Introduction
In about 10% of all reconstructive operations caused by traumatic, resectional or congenital de- fects, bone grafts or bone substitutes are neces- sary. In the course of reconstructive surgery reaming debris has not been mentioned yet, be- cause it was designated as a necrotic material, which does not contain vital cells.
Autogenic bone grafting is still regarded as the golden standard to treat large bone defects. How- ever, collection of the autogenic material from the iliac crest requires a second site of surgery, which can be associated with complications such as pain, infections and haematoma [1, 2].
Bone substitutes comprise a wide range of ma- terials. Based upon their chemical and physical features, these materials closely mimic the miner- al phase of natural bone. These biomimetic prop- erties promote the process of bone healing. How- ever, in contrast to the natural bone, anorganic bone substitutes reveal no signs of biological ac- tivity due to the lack of growth or differentiation factors and cells, whose participation in the pro- cess of bone healing is of great functional impor- tance [3±5].
Therefore, new strategies have been developed to treat large bone defects. These innovations of tissue engineering are based upon the knowledge about complex interactions between osteoinduc- tive factors and cells in vivo. Consequently, bone healing is improved by using implants in combi- nation with autogenic or recombinant growth fac- tors and/or cells [6].
The clinical use of autogenic cells requires their collection prior to their expansion in vitro before growth on a suitable matrix can be carried out [6]. Cancellous bone [6±8] and bone marrow [9]
represent sources of osteoblastic cells and multi- potential mesenchymal stem cells (MSCs) [10±14].
Because of their differentiation capacity, MSCs have a therapeutic potential that is already used regarding the composition of newly developed bone graft substitutes [10].
Up to now, discussion of osteoregenerative ef- fects of reaming debris has been controversial. It has been assumed that the cells undergo necrosis during the reaming procedure due to temperature elevations [15] and the influence of mechanical forces [15±19]. In contrast, experimental studies in the sheep and a few clinical studies in humans refer to the osteoinductive potential of reaming debris in vivo. Additionally, in vitro culturing of ovine reaming debris [19] has proven the vitality of the cells and has confirmed their osteoblast-like properties. In accordance with these findings, clinical studies have documented that fractures of long bones showed distinctly shorter times of healing when treated with reamed nails, than frac- tures treated with non-reamed nails [20±25].
Although the cause of this phenomenon has still to be clarified, it has been assumed that reaming creates a high potential of regeneration caused by the osteoinductive effects of blood and bone mar- row, whereas dispersed bone fragments promote healing in an osteoconductive manner [26]. How- ever, it has been taken into consideration that reaming of bones bears the risk of fat embolism or it can cause the adult respiratory distress syn- drome.
Although there is no final scientific evidence for the assumption that reaming debris has stimu- latory effects on bone healing, it might represent an autogenous source of vital cells, which could be used in the scope of bone tissue grafting. With special regard to this potential clinical signifi- cance, cells of human reaming debris have been investigated in the present in vitro study. The pur- pose was to prove vitality of the cells by means of cell culturing and by transmission electron micro- scopy and ± if so ± to characterise the viable cells phenotypically by means of morphology, anti- genic properties and their capacity to differentiate in multiple cell lineages.
Differentiation Capacity and Characterisation of Cells Growing Out of Human Reaming Debris
K. Trinkaus, S. Wenisch, C. Siemers, R. Schnettler
Materials and Methods
Reaming debris (n=13) was collected from the reamer head and was immediately transferred into F12K medium. In the laboratory each speci- men was divided into four parts. The first part was used for transmission electron microscopic examination (Zeiss EM 109). These specimens ± weighing 1±1.5 g ± were embedded in Epon (Ser- va, Heidelberg).
The second part was placed directly into the culture dish. The standard medium contained F12K medium, 20% fetal calf serum (FCS), 0.05 U/ml penicillin and 0.05 lg/ml streptomycin. The third part was washed in Hank's balanced salt so- lution until the components of blood and fat were macroscopically removed. The fourth part of each specimen was treated with collagenase for 2 h at 378C before it was placed in the culture dish. The cells were placed in the standard medium until they were grown to confluence. The medium was replaced every week. Then the cells were trans- ferred into 24-well dishes.
