HISTOLOGIC AND IMMUNOHISTOLOGIC ANALYSES

Hematoxylin-phloxin-saffron (HPS) staining was performed on 5 µm deparaffinated slices. The hematoxylin is a basic dye that stains the nucleus of the cells blue. The phloxin is an acid dye staining the cytoplasm pink, and the saffron stains the connective tissue orange.

The samples were also embedded in Tissue Teck; 7 µm slices were next fixed in isopropanol and then stained with a solution of Oil Red O (Sigma), an adipocyte-specific dye which stains the lipid-filled vacuoles red.

Slices 5 µm thick were deparaffinated and then incubated with anti-mouse collagen I antibodies. We used anti-mouse IgG antibodies coupled to peroxidase to detect the immunocomplexes, using diaminobenzidine as the substrate. Counterstaining was performed using Harris hematoxylin. This labeling enabled us to detect cells of mouse origin. Human skin was used as a negative control.

Other slices were incubated with anti-human collagen I specific primary antibodies, followed by a secondary antibody coupled to a fluorochrome to reveal the binding sites of the primary antibody. These could thereby be observed under a fluorescent microscope to visualize collagen fibers of human origin.

HISTOLOGIC AND IMMUNOHISTOLOGIC RESULTS

Eight weeks after implantation, all 10 mice had a left parascapular growth corresponding to the site of transplantation of the adipose tissue. Histologic analysis demonstrated the presence of mature adipocytes stained by Oil Red O. The HPS staining revealed a considerable increase in density of the extracellular matrix surrounding the adipose tissue and between it and the dermis. The dermis also showed an increased thickness after injection of adipose tissue. This increase in density of the extracellular matrix was made up of collagen I fibers, an absence of which could be noted on the control noninjected side.

Histologic slices of control skin (A) and skin with grafted adipose tissue (B) are shown. Note the thickening of the dermis and the abundance of extracellular matrix between skin and adipose tissue, as well as that surrounding the fat lobules after the transfer of adipose tissue. Immunohistochemistry, using an anti-mouse collagen I antibody to label in brown the extracellular matrix, shows that the mass of collagen fibers are of murine origin (C and D). The negative control (human skin) shows an absence of labeling at this level (E). Adipocyte viability is confirmed by Oil Red O staining (F).

Immunohistochemical analysis using a labeling technique with anti-mouse collagen I allowed us to demonstrate the origin of the newly formed cells. The neosynthesized extracellular matrix was found to be of murine origin and not human, confirmed by the human skin negative control. Concordantly, immunofluorescence using an anti-human collagen I antibody confirmed the murine origin of collagen fibers, since no labeling was found in the mice.

This immunofluorescent slice of mouse skin following fat tissue grafting (A) shows the absence of human collagen fibers after labeling with an anti-human collagen I–specific antibody. In contrast, the human skin (B) shows a considerable density of human collagen fibers, stained green.

CONCLUSIONS

The results of our experimental study in mice agree with clinical observations made in patients treated by the same technique. Adipose tissue grafts survive and have a volume-increasing effect. Moreover, the presence of human adipose tissue increases murine dermis thickness at the recipient site by induction of collagen neosynthesis, likely as a result of the introduction of stem cells, growth factors, and cytokines. This collagen, mostly type I but also types V and VI, is of murine origin. Our study also demonstrated that adipose tissue can stimulate production of collagen by the host cells at the recipient site. This is supported by other studies showing that ADSCs present in adipose tissue promote proliferation and migration of fibroblasts and activation of extracellular matrix synthesis. ¹³ – ¹⁶ However, our study did not allow us to determine whether this induction is directly caused by the adipose tissue or is the result of an indirect passage through an intermediary inflammatory reaction. The presence of the adipose tissue graft causes an increase in local vascularization, resulting in a new physiologic equilibrium at the recipient site. Together, these observations demonstrate active filling properties of adipose tissue. Further studies are needed to show the long-term effects of fat grafting at the recipient site.

METHOD

Human SVF was isolated from lipoaspirate obtained from a randomly chosen donor who was undergoing liposuction. Adipose tissue was digested with collagenase and the reaction stopped by adding Dulbecco’s Modified Eagle’s Medium (DMEM). Floating adipocytes were discarded, and cells from the stromal vascular fraction (SVF) were pelleted, rinsed, centrifuged, and counted before being mixed with noncrosslinked hyaluronic acid, Cytocare 532 (Cyt 532), for injection into nude mice. Cyt 532 is a noncrosslinked hyaluronic acid and a rejuvenating complex containing 50 components. It was used as a vehicle in our study because previous published work ²² showed that it was the best culture medium for ADSCs. Crosslinked hyaluronic acid prevents stem cells from multiplying and differentiating.

Freshly isolated cell suspensions containing ADSCs were cultured in ADSC medium until 80% confluence was reached. At confluence, cells were trypsinized and centrifuged. A portion of the ADSCs was obtained using the technique described earlier and these were frozen. The cells were thawed 1 week before surgery and cultured in ADSC medium.

Sixteen male nude 42-day-old mice were anesthetized, and surgery was performed under standard sterile conditions. Two circular full-thickness wounds 12 mm in diameter were created on the backs of the mice and a ring of silicone was sutured around the wound to prevent skin retraction. Each healing wound received 1.10 ⁶ ADSCs, or SVF with half the cells intradermally injected around the wounds into four injection sites, and the other half applied to the wound surface. Each nude mouse received a different treatment for each wound. Cyt 532 was used as vehicle for the cells.

Two sets of experiments were designed, comprising 32 wounds on 16 mice (two wounds per mouse). Five excisional wounds were carried out in each group as follows:

1. 
   

   Negative control (spontaneous healing)

   
   
2. 
   

   Vehicle alone (Cyt 532)

   
   
3. 
   

   Cultured frozen ADSCs (c.fr.ADSC + Cyt 532) cryopreserved at passage 1, then thawed

   
   
4. 
   

   SVF (SVF + Cyt 532)

   
   
5. 
   

   Cultured ADSCs (passage 2) (c.ADSC + Cyt 532) without cryopreservation before injection

Every day the mice were observed and digital images taken. The wound area was measured by tracing the wound margin and calculating the pixel area using image analysis software. The percentage of wound healing was calculated using reepithelialization surface measurements.

At day 27, all scar tissue was harvested with a rim of healthy normal skin for histologic analyses. Tissue samples were fixed, embedded in paraffin, and stained. Inflammation was evaluated on a semiquantitative scale. Sections were examined in a blinded fashion by two independent observers. Perfusion evaluation was done by means of laser Doppler imaging on day 27, when all wounds had closed and biomechanical tests were conducted to evaluate the properties of the scarring mouse tissue.