Post-mortem peripheral blood samples were collected from the femoral or subclavian arteries of 21 cadavers (7 females and 14 males) who underwent autopsy examination between 2 and 9 days after death. The causes of death ascertained through the autopsies, supplemented by available circumstantial data, were: road accidents (4 subjects); self-induced dynamics (4 subjects); endogenous pathological processes (4 subjects); and overdose of illicit substances (9 subjects). At the time of autopsy, information regarding seropositivity for HBV, HCV, and HIV-1 were only available for four of the corpses: one was positive for HBV (even though the documentation in possession disclosed no specific indication for the detection of HBsAg or HBsAb), one for HCV, one for HIV-1, and one for all three viruses. No data were available regarding any possible antiviral treatment received during the patients’ lifetimes. All data regarding the cohort are summarized in Table 1. Blood samples collected post-mortem were identified by progressive numbering from 1 to 21 to ensure complete anonymity. The samples were stored at –20 °C until analysis.

Viral genomes of HBV, HCV, and HIV-1 were extracted from post-mortem whole blood samples by manual protocols and fully automated procedures.

Manual DNA and RNA extraction from whole blood samples were carried out using a solid-phase extraction approach, without any sample dilution. Generally, viral nucleic acid extraction protocols are validated on patients’ serum or plasma samples [33, 34, 35, 36, 37, 38, 39, 40].

For DNA extraction, we used the NucleoSpin® Blood kit (Macherey-Nagel) and slightly adjusted the manufacturer’s instructions to increase DNA purity and recovery [41].

More specifically, the incubation time with Proteinase K was prolonged to 1 hour, vortexing several times to improve cell lysis and DNA recovery from the post-mortem blood samples. Additionally, to enhance the final DNA concentration, this was recovered in a final volume of elution buffer which was half of that recommended. Briefly, 200 $\mu$L of buffer B3 and 25 $\mu$L of proteinase K were added to 200 $\mu$L of each post-mortem whole blood sample. After incubation at 70 °C for 1 hour, 210 $\mu$L of ethanol (96–100%) was added to each sample and vortexed for a few seconds. The lysate was then loaded into the NucleoSpin Blood Column placed in a collection tube to allow DNA absorption to the spin silica membrane column during the centrifugation at 11000 $\times$ g for 1 minute. The spin column was washed twice, first with 500 $\mu$L of buffer BW and subsequently with 600 $\mu$L of buffer B5. The flow-through collected into the collection tube was discarded at each step. After 1 minute of room temperature incubation of 50 $\mu$L of preheated (70 °C) elution buffer BE, the purified DNA was eluted from the spin column by centrifuging at room temperature for 1 minute at 11000 $\times$ g.

For RNA extraction, we used the NucleoSpin® RNA Blood kit (Macherey-Nagel) with minor modifications to the procedure specified in the manufacturer’s instruction to increase the RNA yield [42].

In particular, since the proteinase K-mediated lysis was carried out at room temperature, it was extended to 30 minutes to improve blood cell lysis and, therefore, to enhance RNA recovery from the post-mortem blood samples. In addition, the elution time of the RNA from the column was prolonged and the final volume of RNA recovery was adjusted to achieve a higher concentration of RNA.

In brief, 200 $\mu$L of lysis buffer DL and 5 $\mu$L of Proteinase K were added to 200 $\mu$L of each post-mortem whole blood sample. The samples were placed on a shaker (TS-100 Thermo-Shaker Biosan (LV)) at 1400 rpm for 30 minutes at room temperature. Then, 200 $\mu$L of 70% ethanol was added to the lysate of each sample and, after accurate mixing, was transferred to the NucleoSpin RNA blood column placed in a collection tube, and centrifuged at 11000 $\times$ g for 30 seconds. Membrane desalting buffer MDB (350 $\mu$L) was added to each spin column and centrifuged at 11000 $\times$ g for 30 seconds. After this step, 95 $\mu$L of DNase was added to each spin column and kept at room temperature for 30 minutes to achieve complete DNA digestion. The spin column was washed three times: once with 200 $\mu$L of buffer RB2 to inactivate the DNase, and twice with 600 $\mu$L and 250 $\mu$L of RB3 buffer, respectively. At each phase, the flow-through gathered in the collecting tube was discarded. The three washes were repeated twice to ensure the elimination of any hemolysis products. To eluate the RNA, 30 $\mu$L of RNase-free water was added to the spin column and centrifuged at room temperature for 30 seconds at 11000 $\times$ g.

