New High Throughput Approaches for Culturing Soil Microbes from Shallow Subsurface Soils

Home , Jatrophihabitans, Thermoleophilia

Item Type text; Electronic Thesis

Authors Custer, Joy

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 05/10/2021 09:42:41

Link to Item http://hdl.handle.net/10150/641736 NEW HIGH THROUGHPUT APPROACHES FOR CULTURING SOIL MICROBES FROM SHALLOW SUBSURFACE SOILS

Joy Custer

A Thesis Submitted to the Faculty of the

DEPARTMENT OF ENVIRONMENTAL SCIENCE

In Partial Fulfillment of the Requirements

For the Degree of

MASTER OF SCIENCE

In the Graduate College

THE UNIVERSITY OF ARIZONA

2020

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Master’s Committee, we certify that we have read the thesis prepared by: Joy Custer titled:

and recommend that it be accepted as fulfilling the thesis requirement for the Master’s Degree.

______Date: ______May 6, 2020 Paul Carini

______Date: ______May 6, 2020 Malak Tfaily

______Date: ______May 6, 2020 Julia W Neilson

Final approval and acceptance of this thesis is contingent upon the candidate’s submission of the final copies of the thesis to the Graduate College.

I hereby certify that I have read this thesis prepared under my direction and recommend that it be accepted as fulfilling the Master’s requirement.

______Date: ______May 6, 2020 Paul Carini T hesis Committee chair

Environmental Science

Table of contents

List of Figures………………………………………………………………………………….4

Abstract………………………………………………………….…………………………...…5

Chapter 1: Introduction….…………………………………………………………………...6

Chapter 2: Influence of substrate concentration on culturability of heterotrophic soil microbes isolated by high-throughput dilution-to-extinction cultivation…….10

SUMMARY OF RESEARCH…………………………………………………………10

MY CONTRIBUTION………………………………………………………………….11

HOW THE RESEARCH CONTRIBUTES TO THE FIELD………………………..15

Chapter 3: Separation of viable microbes from soil matrices……………………….17

ABSTRACT…………………………………………………………………………….17

INTRODUCTION………………………………………………………………………18

METHODS…………………………..………………………………………………….21

RESULTS……………………………………………………………………………....25

DISCUSSION………………………………….……………………………………….27

Chapter 4: Conclusion………………………………………………………………………30

References………………………………………………………………………...………..…34

Appendix A: NycodenzTM Buoyant Density Cell Separation……………………………………………39 1.1 Reagents…………………………………………………………………………..39 1.2 Soil Source Information…………………………………………………………..39

Appendix B: Influence of substrate concentration on culturability of heterotrophic soil microbes isolated by high-throughput dilution-to-extinction cultivation…….40 SUPPLEMENTARY MATERIAL……………………………………………………..55

Appendix C: JGI CTAB/NaCl DNA Extraction Protocol...……………………………..61

Appendix D: Phase Lock Gel DNA Extraction Protocol……………………………….65

List of figures

Figure 2.1: Images of buoyant density centrifugation tubes of a surface soil and subsurface soil after centrifugation. …………………………………………………………13

Figure 2.2: Image of cultures subcultured from positive growth chambers into 25 mL flasks…………………………………………………………………………………………….15

Figure 3.1: Flow chart describing the high-throughput dilution-to-extinction process…18

Figure 3.2: Experimental design for the NycodenzTM buoyant density cell separation and subsequent fractionation and screening...... 21

Figure 3.3: Appearance of NycodenzTM density separation tubes before and after centrifugation.…………………………………………………………………………………..22

Figure 3.4: Distribution of NycodenzTM-separated cells from surface and subsurface soils collected at the four Catalina-Jemez Critical Zone Observatory sites.……………………………………………………………………………………...…..…25

Figure 3.5a and Figure 3.5b: Examples of cell counts across replicate buoyant density gradient separations of subsurface soil samples from Mt. Bigelow.………………….….28

Abstract

My thesis research directly addresses the need for new microbial cultivation methods for soil oligotrophs. I developed a cell separation protocol to extract viable microbes from shallow subsurface soils and used the extracted cells for high-throughput dilution- to-extinction. We reported the impacts of substrate concentration on culturability of microbes inhabiting a shallow subsurface soil in a conifer forest in Tucson, AZ.

Substrate concentration significantly influenced the culturability of Actinobacteria from these locations. Our results indicate that the dilution-to-extinction method, combined with the optimized method for separating viable cells from soil, is suitable for the isolation of oligotrophic microbes from shallow subsurface soils.

Chapter 1: Introduction

Microbes are globally abundant and ubiquitous in soils, where they play crucial roles in Earth’s biogeochemistry, soil fertility, and contaminant remediation [1–4]. For example, new sustainable agricultural practices include the use of microbes because of their ability to increase plant nutrient uptake, growth, and nitrogen fixation [1, 5].

