Immunohistochemical expression of HMGB1 and related proteins in the skin as a possible tool for determining post-mortem interval: a preclinical study

Determining the post-mortem interval (PMI) is one of forensic pathology’s primary objectives and one of its most challenging tasks. Numerous studies have demonstrated the accuracy of histomorphology and immunohistochemical investigations in determining the time of death. Nevertheless, the skin, a robust and easy-to-remove tissue, has only been partially analyzed so far. By studying 20 adult male mice, we tried to determine whether post-mortem immunohistochemical detection in the skin of HMGB1 proteins and associated components (Beclin1 and RAGE) could be used for this purpose. We discovered that nuclear HMGB1 overexpression indicates that death occurred within the previous 12 h, nuclear HMGB1 negativization with high cytoplasmic HMGB1 intensity indicates that death occurred between 12 and 36 h earlier and cytoplasmic HMGB1 negativization indicates that more than 48 h have passed since death. RAGE and Beclin1 levels in the cytoplasm also decreased with time. The latter proteins’ negativization might indicate that more than 24 and 36 h, respectively, have passed from the time of death. These indicators might potentially be helpful in forensic practice for determining the PMI using immunohistochemistry.


Introduction
One of the main goals of forensic pathology, as well as one of its most difficult tasks, is the determination of time since death or post-mortem interval (PMI).Undoubtedly, this parameter is of crucial importance, especially in relation to criminal investigations, yet many studies on the topic discuss only the time dependency of post-mortem parameters that have little application to real forensic practice [1][2][3].Despite the development in recent years of several new approaches to the assessment of PMI [3,4], traditional methods, based on the study of rigor, algor and livor mortis, are still the most commonly used.These factors, however, are largely the result of physical and chemical processes that occur in the post-mortem period and may be affected by a wide range of individual and environmental factors (i.e.ambient temperature, age, gender and physiological and pathological states).As a result, traditional approaches are often characterized by inaccuracy, a lack of reliability and consequently limitations in their application [1][2][3].
The new approaches aiming to develop more precise PMI estimates vary not only in terms of the biological processes considered but also in terms of their scientific rigour and the data validation methods applied.In a recent review, Gelderman et al. [5] examined, in the light of the Daubert criteria, the reliability of several approaches used to estimate PMI.Of all the approaches considered, only Henssge's nomogram and forensic entomology met the required criteria.The need to overcome the limitations of methodologies currently used for calculating PMI and to reduce the temporal uncertainty, which is generally far too wide, has prompted us to consider immunohistochemistry as a potential approach that may provide more objective data for estimating time of death.
The skin is more resilient and robust than other soft tissues, and it may be able to withstand abiotic and transformative processes better, thereby providing significant data for PMI assessment.Furthermore, skin is readily accessible for sampling, and immunohistochemistry investigations are relatively easy to carry out and do not require advanced investigative procedures.Skin tissue analyses have the potential to provide useful information concerning not only PMI but also lesions and pathologies related to cause of death.
Post-mortem transformation typically causes cell necrosis, which results in the release of the chromatin protein high mobility group box-1 (HMGB1) [15,23,24].
HMGB1 is found in many eukaryotic cells and has an amino acid sequence that is substantially conserved across species.Its biological roles include that of intracellular transcription regulator, and following necrosis, it translocates outside of the nucleus and is released by macrophages.To date, only a few studies have been conducted with the aim of investigating the role of this biomarker in the context of PMI estimation [23,24].
The purpose of this study was thus to examine the possible applicability of post-mortem histological alterations in the skin, in conjunction with the immunohistochemical detection of HMGB1 proteins and related factors (Beclin1 and RAGE), for estimating the time elapsed since death, using 20 adult male mice.

Animal specimens
Twenty adult male albino mice were used in this study (age ranged from 8 to 9 months).The mice were classified into two groups (10 mice per group) based on the time and day of post-mortem skin biopsy collection.All 20 animals were dissected to obtain full-thickness skin samples at different intervals (0, 12-, 24-, 36-and 48-h post-mortem).Therefore, 4 animals for each time point were selected.

