Thoughts on BM homing assays
The below comments were originally prepared as a part in the rebuttal argument to protect our papers. They are now shown below with some modifications.
We believe that “BM homing of HSCs” has to have a clear definition in the field of HSC research. Before discussing what is the best way to assess “BM homing”, we should share a common view on what “BM homing” should mean and what technical limitations we currently have.
l In a traditional setting of mouse experiments, does “homing” simply reflect presence of test cells in the whole cell population obtained from recipient BM samples at the indicated times (like 4 hours, 12 hours, 24 hours)?
Ø Test cells may be still located within capillary vessels. This means that they are still sitting outside marrow cavity, thus should not represent “BM homing”.
Ø Test cells may move somewhere else in another few hours, or even a few days later. Should we still recognize these events as “BM homing”?
l If the aim to assess “BM homing” is to correlate that estimation and the later hematopoietic reconstitution outcomes, the above ambiguous definition may not be justified because the estimation would involve the events, which should be regarded as only “transient BM (or even outside marrow) residence”, thus failing to contribute to long-term hematopoiesis.
l If the aim of assessment of BM homing is to predict subsequent hematopoietic reconstitution derived from test HSCs, such “homing” has to be “productive”. In other words, “BM-homed HSCs” defined with the aim to predict hematopoiesis should not move away from the initial site of residence, should self-renew on sites, then should contribute, at least to some extent, to subsequent peripheral hematopoietic reconstitution.
However, at present, we must say that to track whole life history of single HSCs and their progenies in vivo, from i.v. injection to long-term (at least >12 weeks) hematopoiesis, is impractical, even when we have much improved technologies comparing with 10-20 years ago, such as in vivo imaging of live animals, methods to make mice transparent, and advanced microscopy techniques.
So, the points we think important are:
Although “BM homing” is assessed in many different ways as the earliest events in the process of hematopoietic reconstitution following HSC transplantation, the estimated values may not necessarily have significant meanings in prediction of subsequent hematopoietic reconstitution outcomes, unless we make a breakthrough in development of analysis methods.
Here we are not denying the importance of homing assays. If we find 10-20 times more HSCs residing in recipient BM samples 24 hours after injection, we may be able to expect better transplantation outcomes. If improved imaging analysis support that finding, confirming that many test HSCs are actually localized within a marrow cavity, such expectation will increase likelihood. One may track hematopoietic reconstitution kinetics derived from test HSCs sequentially, like we did in this paper, so that one can confirm that the initial increase in HSCs at BM residence actually most likely leads to subsequent enhancement in donor cell-hematopoiesis (chimerism formation in the case of competitive repopulation assays).
Rationale for the choice of our assays to assess BM homing (transient residence?) of purified HSCs / HSPCs
l We thought it preferable to use doses of test cells within “ordinal” ranges when translated into the case of whole bone marrow transplantation.
Ø For purified HSCs (loss-of-function with Cxcr4-KO mice) or cultured HS(P)Cs (gain-of-function): 200-400 cells per recipient (correspond to ~0.5-1.0 x 107 WBM cells)
Ø Purified HSPCs (loss-of-function with AMD3100-antagonism) or cultured HSPCs (gain-of-function): 10,000-50,000 cells per recipient (correspond to ~XX-XX x 107 WBM)
l We thought it most straightforward for assessment of BM homing to inject a pure HSC/HPC population as test cell materials with their progenies excluded.
l We also thought it most reliable to assess homing events by directly enumerating the cells present in recipient BM as long as sensitivity allowed, with assured evidence that they are of test cell-origin.
l According to the above consideration, detection was carried out as follows:
Ø For purified HSCs or cultured HS(P)Cs: assessed by the number of CFCs detectable per defined mass of bones (e.g., two femurs + two tibias + pelvic bones) in each recipient. Flow cytometry analysis could not allow reliable detection of homing events in this context (as long as we use the above described # of test cells).
² In the revised manuscript, we have succeeded in detecting reliable numbers of CFCs as early as 16 hrs post infusion. So far we believe this is the maximum sensitivity obtainable with a combination of this nature/dose of test cells and this assay system.
² Note that limitation in this detection method does exist in that recipients must be lethally irradiated. Otherwise, recipient-own HSCs/HSPCs would make much more colonies than those derived from test cells, thus compromising the readouts. So when one wish to examine “BM homing in non-irradiated hosts”, this method cannot be applicable to that aim.
Ø For purified HSPCs or cultured HSPCs: assessed by flow cytometry using test cells identifiable due to their bright fluorescence even when the events are scarce. We considered it inappropriate to score the homed events per fixed numbers of acquired events (e.g., per 106 total events), because we recognized inter-recipient variability in BM cellularity so significant after lethal-dose irradiation. To overcome this issue, we co-injected fixed numbers of “reference” or “indicator” cells together with test cells, both having different fluorescence clearly distinguishable. In this scenario, we estimated homed events as “numbers”, but represented BM homing as the ratio of test/indicator cells in analogy with the competitive repopulation assay, the golden standard for the assessment of reconstitution ability in HSCs. The rational for this assay can be supported by one of pioneering homing studies “Organ-selective homing defines engraftment kinetics of murine hematopoietic stem cells and is compromised by Ex vivo expansion, Blood 1999; 93: 1557”, which demonstrated linearity of the homing assay and no inhibition of HSC/HPC homing by co-injection of excess numbers (up to 108 cells) of another whole BM cell population.
