Thank you for your comments. However, we disagree with your assertion that our conclusions were “pre-ordained by a prior view”, as we got into these studies thinking that pleomorphic structures observed in our negative-stained EM images were bacteria. We, too, were astonished that you would post your comments without recognizing that, in fact, we were able to amplify 16S in our negative control (molecular grade, sterile water, please see Fig. 3B), supporting the presence of contaminants.

We have been interested in red blood cell microparticles (RMPs), which accumulate in the supernatant of stored units of RBCs. The current study was prompted by our initial EM experiments with pelleted material from stored RBC units (up to 42 days at 4C, consistent with standard blood bank procedures). The structures that we initially observed in this material, depicted in Fig. 1A and B, resembled those published by McLaughlin et al. (Journal of Clinical Microbiology, 2002; 40:4771), and, because those authors identified the pleomorphic structures as bacteria, we strongly considered that we were also observing bacteria. However, we also considered that we might be looking at artifact that resulted from processing of RMPs for EM. This led us to perform more electron microscopy experiments, including 3D Cryo-EM/tomography. Cryo-EM/tomography is considered a gold standard for imaging microparticles, bacteria, viruses, and mammalian cells. Under Cryo-EM/tomography, the morphology of the structures appeared to be highly variable but not rod-like (as we had seen with unfixed, negative-stained samples), and they appeared to be filled with a dense granular substance. Further arguing against the identification of these structures as bacteria, in the slices from 3D Cryo-EM/tomography volumes, the structures did not have a cell wall, ribosomes, or other essential components associated with bacteria. We believe that the dense granular material in these vesicles is hemoglobin, which we observed on Western analysis (Fig 4). Also, these structures stained for GPA, a protein found on the surface of RBCs, and they contain miR-451, which is a miRNA highly expressed in RBCs. These data, along with our efforts to identify microbes in a large volume of RBC storage unit supernatant led us to the central conclusion of the paper: the pleomorphic structures in negative-stained electron micrographs were RMPs, not bacteria.

We acknowledge that our analysis for microbes did not include culture techniques described by Damgaard and co-workers (PLoS One, 2015; 10:e0120826); however, we failed to detect bacterial genomic DNA in billions of the vesicles pelleted from stored RBC units. We were able to detect 16S in this pelleted material, but we were also able to detect 16S in our negative control, which was sterile, molecular grade water. Furthermore, we found that the 16S amplified in our pelleted material and molecular grade water belonged to microbes that have been described to be frequent contaminants of laboratory reagents.

Since we had also seen similar pleomorphic structures, but much less frequently, in negative-stained EM of pelleted material from healthy serum (less than 1 structure/ml of serum or plasma), we performed the experiment depicted in Fig. 3C to try to replicate an experiment done previously by McLaughlin and co-workers (referenced above). Using flow cytometry, they had found that their “bacteria” had increased in number in serum that sat for seven days at 30C, indicating growth. Under similar conditions, we did not find evidence of growth. Interestingly, in the study by McLaughlin et al, the authors were unable to isolate or identify the type of bacteria that was “growing” in human serum.

We have examined electron micrographs in one of the reviews you referenced (Potgieter et al., FEMS Microbiology Reviews, 2015; 39:567). These micrographs were presented in support of the presence of bacteria in blood. Apparently, these micrographs were from studies that had been previously published, but the micrographs themselves had not been published. We were curious whether the assertion that certain structures in these images were bacteria was based solely on the way they appeared or was there other confirming evidence that they were bacteria.

In sum, we acknowledge that, based on our findings, we cannot conclude that there are not dormant microbes circulating in healthy human blood. However, we believe we have strong evidence indicating that pleomorphic structures in the supernatants of stored RBC units are not bacteria, dormant or otherwise.

Hopefully, others who review negative-stained EM images of pelleted material from RBC units, plasma, or serum will consider that pleomorphic structures in this material may be microparticles.

Best,

Charlie Searles (csearle@emory.edu)