Histopathology and immunohistochemical study
The skin tissue samples of 20 mice were obtained at 0, 12-, 24-, 36-and 48-h post-mortem (hpm).Mice were sacrificed by cervical dislocation and all the skin biopsies were taken from the dorsal region of the animals.All samples were stained with hematoxylin and eosin (H&E) and anatomical integrity was verified.Subsequently, immunohistochemical analyses were carried out on serial sections, using the following antibodies: Serial sections of 5-7 μm were obtained from each sample, and the most representative sections were selected after observation with H&E staining.All sections were then washed with xylene and fully rehydrated using a sequence of decreasing alcohols before being washed with phosphatebuffered saline (PBS).PBS was used for all subsequent washes and for antibody dilution.Tissue sections were sequentially treated with 3% hydrogen peroxide in aqueous solution and blocked with 6% milk in PBS.The slides were then incubated for 1 h at room temperature with each of the specified antibodies, at a final dilution of 1:100.Following three PBS washes aimed at removing excess antibodies, the slides were incubated with the UltraTek HRP secondary antibody (ScyTek Laboratories, Logan, UT, USA) for 1 h at room temperature.The ABC technique (Vector Laboratories) was applied to all slides for 30 min at room temperature.Diaminobenzidine (ScyTek Laboratories, Logan, UT, USA) was used as the final chromogen, and hematoxylin was used as a contrast agent.For each tissue section, a negative control was generated without the primary antibody.All samples were processed under the same conditions.The cellular expression levels of HMGB1, Beclin 1 and RAGE per field (10X) were calculated under the microscope, with two different observers comparing samples, and they were characterized as follows: score 0 (absent), score 1 (low or moderate) and score 2 (high).An average of 22 fields was observed for each sample.The temporal evolution of the HMBG1 was statistically tested with the Jonckheere's trend test.The level of concordance, expressed as the percentage of agreement between the observers, was 95%.For the remaining specimens, the score was obtained after collegial revision and agreement.

Results
We examined the skin samples of 20 adult male mice with the aim of identifying possible patterns in the post-mortem behaviour of specific proteins, particularly HMGB1, RAGE and Beclin-1.Significant quantitative alterations were found in the expression of the proteins studied, as well as a signal translocation for HMGB1 from the nucleus to the cytoplasm.At time 0 hpm, the expression of HMGB1 was found to be high (score 2) and at nuclear localization.Twelve hours after death, HMGB1 expression was no longer at the nuclear level, but rather at the cellular level, still with high intensity.At 24 and 36 hpm, both the type and degree of expression remained unchanged.However, at 48 hpm, the intensity of expression was reduced relative to the previous time periods, although it was still cytoplasmic; this trend was considered significant (p < 0.05) at the Jonckheere's (p = 5.995*10^-16).Figure 1 shows an example of the HMGB1 expression in the skin epithelium.
Figure 2 shows the temporal evolution of the HMGB1 classes for each subject.The expression patterns observed for RAGE and Bec-lin1 were different.At time 0, RAGE expression was high and cytoplasmic in the skin epithelium.At 24 and 36 hpm, the expression was still cytoplasmic, but it was found to be lower than that observed at time 0. Finally, after 48 h, the protein's expression was completely absent (Fig. 3).
In contrast, at time 0, Beclin1 expression was low in skin appendages (particularly at the level of the sebaceous glands) and almost completely absent in the skin epithelium.At 24 and 36 hpm, Beclin1 expression was high in the appendages and low in the epithelium.Finally, after 48 h, the protein was expressed in neither the appendages nor in the skin epithelium (Fig. 4).
Figure 5 shows the trends of cytoplasmic expression in mice skin and appendages for HMGB1, Beclin1 and RAGE.The intensity of HMGB1 at the cytoplasmic level peaked between 12 and 36 hpm, after which, the marker decreased and was no longer visible after 48 h.The other markers, on the other hand, were always present in the cytoplasm, albeit in varying amounts, depending on the different times of the analyses.The negativization of these markers, which occurred at 36 hpm for RAGE and at 48 hpm for Beclin1, is noteworthy (Fig. 5).
All the results of the immunohistochemical examination, including the different expression levels and the sub-cellular localizations of the proteins analysed at the different time points are presented in Table 1.

