clonogenic output of equivalent aliquots of the cell suspension originally seeded onto the filters.

In some experiments the effect of chemokines [stromal-derived factor-1 (SDF-1) and macrophage inflammatory protein 3 (MIP-3)] on transmigration was tested. Chemokines at stated concentrations were placed in the lower chamber of Transwells at the start of the transmigration assay. Clonogenic assays on the input cell population were set up with similar concentrations of chemokine as the transmigrated cells would be exposed to in the subsequent methylcellulose culture, to rule out any influence of chemokines on colony formation. The effect of growth factors on transmigration was examined by incubating fetal MNCs or MPB CD34+ cells with IL-3 (12 ng/ml), IL-6 (10 ng/ml) and SCF (10 ng/ml) for 3 d, after which cells were washed prior to being used in the transmigration assay.

For experiments on LTC-IC migration, HUVEC-covered Transwell filters were rinsed and placed with fresh medium into wells of a 12-well plate containing irradiated MS-5 cells, which had been seeded at least 3 d previously. Transendothelial migration assays were carried out with fetal MNCs (5 × 104/filter) or MPB CD34+ cells (104/filter) over 20 h, at the end of which filters were removed and transmigrated cells were cultured on the irradiated stroma for 5 weeks, with weekly half-medium exchanges. CAFCs were scored at the end of the 5-week period, and the number of CAFCs generated by transmigrated LTC-IC compared with the number produced by known numbers of input cells (seeded onto MS-5 cells in adjacent wells of the 12-well plate) to obtain the percentage LTC-IC migration.

Statistics.  Statistical analysis was performed using the StatView programme (Macintosh) and data were compared using a two-tailed Student's t-test. A P-value of < 0·05 was considered significant. Data from limiting dilution experiments were analysed using Poisson statistics and the method of maximum likelihood.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

CD34+ cell, colony-forming cell (CFC) and LTC-IC content of fetal blood and liver

CD34+ cells accounted for 2·3 ± 0·7% of nucleated fetal blood cells (mean ± SE, n = 7) and 2·0 ± 0·4% of nucleated fetal liver cells (n = 8). Approximately two thirds of the nucleated cells from both fetal blood and liver were nucleated red cells which do not express CD45. In fetal blood, the proportion of CD45+ cells co-expressing CD34 was 7·5 ± 1·7% (n = 7). In fetal liver samples from late gestational age (> 12 weeks), the percentage of CD45+ cells co-expressing CD34 was 4·2 ± 2·3% (n = 5), while in samples from earlier gestational age (< 12 weeks) it was 25·7 ± 7·5% (n = 11). All fetal blood samples analysed were from late gestational age. In comparison, the percentage of CD45+ cells co-expressing CD34 in mobilized adult peripheral blood (MPB) was 0·8 ± 0·1% and in cord blood 0·7 ± 0·2% (P < 0·01 for both, compared with either fetal tissue). Additionally, the percentage of CD45+ cells co-expressing CD34 is higher in fetal liver samples from earlier gestational age.

Low-density cells from fetal blood and liver were analysed for their LTC-IC content. Table I shows the LTC-IC content of fetal blood, fetal liver and adult mobilized blood (MPB), normalized for the CD34+ content of each sample tested. The fetal samples examined for LTC-IC content were all from late gestational age (> 12 weeks). There is no significant difference in LTC-IC frequencies between fetal and adult sources of haemopoietic cells. Finally, we analysed the colony-forming cells in these populations. Table II shows the frequencies of myeloid and erythroid colonies in fetal blood and liver, compared with cord and adult blood. Fetal liver low-density cells, irrespective of gestational age, produced more erythroid colonies than adult cells (P < 0·01); consequently, the total CFC output was also higher.

Table I.  Long-term culture-initiating cells (LTC-IC) content of fetal and adult CD34+ cells.
 LTC-IC/104 CD34+ cell
Mean ± SERange
  1. Mean ± SE of six experiments with fetal liver, four experiments with fetal blood and eight experiments with adult mobilized blood samples (MPB).

Fetal blood1454 ± 390590–3335
Fetal liver1015 ± 249385–2178
Adult MPB727 ± 253370–1250
Table II.  Progenitor output of fetal, cord and adult mobilized blood CD34+ cells.
 Progenitors/104 CD34+ cells
Fetal bloodFetal liver (< 12 weeks)Fetal liver (> 12 weeks)Cord bloodAdult MPB
  • *

    P < 0·01 compared with adult MPB.

