Shuji Kozawa

Detection of diverse aquatic microbes in blood and organs of drowning victims: First metagenomic approach using high-throughput 454-pyrosequencing

Current 454-pyrosequencing technology enables massive parallel sequencing. We used this technology to investigate the diversity of aquatic microbes in 14 specimens (blood and organs) of two drowning victims and in two water samples taken from the discovery sites. The 16S ribosomal RNA (rRNA) genes of microbes, which are often used to identify species (or genera), have nine highly variable regions (V1-V9), each of which is surrounded by conserved regions. Some parts within the conserved regions are common over domains of microbes, such as between bacteria and algae (16S rRNA genes on algal chloroplast genomes). We therefore simultaneously amplified the target regions (V7 and V8) of various microbes in the blood and organs of drowning victims using PCR with custom-designed primers that were based on the conserved regions. We then exhaustively analyzed the PCR products by pyrosequencing using the Genome Sequencer FLX Titanium system (Roche-454 Life Sciences). This approach identified a wide array of bacteria including cyanobacteria and algae including Bacillariophyceae (diatom), Cryptophyceae, Dictyochophyceae, Chrysophyceae and Trebouxiophyceae in the blood and organs of the victims and water at discovery sites. Our data further indicated that when conventional diatom testing of lungs yielded insufficient evidence of water aspiration, the detection of various exogenous microbes by 454-pyrosequencing is very useful to support a conclusion of death by drowning. To the best of our knowledge, this is the first attempt to use a new generation sequencer to investigate diverse aquatic microbes in the blood and closed organs of drowning victims.

Freshwater bacterioplankton cultured from liver, kidney and lungs of a decomposed cadaver retrieved from a sandy seashore: Possibility of drowning in a river and then floating out to sea

A decomposed female body with an open abdomen and pleural cavity washed up on a beach after a powerful typhoon. Autopsy findings could not determine the cause of death because of leaching and putrefaction. Numbers and types of diatoms in organs overall, suggested the aspiration of fresh or brackish water with low salinity. However, this could not be confirmed because of contamination via the open cavities. We simultaneously investigated the presence of bacterioplankton in liver, kidney and lung homogenates using a modification of our reported bacteriological method. The freshwater bacterioplankton Plesiomonas shigelloides was identified in each of these organs, but marine bacterioplankton were undetectable despite the circumstances under which the body was discovered. The presence of freshwater bacterioplankton reinforced the results of the diatom test, and we concluded that this victim had died of drowning in fresh or brackish water with low salinity.

Bioluminescent bacteria have potential as a marker of drowning in seawater: Two immersed cadavers retrieved near estuaries

We detected numerous bioluminescent bacteria in blood samples from two cadavers that had been immersed in estuarine environments. Autopsy, diatomaceous and toxicological findings indicated death by drowning, which agreed with environmental aspects and the findings of police investigations. Bioluminescent bacteria appeared in blood samples cultured on selective agar containing 2%, 3% and 4% NaCl after about 18h. Blood from the left side of the heart, the right side of the heart and the femoral vein generated 7.0 x 10(2), 2.0 x 10(4) and 8.0 x 10(2) cfu/ml of blood (case 1), and 1.8 x 10(4), 1.1 x 10(3) and 2.5 x 10(1) cfu/ml (case 2) of bioluminescent colonies, respectively, in agar containing 4% NaCl. Homologous analysis based on the 16S rRNA gene also identified the bioluminescent colonies as Vibrio fischeri and V. harveyi, which normally inhabit seawater. This simple assay might serve as an additional indicator to support a conclusion of death by drowning together with the diatom test.

