Impact of enzymatic digestion on bacterial community composition in CF airway samples

Abstract

Background: Previous studies have demonstrated the importance of DNA extraction methods for molecular detection of Staphylococcus, an important bacterial group in cystic fibrosis (CF). We sought to evaluate the effect of enzymatic digestion (EnzD) prior to DNA extraction on bacterial communities identified in sputum and oropharyngeal swab (OP) samples from patients with CF.

Methods: DNA from 81 samples (39 sputum and 42 OP) collected from 63 patients with CF was extracted in duplicate with and without EnzD. Bacterial communities were determined by rRNA gene sequencing, and measures of alpha and beta diversity were calculated. Principal Coordinate Analysis (PCoA) was used to assess differences at the community level and Wilcoxon Signed Rank tests were used to compare relative abundance (RA) of individual genera for paired samples with and without EnzD.

Results: Shannon Diversity Index (alpha-diversity) decreased in sputum and OP samples with the use of EnzD. Larger shifts in community composition were observed for OP samples (beta-diversity, measured by Morisita-Horn), whereas less change in communities was observed for sputum samples. The use of EnzD with OP swabs resulted in significant increase in RA for the genera Gemella (p < 0.01), Streptococcus (p < 0.01), and Rothia (p < 0.01). Staphylococcus (p < 0.01) was the only genus with a significant increase in RA from sputum, whereas the following genera decreased in RA with EnzD: Veillonella (p < 0.01), Granulicatella (p < 0.01), Prevotella (p < 0.01), and Gemella (p = 0.02). In OP samples, higher RA of Gram-positive taxa was associated with larger changes in microbial community composition.

Discussion: We show that the application of EnzD to CF airway samples, particularly OP swabs, results in differences in microbial communities detected by sequencing. Use of EnzD can result in large changes in bacterial community composition, and is particularly useful for detection of Staphylococcus in CF OP samples. The enhanced identification of Staphylococcus aureus is a strong indication to utilize EnzD in studies that use OP swabs to monitor CF airway communities.

Keywords: 16s Sequencing; Beta diversity; DNA extraction; Gram positive; Microbiome; Sputum.

Figure 1. Boxplots for individual taxa that change within paired samples.

The distribution of differences in paired samples is shown for phyla (A) and genera (B). Taxa with limited differences between the paired samples are tightly distributed around zero, and those with increased relative abundance after EnzD have positive distributions. Note the median RA for Fusobacteria in sputum is less than 1%. P-values correspond to Benjamini–Hochberg corrected Wilcoxon Signed rank test for paired samples.

Figure 2. Boxplots showing distribution of Morisita-Horn (MH) similarity metric for paired samples in OP and sputum.

Wilcoxon rank sum tests indicate the distributions between the two sample types were statistically different (p < 0.01). The MH metric ranges from 0 to 1 with a value of 1 indicating identical communities, and 0 indicating no overlap between communities. The line indicates the median value and the box ranges from the 25th and 75th percentiles, the whiskers extend to 1.5 times the interquartile range.

Figure 3. Ordination biplot, using 1-MH beta diversity values for OP (A) and sputum (B).

Values closer together are more similar. Paired samples are connected with a line, vectors for the genera with at least 1% RA and p < 0.005 (by permutation test) for R² of the loading.

Figure 4. (A) Scatterplot of relative abundance of Gram-positive taxa in non-EnzD samples versus Morisita-Horn (MH) similarity metric of paired samples. MH metric ranges from 0 to 1 with a value of 1 indicating identical communities, and 0 indicating no overlap between communities. OP samples are shown in red and sputum in blue. (B) Scatterplot of the difference in relative abundance of Gram-positive taxa versus MH of paired samples. OP samples are shown in red and sputum in blue.