  • **

    P < 0·05 compared with adult MPB.

  • Mean ± SE of six experiments with fetal blood (all from late gestation), 11 with fetal liver (five from early and six from late gestation), 10 with cord blood and 10 with adult MPB. MPB, mobilized blood samples; BFU-E, erythroid burst-forming units; CFU-GEMM, mixed-lineage colony-forming units; CFU-GM, granulocyte–macrophage colony-forming units; CFC, colony-forming cells.

BFU-E1822 ± 3323270 ± 368*3795 ± 489*1824 ± 607829 ± 210
CFU-GEMM188 ± 38237 ± 89371 ± 101539 ± 110260 ± 51
CFU-GM1064 ± 2571047 ± 3641587 ± 3011805 ± 324869 ± 260
Total CFC3074 ± 6484554 ± 723**5753 ± 805**4168 ± 7071958 ± 510

Adhesion molecule and CXCR4 expression on fetal CD34+ cells: comparison with cord and adult CD34+ cells

We used flow cytometry to examine the expression of adhesion molecules considered to play a role in homing and/or transendothelial migration of CD34+ cells. Fetal CD34+ cells expressed higher levels of very late antigen-4 (VLA-4, 84·3 ± 4·3% positive for fetal blood, and 89·1 ± 3·8% positive for fetal liver) than cord and adult MPB CD34+ cells (67·3 ± 6·1% and 64·3 ± 9·6%, respectively, P < 0·05, Fig 1). Levels of l-selectin were variable on fetal, cord and adult blood CD34+ cells, and no significant differences were demonstrated among them. Fetal CD34+ cells expressed lower levels of lymphocyte function-associated antigen-1 (LFA-1) than cord and adult CD34+ cells (P < 0·05, Fig 1). There was no difference in expression of adhesion molecules between fetal blood and liver CD34+ cells. We were also unable to detect any differences in adhesion molecule expression between fetal liver CD34+ cells from early versus late gestational ages.

image

Figure 1. Surface expression of adhesion molecules and CXCR4 on CD34+ cells from fetal, cord and adult sources. Cells were dual stained with anti-CD34 and monoclonal antibodies to VLA-4, LFA-1, l-selectin or CXCR4 and analysed by flow cytometry. Isotype controls were used to set positive gates for analysis of antigen expression. Mean ± SD of five fetal blood samples (FB), seven fetal liver samples (FL), seven cord blood (CB) samples and six adult mobilized blood samples (MPB). Data are expressed as mean ± SE percentage of cells showing specific staining for the antigen.

Download figure to PowerPoint

SDF-1 induces chemotaxis of CD34+ cells (Aiuti et al, 1997) and has been shown, together with its cognate receptor CXCR4, to play a role in the engraftment of human CD34+ cells in severe combined immunodeficient/non-obese diabetic (SCID/NOD) mice (Peled et al, 1999). We compared surface levels of expression of CXCR4 on fetal, cord and adult CD34+ cells. Levels of CXCR4 expression on fetal blood and liver CD34+ cells were similar and comparable to those on CD34+ cells in MPB (Fig 1). Once again, the fetal blood samples examined were all from late gestational age; however, in the small number of fetal liver samples analysed (three from early and four from late gestational age), there was no discernible correlation with gestation. In contrast, cord blood CD34+ cells expressed markedly higher levels of CXCR4, both in terms of the percentage of positive cells, as well as the mean cell fluorescence for the whole population (data not shown).

Cell cycle status of fetal CD34+ cells

Gated CD34+ cells from fetal, cord and adult sources were analysed for cell cycle profiles, and Fig 2A shows representative DNA histograms for each cell source. While the vast majority of CD34+ cells in adult mobilized blood and in cord blood were quiescent (97·7 ± 0·7% and 96·7 ± 0·4%, respectively, in G0/G1 phase of the cell cycle), a substantial proportion of CD34+ cells present in fetal blood and liver were in the proliferative phase of the cell cycle (11·0 ± 1·7% and 28 ± 1·1%, respectively, in S+G2M, P < 0·001 compared with cord or adult cells, Fig 2B). The percentage of fetal liver CD34+ cells in S+G2/M was also significantly higher than that for fetal blood CD34+ cells (P < 0·01). In two cases in which paired fetal blood and liver samples were available for comparison, the percentage of CD34+ cells in S+G2M was 14% and 11·9% for blood and 30% and 33·4%, respectively, for liver. Three adult bone marrow (ABM) samples were analysed for cell cycle status: the percentage of CD34+ cells in S+G2/M was 4·3%, 8·2% and 7·2% respectively.

