Stimulation of angiogenesis and osteogenesis in a dynamic co-culture model of bone tissue engineering
Date issued
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License
Abstract
The development of a successful strategy for the generation of bone tissue
engineering constructs in vitro is highly dependent on the integration of a complex
vascular system. For this purpose, co-cultures (CC) of stromal cells with endothelial
cells (EC) have been developed and evaluated in order to stimulate the formation of
vessel-like structures (VLS) ex vivo. In this context, the time for in vitro processing
aiming at tissue formation has to be optimized according to the overall time span and
maximum therapeutic outcome of the chosen strategy. Furthermore, the application
of state-of-the-art technologies such as perfusion bioreactors can additionally support
the cells within a generated construct. Therefore, the present study aimed at
developing a cultivation strategy that permits the generation of prevascularized
osteogenic CC samples in vitro.
An initial evaluation of osteogenesis in mesenchymal stem cell mono-cultures
(MSC MC) found a cell seeding density (CSD) of 20,000 MSC/cm2 to result in the
highest degree of osteogenic differentiation. When evaluating osteogenesis in
MSC MCs on 3D trabecular nickel-titanium alloy (tNiTi) scaffolds in a perfusion
bioreactor it was found that medium perfusion supports the osteogenic differentiation
of MSC. In contrast, medium perfusion as sole stimulus was not sufficient to induce
osteogenesis in human MSC. These results represent a starting point for the
evaluation of a CC system of MSC with EC under comparable dynamic conditions.
However, primary EC are limited in their angiogenic activity and total available
numbers upon expansion in vitro. Therefore, strategies focusing on the stimulation of
osteogenesis and VLS formation in a common CC system were initially performed in
2D. This should allow for an initial screening, the adequate optimization and
evaluation of individual factors of the culture system, as well as the determination of
their respective impact. The initial evaluation of the cell seeding ration (CSR) in CC of
MSC with human dermal microvascular EC (HDMEC) demonstrated a remarkable
effect on the formation of VLS. VLS formation was found to reach a maximum when
a CSR of 3:2 (MSC:HDMEC) was combined with a MSC CSD of 20,000 MSC/cm2.
Interestingly, this CSD also resulted in maximum osteogenesis in MSC MCs.
The following screening studies aimed at examining the combination of proosteogenic
and pro-angiogenic factors in order to simultaneously stimulate
osteogenesis and VLS formation in the CC system. The results clearly indicated that
osteogenic supplements (OS) suppress VLS formation while pro-angiogenic
supplements (HF, heparin+bFGF) suppress osteogenic differentiation by the
incorporated MSC fraction. The interaction between OS and HF indicated that a
sequential strategy involving separate stimulation of osteogenesis in MSC and the
formation of VLS might be superior to a simultaneous stimulation strategy. For this
purpose, two sequential strategies were evaluated. In a first strategy MSC MCs were
primed in osteogenic differentiation medium (ODM) to induce osteogenic
differentiation and subsequently evaluated for their capacity to induce VLS formation in a CC with EC. The results clearly demonstrated that osteogenic priming of MSC in
ODM reduced their potential to induce the formation of VLS in CC with EC, most
likely due to the dexamethasone (Dex)-mediated reduction of VEGF expression by
the MSC. The observed effects were found to be independent of the initial number of
MSC present in the CC or of osteogenic matrix that was formed during the
osteogenic priming process. However, since VEGF expression is required for the
induction of VLS formation in vitro a second sequential cultivation strategy focused
on the induction of VLS formation before osteogenesis by the incorporated MSC
fraction was stimulated. This approach reliably resulted in the generation of CC
samples that contained pre-formed vascular networks embedded within an
osteogenic matrix. Analysis of other factors produced in the model indicated that
inhibition effects on the VLS formation were found upon exogeneous phosphorus
supplementation and that this could be reversed by re-switching the CC to a proangiogenic
medium. At the same time, the formation of osteogenic matrix could be
maintained. Furthermore, the results showed that the observed rapid osteogenic
differentiation of MSC within prevascularized CC is dependent on the presence of but
independent of the type of EC. Similarly, the rapid osteogenic induction could be
obtained when exchanging bone marrow-derived MSC (bmMSC) with adipose tissuederived
MSC (atMSC). Nevertheless, by varying the type of ODM it could be
demonstrated that the outcome of the second sequential cultivation strategy does not
represent an artifact induced as a response to a specific differentiation medium.
Altogether, the present study established a reproducible strategy for the generation
of a prevascularized CC containing an osteogenic matrix within a well-defined time
frame of less than 14d. At the same time, the evaluation of the opposite sequential
strategy or the simulateous stimulation strategy revealed reducing or inhibitory
effects of medium components on either vascularization and angiogenesis or
osteogenesis. Together with the findings of enhanced osteogenesis in MSC MCs on
3D scaffolds by medium perfusion the present study defined conditions applicable for
the transfer of the promising sequential cultivation strategy into a 3D dynamic
environment.