After manual extraction of viral nucleic acids, the DNA or RNA extracted from each sample was assessed by spectrophotometric measurements using the NanoDrop One Microvolume UV-Vis spectrophotometer (Thermo Fischer, Waltham, USA). The total DNA and RNA amounts in the samples were determined as between 78.4 and 621.4 ng/$\mu$L and 24.9 and 648.7 ng/$\mu$L, respectively.

Automated viral nucleic acids extraction was performed using the COBAS system. This system is a fully automated platform that simultaneously performs the extraction and purification of total nucleic acids from the specimens within the sample processing module, amplification, and the detection of extracted nucleic acids and their quantification by real-time polymerase chain reaction (PCR) assay. By using two different fluorescent dyes, one labeled to the specific viral genome probe and the other to the internal control (IC) probe, the COBAS system allows the simultaneous detection of the viral genomes and verification that the entire process has been well performed. Because the COBAS method had not yet been used on whole blood samples collected from cadavers, before sample processing, dilutions varying from 1:5 to 1:10 with phosphate-buffered saline (PBS), optimized for each post-mortem blood sample (Table 2) according to their degree of hemolysis [13], were performed to reduce possible interference with inhibitors of the PCR process due to post-mortem modification sample phenomena [13, 26, 27]. All the samples were tested once (n = 1). In case of inconclusive results, the test was repeated (n = 2).

The extraction of viral nucleic acids is based on the use of magnetic bead technology: after cell lysis has been carried out with a lysis buffer containing Proteinase K and chaotropic salts, the nucleic acids bind to the surface of magnetic glass beads, and cell debris is effectively removed from the lysate by several washing steps.

Quantitative analysis of HBV DNA, HCV RNA, and HIV-1 RNA was carried out using the COBAS 6800 system (Roche, Switzerland), an accurate in vitro viral nucleic acid quantification platform validated on plasma and serum samples. The COBAS system provided for the fully automated extraction, amplification, and detection of the viral nucleic acid target, and was able to simultaneously detect viral genomes and verify the analytical performance of the entire process by means of probes using the TaqMan technology and the use of an IC probe [43]. The quantitative result indicated the degree of active viral load and was expressed in standardized international units per milliliter (IU/mL) or copies/mL.

Three specific kits were employed to amplify and quantify these three viruses by following the manufacturers’ instructions: COBAS® AmpliPrep/COBAS® TaqMan® HBV Test, v2.0 kit [44], COBAS® AmpliPrep/COBAS® TaqMan® HCV Quantitative Test, v2.0 kit [45] and COBAS® AmpliPrep/COBAS® TaqMan® HIV-1 Test, v2.0 [46].

The COBAS® AmpliPrep/COBAS® TaqMan® HBV Test, v2.0 kit can detect all eight genotypes of HBV (A-H) with a limit of detection (LoD) of 20 international units (IU)/mL (1 IU = 5.26). The COBAS®AmpliPrep/COBAS® TaqMan® HCV Quantitative Test, v2.0 kit is characterized by a LoD of 15.3 IU/mL, where 1 IU generally corresponds to 2.7 copies/mL and can detect genotypes 1 to 6. The COBAS HIV-1 v2.0 assay using two TaqMan probes simultaneously amplifies and detects two highly conserved regions (gag gene and LTR region) in the HIV-1 genome. The LoD was determined at 20 copies/mL.

Every kit displayed one negative sample and two specific positive samples, characterized by a high and a low number of copies respectively. In all samples, a non-viral Quantitation Standard (QS) was added, in known copy numbers, in the preparation step as an internal control to monitor the whole procedure from the extraction of the nucleic acid to its quantification.

Hepatitis B virus (HBV) DNA was revealed by qualitative endpoint real-time PCR using a fluorescent DNA-binding dye called BRYT (Promega, Madison, WI, USA) and a set of forward and reverse primers designed to target positions 379 to 476 in the highly conserved region of the HBV S gene encoding the surface antigen (HBsAg), as previously described [47]. The primer sequences were: HBVF 5${}^{\prime }$-GTGTCTGCGGCGTTTTATCA-3${}^{\prime }$ and HBVR 5${}^{\prime }$-GACAAACGGCAACATACCTT-3${}^{\prime }$. After the amplification process, an amplicon of 98 base pairs (bp) was obtained. The amplification mix consisted of 100 ng DNA/reaction, 12.5 $\mu$L of 2X GoTaq® qPCR Master Mix (Promega), 0.5 $\mu$L of each primer (10 $\mu$M), and distilled water to achieve a final volume of 25 $\mu$L. The amplification was carried out at 95 °C for 2 minutes, followed by 40 cycles of 95 °C for 5 seconds, 60 °C for 30 seconds, and 72 °C for 5 seconds, with a final extension at 72 °C for 7 minutes in a CFX96 Real-Time PCR Detection System (BioRad, Hercules, CA, USA). Melting curve analysis was performed at the end of the qPCR amplification by raising the temperature from 60 °C to 95 °C, in increments of 0.5 °C/s. The software used for amplification acquisition was CFX Manager (BioRad).