image

Figure 2. Cell cycle analysis of CD34+ cells from fetal blood (FB), fetal liver (FL), cord blood (CB) and adult mobilized peripheral blood (MPB). Cells were stained with anti-CD45 and anti-CD34 monoclonal antibodies, permeabilized with saponin/paraformaldehyde as outlined in Materials and methods, and the DNA stained using TOPRO. (A) Representative DNA histograms from each cell source. Percentages of cells in S + G2/M are indicated for each sample. (B) Mean ± SD of the percentage of CD34+ cells in S + G2/M phase of the cell cycle from fetal blood (n = 6), fetal liver (n = 8), cord blood (n = 12) and adult blood (n = 20).

Download figure to PowerPoint

Transendothelial migration of fetal haemopoietic cells

When tested in the Transwell assay, adult progenitors demonstrated minimal levels of spontaneous migration, with little difference between bone marrow-derived and peripheral blood progenitors (CFU-GM/CFU-GEMM migration was 1·0 ± 0·3% for MPB and 1·2 ± 0·3% for ABM). Similarly, cord blood progenitors showed little spontaneous migratory activity (1·6 ± 0·6%, P < 0·05). In contrast, freshly isolated fetal progenitors demonstrated significant levels of transmigration (7·0 ± 1·8% and 4·6 ± 0·8% migration of CFU-GM/GEMM from fetal blood and liver, respectively, P < 0·01, Fig 3A). Similar results were obtained when transmigration of BFU-E was analysed (Fig 3A).

image

Figure 3. Transendothelial migration of fetal, cord and adult haemopoietic cells. Transendothelial migration assays were carried out essentially as described in Materials and methods. For these assays, CFU-GM and CFU-GEMM were enumerated together and referred to as CFU-GM. (A) Transmigration of CFU-GM/GEMM and BFU-E from fetal blood (FB, n = 6), fetal liver (FL, n = 11), cord blood (CB, n = 8), adult mobilized blood (MPB, n = 7) and adult bone marrow (ABM, n = 4). Mean ± SE. (B) Effect of gestation period on transmigration of fetal progenitors. Left: Percentage transmigration of CFU-GM/GEMM in fetal liver samples from early (< 12 weeks, n = 10), and late (> 12 weeks, n = 8) gestation, and in fetal blood samples from late gestation (n = 7), mean ± SE. Right: Individual results of CFU-GM/GEMM transmigration from three paired fetal liver and blood samples from late gestation. (C) Percentage transendothelial migration of LTC-IC in adult (MPB, n = 4) and fetal samples (n = 4). Mean ± SE.

Download figure to PowerPoint

To explore the possible influence of gestational stage on the transmigratory capacity of fetal progenitors, we compared the transmigration of CFU-GM/GEMM from fetal liver cells up to and including 12 weeks of gestation with that of cells obtained from later gestational stages. As seen in Fig 3B, progenitor migration was higher in ontogenically earlier cells (6·9 ± 0·9%) than in cells from later gestational stages (3·1 ± 0·7% n = 6, P < 0·05). Similar results were obtained for BFU-E migration. There were insufficient fetal blood samples from early gestation (< 12 weeks) to make a similar comparison; however, late gestation (> 12 weeks) fetal blood samples were analysed and compared with late gestation fetal liver samples. The transmigration of CFU-GM/GEMM from fetal blood was significantly higher than that of progenitors from fetal liver (11·8 ± 3·8% versus 3·1 ± 0·7%, P < 0·01, Fig 3B). In three instances, we were able to compare the migratory activity of CFU-GM/GEMM from paired fetal blood and liver samples (Fig 3B). In each of these, the percentage of migration was higher in progenitors obtained from blood than from the liver.

