Authorea

chuckpr final edits for submission almost 9 years ago

Commit id: 4fa9f34d0ebc46dc569a0abc0e184b29ed8501d5

deletions | additions

predominantly \textit{Firmicutes} at day~1 followed by \textit{Bacteroidetes} at day~3 and then \textit{Actinobacteria} at day~7. These dynamics of $^{13}$C-labeling suggest labile C traveled through different trophic levels within the soil bacterial community. Microorganisms In contrast, the microorganisms that metabolized cellulose-C increased in relative abundance over the course of the experiment later (after 14 days) with the highest number of OTUs exhibiting evidence for $^{13}$C-assimilation after 14 days. Microbes Microorganisms that metabolized cellulose-C belonged to cosmopolitan soil lineages that remain uncharacterized including \textit{Spartobacteria}, \textit{Chloroflexi} and \textit{Planctomycetes}. Using an approach that reveals the C assimilation dynamics of specific microbial lineages we describe

responders changed between days~1,~3 and~7 and few OTUs appeared $^{13}$C-labeled in the $^{13}$C-xylose treatment after day~7. In the $^{13}$C-cellulose treatment, $^{13}$C-labeled OTUs were few in the beginning of the experiment and most abundant on day~14 days~14 and~30. Finally, few (8~of~104) OTUs appeared to metabolize both xylose and cellulose indicating most $^{13}$C-responders had distinct activity and that cellulose responders grew in succession to xylose responders. activity. % Fakesubsubsection: Correlations between community composition The ecological characteristics of microorganisms are often inferred from

assimilate C from multiple sources. Xylose responders assimilated xylose-C into DNA within~24 hours and had low $\Delta\hat{BD}$ relative to cellulose responders suggesting xylose was not the sole C source used for growth. Xylose represented 15\% of the amendment and~3.5\% and~3\% of total soil C. Xylose responders often included the most abundant OTUs within the non-fractionated DNA and had high estimated \textit{rrn} copy number relative to cellulose responders. However, to some degree, high \textit{rrn} gene copy number may inflate

estimated \textit{rrn} copy number than xylose responders. The majority of cellulose responders were not close relatives of cultured isolates although a number of cellulose responders shared high SSU rRNA gene sequence identity with cultured \textit{Proteobacteria} (e.g. \textit{Cellvibrio}), . \textit{Cellvibrio}). We identified cellulose responders among phyla such as \textit{Verrucomicrobia}, \textit{Chloroflexi}, and \textit{Planctomycetes} -- common soil phyla whose functions within soil communities remain unknown. % Fakesubsubsection: Verrucomicrobia comprise \textit{Verrucomicrobia} represented 16\% of the cellulose responders. \textit{Verrucomicrobia} are cosmopolitan soil microbes microorganisms \citep{Bergmann_2011} that can make up to 23\% of SSU rRNA gene sequences in soils \citep{Bergmann_2011} and 9.8\% of soil SSU rRNA \citep{Buckley_2001}. Genomic analyses and laboratory experiments show that various isolates within the

% Fakesubsubsection: Responders did not necessarily Responders did not necessarily assimilate $^{13}$C directly from $^{13}$C-xylose or $^{13}$C-cellulose. In $^{13}$C-cellulose but, in many ways, knowledge of secondary C degradation and/or microbial biomass turnover may be more interesting with respect to the soil C-cycle than knowledge of primary degradation. The response to xylose suggests xylose-C moved through different

\textit{Actinobacteria} and \textit{Bacteroidetes} xylose responders consumed waste products generated by primary xylose metabolism (e.g. organic acids produced during xylose metabolism). These latter two hypotheses cannot explain the sequential loss of $^{13}$C-label, $^{13}$C-label in combination with the abundance dynamics in non-fractionated DNA, however. If trophic transfer caused the activity dynamics, at least three different ecological groups exchanged C in~7 days. Models of the soil C cycle often exclude trophic interactions between soil bacteria (e.g.

traits that allow organisms to compete for a given substrate as it occurs in the soil. For instance, fast growth and rapid resuscitation allow microorganisms to compete for labile C which may often be transient in soil. Hence, life history traits may constrain the diversity of microbes microorganisms that metabolize a given C source in the soil under a given set of conditions.