Additionally, microbes are capable of contaminant degradation [3]. In addition to these interactions at the soil surface, microbes in the subsurface participate in terrestrial system processes such as the turnover of natural organic carbon and soil formation [2,

6]. These subsurface environments typically contain lower nitrogen, phosphorus, and organic carbon than surface soils, and they experience less fluctuation in temperature and moisture [7–12]. While previous studies of soil microbiomes have primarily focused on surface soils, the microbial taxa present in subsurface soil environments are often distinct from those at the surface and underrepresented in microbial culture and genome databases [13, 14].

Soil microbial communities are tremendously diverse with tens of thousands of unique phylotypes in a single gram of soil [15, 16]. Our ability to infer specific environmental roles for individual microbial taxa is hindered by our inability to grow and study the majority of this microbial diversity under defined laboratory conditions. A recent DNA-based molecular study showed that although soil microbial communities are diverse, only ~2% of the total microbial taxa detected in soils account for almost half of

16S RNA gene sequences recovered from soil microbial communities worldwide [16].

Of these abundant and ubiquitous microbes, ~82% have no representative genome sequence match at ≥97% 16S rRNA gene sequence identity. While 16S rRNA gene

sequence information can provide insight into microbial ecology and how microbes are distributed, it is limited in its ability to identify variations between strains and the presence of novel genes [17, 18].

An important aspect to understanding the vast soil microbial diversity is to culture and study microbes in pure laboratory culture. It is estimated that one-quarter of microbial cells on Earth belong to phyla with no cultured relatives, so cultivating these organisms could provide key insights into ecosystem functions, novel cellular metabolisms, and interpretation of metagenomic datasets [19, 20]. Genomes, transcriptomes, and proteomes of cultured cells in controlled experiments can illuminate ecologically-relevant characteristics of microbial cells [21]. These insights into cell function can generate information about the emergent principles of the cells themselves as well as the communities in which they exist. Molecular data can be used to inform which taxa should be targeted for cultivation and why they are important [21]. Many of the currently uncultivated microbes likely have novel biosynthetic pathways and unknown biochemical features [22]. Uncultured microbes are expected to offer innovative solutions for biotechnology, agriculture, and public health [23].

One possible explanation for why a large portion of this microbial biodiversity is recalcitrant to cultivation attempts is that many uncultivated lineages may be so-called

‘oligotrophs.’ Oligotrophy refers to the paradoxical physiology of microbes that grow optimally when nutrient availability is low. Oligotrophs dominate non-host-associated microbial environments globally [24]. These cells tend to be small, slow growing, and have unusual or combinatorial nutrient requirements [25–27]. In many cases, oligotrophic microbes are inhibited by high carbon content in routine microbial growth

media, possibly resulting in ‘substrate accelerated death’ [28]. Additionally, microbes may not grow on petri plates because of competition with fast-growing strains on the plate or because of the production of antibiotics by rapidly growing strains [29, 30].

Another potential explanation is that oligotrophs may rely on metabolic processes that do not adapt well when taken from a low-nutrient environment and are placed into a high-nutrient environment [30]. For example, high nitrogen and phosphorus levels shift bacterial communities to favor taxa that thrive in nutrient-rich conditions over oligotrophic taxa [31].

There are three types of oligotrophs that can be cultivated: 1) cells that grow on nutrient-poor medium but cannot be subsequently recultivated; 2) microbes that grow on nutrient-poor medium but can be subsequently recultivated on nutrient-rich medium; and 3) microbes that require special nutrient-poor medium for both isolation and recultivation [32]. Although the mechanisms that delineate these categories are not well understood, it is accepted that these groups of oligotrophs contain more similarities than merely being able to thrive under nutrient-poor conditions. For example, they tend to grow slower and to lower final cell densities than microbes cultured on higher-nutrient growth medium. Because of this, typical methods for evaluating growth such as colony enumeration, microscopy, and turbidity, are more challenging to implement while monitoring oligotrophic growth [20].

One approach to culturing abundant oligotrophic microbes is called ‘dilution-to- extinction’ culturing. The dilution-to-extinction approach is based on the growth of cells in small volumes of low-nutrient media which selects for the most abundant microbes instead of the most nutrient-tolerant [33, 34]. The growth media used in these

experiments contain substrate concentrations that are typically several orders of magnitude lower than common media formulations [34]. The combination of highly sensitive detection of microbial growth at low nutrient concentrations together with high- throughput screening techniques (including arrayed microscopy or flow cytometry) allows for quick detection of oligotrophic growth [33, 35, 36]. While dilution-to-extinction culturing has previously been used to isolate microbes from marine environments, it is more difficult to use for soil samples because soil and sediments are among the most complex of all environmental samples, often containing tens of thousands of uncultivated microbial taxa in a multidimensional-organic matrix [37]. To isolate soil oligotrophs by dilution-to-extinction, the microbial cells must first be separated from the soil particles [38]. This separation step is essential for removing cells adhered to particles made of organic matter and minerals either directly or through electrostatic interactions [39]. Additionally, the presence of soil aggregates makes it difficult to accurately quantify cells—a key component of dilution-to-extinction cultivation [40].