Finally, we carried out a limited series of experiments on the transendothelial migration of cells with LTC-IC potential. Transendothelial migration of LTC-IC in growth factor-activated adult MPB CD34+ cells over 20 h was 12·1 ± 1·4% (n = 4, Fig 3C). In contrast, fetal LTC-IC in freshly isolated fetal samples demonstrated higher levels of migration (44%, 24·5% for two fetal liver samples, and 21·6%, 39·5% for two fetal blood samples examined). Figure 3C gives the data as mean ± SE of all four fetal samples. All fetal samples examined were from late gestational age.

Effect of chemokines and growth factors on the transmigration of fetal cells

We compared chemokine responses in fetal and adult progenitors, using the transendothelial migration assay. Freshly isolated adult progenitors showed minimal responses to SDF-1, with no real differences between bone marrow and peripheral blood cells. In contrast, freshly isolated cord and fetal progenitors showed significant migratory response to SDF-1 (P < 0·01 for each, Fig 4A). Results for fetal BFU-E were similar to those for CFU-GM/GEMM (Fig 4A). While freshly isolated MPB progenitors did not respond to SDF-1, these cells, after 3 d stimulation with growth factors (IL-3, IL-6 and SCF), were able to migrate to SDF-1 (Fig 4B). Similar results were obtained for adult bone marrow-derived CD34+ cells. We investigated whether growth factor stimulation would similarly augment the chemokine response of fetal cells. As seen in Fig 4B, prior incubation with growth factors markedly enhanced the transendothelial migration of fetal progenitors to SDF-1, from 12·8 ± 1·7% and 12·1 ± 2·2%, respectively, for fetal blood and liver, to 42·7 ± 4·3% and 45·7 ± 9·4% (n = 4, P < 0·01, Fig 4C), levels which far exceeded those for adult cells.

image

Figure 4. Effect of chemokines on the transendothelial migration of fetal progenitors. (A) Left: effect of SDF-1 on transmigration of CFU-GM/GEMM in freshly isolated fetal blood (FB, n = 6) and liver (FL, n = 8), cord blood (CB, n = 8), and adult samples (MPB n = 7, ABM n = 4). Right: corresponding results for BFU-E migration in fetal samples. Mean ± SE. (B) SDF-1 directed transmigration and effect of preincubation with growth factors. CFU-GM/CFU-GEMM from FB, FL or MPB were tested for their migration to SDF-1, before (fresh) and after 3 d incubation with growth factors (GF-stim). Mean ± SE of four samples for each. (C) Effect of MIP-3β on the transmigration of myeloid progenitors from FL and MPB. The results for both freshly isolated MPB, and growth factor stimulated MPB, are shown. Mean ± SE of four samples for each.

Download figure to PowerPoint

Finally, we tested the transmigratory response of fetal progenitors to MIP-3β (a chemokine which we have previously demonstrated to exert a chemotactic effect on cord blood progenitors (Yong et al, 1999b). Freshly isolated fetal CFU-GM/GEMM showed increased transmigration towards MIP-3β (from 4·4 ± 0·9% to 9·9 ± 2·2%, P < 0·05, n = 4) while adult progenitors did not, whether freshly isolated or following growth factor activation (Fig 4C). Neither fetal nor adult BFU-E showed any significant response to MIP-3β (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The establishment of haemopoiesis in diverse anatomical locations during embryonic and fetal life means that multipotent haemopoietic cells must circulate and continually migrate into extravascular fetal tissues. Early studies have reported the presence of haemopoietic progenitors in fetal blood, the numbers correlating inversely with gestational age, as would be expected (Hann et al, 1982; Linch et al, 1982). Such cells, in keeping with their physiological role and ontogenic status, may possess superior homing and migratory properties. Our findings confirm the presence of circulating haemopoietic cells in the fetus and also indicate that, at least in the liver, the percentage of CD45+ cells co-expressing CD34 is highest in early gestation. When normalized for CD34+ content, fetal haemopoietic cells had LTC-IC and progenitor frequencies which were similar to cord and adult haemopoietic cells. The striking difference between fetal and cord/adult CD34+ cells lies in their migratory and cell cycle behaviour. Fetal LTC-IC and progenitors demonstrated significantly higher levels of migration across endothelium, which were augmented further by prior growth factor stimulation, and the presence of chemokines (SDF-1 and MIP-3β). In addition, fetal blood CD34+ cells were actively in cycle, again in contrast to cord and adult blood haemopoietic cells.

