Page 10 - Read Online
P. 10

Tejiram et al. Plast Aesthet Res. 2025;12:9  https://dx.doi.org/10.20517/2347-9264.2024.109  Page 5 of 16

               Sample collection
               Given the systemic inflammatory response that follows burn injury, there is potential for multiple
               microbiomes to be affected on a given host. Among organs, gut permeability increases in severe burn injury
               and may provide a nidus for gram-negative bacterial translocation. Although a host may harbor diverse
               microbiomes, an interrelationship can exist among them due to the inflammatory processes that follow
               burn injuries, which may make them comparable to one another. Sampling areas such as the buccal
               microbiome can be an indicator of gut microbiome taxonomy or behavior. To observe this over time,
               biological samples including wound swabs, buccal swabs, and blood samples were taken at serial time points
               from each patient: intraoperatively prior to wound bed preparation (pre-excision), post-excision, during the
               initial postoperative dressing takedown (1-3 days post grafting), and follow-up (7-10 days post discharge).
               Digital photography was also performed at each time point.

                                                                          TM
                                                                                       TM
               Swabs of the wound bed and donor site were obtained using BD BBL  CultureSwab  (Becton Dickinson,
               Franklin  Lakes,  NJ).  Swabs  were  deposited  in  iSwab  microbiome  collection  tubes  (Mawi  DNA
                                                                              [26]
               Technologies, Hayward, CA) and stored at -20 °C for 16S rRNA analysis . Buccal swabs were collected,
               deposited in iSwab microbiome collection tubes (Mawi), and stored at -20 °C for 16S rRNA analysis.

               DNA isolation from iSwabs
               The ZymoBIOMICS DNA/RNA Miniprep kit (Zymo Research, Irvine, CA) was used to extract DNA from
               samples per the manufacturer’s protocol and eluted using 50 µL of DNase/RNase-free water. DNA
               concentrations in extracts were then quantified using an Invitrogen Qubit 4 Fluorometer and 1X Qubit
               dsDNA High Sensitivity Assay Kit (ThermoFisher Scientific, Waltham, MA). All 16S rRNA illumina-tag
                                                                                               [27]
               PCR reactions were performed on DNA extracts per the Earth Microbiome Project’s protocol . Negative
               controls (PCR grade nuclease-free water) were processed along with DNA extracts for PCR amplification.
               PCR products and negatives were pooled. Gel purification was done with the pool and a 2% agarose gel
               using the QIAquick Gel Purification Kit (Qiagen, Frederick, MD). The purified pool was subsequently
               assessed for quality using an Agilent 2100 BioAnalyzer and Agilent DNA High Sensitivity DNA kit (Agilent
               Technologies, Santa Clara, CA). The purified pool was stored at -20 °C and then shipped to Laragen Inc.
               (Culver City, CA), which sequenced it using an Illumina MiSeq with v2 chemistry to generate 250 base pair
               paired-end reads.

               16S rRNA data pre-processing
               QIIME2 was used to process and analyze raw data . Cumulative expected error for each position was
                                                            [28]
               determined with VSEARCH . QIIME2’s implementation of the DADA2 pipeline was used to subject raw
                                       [29]
               sequence data to quality filtering, pairing, chimera removal, and ASV grouping. Based on VSEARCH’s
               calculated expected errors, forward reads were truncated at a length of 211, with a maximum expected error
               of 0.5, and reverse reads were truncated at a length of 196, with a maximum expected error of 0.5. This
               produced a table of ASVs and a file containing one representative sequence for each ASV.

               Taxonomy was assigned to the representative sequences with a Naive Bayes classifier implemented in
               QIIME2 and a pre-trained Silva 138 database containing 515F/806R sequences . A rooted phylogenetic
                                                                                    [30]
               tree was created with the representative sequences as well, with QIIME2’s implementation of MAFFT and
               FastTree [31,32] . The decontam package in R (R Core Team 2020) was used to identify likely contaminants
               based on the 11 negatives [33,34] . The ASVs identified as contaminants and negatives were then removed from
               the table. ASVs identified as mitochondria or chloroplasts were also filtered out on the basis that they likely
               represented eukaryotic contamination, instead of true bacterial signal. Lastly, samples with fewer than 1,000
               sequences after all filtering steps were removed.
   5   6   7   8   9   10   11   12   13   14   15