In this study, both WT and CS adipose-derived cells showed low levels of ALP activity, regardless of time in culture or BMP4 stimulation. These results are consistent with the majority of the reports showing a limited ability for adipose-derived cells to express ALP after exposure to osteogenic medium27, 39-42
and support the previous observation that rabbit adipose-derived cells did not show a marked ALP response to BMP4 stimulation.43
Although adipose cells showed relatively poor ALP expression, they did react similarly between WT and CS rabbits. This similarity among adipose cells from rabbits with different diagnoses suggests that adipose cells may be suitable candidates for designing cell-based therapies that work consistently among different patients.
ALP activity has been consistently demonstrated in bone marrow-derived cells.3, 39-42, 44
The data presented here suggest that rabbit bone marrow-derived cells did not have significantly increased ALP activity following stimulation with 50ng/ml BMP4, presumably due to the high baseline ALP activity (unstimulated). Bone marrow-derived cells were different between CS and WT rabbits. CS bone marrow cells showed higher ALP activity compared to WT bone marrow-derived cells in all treatment conditions.
Muscle-derived cells showed relatively low baseline (0ng/ml BMP4) ALP activity and demonstrated an approximate 4-fold increase in ALP activity after 7 days of stimulation with BMP4. Similar results have been noted in mouse muscle-derived stem cell populations.45, 46
These characteristics make muscle-derived cells an attractive target for bone tissue engineering applications because it would be beneficial to have cells that are highly responsive to a stimulating factor (in this case, BMP4). However, there were significant differences noted between CS and WT muscle-derived cells. Craniosynostotic muscle cells had higher mean ALP expression than WT cells in all conditions studied. Therefore, the reaction of muscle-derived cells to a specific treatment may not be consistent enough to use as a basis for a cellular therapy.
Comparison among the different tissues showed a significant phenotype by diagnosis interaction suggesting that cells derived from CS rabbits were not related to each other in the same way as cells from WT rabbits. In the cells derived from WT rabbits, muscle-derived cells had the highest overall ALP expression. Bone marrow-derived cells had the highest ALP expression in cells derived from CS rabbits. Both types of animals had similarly inactive adipose-derived cells. Therefore, it was not found that all cells from CS rabbits were simply more ALP-positive.
Together these data suggest there is not a uniform alteration in the CS progenitor cells that would manifest as increased osteogenic response in every cell population tested. Rather, 1) there are cells within specific tissues that act differently in the CS rabbits, and 2) the isolated CS phenotype plays some role in the relationship among cells from specific tissues.
The distinctions between the cells derived from CS rabbits and WT rabbits are important because they offer insight into larger questions such as 1) growth factor and scaffold therapies developed using normal cells may not function correctly when used in a patient with a seemingly unrelated pathology, 2) if stem cell function is affected, other processes such as healing after injury may also be affected, and 3) because of variability, it may be very difficult to identify appropriate sources for cells within any given patient population for consistent cell-based therapies.
Nonsyndromic CS is characterized by an isolated pathology, meaning that there is a fusion of bones in the skull but there are no other related pathologies. In cases of nonsyndromic disease, we would expect to see no alterations in cells from tissues that have no pathology in the patient. Non-cranial bone growth (from metacarpal bone growth measurements) has been found to be the same between CS and WT rabbits.33
The results presented here suggest that CS and WT rabbits have cells in their bone marrow and muscle that do not react similarly to BMP4 stimulation. This unexpected result begs the larger stem cell biology question, “How many of the diseases that are characterized as “nonsyndromic” have related but currently unidentified stem cell issues?” Most studies of animal models of human disease are to better understand the pathology of interest and have a focus limited to the target organ or tissue. In our rabbit model of craniosynostosis, we have identified differences in progenitor cells from tissues that were thought to be unaffected in the animals. It is necessary to continue studying the effects that a disease has on an individual's progenitor cells.31
It is imperative to isolate postnatal progenitor cells from currently available animal models (knock-out, knock-in, deficient mutants, etc) to confirm whether seemingly isolated pathologies might influence stem cell activity.
Results presented here also shed light on the issue of variability. We found high levels of variability among cells derived from postnatal rabbit tissues. It was interesting to observe such variability within the relatively closed population, genetically speaking, of CS NZW rabbits. Other groups have also reported high variability in rabbit bone marrow-derived cells.47
If such variability exists among NZW rabbits, the more genetically diverse human population is bound to exhibit higher variability. In fact, variability within mesenchymal stem cells has been identified in patient-derived populations.48-50
Therefore, we are confident that such variability is endemic in primary cell isolations and not an artifact of isolation technique. With this in mind, caution should be used when interpreting results of studies on human tissues that involve very small sample sizes.51-54
The results presented in this report are strengthened by the sample size used for each analysis. The large sample sizes of either WT or CS rabbits allowed for a better understanding of cellular characteristics. High cellular variability between individuals is a major problem that plagues the development of effective cell-based therapies. Much of the focus in tissue engineering so far has been in identifying cells, scaffolds, or growth factors that lead to specific tissue regeneration. However, high variability between patient responses to therapy, either due to variability in individual patient healing (innate stem cell differences) or to the effects of different diseases (disease-specific changes in stem cells), would complicate designing therapies that consistently achieve the desired results. One solution to this issue may be to focus on developing patient-based, customized therapies. To that end, it may be helpful to identify cells within a patient population that may serve as an in vitro proxy to test therapies and give insight into how a therapy may function when applied in vivo.
Overall, the results presented here do not support the hypothesis that cells from unaffected tissues in CS and WT rabbits react similarly to BMP4 stimulation. They do, however, suggest that adipose-derived cells are not as ALP-positive or as BMP4-responsive as their muscle- or bone marrow-derived counterparts. The results also suggest that outcomes of a particular therapy may, in part, depend on the tissues being treated. Finally, due to the high level of variability observed between donors, it may be beneficial to focus future research on the development of customized therapies, rather than traditional “off-the-shelf” therapies.