In this thesis, I developed cell separation protocols to detach viable cells from shallow subsurface soil samples using buoyant density centrifugation and used these separated cells to inoculate dilution-to-extinction experiments in a defined growth medium. I screened these cultures for growth using high-throughput flow cytometry. I calculated the differences in culturability across microbial growth medium formulations.

This is the first such application of high-throughput dilution-to-extinction methods for soil samples.

Chapter 2: Influence of substrate concentration on culturability of heterotrophic

soil microbes isolated by high-throughput dilution-to-extinction cultivation

SUMMARY OF RESEARCH

Here we use high-throughput dilution-to-extinction to isolate shallow subsurface microbes from a conifer forest in Arizona, USA. We hypothesized that the concentration of heterotrophic substrates in microbiological growth medium would affect which microbial taxa were culturable from these soils. To test this, we diluted extracted cells into two custom-designed and defined growth media that differed only by a 100-fold difference in the availability of amino acids and organic carbon. Across both media, we isolated a total of 133 pure cultures—all of which were classified as Actinobacteria or

Alphaproteobacteria. The substrate availability dictated which actinobacterial phylotypes were culturable but had no significant effect on the culturability of Alphaproteobacteria.

We isolated cultures that were representative of the most abundant phylotype in the soil microbial community (Bradyrhizobium spp.) and representatives of five of the top 10 most abundant Actinobacteria phylotypes, including Nocardioides spp., Mycobacterium spp., and several other phylogenetically divergent lineages. Flow cytometry of SYBR green-stained cells showed that cultures isolated on low-substrate medium had significantly lower nucleic-acid fluorescence than those isolated on high-substrate medium. These results show that the concentration of substrates in culture medium influences the culturability of specific microbial lineages and that dilution-to-extinction is an effective method to isolate abundant soil microbes.

MY CONTRIBUTION

My contributions to the paper titled “Influence of substrate concentration on the culturability of heterotrophic soil microbes isolated by high-throughput dilution-to- extinction cultivation” now published in mSphere (Bartelme R, J Custer et al., 2020;

Appendix 2), are as follows:

I separated viable cells from freshly collected soils and enumerated the separated cells on the Guava EasyCyte flow cytometer. Using these cell counts, I inoculated dilution-to-extinction experiments across different medium concentrations and pH values (pH 4, 7, and 11). After incubation, I screened the experiments for growth using high-throughput flow cytometry at 4 and 11 weeks. No wells showed growth in pH 4 media and only minimal growth was observed at pH 11, therefore our publication focused on pH 7 samples only. I calculated the percent culturability of the experiment from the number of positive wells relative to the number of inoculated wells. I contributed to data organization and manuscript preparation.

I separated viable cells from freshly collected soils and enumerated the separated cells on the Guava EasyCyte flow cytometer. Using these cell counts, I inoculated dilution-to-extinction experiments across different medium concentrations and pH values (pH 4, 7, and 11). For reference, the pH of the subsurface soil locations ranged from 6.26 (Table 1.2.2 in Appendix 1). After incubation, I screened the experiments for growth using high-throughput flow cytometry at 4 and 11 weeks. No wells showed growth in pH 4 media and only minimal growth was observed at pH 11, therefore our publication focused on pH 7 samples only. I calculated the percent culturability of the experiment from the number of positive wells relative to the number of inoculated wells.

I contributed to data organization and manuscript preparation. A brief summary of the methods I used follow. The full details can be found in the Methods section in Appendix

Cell separation: I modified the cell extraction method reported by Khalili et al., 2019 to extract viable cells [40]. The main major modification was to exclude glutaraldehyde, a fixative that kills cells, since the cells were to be used for downstream cultivation. In brief, I prepared the soil-buffer slurry (137.5 mM NaCl, 26.78 mM tetrasodium pyrophosphate, and 0.27% (v/v) Tween 80), vortexed, and carefully layered it over 80%

(w/v) NycodenzTM solution in 50 mM tetrasodium pyrophosphate (see Appendix 1.1 for

1.1). I centrifuged the tubes and extracted 3 × 0.5 mL aliquots from the resulting buoyant density cell separation. Because there is little organic matter in shallow subsurface soils, the location of cells in the buoyant density centrifugations was difficult to identify visually. To ameliorate this, we compared the appearance of the buoyant density centrifugation tubes after separation of shallow subsurface soils (with no visible organic matter) to a surface soil with visible organic matter (Figure 2.1). The soil with high organic matter content shows a delineation between the NycodenzTM and supernatant after centrifugation at ~25 mm from the bottom of the tube, which is where we obtained our samples for cultivation (Figure 2.1). After extracting cells from the buoyant density separation, I washed the extracted cells and resuspended the cell pellets in 137.5 mM NaCl.

