Connective tissue fibroblast properties are position-dependent during mouse digit tip regeneration

Abstract

A key factor that contributes to the regenerative ability of regeneration-competent animals such as the salamander is their use of innate positional cues that guide the regeneration process. The limbs of mammals has severe regenerative limitations, however the distal most portion of the terminal phalange is regeneration competent. This regenerative ability of the adult mouse digit is level dependent: amputation through the distal half of the terminal phalanx (P3) leads to successful regeneration, whereas amputation through a more proximal location, e.g. the subterminal phalangeal element (P2), fails to regenerate. Do the connective tissue cells of the mammalian digit play a role similar to that of the salamander limb in controlling the regenerative response? To begin to address this question, we isolated and cultured cells of the connective tissue surrounding the phalangeal bones of regeneration competent (P3) and incompetent (P2) levels. Despite their close proximity and localization, these cells show very distinctive profiles when characterized in vitro and in vivo. In vitro studies comparing their proliferation and position-specific interactions reveal that cells isolated from the P3 and P2 are both capable of organizing and differentiating epithelial progenitors, but with different outcomes. The difference in interactions are further characterized with three-dimension cultures, in which P3 regenerative cells are shown to lack a contractile response that is seen in other fibroblast cultures, including the P2 cultures. In in vivo engraftment studies, the difference between these two cell lines is made more apparent. While both P2 and P3 cells participated in the regeneration of the terminal phalanx, their survival and proliferative indices were distinct, thus suggesting a key difference in their ability to interact within a regeneration permissive environment. These studies are the first to demonstrate distinct positional characteristics of connective tissue cells that are associated with their regenerative capabilities.

Figure 1. Isolation of connective tissue cells from the P2 and P3 regions.

A, B) Histological section of the P2 (A) and P3 (B) bone after dissection. Loose connective tissue is attached to the surface of the intact bone. C, D) After enzymatic digestion the majority of loose connective tissue was digested off the bone while tissues within the bone marrow are still present. E) DiI labeled cells (red) within the connective tissue (CT) dorsal to the P3 skeletal element (B) are clustered in the amputation stump 2 days post-injection. BM, bone marrow. F) Blastema stage regenerate at 13 DPA showing DiI labeled cells (red) scattered throughout the blastema but not overlapping with Osteocalcin immunohistochemical labeling (green) to identify regenerating osteoblasts. The blastema is contiguous with the connective tissue (CT) and bone (B) of the stump. Scale bars  = 200 µm.

Figure 2. Flow cytometric profiling and RT-PCR analysis of P2 and P3 cells.

A) Flow cytometry analysis identify a similar signature for P2 (bottom panels) and P3 cells (top panels). Analyses were performed in series on the same samples, P1-P4 denote the order of the measurements. In the scatter plots, both cell lines are positive with a single peak for Sca-1, CD29, V-CAM and negative for CD45, C-kit, CD49f, CD104 and CD34. This signature is similar to other characterized MSC lines. B) RT-PCR results from both P3 and P2 cells expanded for 30 days compared with established cell lines. Digit regeneration marker genes Bmp4 and Msx1 are stably expressed in P3 cells and variably expressed in P2 cells. Stem cell marker genes are differentially expressed in P2 and P3 cells. Oct4 is not expressed in P2 and P3 cells. Rex1 is minimally expressed in P2 and P3 cells. Sox2 is minimally expressed in P2 cells and in early passage P3 cells but strongly expressed in the P3 cell line.

Figure 3. Position-specific characterization of P2 and P3 cells.

