In the era of COVID-19, accurate autopsies are crucial to determine the cause of death in decedents who test positive for SARS-CoV-2. Given the substantial pathological commonalities of cause of death from or with SARS-CoV-2, we investigated the possibility of a microbial fingerprint in organs from SARS-CoV-2 deaths. A human corpse subsists as a specialized disturbance habitat that selects for a distinct thanatomicrobiome structure capable of decomposing the host depending on the cause of death and abiotic and biotic factors surrounding the death (Can et al., 2014; Javan et al., 2016; Kaszubinski et al., 2020). SARS-CoV-2 binds to angiotensin-converting enzyme 2 (ACE2) receptors present on the surface of various cells in the body and can negatively affect essentially all organs of the host. Before death, there is strong evidence that COVID-19 affects the brain structure (Douaud et al., 2021), heart (Delorey et al., 2021), lungs (Elezkurtaj et al., 2021), and gut microbiome (Yeoh et al., 2021). Many postmortem molecular questions regarding the pathophysiology of COVID-19 infection have not been elucidated yet.

Brain imaging has demonstrated degenerative spread of COVID-19 via neuroinflammatory events (Yachou et al., 2020), olfactory pathways involving anosmia (Gori et al., 2020), or loss of sensory input of taste (Ludwig et al., 2022). After death, the brain demonstrates changes due to hypoxia, increased carbon dioxide levels, and cytokine storm (Jain, 2020; Douaud et al., 2021). Furthermore, the principal cause of death in SARS-CoV-2 infection is respiratory failure but cardiac indications also contribute largely to mortality. Studies have shown that abnormal echocardiography were present in up to 55% of all COVID-19 infected cases (Dweck et al., 2020). The contribution of the co-infection of SARS-CoV-2 and other microorganisms in cardiomyocytes remains unclear.

Bacteria and/or bacterial products often have direct and indirectly interactions with viruses that aid in pathogenicity. For example, respiratory syncytial virus (RSV) interacts with Streptococcus pneumonia, Pseudomonas aeruginosa, and Haemophilus influenzae to increase bacterial invasiveness and increase host cell adhesion molecules (Lian et al., 2022). Likewise, intracellular overgrowth of bacteria contributes to intracytoplasmic organelle damage which causes decreased viral antigen production during co-infection (Di Biase et al., 2000). Normal oxidative and inflammatory molecular pathways in the brain, heart, and lungs can drastically change during COVID-19 illness which could potentially allow proliferation of microorganisms and subsequent damage to organs. The current study seeks to ascertain the possibility of a microbial fingerprint in the brain, heart, and lungs related to SARS-CoV-2 death (Figure 1).

Microorganisms deliver specific fingerprints that differ from person to person. Regarding distinguishing microbes, often it is crucial to differentiate them between their taxonomic classes, which is called microbial fingerprinting. This study demonstrates the potentiality of a microbial fingerprint and its use in crime scenes parallel to conventional fingerprinting. For example, can postmortem microbial fingerprint characterization be used in medicolegal investigations to link physical evidence to the criminal or victim (Fierer et al., 2010; Aaspõllu et al., 2011; Nishi et al., 2015)?

The distinct thanatomicrobiome profiles in organs from human cadavers that died of or with SARS-CoV-2 may help guide the application of forensic microbiology tools to establish the real cause of death. The H₀ (null hypothesis) of our study is that there are no differences in the microbiome of COVID-19 and non-COVID-19 cadavers. Thus, the H_A (alternative hypothesis) is that there are differences in the microbiome of COVID-19 and non-COVID-19 cadavers. To test the alternative hypothesis, we used a two-pronged strategy to first, establish an in vitro model of detection of microbial diversity in brain, heart, and lung tissue microbiota from 40 human cadavers, half of them infected with SARS-CoV-2.

The chi-squared test was used to determine if the deviations in the data are caused by sampling or experimental error. Deviations may be caused by random chance or if there are differences in our data that are due to a statistically significant difference. If they were caused by a statistically significant difference, it gives evidence to support or reject our hypothesis. Generally, when approaching a chi square test, you start with the null hypothesis that there is no difference in the data; that is, that any differences are in fact due to sampling errors, experimental errors, or by chance errors. So, by calculating chi-squared values generate statistical support for whether to reject or accept the hypothesis. The second approach involved adjusting all p-values for multiple comparisons with the FDR algorithm to control the number of false discoveries in those tests that result in a discovery (i.e., a significant result). It has greater ability (i.e., power) to find truly significant results.

The results demonstrated that microbial profiles vary significantly during COVID-19 infection as the corpse decomposes. When comparing non-COVID-19 infected samples versus COVID-19 infected samples, there is statistical difference between the brain, heart, and lung from the infected group only. Among the three organs, microbial richness was significantly higher in the non-COVID-19-infected tissues compared to the infected tissues. Further, Unifrac distance analyses demonstrated that there was more variance between control and COVID-19 groups in weighted analysis than unweighted; both were statistically different. Also, unweighted Bray-Curtis principal coordinate analyses showed a near distinct two-community structure: one for control and the other for the COVID-19-infected group. Both unweighted and weighted Bray-Curtis showed statistical differences. These results facilitated the defining of microbiome profiles in COVID-19 decedents that could be identified as taxonomic biomarkers effective for predicting the occurrence, dysbiosis, and evolution of the virus.