² In the experiments shown in revised manuscript, the earliest time point evaluable was 4 hrs post infusion, although detectable events remained limited with the condition used (17,000 test cells and 106 indicator cells). At later times, 15 and 24 hrs, detectable events steadily increased, we thus judged this assay suitable for the assessment of “early” or “short-term” homing with relatively limited numbers of HSPCs used as a test cell source.
Justification of the idea that our data do not necessarily contradict previous publications
We appreciate the comment meaning “the data here is not convincing enough to support the conclusion, which contradicts previous publications ”. We believe that one major contradiction pointed out is regarding our conclusion “CXCR4 expression is unimportant for homing of mouse HSPCs”, namely the one drawn from our loss-of-function experiments. This actually prompted us to re-evaluate other groups’ published data with the highest attention to detail. Consequently, we have reached the idea that because of significant difference between assay systems, our data will not necessarily conflict with the previous observations. We listed such examples below by referring to the data shown in three important papers that have frequently been cited by so many researchers.
1 Bonig H, Priestley GV, Papayannopoulou T. Hierarchy of molecular-pathway usage in bone marrow homing and its shift by cytokines. Blood. 2006;107(1):79-86.
To desensitize the cells to SDF-1, they used the method “adult BM cells were incubated in PBS/BSA 0.5% with AMD3100 (100 mg/mL) for 15 minutes at 37°C prior to transplantation and coinjected with 100 mg/mL AMD3100/recipient, as previously described,18 except that a higher dose of AMD3100 was used by us”. The way of BM homing assessment (3 hours) was described as: “CFU-C number homed per femur was calculated from the number of CFU-Cs in the colony assays and expressed as fraction of the input CFU-C/femur”, thus significantly differing from the way we took. Most importantly, however, one of their conclusions shown in Figure 1B in this paper was “Three hour BM homing of fresh WT BM cells was not affected by AMD3100 incubation/coinjection”. They also showed that “Three-hour BM homing of SCF-treated WT BM cells was significantly reduced by AMD3100 incubation/coinjection”, with these observations constituting their central findings supporting the idea that dependence on SDF-1/CXCR4 axis of HSC/HPC homing to BM could vary according to cell types and cell status. As we used fresh HSPCs to see if AMD3100-antagosism alter BM homing, the results now shown in the revised manuscript can be regarded to be in complete match with their observation.
2 Christopherson KW, 2nd, Hangoc G, Mantel CR, Broxmeyer HE. Modulation of hematopoietic stem cell homing and engraftment by CD26. Science. 2004;305(5686):1000-1003.
This epoch-making paper also utilized AMD3100 to antagonize CXCR4 as it reads “Treatment of 10 x 106 LDBM donor cells with CXCR4 antagonist AMD3100 for 15 min before transplantation reversed increases in homing efficiency of CD26-inhibited or deleted Sca-1+lin- cells (Fig. 1D)”, and has been cited as ref. 18 in the above Blood paper. We believe that as shown in this Fig. 1D the main emphasis should lie on the dependence of enhanced BM homing by CD26-inhibition on SDF-1/CXCR4 axis in HSC/HPCs. In the same figure, they showed graphic representation containing multigroup histograms and described the results as “Increased homing efficiency of Sca-1+lin- C57BL/6 HSC within LDBM cells noted with Diprotin A treatment or with CD26-/- cells is reversible by their treatment with CXCR4 antagonist AMD3100 for 15 min before transplant. AMD3100 also reduces homing efficiency of C57BL/6 donor cells in the absence of CD26 inhibition”, the latter conclusion thus may conflict with our results. We, however, belieave it acceptable considering significant difference in two assay systems: whereas we treated fresh HSPCs alone with AMD3100 and tracked their in vivo fate directly, they preincubated low-density (LD) BM donor cells and assessed homing efficiency of Sca-1+lin- cells “within” LDBM cells. The way they took was as follows: “The number of Sca+lin- donor cells in the context of BM Sca-1+lin- recipient cells was calculated by measuring CD45.1+ and CD45.2+ on Sca-1+lin- LDBM cells isolated from the recipient mice 24 hours post transplant, believed to be adequate time for HSC homing to have occurred”. We have to emphasize that we do not intend to belittle the importance of their pioneering work. We rather totally agree with their statement “The use of LDBM cells more accurately represents clinical protocols than does the use of sorted Sca-1+lin- HSCs”. Instead, we would just like to note that different approaches may yield some conflicting results, but will eventually compensate each other to expand our knowledge in the research field, culminating into clinical translation for the human health.
3 Hoggatt J, Singh P, Sampath J, Pelus LM. Prostaglandin E2 enhances hematopoietic stem cell homing, survival, and proliferation. Blood. 2009;113(22):5444-5455.
This is another paper that is indispensable for our reseach field. They also used AMD3100 as read like “To evaluate the role of CXCR4 in homing, Linneg CD45.2 cells were treated with vehicle or 1 mM dmPGE2 plus 10 mM AMD3100, 2 x 106 treated cells injected into lethally irradiated CD45.1 mice, and homed SKL cells analyzed 16 hours after transplantation”. In analogy to the above Science paper, the major conclusion drawn by this experiment, at least to us, is likely that “incubation with AMD3100 abrogated the improved homing efficiency of dmPGE2-pulsed cells (Figure 5D)”, although they also stated “incubation of vehicle or dmPGE2-pulsed cells with AMD3100 significantly reduced SKL cell homing”. But again in analogy to the former case, significant difference in assay systems will likley explain somehow conflicting results: they treated Linneg cells with AMD3100, assessed homing efficiency of HPCs “within” Linneg cells in a way similar to that taken by Dr. Broxmeyer’s group.
1