Microbial Functional Responses to Cholesterol Catabolism in Denitrifying Sludge

Abstract

The 2,3-seco pathway, the pathway for anaerobic cholesterol degradation, has been established in the denitrifying betaproteobacterium Sterolibacterium denitrificans. However, knowledge of how microorganisms respond to cholesterol at the community level is elusive. Here, we applied mesocosm incubation and 16S rRNA sequencing to reveal that, in denitrifying sludge communities, three betaproteobacterial operational taxonomic units (OTUs) with low (94% to 95%) 16S rRNA sequence similarity to Stl. denitrificans are cholesterol degraders and members of the rare biosphere. Metatranscriptomic and metabolite analyses show that these degraders adopt the 2,3-seco pathway to sequentially catalyze the side chain and sterane of cholesterol and that two molybdoenzymes-steroid C25 dehydrogenase and 1-testosterone dehydrogenase/hydratase-are crucial for these bioprocesses, respectively. The metatranscriptome further suggests that these betaproteobacterial degraders display chemotaxis and motility toward cholesterol and that FadL-like transporters may be the key components for substrate uptake. Also, these betaproteobacteria are capable of transporting micronutrients and synthesizing cofactors essential for cellular metabolism and cholesterol degradation; however, the required cobalamin is possibly provided by cobalamin-de novo-synthesizing gamma-, delta-, and betaproteobacteria via the salvage pathway. Overall, our results indicate that the ability to degrade cholesterol in sludge communities is reserved for certain rare biosphere members and that C25 dehydrogenase can serve as a biomarker for sterol degradation in anoxic environments. IMPORTANCE Steroids are ubiquitous and abundant natural compounds that display recalcitrance. Biodegradation via sludge communities in wastewater treatment plants is the primary removal process for steroids. To date, compared to studies for aerobic steroid degradation, the knowledge of anaerobic degradation of steroids has been based on only a few model organisms. Due to the increase of anthropogenic impacts, steroid inputs may affect microbial diversity and functioning in ecosystems. Here, we first investigated microbial functional responses to cholesterol, the most abundant steroid in sludge, at the community level. Our metagenomic and metatranscriptomic analyses revealed that the capacities for cholesterol approach, uptake, and degradation are unique traits of certain low-abundance betaproteobacteria, indicating the importance of the rare biosphere in bioremediation. Apparent expression of genes involved in cofactor de novo synthesis and salvage pathways suggests that these micronutrients play important roles for cholesterol degradation in sludge communities.

Keywords: activated sludge; biodegradation; cholesterol; cobalamin; metatranscriptome; molydoenzyme; rare biosphere.

FIG 1

The 2,3-seco pathway of anaerobic cholesterol metabolism in Stl. denitrificans. Enzymes functionally characterized in previous studies are bolded. AcmA, cholesterol dehydrogenase/isomerase; AcmB, cholest-4-en-3-one-Δ1-dehydrogenase; C25DH, cholesterol C25 dehydrogenase; ACS, acyl-CoA synthetase; ACAD, acyl-CoA dehydrogenase; ECH, enoyl-CoA hydratase; ALD, aldolase; ALDH, aldehyde dehydrogenase; Δ4R, ketosteroid-Δ4-reductase; DH, dehydrogenase; AtcABC, 1-testosterone dehydrogenase/hydratase.

FIG 2

The consumption of cholesterol and nitrate in (A) sludges treated with nitrate only, (B) sludges treated with cholesterol only, and (C) sludges treated with nitrate and cholesterol. Data are shown as means of results from three independent measurements with standard deviations.

FIG 3

UPLC-APCI-MS analysis of steroid metabolites extracted from sludges treated with cholesterol and nitrate. The log-transformed intensity of each identified metabolite was determined for three independent replicates at each incubation time. The original mass spectrum of each metabolite is listed in Fig. S1.

FIG 4

Principal-coordinate analysis (PCoA) for determination of similarities between bacterial communities based on UniFrac distance matrix data (OTU level) from the original sludges and from sludges treated with cholesterol and nitrate (SCN), nitrate only (SN), and cholesterol only (SC) at each incubation time (day 0 to day 16). Colors represent the Shannon diversity index of each data point.

FIG 5

Temporal changes in OTUs classified as genus Sterolibacterium from sludges treated with cholesterol only (SC), cholesterol and nitrate (SCN), and nitrate only (SN) during mesocosm incubation (day 0 to day 16). Data are shown as averages of triplicates.

FIG 6

Taxonomic contribution of each KEGG pathway involved in (A) chemotaxis, (B) type IV pili, (C) flagellar, (D) uroporphyringonen and ALA synthesis, (E) cobalamin de novo synthesis, and (F) cobalamin salvage synthesis in SCN (blue) and SN (orange) communities. Each arm of each spider chart indicates the mean of log-transformed counts per million (CPM) for a given bacterial phylum. The −2 value represents a mean CPM value of 0 before log transformation. AQ, Aquificae; BP, Betaproteobacteria; DP, Deltaproteobacteria; GP, Gammaproteobacteria; FM, Firmicutes; TM, Thermotogae. Other phyla denoted by two-letter codes are listed in Table S4.

FIG 7

Volcano plot of significantly expressed transcripts involved in anaerobic cholesterol degradation between SCN and SN samples. Up to 257 transcripts involved in cholesterol uptake, A-ring modification, C25 hydroxylation, side chain degradation, and sterane degradation were identified, and only two transcripts annotated as C/D-ring degrading were expressed in SN samples.

FIG 8

(A) Maximum likelihood tree of 16S rRNA gene of steroid-degrading anaerobes and Sterolibacterium-like OTUs. Only enriched OTUs are highlighted. (B) AtcA and its homologues and other molybdopterin-containing xanthine oxidase family members. (C) Alpha subunit of C25DH and its homologues. Subunits (A1 to A4) displaying specificity for side chain substrates are highlighted in blue. Branch support of higher than 50% of bootstrapping time is shown. Details for all reference sequences are provided in Table S3.