Figure 2 – Non-essential nutrients stimulate metabolism and manipulate the microbiome composition. A) Pressure generated in microcosms from Well D & G when treated with various nutrient supplements; T-tests performed with respect to None. B) Diversity score (Pielou’s evenness) of microcosms after treatment with micronutrients (days 3, 4, 5 and 7). C) Unweighted Unifrac PCoA plot of Well D microcosms after micronutrient supplementation; shapes represent days, colors indicate treatment group: spheres = day 1, diamonds = day 3, star = day 4, rings = day 5, cones = day 7. Error bars represent standard deviation, n=3; * = p < 0.05, ** = p < 0.005, and *** = p < 0.001.
While the fate of nitrate is ultimately determined by the microbial community, we found that nitrate and molybdate supplements shifted the microbial community composition in different ways. Nitrate, molybdate, and the nitrate-molybdate combination treatments provide a selective pressure that selects for or enriches specific species reducing the alpha diversity (Shannon index) of the cultures from Well D (Figure 2B ). However, the treatments do not enrich for the same species - the beta diversity of the communities as estimated by principal coordinate analysis (PCoA) analysis of the 16S rRNA gene composition (Figure 2C ) reveal distinct groups of data points. Moreover, the time evolution of these communities diverge over the first three days (spheres to diamonds) as a result of their different selective pressures. However, the combined molybdate and nitrate treatment converges on that of molybdate alone (Figure 3C ) suggesting that the selection pressure applied by molybdate is stronger than that of nitrate and, ultimately, determines microbial metabolism and community structure. Hydrogen sulfide inhibition in the literature is more sensitive to molybdate than nitrate42,43, adding further support for the strong selective pressure of molybdate.
To determine which taxa were specifically selected against and enriched in these microcosms, we evaluated the 16S rRNA gene composition of the microbial community. Interestingly, we found that after Day 1 of cultivation, there was little effect across the treatments with only a slight increase in theEnterobacteriaceae in the molybdate and combined treatments (Supplemental Figure 4 ). After Day 3, however, the treatments varied noticeably from the untreated microcosms (Figure 3A ). For example, nitrate microcosms saw a bloom of Sulfospirillium in days 3 and 4, while the molybdate and combined treatments each saw a bloom in Enterobacteriaceae sp. and Lachnoclostridiumsp., which are MEOR-relevant bacteria. ManySulfospirillium species contain both nitrate- and sulfur-reducing pathways, so it is reasonable to expect that the nitrate substrate allowed these species to outgrow organisms that could not use this nutrient. Similarly, a few Enterobacteriaceae have been found to reduce molybdate44,45 which may have provided a similar advantage to these microbes. At the same time, molybdate is structurally similar to sulfate and has been shown to inhibit the growth of H2S producers43 which may account for the absence of Desulfovibrio , Dethiosulfovibrio ,Sulfospirillium , and other sulfate reducing bacteria in the molybdate and combined treatment microcosms. After 7 days, the cultures primarily consisted of members of Bacteroides andLachnoclostridium regardless of treatment suggesting that the community may have shifted to scavenge or use other nutrient sources to survive. Similar progressions of microcosms toward primarily Bacteroidetes and Clostridia have been seen in previous studies36 suggesting these groups of bacteria have the advantage over other organisms in anaerobic consortia when the initial nutrients have been depleted at the end of the cultivation. Additionally, no known H2S reducing bacteria were detected with molybdate and combination treatments suggesting that the effects of the molybdate were effective over this whole time-course of the cultivation. In contrast, Sulfurospirillium species were present at several timepoints in the nitrate microcosms. In a previous H2S inhibition study, molybdate was also found to be the strongest inhibitor of H2S reducing bacteria43 from marine enrichment cultures. Where previous findings suggested that the addition of these inhibitors had no effect on the composition of the microbial community,42 we find that they can drastically shape the community composition and metabolism. However, our findings that Well D and G (Supplemental Figure 4) respond differently to the nitrate and molybdate inhibitors further support previous assertations that the efficiency of H2S inhibitors depends on the composition of the microbiome.42,43 While changes in microbial community composition indicate a response to our top-down engineering, they are not definitive of metabolic output, which must be characterized directly.