A number of previous reports have characterized the haemopoietic cells obtained from fetal liver; however, in these studies, fetal cells were obtained from pregnancies in late second trimester (Huang et al, 1998; Roy & Verfaillie, 1999). Our report has focused on fetal blood as well as liver cells obtained largely from the first trimester of pregnancy, and hence provides unique insights into the behaviour and function of haemopoietic cells found very early in ontogeny. The CD34+ content of CD45-positive cells in fetal blood reported here was similar to that in previous studies, one of which used a lymphoid gate on flow cytometric forward and side scatter plots (Thilaganathan et al, 1994; Roy & Verfaillie, 1999). Clonogenic cell frequencies obtained in this present study also correlated well with an earlier report from our group (Linch et al, 1982). Fetal liver cells generated a greater proportion of erythroid colonies than adult cells, hence the total clonogenic output per CD34+ cell was also higher (Table II).

The increased numbers of cycling CD34+ cells in fetal blood and liver may simply reflect the enhanced proliferative status and shorter cell cycle transit time of fetal haemopoiesis in general. The higher proportion of cycling CD34+ cells in extravascular fetal haemopoietic tissue, when compared with circulating CD34+ cells, is reminiscent of the situation in the adult in which the vast majority of mobilized peripheral blood CD34+ cells are in G0/G1 phase of the cell cycle, while up to 15% of bone marrow CD34+ cells are cycling (in S+G2/M) (our data and Fruehauf et al, 1998). The effect of tissue localization on cell cycle status may be related to the absence of proliferative stimuli in the circulation. The precise relationship between cell cycling and migration is unclear, and may differ between ontogenically distinct sources of haemopoietic cells. Fetal blood progenitors, which contain a lesser proportion of cells in S+G2/M, nevertheless migrated more efficiently than their fetal liver counterparts. In the adult, bone marrow CD34+ cells were actively cycling but were no more migratory than peripheral blood CD34+ which were largely quiescent.

The superior migratory activity of fetal haemopoietic cells was evident in both LTC-IC and progenitors, indicating that enhanced migration is an intrinsic property of fetal haemopoietic cells at all maturational stages. Our observation that spontaneous levels of transmigration were highest in the most ontogenically primitive cells is in keeping with the physiology of these cells which must continually seed and establish haemopoiesis in different tissues, particularly in early fetal life. It is interesting that the migration capacity was higher in fetal blood progenitors, an observation which may simply reflect the preferential extravasation of more migratory cells into the vascular compartment. Similarly, in the adult, G-CSF-mobilized CD34+ cells were found to migrate more readily across Matrigel membranes than bone marrow CD34+ cells (Janowska-Wieczorek et al, 1999). The high levels of migration in response to SDF-1 may also reflect an intrinsically greater migratory potential, rather than an enhanced chemokine responsiveness. Indeed the levels of CXCR4 expression on fetal CD34+ cells are similar to those on adult CD34+ cells and both were lower than levels on cord blood CD34+ cells. The high expression of chemokine receptors on cord blood cells has recently been confirmed by another group (Rosu-Myles et al, 1999) and may be related to stimulation by cytokines released during labour and delivery. A recent report from another group confirmed that expression of CXCR4 on fetal and adult CD34+ cells was similar (Aiuti et al, 1999).

Transendothelial migration is mediated by cell surface adhesion molecules, although the relative importance of different receptor–ligand pairs may depend upon the particular assay conditions used. We have previously shown that the transendothelial migration of haemopoietic cells was mediated by the adhesion molecules CD11a/CD18 (LFA-1) and CD31 (PECAM-1) (Yong et al, 1998), while others have reported a role for E-selectin/ESL-1 ligand (Naiyer et al, 1999). In this latter study, the reported levels of adult LTC-IC migration were similar to our data presented here. High expression of VLA-4 on fetal CD34+ cells may be relevant to their migratory and engraftment properties. Although there is evidence from animal models that VLA-4 plays a part in the engraftment of both murine and human progenitors (Papayannopoulou et al, 1995; Peled et al, 2000), in vitro studies have yielded conflicting reports on the role of VLA-4 in the transendothelial migration of CD34+ cells (Yong et al, 1998; Naiyer et al, 1999). LFA-1 expression was lower on fetal CD34+ cells, a finding which may reflect a more primitive state, as LFA-1 levels increase with haemopoietic cell maturation (Campana et al, 1986). Our data on LFA-1 expression on fetal cells are in accordance with a recent study on fetal liver cells (Roy & Verfaillie, 1999); however, those authors also reported that fetal liver CD34+ cells expressed higher levels of l-selectin, a finding which we did not confirm. Roy & Verfaillie (1999) also examined the differential adhesive behaviour of fetal liver, cord blood and adult BM CFC, with the notable finding that fetal CFC adhered significantly more to collagen IV, and that this was mediated by VLA-2. It is not clear, however, from the results of either study, that any specific pattern of adhesion molecule expression or function can account for the distinct migratory features of fetal haemopoietic cells reported here.