\citep{wieder_2014a}. Including these functional types improved predictions of C storage in response to environmental change. We identified microbial lineages engaged in labile and structural C decomposition that can be defined as copiotrophs or oligotrophs, respectively. Our results suggest greater and or faster turnover for copiotroph biomass relative to oligotroph biomass, and that the copiotroph-oligotroph dichotomy leaves out guilds Additionally, we show thatmay play important roles in soil-C cycling. That is, soil-C may travel through multiple bacterial trophic levels where each C transfer represents an opportunity for C stabilization in association with soil minerals or C loss by respiration. Our understanding of soil C dynamics will likely improve as we develop a more granular understanding of the ecological diversity of microorganisms that mediate C transformations in soil. \subsection{Conclusion} % Fakesubsubsection: Microorganisms sequester atmospheric C Microorganisms govern\\ C-transformations in soil influencing climate change on a global scale but we do not know the identities of microorganisms that carry out specific transformations. In this experiment microbes microorganisms from physiologically uncharacterized but cosmopolitan soil lineages participated in cellulose decomposition. Cellulose responders included members of the \textit{Verrucomicrobia} (\textit{Spartobacteria}), \textit{Chloroflexi},

and \textit{Actinobacteria} likely became labeled by consuming $^{13}$C-labeled constituents of microbial biomass either by saprotrophy or predation. Our results suggest that cosmopolitan \textit{Spartobacteria} may degrade cellulose on a global scale, decomposition of labile plant C may initiate trophic transfer within the bacterial food web, and life history traits may act as a filter constraining the diversity of active microorganisms relative to those with the genomic potential for a given metabolism.

\citep{Saha2003}. Xylose is often the most abundant sugar in hemicellulose, comprising as much as 60-90\% of xylan in some plants (e.g hardwoods \citep{Spiridon2008}, wheat \citep{Sun2005}, and switchgrass \citep{Bunnell2013}). Microbes Microorganisms that respire labile C in the form of sugars proliferate during the initial stages of decomposition \citep{Garrett1951,Alexander1964}, and metabolize as much as 75\% of sugar C during the first 5 days \citep{Engelking2007}. In contrast,

\citep{Garrett1963,Bremer1994} followed by slow growing organisms targeting structural C such as cellulose \citep{Garrett1963}. Evidence to support the degradative succession hypothesis comes from observing soil respiration dynamics and characterizing microbes microorganisms cultured at different stages of decomposition. The degree to which the succession hypothesis presents an accurate model of litter decomposition has been questioned \citep{AnneliseHKjoller2002,Frankland_1998,Osono_2005} and it's clear that we

\citep{Cavigelli2000}, nitrification \citep{Carney2004,Hawkes2005,Webster2005}, methanotrophy \citep{Gulledge1997}, and nitrogen fixation \citep{Hsu2009}). However, the lack of diagnostic genes for describing soil-C transformations has limited progress in characterizing the contributions of individual microbes microorganisms to decomposition. Remarkably, we still lack basic information on the physiology and ecology of the majority of organisms that live in soils. For example, contributions to soil processes remain uncharacterized for

\citep{Buckley_2007,9780408708036,Holben1995,Nusslein1999}. As a result, most applications of SIP have targeted specialized microorganisms such as methanotrophs \citep{radajewski2000stable}, methanogens \citep{lu2005}, syntrophs \citep{lueders2004}, or microbes microorganisms that target pollutants \citep{derito2005}. Exploring the soil-C cycle with SIP has proven to be more challenging because SIP has lacked the resolution necessary to characterize the specific contributions of individual microbial groups to the decomposition of

Yan, New York. Soils were sieved (2 mm), homogenized, distributed into flasks (10 g in each 250 ml flask, n = 36) and equilibrated for 2 weeks. We amended soils with a mixture containing 2.9 mg C g$^{-1}$ soil dry weight (d.w.) and broughtexperimental soil to 50\% water holding capacity. By mass the amendment contained 38\% cellulose, 23\% lignin, 20\% xylose, 3\% arabinose, 1\% galactose, 1\% glucose, and 0.5\% mannose. 10.6\% amino acids (Teknova C9795) and 2.9\% Murashige Skoog basal salt mixture which contains macro and