In parallel, the amplification of a highly conserved region of the C-C chemokine receptor type 5 (CCR5) gene located on chromosome 3 was performed for each sample using the following set of primers: CCR5F: 5${}^{\prime }$-ATGATTCCTGGGAGACGC-3${}^{\prime }$ and CCR5R: 5${}^{\prime }$-AGCCAGGACGGTCACCTT-3${}^{\prime }$ [48]. The CCR5 amplicon size was 81 bp. The amplification mix was the same as reported above. The amplification was carried out at 95 °C for 2 minutes, followed by 40 cycles of 95 °C for 5 seconds, 60 °C for 30 seconds, 72 °C for 5 seconds, and 76 °C for 5 seconds, with a final extension at 72 °C for 7 minutes, in a CFX96 Real-Time PCR Detection System (BioRad). Single fluorescence detection was performed at 76 °C to reveal the samples’ positivity and rule out eventual interference (i.e., primer-dimer artifacts). Melting curve analysis was performed at the end of the qPCR amplification by raising the temperature from 60 °C to 95 °C, in increments of 0.5 °C/s. CCR5 gene amplification was used as an internal control to exclude any interference with or inhibition of the PCR process due to haemolysis or degradation phenomena resulting from post-mortem blood samples and, at the same time, to allow the evaluation of the number of nucleic acids present in the sample.

HBV-negative and HBV-positive blood samples were used as negative and positive HBV controls, respectively.

The total RNA purified from post-mortem whole blood samples was reverse-transcribed to generate complementary DNA (cDNA) by using the ImProm-II™ Reverse Transcription System kit (Promega) [49], according to the manufacturer’s instructions.

Briefly, reverse transcription (RT) reaction mix was prepared for each sample by adding ImProm-II™ 5X reaction buffer, 2.5 mM MgCl${}_2$, dNTP (0.5 mM for each deoxynucleotide), 5U of ImProm-II™ Reverse Transcriptase and nuclease-free water to a final volume of 15 $\mu$L. An aliquot of extracted RNA (4 $\mu$L) from each sample and random primers (0.5 $\mu$g/reaction) were mixed with the RT mix for a final reaction volume of 20 $\mu$L per reaction. The reverse transcription reaction was started with the annealing step at 25 °C for 5 minutes, then the cDNA synthesis continued at an extension temperature of 55 °C for 60 minutes and the reaction vessel was finally incubated at 70 °C for 15 minutes to heat inactivate the ImProm-II™ Reverse Transcriptase.

HCV cDNA was then used for the qualitative endpoint RT-PCR assay using two specific 5${}^{\prime }$-non-coding regions (5${}^{\prime }$-NCR) HCV primers: 5${}^{\prime }$-GTCTAGCCATGGCGTTAGTATGAG-3${}^{\prime }$ (77–100; HCVP1) and 5${}^{\prime }$-ACCCTATCAGGCAGTACCACAAG-3${}^{\prime }$ (302–280; HCVP2) [33], which allowed the amplification of a specific fragment of 226 bp.

The amplification mixture was prepared with 12.5 $\mu$L of 2X GoTaq® qPCR Master Mix (Promega), 0.5 $\mu$L of each primer (10 $\mu$M), 5 $\mu$L of cDNA, and distilled water for a 25 $\mu$L final volume. The amplification program was set up as previously described [33, 50] in a CFX96 Real-Time PCR Detection System (BioRad). Melting curve analysis was performed at the end of the qPCR amplification. HCV-negative and HCV-positive blood samples were used as negative and positive HCV controls, respectively.

HIV-1 belongs to the Retroviridae family and exhibits a peculiar replication cycle in which its genomic RNA is retro-transcribed into proviral DNA by a reverse transcriptase enzyme and then integrated into the host cell genome [51, 52, 53]. This particular feature indicates two distinct approaches to recognizing the presence of the HIV genome in clinical samples: the first is the detection of HIV-1 virions with RNA genome in plasma (described above) and the second is the detection of HIV-1 proviral DNA integrated into the genome of infected cells (usually CD4${}^{+}$T lymphocytes and monocytes/macrophages) in peripheral blood [52].