Discussion
The traditional methods of forensic practice are still commonly utilized for estimating PMI.However, these methods are limited in their applicability as the indicators rigor, livor and algor mortis are derived primarily from post-mortem physical and chemical processes and are susceptible to modification by a range of intrinsic and extrinsic factors.This often leads to inaccuracies in interpretation.Any anomalous or unusual course that impacts one or more cadaveric phenomena, which might also mutually impact each other, will have chronological implications and increase the likelihood of inaccuracy.As time passes, the assessment of PMI becomes increasingly approximate.
In recent years, several experimental approaches have been explored to establish a more precise method for assessing PMI.These techniques differ both in nature and in terms of their scientific rigour.A review by Gelderman et al. [5] investigates the reliability of several approaches that are reported in 94 recent Dutch judicial proceedings in which time of death was requested.The methods employed to determine PMI are evaluated in the light of Daubert's criteria.Only algor mortis, as measured by the Henssge nomogram, and forensic entomology provided scientifically sound results that met Daubert's criteria.
Histopathology has played an important role in the forensic field and is usually primarily used to establish the cause of death .Over the years, however, numerous novel approaches have been developed for estimating PMI, with some addressing the domains of histology and immunohistochemistry. Table 2 provides an overview of the literature on the topic.Ideally, a complete evaluation of all measurable changes in the process of human tissue and organ decomposition would reveal the presence of a distinct sequence, which could be used to calculate PMI.However, in practice, this is hard to recognise.Wehner et al. [6][7][8][9] investigated whether positive immunoreactions to various antigens (i.e.insulin, glucagon, thyroglobulin and calcitonin) correlated with PMI, assuming that antigen structure changes post-mortem and that the efficacy of protein denaturation staining will decrease with increasing PMI.Among studies that have    investigated human tissue and organs, some have focused on the macroscopic and histological changes occurring, post-mortem, in the skin, and these have occasionally been used to determine the time of death under varied settings.For instance, Kovarik et al. [11] investigated the macroscopic and microscopic appearance of the skin in three deceased individuals during the first post-mortem week, under specific climatic conditions.For each case, skin biopsies were obtained from four different sites at 12-to 24-h intervals.The authors observed three major histological changes, including isolated dermal-epidermal separation, eccrine duct necrosis and dermal degeneration, which could potentially be used to estimate PMI in the early post-mortem period.Studies have also shown that skin tissue can provide important information on lesions and diseases, which are frequently associated with both cause of death and PMI.Because the skin is easier to sample and more resilient than other soft tissues, it has the potential to challenge abiotic and transformative factors, thereby providing significant data for measuring PMI.However, only a few studies dealing with the use of skin markers for the determination of PMI have been reported in the literature.Furthermore, as previously mentioned, individual and environmental variables commonly influence PMI and, as a result, restrict the applicability of these approaches.The present study therefore considers animal studies, as animal models are often able to illustrate the basic principles of biological processes.El-Din et al. [15] investigated the possible association between post-mortem skin abnormalities and HMGB1 changes as well as PMI, through the immunohistochemical staining of serum and skin, using both animal (40 adult male albino rats) and human (40 cases) samples.Human forensic autopsies were performed within the first 24 hpm on deceased individuals with a known time of death, whereas mice were dissected at 0, 3, 6, 12 and 24 hpm.The authors identified a specific pattern of HMGB1 expression in mice skin tissue: at 0 hpm, HMGB1 showed a weak positive immunoreaction, a mild positive reaction at 3 and 6 hpm, and moderate and strong positive immunoreactions at 12 and 24 hpm, respectively.These findings revealed a considerable timedependent increase in serum HMGB1 levels, as well as its overexpression in immunohistochemically stained skin tissue, suggesting that HMGB1 might be a reliable postmortem marker.
In this study, we investigated the possibility of estimating PMI using changes in the immunohistochemical expression of the HMGB1 protein and its related factors, Beclin1 and RAGE.HMGB1 is a nuclear, non-histone DNA-binding protein, characterized by high electrophoretic mobility on polyacrylamide gels.In mammals, there are four distinct HMGBs (HMGB1-4).However, HMGB1 is by far the most prevalent and ubiquitously expressed.It is a highly mobile protein found inside the nucleus of cells, where it modulates chromatin structure and increases the accessibility of binding sites to regulatory elements, such as transcription factors and nucleosomes [31][32][33][34].
HMGB1 comprises 214 amino acid residues, with a sequence identity of 99% in mammals.The molecule is divided into two positively charged DNA-binding domains, known as HMG-box A and box B, as well as a strongly negatively charged C-terminal tail, containing 30 glutamic and aspartic acid residues that are repeated 30 times.Lysine residues account for 20% of the amino acids in the whole molecule.Because of the sequential order and composition of a high number of negatively and positively charged amino acids, HMGB1 is a unique bipolarly charged molecule, a feature that earned it the additional designation "amphoterin" [31][32][33][34][35].
When HMGB1 is passively released after cell death or actively secreted into the extracellular space, it becomes a strong mediator of inflammation.Nuclear HMGB1 translocates from the nucleus to the cytoplasm in response to cell activation or injury, where it participates in inflammasome activation and pyroptosis, promoting autophagy by binding to Beclin1 and inhibiting apoptosis.Although numerous aspects of these key intracellular activities have been clarified, there is still more to learn about HMGB1 biology, which is vital for both cell survival and cell death.