A) Hematoxylin and eosin stained sections of skin equivalents generated from P2 cells co-cultured with human keratinocytes showing the induction of stratified sheets of differentiating epidermis. B) Hematoxylin and eosin stained sections of skin equivalents generated from P3 cells co-cultured with human keratinocytes showing the induction of aggregate structures that display keratosis similar to that observed from cultures of nail matrix epidermis. C) Whole mount preparations of P3 skin equivalents stained with Rhodanile blue showing the aggregation of keratinocytes induced by P3 cells. D) Collagen gels seeded with P2 cells (top) resulted in a contracted gel phenotypes whereas gels seeded with P3 cells did not display a contraction response. E) Area measurements of collagen gels show that P2 cells induced a cell density dependent contraction response, whereas P3 cells failed to contract the gel at similar seeding densities. F) Co-cultures of P2 cells with human keratinocytes enhanced the contraction response which was measured by determining the width of the collagen gel. Keratinocyte co-cultured with P3 cells induced a contraction response that was of a similar magnitude. All chart measurements are mean ± SEM (n = 3). A and B, scale bar  = 100 µm; C, scale bar  = 300 µm.

Figure 4. P3 cells are regeneration competent after expansion in vitro.

A) LacZ positive P3 cells were injected into the digit tip of SCID mice 1 day prior to amputation and collected at 10 DPA when the regenerate at the blastema stage. LacZ positive cells are present at the injection site in the dorsal connective tissue (*) and are scattered throughout the blastema. B) During the differentiation stage (16 DPA) LacZ positive P3 cells are primarily found in the regenerating connective tissue with small clusters of cells present within the trabeculae of the regenerating bone (arrows). C–E) GFP+ human breast cancer cells injected into P3 connective tissue prior to amputation remained aggregated in the regeneration stump and did not enter the blastema. GFP positive cells (C) shown aggregated in the stump of a 16 DPA regenerate indicating that they survive engraftment. Immunohistochemical co-staining for the endothelial marker vWF (D, E) indicates that these cells differentiate in situ without participating in the regenerative response. A and B, scale bar  = 200 µm; C–E, scale bar  = 100 µm.

Figure 5. P2 cells participate in a regenerative response.

A) Engraftment of LacZ labeled P3 cells into the dorsal P2 region (arrow) prior to amputation shows labeled cells (blue) at 16 DPA participate with non-labeled dermal cells in the wound healing response. B) Control engraftment of LacZ labeled P2 cells into the dorsal P2 region (arrow) prior to amputation shows labeled cells (blue) at 16 DPA also participate with non-labeled dermal cells in the wound healing response. The amputated P2 bone (B) and bone marrow (BM) are outlined in A and B. C) Engraftment of LacZ labeled P2 cells (arrow) into the dorsal P3 connective tissue (CT) prior to digit amputation shows labeled cells (blue) at 10 DPA participating in blastema formation. D) During the differentiation stage (16 DPA) LacZ positive P2 cells (blue) are found in the regenerating connective tissue (CT) and do not associate with the trabeculae of the regenerating bone. The amputated P3 bone (B), bone marrow (BM) and regenerating nail (N) are outlined in the section to demarcate the stump and the regenerate in C and D. A–D, scale bar  = 200 µm.

Figure 6. P3 cells proliferate in the blastema.

A–C) GFP and Ki67 co-immunohistochemical reveal P3 isolated cells engraft and proliferate. A) GFP labeled P3 cells are identified within the digit. B) Ki-67, a marker for proliferation, is expressed in engrafted and endogenous P3 cells. C) Merged GFP and Ki67 expression identifies a fraction of engrafted P3 cells that are proliferating. Double labeled cells shown in C are indicated with arrows in the image set. GFP- red, Ki67- green, DAPI nuclear stain- blue. Scale bar  = 50 µm. D) Quantification of the proliferation indices of P2 and P3 engrafted cells compared to endogenous neighboring cells. Left: In unamputated digits 17 days post engraftment, P3 cells proliferated at a rate similar to neighboring control cells, whereas P2 cells are non-proliferative. Right: P3 cells participating in blastema formation at 10 DPA display a proliferation index that was lower than neighboring endogenous cells but significantly higher than P2 cells within the blastema. All chart values are expressed as means ± SEM. The proliferative index of P3 cells is significantly greater than P2 cells in both studies (**, p-value <0.005).