Page 38 - Read Online

Wang et al. Microbiome Res Rep 2024;3:39 https://dx.doi.org/10.20517/mrr.2024.21 Page 3 of 17

peptides, are implicated in cell-signaling, transport, enzymatic activities, antitoxin systems, pathogenic
colonization resistance, etc. [25-27] . Theoretically, these small proteins present in the intestinal samples can be
identified in metaproteomics data; however, in practice, their identification is often overlooked due to
suboptimal sample preparation and bioinformatics annotation steps, resulting in the loss of these portions
of protein components [28,29] . Sample preparation methods, such as differential centrifugation, are commonly
used to enrich and purify microbial cells from fecal sample debris and remove chemical contaminants that
can impact downstream protein extraction and protein digestion . However, the secreted small
[30]
antimicrobial peptides from either the host or microbes may be lost during differential centrifugation.

It is well recognized that different sample processing methods such as the use of differential centrifugation,
protein extraction methods, and protein digestion methods, and differences in data analysis workflows can
yield different results and metaproteomic insights [30-32] . We have previously reported that physical disruption
using bead beating or ultrasonication is needed to optimally extract proteins from Bacillota (previously
[32]
Firmicutes) . Tanca et al. showed that stool pretreatment by differential centrifugation significantly
impacted metaproteomic observations . However, there is still a need for further evaluation of the bias
[30]
introduced by different steps of the experimental workflow as well as the emerging DIA data acquisition
mode, and how they can contribute to the recovery of otherwise overlooked components, such as small
antimicrobial proteins. Therefore, in this study, we conducted a comprehensive comparison of fecal sample
preparation, protein digestion, data acquisition mode, and bioinformatic workflow using mouse feces to
serve as a guide for metaproteomic experimental design.

METHODS
Mouse feces collection and differential centrifugation
Mouse fecal samples were collected from nine female C3H/HeN mice purchased from Charles River
Laboratories, Senneville, Quebec, Canada. The animal procedures were approved by the Animal Care
Committee in Health Canada and performed in accordance with institutional guidelines. Fecal samples
were combined and crushed into a homogenous powder. This stock fecal sample was used for both
differential centrifugation (DC) and non-differential centrifugation (NC) workflows.

Differential centrifugation of feces was performed according to a previous study . Briefly, 0.5 mL glass
[32]
beads (BioSpec, Cat#11079125) and 7.5 mL cold phosphate-buffered saline (PBS) per gram of fecal sample
were added to the samples, followed by vortexing and centrifugation at 300 g at 4 °C for 5 min to collect the
supernatant. The remaining fecal pellets were extracted three more times with 7.5, 5, and 5 mL cold PBS,
and the fecal pellet at the end of the 4 extractions was discarded. Once pooled, additional debris from the
extractions was removed by three centrifugations at 300 g at 4 °C for 5 min. The supernatant extract was
then spun down at 14,000 g at 4 °C for 20 min to collect the microbial pellet. The microbial pellet was
washed twice with cold PBS by resuspending and centrifuging at 14,000 g at 4 °C for 20 min, then frozen
until use.

Protein extraction, trypsin digestion and desalting
Sample lysis
NC feces were lysed by directly resuspending dry, frozen, crushed fecal pellet in lysis buffer at a ratio of
20 mg dry fecal powder material / 1 mL lysis buffer. DC microbial pellets were lysed by resuspending the
wet microbial pellets with a ratio of 100 mg wet pellets / 1 mL lysis buffer (assuming that 20% of the wet
pellets are dry materials). The same lysis buffer was used for all samples, consisting of 4% (w/v) sodium
dodecyl sulfate (SDS), 8 M urea, and 1× Halt protease inhibitor single-use cocktail (Thermo Scientific,
Cat#78425) in 50 mM Tris-HCl (pH 8). Sample lysates were sonicated using a QSonica Q700 water-chilled