Dilution-to-extinction: I

fixed cells extracted from the

buoyant density

centrifugation tubes with

1.75% formaldehyde and

stained them with SYBR

Green I for 3.5 h at room

temperature in the dark. I

used a Millipore Guava flow

cytometer to enumerate the

cells as described Figure 2.1: Images of buoyant density centrifugation tubes of a surface soil (left) and subsurface soil (right) after centrifugation. elsewhere [25]. I diluted the enumerated cells into Artificial Subterranean Medium (ASM) -high or ASM-low nutrient medium (Appendix 2) to a density of 5 cells mL-1 and aliquoted 1.0 ml of the dilute cell suspension into the wells of polytetrafluoroethylene 96-well microtiter plates (Cowie

Technology, New Castle, DE) so that on average each well contained 5 cells. Plates were covered with plastic lids that allow air circulation and incubated at 16ºC in the dark under aerobic conditions. I screened the dilution-to-extinction plates for growth by fixing

100 µl aliquots with 1.75% formaldehyde and staining them with a 1:4,000 dilution of commercial SYBR Green I stock for 18 h in the dark at room temperature and counting by flow cytometry (EMD-Millipore Guava EasyCyte). I screened plates for growth at 4 and 11 weeks after inoculation. Positive cultures were defined as cultures that

exceeded 1.0 × 104 cells ml-1, which is based on the sensitivity of the flow cytometer.

Actual and theoretical culturability estimates: I estimated culturability by using the equation: V = -ln(1-p)/X, where V is the estimated culturability, p is the proportion of inoculated cultivation chambers that displayed measurable growth (number of chambers positive for growth ÷ total number of chambers inoculated), and X is the number of cells added to each cultivation chamber as estimated from dilutions [35]. I estimated the number of pure cultures (û) as follows: û = -n(1-p) × ln(1-p), where n is the number of inoculated growth chambers and p is the proportion of inoculated wells displaying growth [35].

Culture transfer and storage: I contributed to subculturing growth chambers that were positive for growth into 25 ml of the respective growth medium (ASM-high or ASM- low) in acid-washed, sterile polycarbonate flasks (Figure 2.2). These cultures were incubated at 16ºC with shaking at 50 RPM. At the time of subcultivation, I assigned cultures a unique Arizona Culture Collection (AZCC) number. I contributed to monitoring the subculture flasks for growth for 2 months by counting cell densities with the Guava

EasyCyte Flow Cytometer, as described above. Subcultures displaying growth within two months were cryopreserved in 10% glycerol and stored at -80ºC. If no growth appeared within two months, the cultures were discarded and the assigned AZCC number was retired.

Figure 2.2: Image of cultures subcultured from positive growth chambers into 25 mL flasks.

Culture identification: I contributed to identifying cultures by full length 16S rRNA gene sequencing. Briefly, I filtered 5-10 ml of cell biomass from 25 mL cultures on to

0.2 µm-pore size Supor filters and extracted DNA using a Qiagen PowerSoil DNA extraction kit following the manufacturer's instructions. For these samples, I amplified full-length 16S rRNA genes from the resulting DNA using the 27F-1492R primer set

(27F, 5′-AGAGTTTGATCMTGGCTCAG-3′; 1492R, 5′-ACCTTGTTACGACTT-3′ [41]).

The reaction mix consisted of Promega’s GoTaq HotStart 2× PCR master mix with final concentrations of 0.4 μM 27F and 0.4 μM 1492R primers, and 1 μl of template DNA, in a total reaction volume of 25 μl. PCR reactions were cycled with the following program: 1

× 94°C for 10 min followed by 36 cycles of: 94°C for 45 s, 50°C for 90 s, 72°C for 90 s; and a single 72°C extension for 10 min.

HOW THE RESEARCH CONTRIBUTES TO THE FIELD

Culturing environmental microbes in the lab is crucial for understanding the factors underlying their relative abundance, distribution, and role in soil fertility and biogeochemistry. Subsurface soil environments typically contain low nutrient contents

(see Appendix 1.2 for soil source nutrient information), and microbes that thrive in these environments do not grow well under typical laboratory conditions. By using high- throughput dilution-to-extinction methods to isolate subsurface soil microbes, we show that culturability is greatly impacted by the concentration of growth substrates in the culture medium. This information can be used to design experiments that maximize the culturability of previously uncultured soil microbes.