In order to elucidate their differentiation reper- toire, cells were incubated in osteogenic medium (DMEM low glucose with L-glutamine, 10% FCS, 0.1 lM dexamethasone, 0.05 mM ascorbic acid-2- phosphate, 10 mM b-glycerolphosphate, 0.05 U/ml penicillin and 0.05 lg/ml streptomycin) and in adipogenic induction medium (DMEM high glu- cose with L-glutamine, 1 lM dexamethasone, 0.2 mM indomethacin, 0.01 mg/ml insulin, 0.5 mM 3-isobutyl-1-methyl-xanthine, 10% FCS, 0.05 U/ml penicillin and 0.05 lg/ml streptomycin).
After 3±4 days the adipogenic induction medium was substituted by adipogenic culture medium (DMEM high glucose with L-glutamine, 0.01 mg/
ml insulin, 10% FCS, 0.05 U/ml penicillin and 0.05 lg/ml streptomycin), in which the cells were incubated for 1 week. For chondrogenic differen- tiation, cell pellets were created by centrifugation and were incubated in chondrogenic medium (Cambrex, Apen) for 28 days.
After 3 weeks of incubation within the osteo- genic medium, the cells were stained with Von Kossa and alizarin Red S. Osteocalcin expression was investigated immunohistochemically (Chemi- con, Hampshire, UK). Additionally, cells induced to the osteoblastic phenotype were embedded in Epon for transmission electron microscopy.
The cells incubated within the adipogenic me- dium were stained with Oil Red O.
Cell pellets of the chondrogenic medium were embedded in paraffin and the expression of col- lagen II (Cambrex, Apen, anti-collagen type II, human) was investigated immunohistochemically.
Fluorescence-activated cell sorter (FACS) analy- sis was performed to determine the antigenic properties of the cells incubated within the stan- dard medium. In accordance with the investiga- tions of Deans and Moseley [27], Pittenger et al.
[13] and Hung et al. [28], the following surface markers were selected out of a long list of anti- bodies that are suitable to characterise MSCs:
CD29, CD44, CD90, CD105, CD106, CD71, CD34, and CD45.
For FACS analysis, the cells were harvested from the standard medium, washed in phosphate- buffered saline (PBS) and incubated on ice for 30 min with fluorescein isothiocyanate or phycoery- thrin-conjugated primary antibodies. Then they were resuspended in PBS and examined by FACS analysis.
Results
Transmission electron microscopy of reaming debris demonstrated that few free cells survived the reaming procedure (Fig. 2.4.1). These cells did not exhibit signs of degeneration or necrosis.
However, cells with distinct morphological altera- tions such as condensation of the nuclear chro- matin and the cytoplasm and disintegration of their membranes could also be detected.
By means of cell culturing, cell outgrowth be- came evident in all specimens (n=13). However, the time of appearance of the cells was dependent on the treatment of the reaming debris prior to the incubation within the standard medium.
Those parts of the specimens that were trans- ferred directly into the medium showed cell out- growth after 4±5 days. The cells revealed a large, flattened morphology.
In specimens treated with Hank's balanced salt solution, small spindle-shaped cells started to grow out of the bone fragments after 7 days, whereas in collagenase-treated samples outgrowth of cells could be seen for the first time after 2 weeks.
All in all, within all samples investigated, cell outgrowth started from the bone fragments (Fig.
2.4.2). Interestingly, even bone fragments from reaming debris that had been kept in culture for 6 months were still a source of outgrowing cells.
As shown in Fig. 2.4.3, cells could be also de- tected in regions of the culture dish where bone fragments were lacking, particularly in specimens that had not been treated with Hank's balanced salt solution. This is in accordance with the elec- tron microscopic investigation, which revealed free and intact cells in human reaming debris. It
58 K. Trinkaus et al.
should be mentioned that non-adherent compo- nents such as erythrocytes and bone fragments were removed subsequently due to the exchange of the medium.
Irrespective of the different treatment proce- dures prior to the incubation in the standard me- dium, three types of cells could be seen in all cul- ture dishes. The first were small cells revealing polygonal to round shapes (Fig. 2.4.4) adhered to the culture dish; they could also be seen as free- floating cells within the medium. The second, spindle-shaped cell type was larger in size (Fig.
2.4.5), while the third type was extremely large, of flattened morphology, often binucleate and inten- sively granulated (Fig. 2.4.6).
Fig. 2.4.1.
Reaming debris examined by transmission electron microscope.
aFree and intact cells.
bCells with signs of de- generation such as disintegration of chromatin, cytoplasm and membranes (arrow). N nucleus, C cytoplasm
Fig. 2.4.2.
Cells (arrows) are growing out of the small bone fragments of reaming debris
Fig. 2.4.3.
