Recommendations for Future Work

105 In Chapter 3, we found that by producing methane from thin stillage (i.e., essentially the waste left over after distillation to remove ethanol from fermented “beer”), corn- kernel to ethanol plants could offset a significant amount of the energy they consume. Most of the energy consumed in the plants is for distillation (Chapter 2), which upgrades ethanol from low concentrations in beer (with little value) to a 95% pure form that is valuable as a fuel and chemical feedstock. In Chapter 5, we found that ethanol pushing helped to increase the value of carboxylate products fermented from dilute-acid pretreated corn fiber to n-caproate and n-caprylate. Further, n-caproate and n-caprylate are easy to extract with membrane-based systems, which uses little energy (other than producing alkali chemicals to maintain the pH gradient). This represents a new method to indirectly recover ethanol while converting both the ethanol and lignocellulosics contained in “beer” to a specific end product. The economic and environmental sustainability of this process is unclear, and an assessment should include a full life-cycle analysis.

Lactate chain elongation in thermophilic bioreactors

In Chapter 4, we discovered that lactate was produced as an intermediate during the feeding cycle of the bioreactors producing primarily n-butyrate, because of fermentation of saccharides for pretreatment of corn fiber. Lactate was most likely consumed by a coupling of lactate oxidation to acetate and n-butyrate reduction to produce n-butyrate and n-caproate, respectively (i.e., lactate chain-elongation), perhaps by a member of the genus Thermosinus. In Chapter 5, we found that bioreactor communities were incapable of catalyzing ethanol chain-elongation at thermophilic (55oC) temperatures, and functional metagenomics confirmed that the community was incapable of performing the

106 process at these conditions. Thus, lactate should be evaluated as an electron-pushing substrate to upgrade the product spectrum in thermophilic bioreactors. In that study we correlated n-caproate production with Thermosinus, but it is possible that Thermosinus is only involved in an aspect of the chain elongation. Therefore, the process should be evaluated in lab-scale carboxylate platform bioreactors, coupled with a functional metagenomic analysis to evaluate which factors contribute to lactate chain elongation so that the process can be optimized. Based on our results with ethanol chain elongation, the study should focus on whether product extraction is required for chain elongation and whether or not significant increases in extraction efficiency improve performance.

The role of community evenness in thermophilic bioreactors

Research has already revealed that community evenness is important in maintaining ecosystem function under stress in mixed microbial communities [Werner et al., 2011b; Wittebolle et al., 2009]. In Chapter 4 we discovered that our thermophilic bioreactor communities were relatively uneven, and it is unclear if this is a function of temperature, toxicity of end products, or both, or if an uneven structure is relative and is always a result of thermophilic temperatures. To study the stability of thermophilic systems, future research should implement high-throughput 16S rRNA gene surveys of microbial community structures in bioreactors to evaluate evenness under a range of conditions. One possible scenario would compare four bioreactors fed the same lignocellulosic substrate with key differences: 1. A mesophilic anaerobic digester supplemented with lactate and with optimum conversion of carboxylates to methane, 2. A thermophilic anaerobic digester supplemented with lactate and with optimum conversion of carboxylates to methane, 3. A thermophilic bioreactor supplemented with lactate to

107 accumulate a mix of carboxylates, and 4. A thermophilic bioreactor supplemented with lactate and with high-rate product specific extraction to remove nearly all carboxylates from the system. This scenario would determine whether thermophilic temperatures are cause for uneven communities in bioreactors, if product removal is responsible for improvements in evenness in thermophilic environments, or if anaerobic digestion results in an even community because of the series of metabolic conversions required to completely convert substrate to methane.