Chapter 3: Separation of viable microbes from soil matrices

ABSTRACT

The majority of soil microbial biodiversity is uncultured in the laboratory. In part, the inability to cultivate many microbial lineages may be because these uncultured microbes are oligotrophic and do not grow on typical growth media. While there are several microbial cultivation methods designed for microbes inhabiting oligotrophic marine environments, these methods are challenging to implement in soil. This chapter describes the optimization of a NycodenzTM buoyant density centrifugation method to separate viable cells from soil that are suitable for cultivation approaches that target oligotrophic microbes. Here we show that microbial cells are present—but differentially distributed—throughout the resulting buoyant density profile. The greatest number of cells accumulated at the NycodenzTM interface, regardless of soil source. However, the number of cells at this interface was soil-specific. This suggests that buoyant density centrifugation might be useful to differentially separate for specific microbial populations on the basis of where they migrate in the buoyant density separation. To test this hypothesis in future experiments, we will quantify the types and relative abundances or microbial taxa from individual fractions to determine if different microbial taxa localize at different fractions on the density gradient after centrifugation.

INTRODUCTION

Dilution-to-extinction culturing has resulted in successful culturing of previously uncultured bacteria from marine environments, but these approaches are challenging to implement for soil microbes [34, 35, 39]. The first step in dilution-to-extinction approaches is to accurately count the microbial inoculum (Figure 3.1). Direct counts can be obtained through microscopy or flow cytometry. In contrast to aquatic systems, accurate direct counts of microbes are challenging to obtain from soil because of interferences by the soil matrix (organic and mineral constituents) [42].

Count microbial inocula with flow cytometry or microscopy

Dilute to 1-5 cells per growth chamber in miniaturized cultivation chambers (such as 96-well plate microtiter plates)

Incubate

Screen plates for growth using flow cytometry or microscopy

Cryopreserve cultures

Figure 3.1: Flow chart describing the high-throughput dilution-to-extinction process.

One way to circumvent the difficulty of obtaining direct counts from soil is to separate microbial cells from the soil particles before preparing direct counts. There are two steps in the soil microbial extraction process: dispersion and separation. The dispersion step uses physical or chemical means to dislodge any bacteria that are

encompassed by soil aggregates [39]. Physical dispersion methods include sonication, blending, grinding, or shaking to extricate cells. To increase dislodgement efficiency, sometimes chemical dispersion is used following a physical dispersion method.

Chemical dispersion involves using a chemical dispersion agent such as Chelex to dislodge cells from soil particles [39].

Once the soil matrix and cells are dispersed, they must be separated from each other. An example of a separation method is densitometry, which separates cells by their density. During density gradient centrifugation, cells migrate through a density gradient until they reach a density that matches their own. A similar approach is buoyant density centrifugation which separates cells by ‘floating’ them on the surface of a dense medium while denser soil particles pellet through the separation medium. NycodenzTM is an effective non-toxic density gradient/buoyant density centrifugation medium that can be used to separate microbial cells from soil [43–45]. NycodenzTM is less viscous than other gradient mediums such as metrizamide and sucrose solutions, and it is relatively easy to remove from samples in comparison to other gradient media [44]. However, this method results in significantly lower biodiversity of extracted cells than is present in the original soil, and taxonomic bias in which cells are preferentially extracted can occur

[46].

Here, we optimize the NycodenzTM buoyant density centrifugation method to extract viable cells from assorted soil samples. The key distinction between our method and previous studies is that we are separating viable cells instead of fixed dead cells.

Our hypothesis is that different soil microbial taxa will partition at different locations in the buoyant density cell separation, owing to subtle differences in cell density. By

deducing the specific location of individual microbial populations in the buoyant density cell separation, we can extract specific fractions that are enriched in microbial populations of interest to use for culturing. In an effort to test our hypothesis, we attempted to extract DNA from the fractions for microbial community analyses. As we describe below, the number of cells in each buoyant density centrifugation fractions did not yield enough DNA for successful PCR.

METHODS

Soil sources: Soils were collected in May of 2018 from field sites within the Santa

Catalina Critical Zone Observatory outside of Tucson, AZ, USA: Mount Bigelow,

Biosphere 2, Oracle Ridge, and Marshall Gulch (Table 1.2.1 in Appendix 1). At each site, we dug a 60 cm pit and collected horizontal soil cores into sterile plastic bags from depths of 0 cm and 50 cm. laboratory, we sieved the soils to 2 mm and stored at 4ºC until cell separation (up to 381 days). Soil characteristics were analyzed for each soil and are summarized in Tables 1.2.1, 1.2.2, and 1.2.3 in Appendix 1.