Conclusions
The only current study in the literature that deals with analysing the trend of HMGB1 at intracellular levels is that by El-Din et al. [15].Although these authors observed a decrease in the cytoplasmic levels of the HMGB1 marker as time passed after death, they did not characterize its translocation from the nucleus to the cytoplasm.However, in our investigation, this translocation was highlighted and seemed to occur around 12 h after death.Thus, the overexpression of nuclear HMGB1 indicates that death occurred within the last 12 h, the negativization of the marker at the nuclear level with a high intensity of the marker at the cytoplasmic level indicates that death occurred between 12 and 36 h previously, while the negativization of the marker at the cytoplasmic level suggests that more than 48 h have elapsed since death.
Beclin1 and RAGE do not appear to have been investigated previously as potential PMI indicators.In this study, we found that the cytoplasmic levels of these proteins (in keratinocytes and annexes) decreased over time.The negativization of RAGE and Beclin1 may imply that more than 24 and 36 h have passed since the time of death, respectively.There is a scarcity of evidence in the literature on post-mortem skin changes relating to time since death.The current study aimed to perform a semi-quantitative analysis of three separate intracellular protein markers: HMGB1, RAGE and Beclin1.Based on the findings of this investigation, these may all be useful markers for estimating PMI using an immunohistochemical approach.Consequently, it may be inferred that these results have laid the groundwork for supporting immunohistochemistry research in the context of thanatochronology.The application of the procedure to a larger case series, taking into account additional PMIs, as well as the implementation of a similar study on human tissues, would be beneficial in generating reliable and practical protocols to be used in a forensic routine.

Key points
1. Immunohistochemistry can be used as a method for determining the PMI 2. Post-mortem skin samples of 20 adult mice were analysed using immunohistochemistry 3. Skin HMGB1 proteins and associated components could help with PMI determination 4. Different expression patterns of the studied proteins correspond to different PMIs

Fig. 2 Fig. 3
Fig. 2 Alluvial graph: the graph shows the temporal evolution of the HMBG1 classes for each subject, at the different postmortem time points (0, 12, 24, 36, 48 h)

Fig. 5
Fig. 5 Trends of cytoplasmic expression in mice skin and appendages

Table 2
Post-mortem histomorphological and immunohistochemical investigation as a tool for the PMI estimation: a review of the literature on this topic