A Novel Pathogen Capturing Device for Removal and Detection

A simple technique that employs an antibody coated polydimethylsiloxane tube is used for effective capturing of bloodborne and foodborne pathogens. By recirculating the entire sample through the antibody coated tube, accumulation of target pathogens is achieved, thereby delivering a higher concentration of pathogens in a small volume. The described method can provide an effective and economical solution to microbiology techniques that rely on enrichment, thereby expediting diagnostics. Using this method 80.3 ± 5.6% of Staphylococcus aureus with a starting concentration of ~107 CFU/mL and 95.4 ± 1.0% of Methicillin-resistant Staphylococcus aureus with starting concentration of ~104 CFU/mL were removed from 5 mL blood in a few hours. This concept was extended to live rats with an induced bloodstream S. aureus infection. A reduction of two orders of magnitude in the bacterial load of the rats was observed within a few hours. The same technique was used to capture a food pathogen, Salmonella typhimurium, with starting concentrations as low as ~100 CFU, from 100 or 250 mL of culture broth within similar timeframes as above. The feasibility for food pathogen testing applications was additionally confirmed by capturing and detecting S. typhimurium in ground chicken and ground beef.

. Full record of bacterial growth profile analyzed by colony counting for: S. aureus capturing, MRSA capturing, MRSA capturing in high throughput condition, and S. typhimurium capturing. Table S1. Bacterial concentration (CFU/mL) in the sample before and after the 5-hour process in 5 mL sample volume and 0.5 mL flow rates. Growth is suppressed using tube capturing. Experiments 4 and 5 in MRSA were performed together with a shared control. Table S2. Bacterial concentration (CFU/mL) in the sample before and after by 5-hour of 20 mL sample volume at 5 mL flow rates.

S1.1. Preliminary confirmation -PCR detection of MRSA from blood
The capturing of bacteria by the antibody conjugated tube and subsequent lysis and detection by RT-PCR were confirmed in MRSA spiked sheep blood. 0.5 mL of 10 CFU/mL serial diluted MRSA (strain 43300) were spiked into 4.5 mL sheep blood (purchased from Hemostat Laboratories, Dixon, CA) supplemented with Na-citrate anticoagulant and pure tryptic soy broth (TSB), making 5 mL solution containing 5 CFU of MRSA. A 120 cm PBP2a antibody coated tube was connected to a vial containing the blood. The blood was circulated for 5 hours at 37 ˚C. Then, the tube was rinsed with 10 mL DI water. MRSA DNA was extracted from the tube by heating the tube filled with DI water to 100 ˚C for 10 min. RT-PCR was run with Qiagen mericon S. aureus kit. 10 µL of the extracted DNA sample was mixed with 10 µL of mastermix (reconstituted mericon assay). The sample was heated to 95 ˚C. Then, up to 70 cycles were performed (each cycle included denaturation at 95 ˚C for 15 sec, annealing at 60 ˚C for 15 sec, and extension at 72 ˚C for 10 sec). As shown in supplementary Fig

S2.1. Method: Blood analysis by complete blood counting with differential test
During in vivo studies, about 1 mL of rat blood samples were collected into vacutainer tubes with EDTA at the following time points: before injection of bacteria and at the end of 8 hours after infection (with and without tube capturing). The blood samples were shipped to a commercial veterinary diagnostics company (Antech Diagnostics, TN) for complete blood count with differential (CBC Differential) analysis. CBC differential analysis was performed for 8 rats: 4 control and 4 that underwent the tube capturing process. The results were statistically analyzed using T-test (two tailed, not paired) between the values before infection and the values 8 hours after infection for the group that underwent tube capturing and the control group. Capturing was initiated 3 hours after infection and the process lasted for 5 hours. By comparing the blood results before infection and 8 hours post infection, the impact of the infection can be assessed.
By comparing the group that underwent tube capturing vs the control group we can assess if the tube capturing process affected the rats. Statistically significant results are identified and discussed. The procedures are summarized in the diagram below.