Studies have shown that the gut microbiome composition is significantly altered in patients with COVID-19 compared to non-COVID-19 cases regardless of whether patients had been treated with medication (Yeoh et al., 2021). In the current study, Staphylococcus aureus was the most abundant bacteria found in COVID-19-infected lung tissue (Supplementary Figure S1). Studies have demonstrated that the gut microbiome was altered in COVID-19 infections with an enrichment in opportunistic pathogens (Wang et al., 2021). For example, S. aureus generally have higher abundance in the lungs of COVID-19-infected patients Thus, S. aureus is commonly found in hospital environments for its risk of deadly outcomes such as endocarditis, bacteremia, sepsis, and death. In past viral pandemics, S. aureus has been the principal cause of secondary bacterial infections, significantly increasing patient mortality rates (Adalbert et al., 2021). The predominance of S. aureus co-infections occurring after patient admission for COVID-19 infection is likely associated with patient interventions identified as intubation and mechanical ventilation, central venous catheter placement, and corticosteroids. In the current study, Gram-positive S. aureus, which belongs to the Firmicutes phylum, was shown to be enriched in all cases, especially COVID-19-infected lungs (Figure 2). The gut area has the largest absolute decomposition burden that spreads to the proximate organs, such as the liver and spleen, and extends to the distal organs, such as the heart and brain, depending on the cause of death (Javan et al., 2019). Likewise, the opportunistic bacteria, Corynebacterium, was the next highest in abundance in the COVID-19-infected lungs. Corynebacterium sp. is well known to be a pathogen in lower respiratory tract infection. Studies have reported that ventilator-associated complications (VACs) in COVID-19 patients were due to Corynebacterium sp. (Ogawa et al., 2022).

Regarding inflammatory responses that characterize COVID-19 infections, the immune system ceases within 24 h after death. Due to Italian laws, the minimum postmortem interval (PMI) for cases in the current study is 24 h. Therefore, due to the extended PMIs, inflammatory components were undetectable in the criminal case cadavers used in this study. Taxa that proliferate in COVID-19 deaths show reduced microbial diversity and an enrichment for microorganisms that can resist inflammatory responses more effectively than others (Hussain et al., 2021). Previous antemortem studies demonstrated that the gut microbiome of non-COVID-19 infected patients had higher abundance of anti-inflammatory bacteria Lachnospiraceae, Roseburia, Eubacterium, and Faecalibacterium prausnitzii compared to the microbiome of patients with COVID-19 (Reinold et al., 2021; Wang et al., 2022). In the current study, as expected these bacteria were not enriched in postmortem brain, heart, and lung samples of COVID-19 cases.

The highest percentage of bacteria in heart tissue on the genera level was Enterobacteriaceae, which occurred in both groups (Supplementary Figure S2). From the 12 samples of the COVID-19 group, three showed increased relative abundance of specific genera, Enterococcus, Enterobacter, and Lactobacillus. From the 19 samples of the control group, four showed increased relative abundance of Hathewaya, Clostridium, Paeniclostridium, and Morganella. Escherichia and Shigella were the most abundant bacteria found in COVID-19-infected heart tissue. The highest percentage of bacteria in brain tissue on the genera level was Enterobacteriaceae (Escherichia and Shigella) in both the control and infected groups (Supplementary Figure S3). E. coli are known to translocate from the blood to the central nervous system without apparent damage to the blood–brain barrier, which indicates a transcytosis process (Kaper et al., 2004). Enterobacter species are increasingly a cause of nosocomial meningitis among neurosurgery patients. In community-acquired infections, Enterobacter was isolated in one of the nine cases of meningitis caused by Gram-negative bacilli (E. coli four times, Klebsiella species three times, and Proteus once) and in five of the 57 episodes of nosocomial meningitis (E. coli 17 times, Klebsiella species 13 times, Pseudomonas species six times, and Acinetobacter species six times). Morganella morganii is a Gram-negative aerobe, found often as intestinal commensal. It is commonly implicated in urinary tract infections and pyogenic infections, but rarely causes CNS infections especially brain abscess.

Of interest, Deblur analyses showed that Firmicutes (which included two Clostridium species) were predominant among all organs from both groups. The detection of Clostridium species is accounted for by the Postmortem Clostridium Effect (PCE) that distinguishes the rapid proliferation of the species in decaying internal body sites (Javan et al., 2017; Lutz et al., 2020; Javan et al., 2022). These bacteria have adaptive properties that facilitate persistence in anoxic and hypoxic environments that persist until the skin ruptures. Thus, a future research question might include, “Could any of these postmortem bacteria, specifically Clostridium, be biomarkers across thanatomicrobiome communities derived from different locations in the body?”