Detection of bacterioplankton in immersed cadavers using selective agar plates

We measured bacterioplankton in blood from cadavers retrieved from the sea (n=12), near estuaries (n=4), rivers (fresh water, n=8) and from bathtubs (n=4) as well as from non-drowned victims (n=10) discovered near aquatic environments. Blood from 11 victims drowned in seawater developed bioluminescent and/or blue colonies (oxidase test positive) on selective media containing 2-4% NaCl. Homology analyses of the 16S rRNA gene showed that all of them were marine bacteria (genera: Photobacterium, Vibrio, Shewanella, Psychrobacter). Blood from all victims drowned in rivers generated blue colonies on plates containing 3%, but not 4% NaCl. Homology analyses showed that the blue colonies were generated from bacteria that inhabit fresh water (Aeromonas). None of the blood samples from victims that drowned in bathtubs generated bioluminescent and blue colonies. However, all cadavers contained bacteria that produced unstained colonies (Staphylococcus, Bacillus, Enterobacter, Escherichia, etc.). Among non-drowned victims, blood from two gave rise to blue colonies on plates containing < or =3% NaCl (Pseudomonas). Of the cadavers found near estuaries, bioluminescent and blue colonies developed from two of them on media containing 2-4% NaCl (Photobacterium, Vibrio, Listonella), but not from two others on plates containing 4% NaCl (at < or =3%; blue colonies, Aeromonas; unstained colonies, Citrobacter, Vagococcus, Proteus, Enterobacter). These results suggested that the presence of numerous bacterioplankton in immersed cadavers could support a conclusion of death by drowning.

A new molecular approach to help conclude drowning as a cause of death: Simultaneous detection of eight bacterioplankton species using real-time PCR assays with TaqMan probes

We developed a novel tool for concluding drowning as a cause of death. We designed nine primer pairs to detect representative freshwater or marine bacterioplankton (aquatic bacteria) and then used real-time PCR with TaqMan probes to rapidly and specifically detect them. We previously cultured the genus Aeromonas, which is a representative freshwater bacterial species, in blood samples from 94% of victims who drowned in freshwater and the genera Vibrio and/or Photobacterium that are representative marine bacteria in 88% of victims who drowned in seawater. Based on these results, we simultaneously detected eight species of bacterioplankton (Aeromonas hydrophila, A. salmonicida; Vibrio fischeri, V. harveyi, V. parahaemolyticus; Photobacterium damselae, P. leiognathi, P. phosphoreum) using three sets of triplex real-time PCR assays and TaqMan probes labelled with fluorophores (FAM, NED, Cy5). We assayed 266 specimens (109 blood, 157 tissues) from 43 victims, including 32 who had drowned in rivers, ditches, wells, sea or around estuaries. All lung samples of these 32 victims were TaqMan PCR-positive including the lung periphery into which water does not readily enter postmortem. On the other hand, findings in blood and/or closed organs (kidney or liver) were PCR-positive in 84% of the drowned victims (except for those who drowned in baths) although the conventional test detected diatoms in closed organs in only 44% of the victims. Thus, the results of the PCR assay reinforced those of diatom tests when only a few diatoms were detectable in organs due to the low density of diatoms in the water where they were found. Multiplex TaqMan PCR assays for bacterioplankton were rapid, less laborious and high-throughput as well as sensitive and specific. Therefore, these assays would be useful for routine forensic screening tests to estimate the amount and type of aspirated water.

Detection of marine and freshwater bacterioplankton in immersed victims: Post-mortem bacterial invasion does not readily occur.

We previously applied our method of detecting marine or freshwater bacterioplankton (bacteria) in the blood of immersed victims as a marker of drowning. However, we did not confirm the absence of post-mortem bacterial invasion during immersion. Here we examined the nature of bacterioplankton in blood samples from 21 immersed and 4 non-immersed cadavers. We found only freshwater bacterioplankton in the blood of two victims that were retrieved from the sea or an estuary inhabited by marine bacterioplankton even though one victim was highly putrefied. The results of diatom testing suggested that these two victims had drowned in fresh or brackish water with low salinity and then flowed out to the estuary or the sea. Two others were submerged in water, but representative bacterioplankton were undetectable in their blood although one victim was highly putrefied. Autopsy findings and the results of diatom tests did not indicate that the cause of death was drowning. As in previous studies, we identified freshwater bacterioplankton in the blood of seven other victims that had drowned in freshwater, marine bacterioplankton in the blood of four victims that had drowned in seawater and none in four victims found on land that had died by means other than drowning. Bacterioplankton in the blood of drowned victims appears to reflect the type of water aspirated and blood does not become easily contaminated with bacteria post-mortem even in decomposed bodies.