In conclusion, we have demonstrated that, when compared with cord or adult cells, fetal haemopoietic cells in blood and liver display superior transmigratory activity which is not directly attributable to adhesion molecule phenotype, or to chemokine receptor expression. The enhanced migratory activity of fetal haemopoietic cells may be a prerequisite for the sequential establishment of haemopoiesis in separate organs and tissues throughout fetal life.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

We are grateful to S. Ings and M. Watts for provision of purified CD34+ cells from apheresis products. We also wish to thank all staff on the Labour ward, University College Hospital, London, for facilitating our collection of cords and cord blood samples.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  • Aiuti, A., Webb, I.J., Bleul, C., Springer, T. & Gutierriz-Ramos, J.C. (1997) The chemokine SDF-1 is a chemoattractant for human CD34+ haemopoietic progenitor cells and provides a mechanism to explain the mobilisation of CD34+ progenitors to peripheral blood. Journal of Experimental Medicine, 185, 111120.
  • Aiuti, A., Tavian, M., Cipponi, A., Ficara, F., Zappone, E., Hoxie, J., Peault, B. & Bordignon, C. (1999) Expression of CXCR4, the receptor for stromal cell-derived factor-1 on fetal and adult human lymphohematopoietic progenitors. European Journal of Immunology, 29, 18231831.DOI: 10.1002/(SICI)1521-4141(199906)29:06<1823::AID-IMMU1823>3.0.CO;2-B
  • Bhatia, M., Awaraji, C. & Gallacher, L. (1999) Cell cycle status of human fetal and cord blood cells capable of multilineage repopu- lation in NOD/SCID mice. Blood, 94, 654 (Abstract).
  • Campana, D., Sheridan, B., Tidman, N., Hoffbrand, A.V. & Janossy, G. (1986) Human leucocyte-function-associated antigen on lympho-haemopoietic precursor cells. European Journal of Immunology, 16, 537545.
  • Fruehauf, S., Veldwijk, M.R., Kramer, A., Haas, R. & Zeller, W. (1998) Delineation of cell cycle state and correlation to adhesion molecule expression of human CD34+ cells from steady state bone marrow and peripheral blood following G-CSF supported chemotherapy. Stem Cells, 16, 271279.
  • Gothot, A., Van Der Loo, J.C.M., Clapp, D.W. & Srour, E.F. (1998) Cell cycle related changes in repopulating capacity of human mobilised peripheral blood CD34+ cells in Non-Obese Diabetic/Severe Combined Immune-deficient mice. Blood, 92, 2641–2649.
  • Hann, I.M., Bodger, M.P. & Hoffbrand, A.V. (1982) Development of pluripotent hematopoietic progenitor cells in human fetal blood. Blood, 62, 118123.
  • Hendrikx, P.J., Martens, A.C.M., Hagenbeek, A., Keij, J.F. & Visser, J.W.M. (1996) Homing of fluorescently labeled murine hematopoietic stem cells. Experimental Hematology, 24, 129140.
  • Huang, S., Law, P., Young, D. & Ho, A.D. (1998) Candidate hematopoietic stem cells from fetal tissues, umbilical cord blood vs. adult bone marrow and mobilized peripheral blood. Experimental Hematology, 26, 11621171.
  • Itoh, K., Tezuka, H. & Sakoda, H. (1989) Reproducible establishment of hemopoietic supportive stromal cell lines from murine bone marrow. Experimental Hematology, 17, 145149.
  • Janowska-Wieczorek, A., Marquez, L.A., Nabholtz, J.-M., Cabuhat, M.L., Montario, J., Chang, H., Rozmus, J., Russell, J.A., Edwards, D.R. & Turner, A.R. (1999) Growth factors and cytokines upregulate gelatinase expression in bone marrow CD34 cells and their transmigration through reconstituted basement membrane. Blood, 93, 33793390.