Experimental design overview: We blended the soil samples collected from 50 cm at each site in a cell extraction buffer. This soil slurry was split into three equal portions and carefully layered over NycodenzTM for buoyant density centrifugation

(Figure 3.2, number 2). Thus, three technical replicate buoyant density cell extractions were prepared from each soil sample. The replicates were then centrifuged simultaneously (Figure 3.2, number 3). Each resulting buoyant density cell separation was fractionated in 500 µL fractions (Figure 3.2, number 4). Each fraction was washed

(Figure 3.2, number 5) and counted using flow cytometry (Figure 3.2, number 6).

Figure 3.2: Experimental design for the NycodenzTM buoyant density cell separation and subsequent fractionation and screening.

NycodenzTM buoyant density centrifugation protocol: I measured buffer and

NycodenzTM amounts by mass instead of volumes to obtain more consistency across replicate samples. I prepared the soil slurry by combining 0.5 g soil in a cell extraction buffer consisting of 40 g saline solution plus 4.8 g of TSP-Tween solution. I blended the soil-buffer solution with an immersion blender for two minutes on ice. After blending, samples were allowed to settle for five minutes on ice. I layered 15 g aliquots of the soil- buffer solution over 14.3 g of the NycodenzTM solution in sterile 50 ml polycarbonate centrifuge tubes (Figure 3.3a) and centrifuged the tubes at 17,000 × g for 2 h at 16°C in a JA-17 rotor in a Beckman Coulter Avanti J-25 centrifuge. The resulting tube contained a large supernatant fraction and a pellet (Figure 3.3b).

Figure 3.3: Appearance of NycodenzTM buoyant density separation tubes before (a) and after (b) centrifugation. The supernatant contains the separated viable cells which will be used for further analysis and culturing.

Buoyant-density fractionation protocol: I removed tubes very carefully from the centrifuge to minimize disruptions to the buoyant density cell separation. I transferred

500 µL fractions of the NycodenzTM density gradient, from the top of the tube, to 1.5 mL microcentrifuge tubes containing 1 mL of the TSP-50 buffer (Appendix 1.2). A total of

48 × 500 µL fractions were collected from each buoyant density centrifugation tube. We performed three replicate buoyant density centrifugations per soil sample for a total of

144 samples per soil (48 fractions × 3 replicates = 144 samples) These fractions were briefly vortexed and centrifuged at 17,000 × g for 25 minutes. I extracted the supernatant completely (without disrupting the pellet) and resuspended the cell pellet in

800 µL of 0.9% saline solution. We further partitioned the cells resuspended in 0.9% saline for flow cytometry and microbial community analysis.

Flow cytometry of gradient fractions: We fixed 200 µL fractions with 0.9% formaldehyde and 2.4% SYBR Green I (prepared in in 0.2 µm-filtered Tris-EDTA [TE] buffer). Stained fractions were incubated overnight (~12-18 h) in the dark at room temperature. We ran the samples on a Guava EasyCyte Flow Cytometer with the following PMT voltages: Side Scatter 999V, Green Fluorescence 716V, Yellow

Fluorescence 250V, and Red Fluorescence 250V.

DNA extraction protocols: I attempted to extract DNA from individual fractions using a variety of approaches, including: i) the Joint Genome Institute (JGI) CTAB/NaCl protocol (Appendix 3); ii) DNeasy PowerSoil kit (according to the manufacturer’s instructions) ; iii) DNeasy blood and tissue kit (according to the manufacturer’s instructions); and iv) a phase lock gel protocol (Appendix 4).

Amplification of 16S rRNA gene amplicons: To profile the composition of the microbial communities in each fraction, I amplified 16S rRNA gene amplicons from the extracted DNA using the following PCR reaction mix (per reaction): 12.5 µL of GoTAQ

Green master mix, 9.5 µL nuclease-free water, 0.5 µL of each 10 mM primer 515-f primer (5’-GTGYCAGCMGCCGCGGTAA-3’) and 806-r primer (5’-

GGACTACNVGGGTWTCTAAT-3’) and 1 µL extracted DNA for use as template for the reaction [47, 48]. We included a positive control (Escherichia coli genomic DNA) and negative control (nuclease-free water), as appropriate. PCR reactions were cycled with the following program: 1 × 94°C for 3 min (the ‘hot start’ reaction), followed by 35 cycles of: 94°C for 45 s, 50°C for 60 s, 72°C for 90 s; and a single 72°C extension for 10 min

[45, 49]. After amplification, PCR products were verified on a 1% agarose gel stained with SYBR Green-I.

RESULTS

Across all soils, the distribution of cell densities in the buoyant density separation ranged from levels below our limit of quantification (≤1×104 cells mL-1) to 4.84×106 cells mL-1. Microbial cells were present throughout the NycodenzTM density gradient for all

four soil sampling locations

(Figure 3.4). The highest cell

densities appeared between

fractions 30 and 35 in all

soils (Figure 3.4). From

largest to smallest, the

number of total extractable

cells (reported as mean ±

standard deviation, n=3) in

individual subsurface soil

was Bigelow (1.41×107 ±

2.69×106)> Marshall

(1.25×107 ± 1.81×106) >

Biosphere 2 (1.19×107 ±

5.18-106) > Oracle Ridge

(3.1×106 ± 1.89×106).

