Cecilie Bækgård: Cellular and molecular signatures for forensic age assessment of granulation tissue

Phd student: Cecilie Bækgård
Email: cecb@sund.ku.dk

www.linkedin.com/in/cecilie-bækgård-8014aa215

Date defended:  10 September 2026

THE PROJECT
Every year 50-60 veterinary forensic cases of potential animal neglect are submitted for investigation at the University of Copenhagen. Around 80% of these cases concern pigs, and approximately a third of these are skin wounds. An accurate wound age estimate is crucial, as it will give an indication of the degree of neglect, i.e. how long the animal has suffered from the wound. During wound healing granulation tissue, which is a highly vascularized and cell-dense, is formed. Current methods of wound age estimation is based on the thickness of this granulation tissue and histological assessments of the tissue. However, granulation tissue thickness has been shown to peak on day 10 in experimental wounds and histological assessments are, in part, based on subjective opinions and rely on the level of experience of the examining veterinary pathologist. Often the wound age estimate is given as “several days”, “several weeks” or “several months”. Therefore, there is a need to obtain more accurate and objective wound age estimates.

Additionally, obtaining an accurate age estimate is further complicated by the poor quality of veterinary forensic samples. The pig has often been through the slaughter process where the skin is singed, scraped and scalded, hereby removing the other layer of the skin. Furthermore, the relevant parts of animal are frozen for an extended period of time prior to forensic investigation.

THE PURPOSE
The project aimed to identify age-dependent cellular and molecular markers in experimental granulation tissue aged 5-35 days by applying different laboratory approaches (gross- and histopathology, immunohistochemistry combined with digital pathology, flow cytometry, qPCR, MALDI-MSI lipidomics and LC-MS proteomics).

Additionally, the purpose of the project was to apply some of these markers to veterinary forensic samples of unknown age to test wound age estimation.

THE RESULTS
In one paper, we applied a multidisciplinary approach. Using gross pathology, we observed granulation tissue from day 5 and a peak in granulation tissue thickness on day 10. Histology was used to identify several time-dependent indicators of age, including infiltration of neutrophils and macrophages, re-epithelialization and granulation tissue presence and maturation. Using mmunohistochemistry combined with digital pathology we identified time-dependent expression of the cellular marker CD45. Flow cytometry was used to show time-dependent expression of the cellular marker CD105.

In another paper, we used qPCR to establish a panel of robust primers that could be used even on highly degraded RNA. The primers were designed based on genes known to play a role throughout the wound healing period. This panel was tested on intentionally degraded samples, and a final panel of 24 robust primers were tested on experimental granulation tissue and veterinary forensic samples of unknown age. The panel could distinguish the experimental samples according to age, however, gene expression of the forensic samples diverged from the experimental samples, making an age estimate unfeasible.

In the last paper, MALDI-MSI lipidomics was used to investigate spatial and temporal expression of lipids in granulation tissue of three different ages. Additionally, LC-MS proteomics was utilized to identify temporal expression of proteins of the same experimental samples. No significant patterns were observed in the spatial expression of the identified lipids. However, MALDI-MSI and LC-MS identified temporal expression of several lipids and proteins, respectively. In total, 34 lipids and 97 proteins were discriminatory for age.

THE FUTURE
Many of the methods utilized in this PhD project, including immunohistochemistry, flow cytometry and qPCR, are sensitive to sample quality. To implement these methods in future veterinary forensic investigations, current protocols for how veterinary forensic cases are handled and sampled needs to be revised.

Additionally, the data and results obtained in this project will be implemented in a Bayesian network model. This type of model utilize experimental variables and probabilities to give an age estimate of a veterinary forensic wound. Furthermore, it will be able to tell which part of a data set contributed the most to the wound age estimate.