  • Jones, H.M., Jones, S.A., Watts, M.J., Khwaja, A., Mills, W., Fielding, A., Goldstone, A.H. & Linch, D.C. (1994) Development of a simplified single apheresis approach for peripheral blood progenitor cell transplantation in previously treated patients with lymphoma. Journal of Clinical Oncology, 12, 16931699.
  • Linch, D.C., Knott, L.J., Rodeck, C.H. & Huehns, E.R. (1982) Studies of circulating hemopoietic progenitor cells in human fetal blood. Blood, 59, 976979.
  • Naiyer, A.J., Jo, D.-Y., Ahn, J. & Rafii, S. (1999) Stromal derived factor-1-induced chemokinesis of cord blood CD34+ cells (long term culture initiating cells) through endothelial cells is mediated by E-selectin. Blood, 94, 40114019.
  • Papayannopoulou, T., Craddock, C., Nakamoto, B., Preistley, G.V. & Wolf, N.S. (1995) The VLA-4/VCAM-1 adhesion pathway defines contrasting mechanisms of lodgement of transplanted murine haemopoietic progenitors between bone marrow and spleen. Proceedings of the National Academy of Sciences of the United States of America, 92, 96479650.
  • Peled, A., Petit, I., Kollet, O., Magid, M., Ponomaryov, T., Byk, T., Nagler, A., Ben-Hur, H., Many, A., Shultz, L., Lider, O., Ronen, A., Zipori, D. & Lapidot, T. (1999) Dependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR4. Science, 283, 845848.DOI: 10.1126/science.283.5403.845
  • Peled, A., Kollet, O., Ponomaryov, T., Petit, I., Franitza, S., Grabovsky, V., Slav, M.M., Nagler, A., Lider, O., Alon, R., Zipori, D. & Lapidot, T. (2000) The chemokine SDF-1 activates the integrins LFA-1, VLA-4 and VLA-5 on immature human CD34 cells: role in transendothelial/stromal migration and engraftment of NOD/SCID mice. Blood, 95, 32893295.
  • Peters, S.O., Kittler, E., Ramshaw, H.S. & Quesenberry, P.J. (1996) Ex vivo expansion of murine marrow cells with interleukin-3 (IL-3), IL-6, IL-11 and stem cell factor leads to impaired engraftment in irradiated hosts. Blood, 87, 3037.
  • Rosu-Myles, M., Khandaker, M., Wu, D., Fellows, F., Kelvin, D. & Bhatia, M. (1999) The expression of chemokine receptors on primitive human hematopoietic cells. Blood, 94, 477 (Abstract).
  • Roy, V. & Verfaillie, C.M. (1999) Expression and function of cell adhesion molecules on fetal liver, cord blood and bone marrow hematopoietic progenitors: implications for anatomical localization and developmental stage specific regulation of hematopoiesis. Experimental Hematology, 27, 302312.
  • Thilaganathan, B., Nicolaides, K.H. & Morgan, G. (1994) Subpopulations of CD34-positive haemopoietic progenitors in fetal blood. British Journal of Haematology, 87, 634636.
  • Wilpshaar, J., Falkenburg, F.J.H., Kanhai, H.H.H., Breese, R. & Srour, E.F. (1999) Engraftment potential of human fetal liver and bone marrow CD34+ cells residing in different phases of cell cycle. Blood, 94, 34 (Abstract).
  • Yong, K., Fahey, A., Thomas, N.S.B., Inge, S., Watts, M. & Linch, D.C. (1999a) Transendothelial migration of CD34+ requires growth factors activation and occurs predominantly in G0/G1 phase of the cell cycle. Blood, 94, 40 (Abstract).
  • Yong, K., Fahey, A., Reeve, L., Nicholls, C., Thomas, N.S.B., Pizzey, A., Ings, S.J., Watts, M.J. & Linch, D.C. (1999b) Cord blood progenitor cells have greater transendothelial migratory activity and increased responses to SDF-1 and MIP-3β compared with mobilized adult progenitor cells. British Journal of Haematology, 107, 441449.
  • Yong, K.L., Watts, M., Thomas, N.S.B., Sullivan, A., Ings, S. & Linch, D.C. (1998) Transmigration of CD34+ cells across specialised and non-specialised endothelium requires prior activation by growth factors and is mediated by PECAM-1. Blood, 177, 14651474.