Radiation effects on haematopoietic stem cells in vitro: possible role of stromal niches in the stem cell hierarchy. J.G. Sharp', D.A. Crouse', J.D. Jackson2, C.M. ...
Br. J. Cancer (1986), 53, Suppl. VII, 133-136
Radiation effects on haematopoietic stem cells in vitro: possible role of stromal niches in the stem cell hierarchy J.G. Sharp', D.A. Crouse', J.D. Jackson2, C.M. Schmidt2, E.K. Ritter2, G.C. Udeaja2 & S.L. Mann2 Departments of 'Radiology and 2Anatomy, University of Nebraska Medical Center, 42nd and Dewey Avenue, Omaha, Nebraska 68105, USA. An increasing variety and number of therapeutic protocols involve the irradiation of large proportions of the sites of haematopoiesis in some patients. A long term consequence of such exposure is the occasional induction of malignancies. These are often but not exclusively second malignancies which are frequently of haematopoietic origin (Penn, 1981). The relative contribution of radiation effects on haematopoietic stem cells (HSC) or microenvironmental stromal cells (MSC) to these long term post-exposure observations remains uncertain (Testa et al., 1985). This is partly because of the increasing realization that the behaviour of stem cells and stromal cells is not independent (Spooncer et al., 1985). Increasingly, it is evident that stromal cells are involved in the regulation of haematopoiesis, particularly via the more primitive members of the HSC hierarchy (Sharp et al., 1985). The ability to maintain HSC in long term culture (LMTC) when in contact with a requisite stroma (Dexter et al., 1977) appeared to offer the opportunity to dissect the differential effects of radiation. on HSC versus MSC. However, this methodology leads to some peculiar radiobiological results which potentially occur because of the effects of HSCMSC interactions. We briefly describe these effects together with some experiments which attempt to elucidate the nature of HSC-MSC interactions.
Materials and methods
These studies employed C57BI/6J mice from Jackson Laboratories. The LTMC were established and maintained by the method originally described by Dexter and colleagues (1977) but supplemented with hydrocortisone. Irradiation of LTMC adherent layers was performed using 60Co y rays at a doserate of 0.5 Gy min- 1. The spleen-colony assays were performed on intravenously injected LTMC cell suspensions by the method originally described by Till and McCulloch (1961) except that in some experiments colonies were evaluated at day 8 or 12 after irradiation. The colonies were sized and colonies smaller than 0.5 mm diameter excluded
from the analysis. CFU-S recipients were irradiated whole body with y rays to a total dose of at least 8.5Gy. In one series of experiments organ cultures were prepared from scraped adherent layer cells, cultured overnight on Nuclepore filters (0.4pm) and transplanted under the kidney capsule of either intact or irradiated recipients. Spleen colonies were evaluated after 8 or 12 days in a manner analogous to that following i.v. administration. Samples of organ cultures and grafted material were fixed in neutral buffered formalin or 2% glutaraldehyde for routine light or electron microscopic evaluation respectively.
Results and discussion
When the initially established LTMC adherent layer
was irradiated with low doses of y irradiation
before refeeding with bone marrow, the supernatant HSC assayable in the day 8 spleen colony (d8 CFU-S) assay declined approximately exponentially in a dose dependent fashion for the first 4 weeks after irradiation and refeeding (data not shown). Although subsequently there was some recovery of supernatant d8 CFU-S numbers, a dose dependent deficit was still evident after 10 weeks in culture. A quasi-survival curve of supernatant d8 CFU-S as a function of dose is shown in Figure 1. If these data are presented in this way and interpreted in the manner that others have employed with similar data in other systems (Hagemann et al., 1971), it would suggest a multiplicity based on a continually bending curve and an initial 'Do' of 0.3 Gy. Although the observation of large effects at very low doses to the LTMC adherent layer is real and has been observed by others, the 'Do' is unlikely to be a characteristic of d8 CFU-S in this system. The supernatant d8 CFU-S arise from more primitive HSC precursors located in the adherent layer (Mauch et al;, 1980; Crouse et al., 1984). We have previously suggested that possibly these primitive HSC reside in 'niches' (Crouse et al., 1984) similar to niches proposed by Schofield (1978) on the basis
© The Macmillan Press Ltd., 1986
J.G. SHARP et al.
134
argued that if such niches exist, then a fixed area of tissue culture flask should only be able to support limited number of stem cells, that is, the number of adherent layer CFU-S should be saturable with stem cells even with repeated refeedings with fresh bone marrow. This is, in fact, the case (data not shown). In an attempt to identify the nature of the interaction between niche forming cells, presumed to be MSC, and stem cells we have compared the outcome of stem cell assays in which adherent layer cells were disassociated into a single cell suspension and injected i.v. into lethally irradiated recipients versus grafting of organ cultures of adherent layer cells which maintain HSC-MSC constacts intact. The results are shown in Table I. Whereas approximately equal numbers of day 8 and day 12 colonies are formed following i.v. administration of adherent layer HSC, fewer day 8 colonies and greater numbers of day 12 colonies were formed following grafting when HSC-MSC contacts were maintained. This is particularly evident when the ratio of day 12 to day 8 colonies is compared (Table I). We have shown elsewhere (Sharp et al., 1985) that d12 CFU-S are more primitive members of the HSC hierarchy than d8 CFU-S. Additionally, they appear to be formed by more radioresistant cells although not to the extent exhibited by cells forming endogenous colonies at day 14 (Sharp et al., 1985). Thus the simplest, although not the only interpretation of this observation is that some MSC function to preserve the primitiveness of at least part of the HSC hierarchy. When this interaction is disturbed, perhaps by low doses of radiation, the primitive HSC in the adherent layer may fail to continue to give rise to more differentiated supernatant d8 CFU-S. Much higher doses of radiation, a a
20
C1. 0
\
c
0-o
U)
II
1
0.05
0.25
1.0
Dose to adherent layer (Gy)
Figure term
1
Supernatant
marrow
irradiated
with
d8
cultures
various
CFU-S
whose
doses
assayable
adherent
of
y
in
long
layer
was
irradiation
im-
mediately before re-feeding with bone marrow 4 weeks
previously.
of in vivo studies. Further, we have suggested that the low dose radiation compromises
the function-
ality of the stem cells in these niches (Grouse et at., 1984).
The
multiplicity
interaction
between
suggested
adherent
that
a
critical
layer stem cells
and
other cell(s) (MSC?) might be important (Cohen et at.,
1982).
In
demonstrate
this
regard,
the existence
we
have
of 'niches'
attempted in
vitro.
to We
Table I Spleen colony number, 8 or 12 days after transplantation of long term marrow culture adherent cells as organ cultures (o.c.) or intravenously (i.v.) to recipients Cells
Endogenous LTMC supernatant LTMC adherent LTMC adherent
Day 8
Day 12
Ratio Day 12C Day 8
0.2+0.l1
1.9+0.8d
9.5
i.v.
5.6+0.6b
1.1+ 1.2bd
0.2
i.v.
4.6+0.5b
4.4+2.1b
1.0
o.c.
0.8+0.3b
.8.1+2.7b d
10.1
Route
aMean+s.e. 10 or more mice. bCorrected for endogenous colony number with compounding of errors. cDay 12 CFU-S are more primitive cells than d8 CFU-S. For purposes of comparison, this ratio is 1.0 for adult bone marrow and 3.0 for foetal liver cells. 'Significantly different (P