S2.2. Result: CBC differential analysis
The results are summarized in supplementary table S3 and S4. Blood samples from infected and cannulated rats (n=4) that underwent the capturing process were compared to control groups of infected rats (n=4) without cannulation and not treated by capturing process.
Both groups showed similarly significant reduction in white blood cell (WBC) count. However, this reduction was observed from both captured and control groups without any statistically significant difference. The reduction in WBC count may be attributed to the effect of bacterial infection. Significantly elevated or decreased WBC count can be caused by sepsis 1,2 , indicating the rats were in a septic condition. Differential analysis exhibited an elevation in neutrophil (acting against bacterial infection) count ratio from ~ 12 % to ~ 45 % in captured and from ~ 11 % to ~ 40 % in control and the decline in lymphocyte (acting against viral infection) count ratio from ~ 86 % to ~ 50 % in the captured and from ~ 86 % to ~ 45 % in the control. This trend is an indication of bacterial infection 3,4 . Red blood cell (RBC) count in the treated rats showed slight reduction after capturing, while RBC count in controls showed slight increase. This effect was also confirmed significant in the captured vs. control comparison. The reduction in RBC count is due to either temporary blood loss because of the extracorporeal circulation or due to minor dilution induced by the heparin injection (about 1 mL over 5 hours) to prevent potential clotting in the catheter and tube (only the captured group needed heparin, note that the total blood volume of rats is about 25 mL). However, this effect was still tolerable because the overall RBC counts are within the reference range (provided by Antech Diagnostics). Hemoglobin level and hematocrit value in the before bacterial injection captured group vs. before bacterial injection control group showed a meaningful difference. This difference was observed in healthy rats before infection, and might be due to the implanted catheter. Reduction in hemoglobin concentration and hematocrit in the rats that underwent capturing can be attributed to the Heparin injection. Controls showed statistically significant increase in mean corpuscular volume (MCV) and increase in mean corpuscular hemoglobin concentration (MCHC). These changes were not observed in the captured group. However, comparison between the captured vs. control groups showed that these two changes were not statistically significant.

immunoassay
The capturing performance in ground chicken was also confirmed using two lateral flow immunoassay devices, namely Neogen's Reveal 2.0 Salmonella strip and Romer Labs' Rapidchek SELECT Salmonella test strip. Lateral flow immunoassays are low cost, simple, easy to use, and portable (point-of-care) detection methods, but they require considerably longer enrichment time than PCR because of a higher limit of detection (LOD, ~ 10 6 CFU/mL). Ground chicken was prepared as described in the previous section. 25 g of ground chicken were added into a Seward filtered sample bag. 225 mL of tryptic soy broth (TSB, non-selective) were added into the bag making the final total volume ~ 250 mL. The sample was blended manually for 5 minutes. About 25 CFU of S. typhimurium (ATCC 14028) were inoculated in the 250 mL. In the first test the solution was incubated for 14 hours at 42 ˚C. The length of tube was reduced to 40 cm (in a 4-way split). The tube was connected to the filter bag inside the incubator and the sample was circulated at a 6 mL/min flow rate. After 14 hours of capturing, the tube was gently rinsed with DI water and then filled with 1 mL of the Pluriselect detachment buffer solution.
After a 15-min detachment, the solution was collected and centrifuged at 3000 RPM for 5 minutes. The supernatant was carefully removed not to interrupt the pellet. The pellet was   Qualitative only 120 cm * 41.7 min 9 -10 * The tube was split in four parts in order to reduce the flow rate. Table S5. Summary of experimental parameters.
S4.1. The retention time within tube cannot be calculated in a straightforward manner, because (a) the proposed capturing system is based on constant flow of samples; (b) the sample volume differs by application (e.g. 5 mL -20 mL for blood experiments and 100-250 mL for food experiments) while the tube's capacity is same (about 1 mL for both blood and food); (c) the quantity of bacteria is continuously increasing due to their growth during the process; (d) the bacteria's growth pattern differs due differences in sample (blood vs. food), bacteria strains, and initial concentrations. Considering the above, the number of cycles, i.e. the number of times the entire sample volume is circulated through the tube during a given time, is a more appropriate figure of merit.