Figure 3.4: Distribution of NycodenzTM-separated cells from None of the DNA subsurface soils collected at the four Catalina-Jemez Critical Zone Observatory sites. Each point is the cell density of an individual extraction protocols yielded fraction. The lines are loess-fit regressions. Cell densities that fell below the limit of detection by flow cytometry (1 × 104 cells ml-1) are enough DNA for not illustrated and were not used in analyses.

amplification. We speculate that there are several factors that contributed to this. First, I observed precipitates in the majority of the tubes, even after completion of the DNA extraction. This occurred with both the JGI CTAB/NaCl protocol (see Appendix 3) as well as the phase lock gel protocol (see Appendix 4). I suspected that these precipitates were residual salts, so I remade all of the DNA extraction reagents to eliminate the possibility that the reagents were not prepared correctly. I identified a few errors. For example, the SDS I had been using was 20% instead of 10%. After remaking all reagents, I continued to experience issues with precipitates across the protocols. Since these samples were low biomass (less than 5 ×106 cells ml-1) and in low volumes (0.5 mL), the total number of cells in a given sample never exceeded 2.5 ×106 cells total. It was difficult to find a protocol that produced DNA suitable for downstream PCR from low densities such as these. I tried many widely used DNA extraction techniques such as

DNeasy PowerSoil kit, DNeasy blood and tissue kit, and the JGI CTAB/NaCl protocol. In all cases, I was unable to amplify gene products with the PCR methods detailed in the methods section.

DISCUSSION

The development of a reproducible cell extraction method was an iterative process that required several modifications. For example, the first variation of this protocol called for an overnight shaking step of the soil slurry. However, the cell densities of individual fractions varied widely both across fractions in a single buoyant density centrifugation separation and across the three replicates for a single soil. For example, I often observed that two of the three replicates for a given soil had similar distributions of cell counts across the separation, while the third replicate produced a distinct distribution of cell counts (Figure 3.5a). To attempt to ameliorate this reproducibility issue, I abandoned the shaking step and replaced it with a blending step.

I changed this aspect of the protocol to eliminate the possibility that the incongruence between replicates was caused by differential dispersion of microbes—particularly those found in cell aggregates. That is, I wanted to maximize the dispersion of soil to release bacteria trapped within soil aggregates [39]. While the shaking step resulted in gradients with visible organic particles (small aggregates, parts of roots, etc.) throughout the aqueous layer, the blending step helped break down and disperse this organic matter.

The blending step helped improve replicability across replicates (Figure 3.5b).

Figure 3.5: Examples of cell counts across replicate buoyant density gradient separations of subsurface soil samples from Mt. Bigelow conducted with a shaking step (a) and with a blending step (b). A second optimization of the buoyant density cell separation was that I lengthened the centrifugation time from one hour to two hours in length. My rationale for this was that perhaps by allowing buoyant density gradient to centrifuge longer, the three replicates would form more similar cell separations. Additionally, in the original protocol, the 0.5 mL fractions were cleaned with TSP-50, I changed this and instead cleaned them with distilled water in the final protocol. I did this to reduce background noise during the flow cytometry steps that might be affecting the cell counts.

An alternative hypothesis as to why the cell separation was not reproducible across replicates is that the tubes were not carefully removed from the ultracentrifuge, resulting in disturbance of the cell layering prior to fractionation. Since the centrifuge rotor used is a fixed-angle rotor, the cell separation forms at an angle during centrifugation and shifts back to horizontal upon removal. I took additional care in

removing the tubes to limit the movement of the tubes themselves as well as the cell separation. To minimize any movement of the cell separation, I took the tubes out of the centrifuge one at a time and fractionated one by one. I fractionated the tubes directly adjacent to the centrifuge rather than moving them to the main lab area. These modifications minimized the effect of microturbation on the cell separation and greatly improved the reproducibility of the technique.

Most buoyant density centrifugation protocols collect the faint whitish band directly above the NycodenzTM after centrifugation [50–53]. Here we illustrate that high cell densities are found across a wide span of the NycodenzTM separation and that the number of cells extracted using in the NycodenzTM separation is soil-specific, likely due to differences in the amounts of cells present in the original soils themselves (Figure

3.4). Thus, a main finding of this work is that the distribution of cells separated by buoyant density centrifugation spans the entire separation tube. If cells of different taxa can be differentiated by their location in the cell separations, this optimized method could help researchers target specific microbial taxa for cultivation studies, potentially maximizing their chances of isolating that specific organism.