To determine the presence of HIV-1 proviral DNA, we selected forward and reverse primers that anneal to complementary sequences within the long terminal repeat (LTR) region of HIV-1, LTR1-F: 5${}^{\prime }$-TACTGACGCTCGCACC-3${}^{\prime }$ and LTR2-R: 5${}^{\prime }$-TCTCGACGCAGGACTCG-3${}^{\prime }$. These oligonucleotides have been previously described [48, 54] and generate an amplicon of 127 bp in size. The amplification mix consisted of 100 ng DNA/reaction, 12.5 $\mu$L of 2X GoTaq® qPCR Master Mix (Promega), 0.5 $\mu$L of each primer (10 $\mu$M), and distilled water to achieve a final volume of 25 $\mu$L. The amplification was carried out at 95 °C for 2 minutes, followed by 40 cycles of 95 °C for 5 seconds, 60 °C for 30 seconds, 72 °C for 5 seconds, and 80 °C for 5 seconds with a final extension at 72 °C for 7 minutes, in a CFX96 Real-Time PCR Detection System (BioRad). Single fluorescence detection was performed at 80 °C to reveal the sample positivity and rule out eventual interference (i.e., primer-dimer artifacts). Melting curve analysis was performed at the end of the qPCR amplification by raising the temperature from 60 °C to 95 °C in increments of 0.5 °C/s. CCR5 internal control, as described above, and negative and positive HIV-1 controls were used.

The quantitative evaluation of HBV DNA was carried out using the COBAS®AmpliPrep/COBAS® TaqMan® HBV Test, v2.0, which amplifies the pre-core/core (PreC/C) region of the HBV genome. This region is one of the four HBV open reading frames (ORFs) encoding for both the hepatitis B e antigen (HBeAg) and the core protein (HBcAg). Unfortunately, no amplification of the PreC/C region of HBV was observed in 20 out of 21 samples (Table 3), including two samples with ante-mortem HBV positive serology. Even after adequate dilution of the samples, inconclusive results were obtained for sample 16, which did not exceed the amplification of the COBAS system’s internal control.

HCV viral RNA quantification was performed using the COBAS® AmpliPrep/COBAS® TaqMan® HCV Quantitative Test, v2.0 kit. HCV positivity was detected in samples 1 (with positive serology already known during life) and 21 (180 and 6780 IU/mL respectively), as shown in Table 3. In contrast, sample 2 (with a known positive serology) resulted in a negative result for HCV viral RNA quantification in the post-mortem sample.

To determine HIV-1 genomic RNA, the COBAS® AmpliPrep/COBAS® TaqMan® HIV-1 Test, v2.0 was used in properly diluted post-mortem blood samples. This assay detected genomic HIV-1 RNA in 5 out of 21 specimens (numbers 2, 3, 4, 20, and 21) ranging from 780 to 3130 copies/mL (Table 3), for two of which (number 2 and 3) the patients were known to be HIV-1 seropositive during life.

Viral nucleic acid was extracted from post-mortem whole blood samples after minor modifications were made to manufacturers’ instructions and were analyzed using the qualitative end-point real-time PCR method, based on the BRYT molecule that binds dsDNA with the subsequent fluorescence emission.

Detection of HBV DNA in post-mortem blood samples based on qualitative PCR analysis was negative for all 21 specimens, including the two samples with HBV-positive ante-mortem serological status (Table 3). Amplification curve analyses were negative for all samples, indicating the absence of detectable HBV DNA in the blood samples even though the CCR5 amplification control was positive in all specimens, excluding the presence of PCR artifacts.

Positive amplification curves for the HCV RNA genome were detected by qualitative assay in 2 out of 21 blood samples (numbers 1 and 21; as shown in Fig. 1 and Table 3). These results were also confirmed by quantitative COBAS data.

HIV-1 proviral DNA genome was determined by qualitative end-point real-time PCR in post-mortem blood samples, using the same total DNA extraction protocol employed for HBV DNA detection. A positive amplification curve was detected in 5 out of 21 samples (numbers 2, 3, 4, 20, and 21; as shown in Fig. 1 and Table 3), even though HIV-1 ante-mortem seropositivity was only reported for 2 out of 5 samples (numbers 2 and 3).