Bacterial interactions during sequential degradation of cyanobacterial necromass in a sulfidic arctic marine sediment

Summary Seafloor microorganisms impact global carbon cycling by mineralizing vast quantities of organic matter (OM) from pelagic primary production, which is predicted to increase in the Arctic because of diminishing sea ice cover. We studied microbial interspecies‐carbon‐flow during anaerobic OM degradation in arctic marine sediment using stable isotope probing. We supplemented sediment incubations with 13C‐labeled cyanobacterial necromass (spirulina), mimicking fresh OM input, or acetate, an important OM degradation intermediate and monitored sulfate reduction rates and concentrations of volatile fatty acids (VFAs) during substrate degradation. Sequential 16S rRNA gene and transcript amplicon sequencing and fluorescence in situ hybridization combined with Raman microspectroscopy revealed that only few bacterial species were the main degraders of 13C‐spirulina necromass. Psychrilyobacter, Psychromonas, Marinifilum, Colwellia, Marinilabiaceae and Clostridiales species were likely involved in the primary hydrolysis and fermentation of spirulina. VFAs, mainly acetate, produced from spirulina degradation were mineralized by sulfate‐reducing bacteria and an Arcobacter species. Cellular activity of Desulfobacteraceae and Desulfobulbaceae species during acetoclastic sulfate reduction was largely decoupled from relative 16S rRNA gene abundance shifts. Our findings provide new insights into the identities and physiological constraints that determine the population dynamics of key microorganisms during complex OM degradation in arctic marine sediments.© 2018 Society for Applied Microbiology and John Wiley & Sons Ltd


4
according to an established protocol (Pernthaler et al., 2002). The embedding of the filter in agarose was omitted in order to facilitate recovery of cells after hybridization. Endogenous peroxidases were inactivated via treatment with 0.1% H2O2 in 1x PBS for 2 minutes, followed by permeabilization according to Pernthaler et al., 2002. The cells were then hybridized in 300 μl of standard hybridization buffer (Pernthaler et al., 2002), without salmon sperm DNA and E.coli tRNA. The buffers for CARD-FISH and two-step CARD-FISH included 25 or 30% (vol/vol) formamide (Supplementary Table S5) and 1 μl of the probe (50 ng μl -1 ). For standard CARD-FISH the reaction chamber (1.5 mL microcentrifuge tube) was incubated for 3 h at 46°C (Woebken et al., 2012). For two-step CARD-FISH, the filter was first hybridized for 3 h at 46°C with the specific unlabeled oligonucleotide probe (Supplementary Table S5), followed by a second hybridization that was performed with the detector probe and a formamide concentration of 20% for 2 h. Afterwards, filters were washed and incubated for 15 minutes at room temperature in 1x PBS. Tyramide signal amplification in 1 ml amplification buffer (0.1% Blocking Reagent [wt/vol], 10% dextran sulfate [wt/vol], 2 M NaCl and 1x PBS) supplemented with 10 µL of 0.15% H2O2 and 1 µL of OregonGreen® 488 Tyramide solution (1 mg ml -1 in dimethylformamide) was performed in the dark for 15 minutes at 46°C. Filters were washed and incubated again for 15 minutes at room temperature in 1x PBS.

Cell separation from filters
Cells were removed from filters by sonication at a power setting of 20-25% on ice followed by vortexing for 30 minutes at 70% of the maximum speed on a Vortex-Genie 2. Suspended cells were pelleted by centrifugation at 10,621 g for 10 minutes and resuspended in 50 µL 1x PBS:ethanol (1:1 vol/vol).

Differences between 12 C-spirulina and 13 C-spirulina incubations
The prepared 13 C-spirulina and 12 C-spirulina solutions were visibly different in regards to color and quality. Concentrations of VFAs produced from 13 C-spirulina and 12 C-spirulina differed substantially ( Figure 1). In particular, the concentrations of most analyzed VFAs in 12 C-spirulina incubations at day 0 were considerably higher than in 13 C-spirulina incubations ( Figure 1). Accordingly, bacterial community responses differed between incubations with HD 13 C-spirulina and HD 12 C-spirulina (Supplementary Figure S3 and Supplementary Figure S4). Shifts in bacterial community composition and enrichment of responding phylotypes was more pronounced in 13 C-spirulina incubations. These observations suggest that the added 12 C-spirulina was already partly degraded.

Physiological properties of closest relatives of responsive phylotypes
Psychromonas phylotype 7435 has >99% 16S rRNA identity to different Psychromonas species (Supplementary Figure S5) with hydrolytic, fermentative and respiratory capabilities (Kawasaki et al., 2002;Xu, 2003;Riley et al., 2008). Psychrilyobacter phylotype 4749 shares 97.4% 16S rRNA identity with P. atlanticus, which ferments sugars and amino acids to acetate, butyrate, propionate and other products (Zhao et al., 2009). P. atlanticus was shown to metabolize 13 C-spirulina in a previous stable isotope probing experiment with temperate tidal flat sediment (Graue et al., 2012). Members of this genus/species might thus have a more ubiquitous role in complex OM degradation in marine sediments.
The closest relatives of Marinifilum phylotype 4400, M. fragile and M. flexuosum (97.5% 16S rRNA identity) (Supplementary Figure S5), ferment sugars such as glucose to acetate and propionate (Na et al., 2009;Ruvira et al., 2013). Colwellia phylotype 7234 had 100% 16S rRNA sequence identity to C. rossensis, which degrades various carbohydrates, fatty acids and amino acids (Techtmann et al., 2016).  ᵃAccording to Alm et al. (1996) ᵇΔG: Free energy for perfect match hybrid was predicted with Oligocalc (Sanguin et al., 2006) ᶜRDP II probe match was performed with database release 11, Update 5 (September 30, 2016 ) containing 3,356,809 16S rRNA sequences. The search for each probe was restricted to sequences of good quality with data in the respective probe binding region. Coverage is defined as the percentage of sequences in the RDP II target taxon with a full match to the probe sequence. The number of non-target hits indicates the total number of perfectly matching sequences outside the respective RDP II target taxon. ᵈNumber of additional, perfectly-matched phylotypes from this study with >0.1% relative 16S rRNA gene abundance in at least one sample