In vitro study of possible microbial indicators for drowning: Salinity and types of bacterioplankton proliferating in blood §

Numbers and types of bacterioplankton proliferating in blood samples mixed with water of various salinity levels were examined to determine the characteristics of species associated with salinity. Water samples (total n=88) were collected from the midstream of two rivers (freshwater; n=10; salinity <0.05%), from around their estuaries (areas of freshwater, n=20, salinity <0.05%; areas of brackish water, n=20, salinity <0.05-3.1%; areas of marine water beyond the mouths of the rivers, n=28, salinity 2.4-3.3%), and from the coast (areas of marine water; n=10; salinity 3.3-3.5%). Freshwater bacteria were identified in 41 of 42 blood samples mixed with water at ≤1.3% salinity, and the genus Aeromonas, which is universally distributed in freshwater environments, was predominant. Marine bacteria were identified in all of 46 blood samples mixed with water at ≥1.8% salinity, and most comprised the genera Vibrio and Photobacterium that are universally distributed in seawater environments. Aeromonas was undetectable in all blood samples mixed with brackish or sea water at ≥1.8% salinity although they are detectable even in seawater environments. Thus, the present results showed that bacterioplankton capable of proliferating in human blood reflects the salinity of water.

Numbers, sizes, and types of diatoms around estuaries for a diatom test.

We collected 68 fresh, brackish, and seawater samples from various sites around the estuaries of 2 rivers at high and low tides. Seawater flowed approximately 2.4 (salinity, 2.2% at the site) and 1.2 km (1.8%) upstream of the estuaries, but the surface comprised essentially fresh water up to the mouth. Sites contained 69 to 22,200 diatoms/50 mL of water, and the numbers varied by depth and at sites separated by only approximately 1.2 km. Diatoms ranged from 2.8 to 429 &mgr;m (mean range, 16.1-59.2 &mgr;m) in size. Large pennate diatoms populated fresh water areas, and most sedimented before reaching the sea. Numbers of pennate diatoms of less than 20 &mgr;m were decreased in areas of seawater. Numbers of centric diatoms tended to increase nearer the sea, and seawater contained large centric diatoms. Brackish water containing large volumes of seawater was easily discriminated by assemblages of marine diatoms, unlike that containing a little seawater, because marine diatoms could be found in fresh water around estuaries. Tides and the nature of the river often altered diatomaceous assemblages at the same estuarial sites. Caution is recommended for forensic interpretation of aqueous media to deduce drowning sites.

Diatom and Laboratory Tests to Support a Conclusion of Death by Drowning

Among many laboratory tests for supporting a conclusion of death by drowning, the diatom test is often regarded as the gold standard. However, opinions are divided on its usefulness, and thus this test is rarely or never applied in some countries or institutions. Indeed, even when characteristic macroscopic findings of drowning are weak, drowning can often be reasonably concluded as a cause of death by taking circumstantial evidence into consideration and carefully excluding other causes. However, incorporation of the diatom test will further increase the certainty of such a conclusion. When highly decomposed corpses are found in and near water environments, autopsy findings are of limited value, and the cause of death often remains undetermined. However, death that could have been due to drowning in some decomposed corpses can be assumed if the results of diatom testing are interpreted as positive. Pollanen illustrated the contribution of the diatom test to conclude drowning as a means of homicide in a presentation of six cases. The largely burned remains of a teenage girl were found in a parking lot where accelerant was detected. The results of an autopsy did not indicate the anatomical cause of death. However, 5 mL of watery fluid was aspirated from the right maxillary sinus. Over ten types of diatoms were detected in the fluid, and four of them that were also detected in bone marrow established drowning as the cause of death.

An autopsy case of chemical burns by hydrochloric acid

A 34-year-old man was discovered by his coworkers in a tank filled with 35% (w/w) hydrochloric acid. Despite undergoing intensive treatment, he died one and a half days later. An autopsy revealed generalized high tensity, overall grayish brown skin color, heavy gastric submucosal hemorrhage and heavy pulmonary edema. We concluded that death was caused by burn shock due to wide, generalized chemical burn. Microscopic investigation of the burn in the area with grayish brown skin considered coagulation necrosis of full-thickness of the skin (third-degree or deep burn), revealed that the burn was judged to cover the partial thickness of the skin (second-degree or dermal burn). These findings suggest that chemical burn by hydrochloric acid results in a change of skin color due to chemical reaction so that the appearance of the chemical burn is more severe than the degree assigned by histological examination.