Bone marrow-derived cells in oral and dermal wound healing
A scientific essay in Medical Sciences
DOCTORAL THESIS defended in public on 29th of March 2012
Wound healing is a process to restore tissue integrity after damage or injury. This normally beneficial process involves wound contraction to reduce the wound surface. Ultimately scar tissue is formed. The main cells in wound contraction and scarring are the myofibroblasts. Myofibroblasts can be recruited from local tissues surrounding the wound. Also circulating bone marrow-derived cells (BMDCs) can be recruited to a wound and differentiate into myofibroblasts. By reducing the myofibroblast population in wounds and thus the formation of scar tissue clinical opportunities arise. This can be achieved by interfering with the recruitment and differentiation of myofibroblasts. Therefore, it is important to identify the main sources for myofibroblasts. The aims of our study were to investigate whether substantial numbers of myofibroblasts are recruited from the bone marrow in response to wounding, and whether myofibroblast differentiation can be modulated.
Chapter 1 describes the background of the study and provides an overview of the experiments. Myofibroblasts are discussed in the context of wound healing with the focus on BMDCs as a potential source. Differences in wound healing between different tissue types, and between prenatal and postnatal tissues are discussed. The experimental design is ntroduced and the specific aims are summarized.
The results of a literature review on stem cells in tissue turn-over and wound healing in epithelia are presented in chapter 2. The characteristics of the stem cell populations and their niche in the epithelia of different tissues (skin, cornea, lung, and intestine) are discussed. Similar mechanisms were deduced regarding the contribution of stem cells to tissue turn-over and wound healing. A model is proposed identifying stem cells for local maintenance, repair, and bone marrowderived rescue stem cells. According to this model, BMDCs are recruited in response to tissue damage that exceeds the regenerative capacity of the local stem cells.
In chapter 3 matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) are introduced. These enzymes and their inhibitors play a key role in tissue turn-over, wound healing and development by regulating the degradation of the extracellular matrix. Apart from the inhibitory effect of TIMPs on MMPs, additional biological functions of these proteins are discussed. This literature study shows that the balance between MMPs and TIMPs is disrupted in the development of cleft palate, in periodontitis, and also in different types of cancer. The measurement of the MMP and TIMP levels might provide diagnostic markers for several pathologies, and modulation of their balance might offer therapeutic possibilities.
The recruitment of BMDCs to palatal wounds in vivo is investigated in chapter 4. Therefore, rats were lethally irradiated to ablate their bone marrow cells. Subsequent, labeled bone marrow cells were transplanted to create chimeric rats. This model was also used for the studies in Chapters 5 and 6. In the rats, experimental wounds were made on the palate. Fourteen days later, the wounds were analyzed for labeled BMDCs. Statistically significant more BMDCs are recruited to the wounds compared to the unwounded palate. The BMDCs differentiated into myofibroblasts, activated fibroblasts and macrophages. Also endothelial cells originating from the bone marrow were detected. Although significantly more labeled BMDCs and myofibroblasts were detected in the wounds, their numbers were not substantial. Most of the myofibroblasts probably originated from local precursor cells. Therefore, we wondered whether the tissue-type or the wound size is crucial in the recruitment of BMDCs.
In chapter 5, the contribution of BMDCs to palatal and skin wounds of equal size in rats is compared. Fourteen days after wounding, about a twelve-fold increase of BMDCs was recruited to palatal wounds compared to unwounded palate. In contrast, no significant changes were detected in the numbers of BMDCs recruited to the skin wounds, compared to unwounded skin. Significantly more bone marrow-derived myofibroblasts were only detected in palatal wounds compared to unwounded palate. However, the numbers of bone marrow-derived myofibroblasts and also activated fibroblasts were similar but not substantial in both palatal and skin wounds. From these results we concluded that BMDCs are preferentially recruited to oral wounds and not to skin wounds.
To study the effect of wound size, only skin wounds were subsequently investigated. The maximum wound size on the palate of rats is limited. In chapter 6 we tested whether wound size plays a pivotal role in the recruitment of BMDCs in skin wounds. Fourteen days after wounding, the smaller wounds were already closed, whereas the large wounds still were open. The density (i.e. the number of cells per square mm) of BMDCs in the wounds and the adjacent tissues was determined, as well as in normal skin from the flank. After correction for cell density, the recruitment of BMDCs was found to be independent of wound size. Moreover, no significant differences were found by comparing the number of BMDCs in wounded and unwounded skin. The density of myofibroblasts and also of bone marrow-derived myofibroblasts significantly increased with wound size, which can be explained by different stages of wound healing; since the smaller wounds were already closed, wound contraction had probably already stopped. After wound contraction, myofibroblasts disappear by apoptosis. This study confirms the results in Chapter 5 in that BMDCs are not preferentially recruited to skin wounds. Additionally, wound size does not seem to affect the recruitment of BMDCs.
In chapter 7 an approach is described to interfere with myofibroblast differentiation. In a three dimensional collagen gel contraction model the effect of hyaluronan was investigated. Hyaluronan is abundantly present in fetal wounds, which heal without a scar. A significant reduction in contraction was measured in the gels with high concentrations of hyaluronan. In these gels also a lower amount of pro-MMP2 was found. No different amounts were detected for the active forms of MMPs or for TIMPs. Also no differences were detected in mRNA expression of MMPs and TIMPs. On histological and mRNA level, no myofibroblasts could be detected. Therefore, we suggest that the reduction of contraction is caused by the pericellular coating of fibroblasts by hyaluronan.
Chapter 8 is the general discussion of the results summarized above. Also, the experimental design is discussed. Finally, general conclusions and suggestions for future research are given.