Cells also appear between erythrocytes (E) in re- gions of the culture dish without bone fragments. Presum- ably those are the free cells that have survived the reaming procedure as shown in Fig. 2.4.1a
Fig. 2.4.4.
Rapidly self-renewing cell (RS cell)
As the small cell type could also be identified as a non-adherent cell, the medium was centri- fuged. Then the remaining pellet was resuspended and placed back in the culture dish. After incuba- tion for a few days, all three cell types could be seen again within the expanding cell culture.
When the small cells were seeded in numbers less than five per cm
2they did not adhere or propa- gate.
Cells of different passages were transferred into the differentiation media in order to elucidate the differentiation capacity. The results have shown that the expansion through sequential passages did not influence their differentiation repertoire.
However, the cells did not differentiate adequately when plated at high densities.
After 1 week of incubation in the osteogenic me- dium, the majority of cells flattened and assumed a cubic shape. During the second week, the onset of extracellular matrix production became visible (Fig. 2.4.7), and intensive staining by means of Von Kossa, alizarin Red and osteocalcin could be seen after 4 weeks.
Moreover, it could be shown ultrastructurally that the cells induced to the osteoblastic pheno- type (Fig. 2.4.10) produced a matrix composed of collagen fibres, and mineral (Figs. 2.4.8 and 2.4.9).
After 6 days of incubation within the adipo- genic medium, the cells assumed morphological features of adipocytes and were stained positively with Oil Red O.
The cell pellets incubated within the chondro- genic medium revealed strong expression of col- lagen II (Fig. 2.4.11), and in a few regions the for- mation of chondrons could be observed.
The data of FACS analysis showed correspond- ing results even when the cells were harvested from different passages. However, the expression pattern of some surface markers (CD29, CD105, CD106) was obviously dependent upon the cell type (Table 2.4.1).
60 K. Trinkaus et al.
Fig. 2.4.5.
Spindle-shaped cell
Fig. 2.4.6.
Mature mesenchymal stem cells (arrows)
Fig. 2.4.7.
Cells from human reaming debris that have been incubated in osteogenic medium for 3 weeks. The cells grow adherently and the extracellular matrix (EM) lays above the cells
Fig. 2.4.8.
Semithin section of MSCs after incubation in os-
teogenic medium, stained with toluidine blue. The cells (ar-
rows) produce osteoid matrix
Discussion
In accordance with the results of Frælke [19], the present investigation has demonstrated that cells have survived the harsh reaming procedure.
Moreover, the cells have been characterised as MSCs due to their morphological and antigenic features and due to their differentiation capacity.
This result favours the idea of using human reaming debris as a tissue graft. Simple culture conditions enable the expansion of MSC in vitro:
after washing, the reaming debris can be natively transferred into the culture dish instead of isolat- ing the cells by density gradient centrifugation, as is commonly done in the case of bone marrow-
derived MSC [13]. It has been shown that cell out- growth is delayed but not altered when the wash- ing procedure is performed prior to the incuba- tion. Therefore, it might be assumed that rinsing of the samples removes not only non-adherent cells but also growth-promoting factors.
In all samples of reaming debris, cell out- growth started from the bone fragments. This finding is in accordance with the assumption that reaming promotes bone healing in both an os- teoinductive and an osteoconductive pattern [19].
Three kinds of morphologically distinct cells could be seen after culturing. Based upon the re- sults of Colter et al. [29], these cells are termed rapidly self-renewing (RS) cells, spindle-shaped cells and mature mesenchymal stem cells (mMSCs).
RS cells are characterised by their small size (in the present investigation approximately 7 lm), their high nucleus/cytoplasm ratio, and their round to polygonal shape. They adhere to the cul- ture dish, but were also found in suspension [28].
These small RS cells were capable of establish- ing a new population of stem cells when seeded
Fig. 2.4.9.
Ultrastructural view of the osteoid matrix shown in Fig. 2.4.8. Collagen fibres (arrows) and deposits of mineral (arrowheads) can be detected. Inset: typical banded collagen fibres
Fig. 2.4.10.
Ultrastructure of osteoblast-like cells of the os- teogenic medium. The cells have abundant rough ER (rER) and show many mitochondria (arrows) and lipofuscin granules (arrowheads). Inclusion bodies (I) of unknown origin could be detected. N Nucleus
Fig. 2.4.11.
Section of a cell pellet after 3 weeks of incuba- tion in chondrogenic medium. The collagen II is stained immunohistochemically, and is visualised by the brown-col- oured reaction product
Table 2.4.1.