Chain elongation pathways in ethanol-supplemented bioreactors

Previous research demonstrated with selective 16S rRNA gene sequencing techniques (i.e., DGGE followed by manual band selection, isolation, and cloning) that chain elongation with ethanol as the electron pushing substrate at pH 7 was catalyzed by microbial communities dominated by Clostridium kluyveri (26/45 clones had 98% sequence similarity to C. kluyveri) in their bioreactors [Steinbusch et al., 2011]. In Chapter 5, our metagenomic survey indicated that more than one microorganism (including a member of the family Syntrophomonadaceae) might be responsible for chain elongation with ethanol at pH 5.5. If this is true, it is important to reveal the mechanisms of the chain elongation process to help reveal how different operating conditions would affect the process. For example, a distribution of the three metabolic functions ethanol oxidation, intermediate metabolite formation, such as crotonate, and conversion of intermediates to n-caproate would force a reevaluation of what is considered to be an optimum environment. Experiments should use high-throughput 16S rRNA sequencing, possibly combined with standard Sanger sequencing of isolated or enriched microbes or consortia and/or Sanger sequencing of isotope-labeled “heavy” 16S rRNA genes (i.e.,

108 DNA-SIP) to determine with more certainty which microorganisms are responsible for chain elongation. After this analysis, hypotheses can be made about whether or not direct oxidation of ethanol to acetate and hydrogen is necessary for chain elongation, or if other intermediates contribute to syntrophic relationships leading to chain elongation, and further experiments can determine the precise mechanisms in the bioreactors.

Further research on the implications of methanogenesis for chain elongation

In Chapter 5, we found that hydrogenotrophic methanogens could reduce carbon dioxide to methane side-by-side with chain elongation reactions that produce medium- chain carboxylates, even though the hydrogen partial pressure was very low. This finding alters the prevailing view that methanogens must be inhibited to prevent loss of product to competing processes [Agler et al., 2011; Steinbusch et al., 2009]. While we did show that n-butyrate oxidation would be unlikely at our bioreactor conditions, we did not perform experiments to prove that we did not experience some loss of product to oxidation reactions. We also were unable to determine if high-efficiency product extraction would allow the onset of carboxylate oxidation. If necessary, methanogens can be inhibited at both thermophilic (Appendix 2) and mesophilic [Steinbusch et al., 2009; Van Kessel and Russell, 1996] temperatures, but the two prevailing techniques, chemical additives and carboxylate inhibition, add cost and limit bioreactor productivity, respectively. Further, the added benefit of increased carbon recovery would be lost. Thus, experiments should determine with more certainty if a bioreactor pH of 5.5 at mesophilic temperatures prevents carboxylate oxidation so that methanogenesis and chain elongation with ethanol can exist together side-by-side (or if product extraction out- competes carboxylate oxidation). Experiments should also explore our hypothesis that

109 for chain elongation at pH 5.5, hydrogenotrophs are required to maintain a low hydrogen partial pressure for bioreactor functionality (Chapter 5).

Isolation and characterization of chain elongating microbes and/or consortium In parallel to experiments studying optimization of chain elongation with lactate or ethanol at thermophilic and mesophilic temperatures, respectively, isolated microbes or microbial consortia should be cultured and characterized. In thermophilic systems, we expect this to include species within Thermosinus that catalyze chain elongation of acetate and n-butyrate with lactate. Because other related microbes have been described that convert lactate to n-caproate [Marounek et al., 1989], we hypothesize that one microbe may be responsible for this reaction. In mesophilic systems, we have hypothesized that a consortium of microorganisms is responsible for chain elongation with ethanol. Culturing the responsible consortia should be followed by metabolic analysis with stable isotopes to determine which intermediate chemicals are involved. If direct ethanol oxidation to acetate is involved, the ethanol-oxidizing microbe could be isolated on ethanol with artificial hydrogen removal or in culture with a hydrogenotrophic microorganism. Once intermediates are identified, the culture could be fed the intermediate substrate to enrich a specific community member and DNA-SIP could be used with a labeled carbon source to determine with certainty which microorganisms are responsible for chain elongation.

110 APPENDIX 1.

SUPPLEMENTAL INFORMATION FOR: THERMOPHILIC ANAEROBIC