framework \citep{love2014}, to identify OTUs that were enriched in high density gradient fractions from $^{13}$C-treatments relative to corresponding gradient fractions from control treatments (for review of RNA-Seq differential expression statistics applied to microbiome OTU count data see (30)). \citep{McMurdie2014}). We define "high density gradient fractions" as gradient fractions whose density falls between 1.7125 and 1.755 g ml$^{-1}$.Briefly, DESeq2 includes several features that enable robust estimates of standard error in addition to reliable ranking of logarithmic fold change (LFC) (i.e. gamma-Poisson regression coefficients) in OTU relative abundance even with low count OTUs where LFC can often be noisy. Further, statistical evaluation of LFC can be performed with user-selected thresholds as opposed to the typical null hypothesis that LFC is exactly zero enabling the most biologically interesting OTUs to be identified for subsequent analyses. For each OTU, we calculated LFC calculates logarithmic fold change (LFC) and corresponding standard errors error for enrichment in high densitygradient fractions of $^{13}$C treatments relative to control. Subsequently, a one-sided Wald test was used tostatistically assess the statistical significance of LFC values. The user-defined values with the null hypothesiswas that LFC was less than one standard deviation above the mean of all LFC values.P-values were corrected for multiple comparisons using the Benjamini and Hochberg method \citep{benjamini1995}. We independently filtered OTUs prior to multiple comparison corrections on the basis of sparsity prior to correcting P-values for multiple comparisons. The sparsity value that yielded the most adjusted P-values less than 0.10 was selected for independent filtering by sparsity. Briefly, eliminating OTUs were eliminated if they that failed to appear in at least 45\% of high densitygradient fractions for a given $^{13}$C/control treatment pair. These sparse OTUs are unlikely to have sufficient data to allow comparison. P-values were adjusted for multiple comparisons using the determination of statistical significance. Benjamini and Hochberg method \citep{benjamini1995}. We selected a false discovery discoverty rate of 10\% to denote statistical significance. See SI for additional information on experimental and analytical methods.

microorganisms had $^{13}$C-labeled DNA in $^{13}$C-cellulose treatments at days~14 and~30. In contrast, in the $^{13}$C-xylose treatment, the SSU rRNA gene composition of high density fractions varied between days~1,~3,~and~7 indicating that different microbes microorganisms had $^{13}$C-labeled DNA on each of these days. In the $^{13}$C-xylose treatment, the SSU gene composition of high density fractions was similar to control on days~14~and~30 (Figure~\ref{fig:ord}) indicating that $^{13}$C was no longer detectable in

experiment by surveying SSU rRNA genes in non-fractionated DNA from the soil. The SSU rRNA gene composition of the non-fractionated DNA changed with time (Figure~\ref{fig:bulk_ord}, P-value~$=$~0.023, R$^{2}$ $=$~0.63, Adonis test \citep{Anderson2001a}). In contrast, the non-fractionated DNA SSU rRNA gene composition microbial community could not be shown to change with treatment (P-value~0.23, Adonis test) (Figure~\ref{fig:bulk_ord}). The latter result demonstrates the substitution of $^{13}$C-labeled substrates for unlabeled equivalents could not be shown to alter the soil microbial community

84\% of xylose responders (Figure~\ref{fig:xyl_count}) and the majority of these OTUs were closely related to cultured representatives of the genus \textit{Paenibacillus} (Table~\ref{tab:xyl}, Figure~\ref{fig:tiledtree}). For example, ``"OTU.57'' ``OTU.57'' (Table~\ref{tab:xyl}), annotated as \textit{Paenibacillus}, had a strong signal of $^{13}$C-labeling at day~1 coinciding with its maximum relative abundance in non-fractionated DNA. The relative abundance of ``OTU.57'' declined until day 14 and ``OTU.57'' did not appear to be

terminally (NRI:~-1.33, NTI:~2.69). The consenTRAIT clade depth for xylose and cellulose responders was~0.012 and~0.028 SSU rRNA gene sequence dissimilarity, respectively. As reference, the average clade depth is approximately~0.017 SSU rRNA gene sequence dissimilarity for arabinase (another arabinose (arabinose like xylose is a five C sugar found in hemicellulose) utilization as inferred from genomic analyses, and was~0.013 and~0.034 SSU rRNA gene sequence dissimilarity for glucosidase and cellulase genomic potential, respectively \citep{Martiny2013,Berlemont2013}. These

author = {Lynd, Lee R. and Weimer, Paul J. and van Zyl, Willem H. and Pretorius, Isak S.}, date = {2002-09}, year = {2002}, pages = {506--577, table of contents}, {506--577}, } @article{Bradford2008,

Brian C and Sharon, Itai and Frischkorn, Kyle R and Williams, Kenneth H and Tringe, Susannah G and Banfield, Jillian F}, title = {{Community genomic analyses constrain the distribution of metabolic traits across the Chloroflexi \textit{Chloroflexi} phylum and indicate roles in sediment carbon cycling}}, journal = {Microbiome}, year = {2013},

Percentage The metabolization of $^{13}$C-xylose and $^{13}$C-cellulose is indicated by the percentage of the added $^{13}$C remaining that remains in soil over time.$^{13}$C is lost from the soil by microbial respiration.

An organic matter enrichment including C components and nutrients commonly found in plant biomass was We added a carbon mixture with inorganic salts and amino acids (not shown here) to each soil microcosms. microcosm where the only difference between treatments was the $^{13}$C-labeled isotope (in red). At days 1, 3, 7, 14, and 30 replicate microcosms were destructively harvested. Bulk harvested for downstream molecular applications. DNA from each treatment and timepoint (n = 14) was subjected to CsCl density gradient centrifugation and density gradients were fractionated (orange tubes wherein each arrow represents a fraction from the density gradient). SSU rRNA genes from each gradient fraction were PCR amplified and sequenced from gradient fractions sequenced. In addition, SSU rRNA genes were also PCR amplified and sequenced from non-fractionated DNA (representing to represent thebulk soil microbial community). community.

microbial community composition in the soil microcosms changes over time, and variance in non-fractionated DNA is smaller than variance in fractionated DNA. SSU rRNA gene sequences were determined for non-fractionated DNA from the unlabeled control, $^{13C}$C-xylose, $^{13}$C-xylose, and $^{13C}$C-cellulose $^{13}$C-cellulose treatments over time (colors indicate time, different symbols used for different treatments). Distance in SSU rRNA gene composition was quantified with the UniFrac metric. The leftmost panel indicates NMDS of data from both non-fractionated and

Change in relative abundance in non-fractionated DNA over time for xylose responders (13CXPS) and cellulose responders (13CCPS). Each panel represents a responders to the indicated substrate (i.e. cellulose (13CCPS) or xylose (13CXPS)) within the indicated phylum except for the lower right panel which shows all reponders to both xylose and celluose. The abbreviations Proteo., Verruco., and Plancto., correspond to \textit{Proteobacteria}, \textit{Verrucomicrobia}, and \textit{Planctomycetes}, respectively.

enrichment values indicate an OTU is likely $^{13}$C-labeled. Different colors represent different phyla and different panels represent different days. The final column shows the frequency distribution of LFC values in each row. Within each panel, shaded areas are used to indicateLFC plus or minus one standard deviation (dark shading) or two standard devations (light shading) about the mean of all LFC values.

NMDS analysis ordination ofSIP gradient fraction SSU rRNA gene sequence composition reveals sequence composition in gradient fractions shows that it is a function of many factors including fraction density, isotopic labeling, and time. Dissimilarity in SSU rRNA gene compositon sequence composition was profiled quantified using the weighted UniFrac metric. SSU rRNA gene sequencess were surveyed in twenty gradient fractions at each sampling point for each treatment. treatment (Figure~S1). $^{13}$C-labeling of DNA is apparent because the SSU rRNA gene sequence composition of gradient fractions from $^{13}$C and control treatments differ at high density. Each point on the NMDS plot represents one gradient fraction. SSU rRNA gene sequence composition differences between gradient fractions were quantified by the weighted Unifrac metric. The size of each point is positively correlated with density and colors indicate the treatment (A) or day (B).

based on estimated \textit{rrn} copy number (A), $\Delta\hat{BD}$ (B), and relative abundance in non-fractionated DNA (C). The estimated \textit{rrn} copy number of all responders is shown versus time (A). Kernel density histogram of $\Delta\hat{BD}$ values shows cellulose responders hadgenerally higher average $\Delta\hat{BD}$ than xylose responders indicating potentially greater $^{13}$C isotope incorporation into DNA (i.e. greater higher average atom \% $^{13}$C) $^{13}$C in OTU DNA (B). The final panel indicates the rank relative abundance of all OTUs observed in the non-fractionated DNA (C) where rank was determined at day 1 (bold line) and relative abundance for each OTU is indicated for all days by colored lines (see legend). Xylose responders (green ticks) have higher relative abundance in non-fractionated DNA than xylose responders (ticks are based on day 1 relative abundance).

relative to control (represented as LFC) for each OTU in response to both $^{13}$C-cellulose (13CCPS, leftmost heatmap) and $^{13}$C-xylose (13CXPS, rightmost heatmap) with values for different days in each heatmap column. High enrichment values (represented as LFC)in heavy density fractions provide evidence of $^{13}$C-labeled DNA.