Future research is needed to determine distribution of different taxa throughout the observed density gradient. The next step of this project is to find a DNA extraction protocol that is suitable for low volume, low biomass samples. Magnetic beads are popular to use for low biomass samples, so the MAGMax Microbiome Ultra Nucleic Acid

Isolation Kit made by ThermoFisher Scientific might be an ideal option. In addition, a

Bioanalyzer can be used to verify the presence of a faint PCR product, as suggested by the Earth Microbiome Project [47, 48]. Nested PCRs are also possibilities to confirm the

presence of dilute PCR products that are not visible on a gel. Once a successful DNA extraction protocol has been determined, 16S rRNA gene amplicon sequencing and

PCR can be conducted to identify the microbial species present in each fraction.

Additionally, community analysis for fractions from all four sites will be conducted to compare types and relative abundances of microbial taxa extracted across soils and how they are distributed in the cell separations.

CHAPTER 4: CONCLUSION

Little is known about the characteristics and functions of the uncultured microbial majority [54]. This thesis outlines two approaches for separating viable cells from soils and new methods for cultivating subsurface soil microbes, with a focus on microbes that live an oligotrophic lifestyle. First, we sought to answer if substrate concentration in growth medium affects the culturability of microbes. We found that substrate concentration did affect the culturability of Actinobacteria but not Alphaproteobacteria

[14]. Second, I investigated whether microbial taxa were differentially distributed across fractions of buoyant density centrifugation cell separations. I found that cells were distributed throughout the aqueous phase after centrifugation, with elevated concentrations just above the NycodenzTM layer. However, attempts to extract DNA were ineffective, so I was unable to determine which taxa were present in each fraction.

By increasing the culturability of soil microbes under laboratory conditions and determining the location of those microbes throughout the buoyant density cell separation, scientists can learn more about how these organisms interact with each other and how they function in the environment.

Most subsurface soils are highly oligotrophic environments, thus the microbes that inhabit them are adapted to low nutrient concentrations. Laboratory growth media commonly contain high nutrient concentrations. For reasons that are unknown, these high nutrient conditions can inhibit the growth of oligotrophic microbes from oligotrophic environments [25, 28, 53, 55–57]. One way to circumvent some of the challenges associated with culturing oligotrophs is to use alternative approaches that combines extended cultivation times with sensitive detection of microbes. An example of such a

method, dilution-to-extinction, can be used to isolate microbes using low-nutrient media

[14, 34–38]. In many cases, dilute growth media results in higher culturability of novel microbes than substrate-rich growth media [52, 53, 56, 57]. Once cultivated, the microbes can be screened for growth using techniques such as flow cytometry that are suitable for low biomass samples.

This thesis used high-throughput dilution-to-extinction techniques previously used with marine microbes to cultivate shallow subsurface microbes. Since soil samples have organic matter and minerals that are not present in marine samples, it is necessary to include a cell separation step. Many established protocols that use

NycodenzTM buoyant density cell separation only use 1 mL directly above the

NycodenzTM after centrifugation [45, 49–51], but here we show that cells are found throughout the separation tube. Because of this, current research methods could be excluding important taxa or large amounts of cells present in a given soil sample. Once I optimized the NycodenzTM buoyant density protocol for soil samples, we conducted an experiment to determine if substrate concentration has an effect on culturability. Our results show that substrate concentration is an integral part in making some cells grow under laboratory conditions. This new knowledge about the effect of substrate concentration on culturability as well as the optimized NycodenzTM buoyant density cell separation method can be used to enrich for novel bacterial taxa.

The optimized methods described in this thesis could potentially increase the ability to isolate and culture previously uncultivated bacteria such as the candidate phyla radiation. The candidate phyla radiation is a recently announced bacterial expansion of the tree of life. “Candidate phyla” describes that this group lacks cultured

representatives [58]. In other words, the candidate phyla microorganisms lack cultured representatives. This new branch represents more than 15% of all known bacterial diversity and is hypothesized to contain more than 70 different novel bacterial phyla

[59]. Culturing the candidate phyla would result in valuable information regarding the tree of life [58–61]. The utilization of the optimized NycodenzTM buoyant density cell separation protocol could aid in cultivating these heretofore uncultivated and elusive microorganisms.

We show that dilution-to-extinction techniques can be used to cultivate abundant oligotrophic soil microbes. These results are similar to those observed for oligotrophic microbes in other environments [33, 36, 55, 62, 63]. Since the mechanism of substrate- induced growth inhibition is largely unknown, future research is needed to illuminate the molecular mechanisms of this inhibition in order to determine new cultivation approaches for isolating oligotrophic soil microbes. In addition, we show that there are cells distributed throughout the buoyant density cell separation after centrifugation, indicating that current methods might be enriching for certain microbial populations.

Additionally, further research must be conducted to determine if taxa are distributed differentially throughout the NycodenzTM gradient. If taxa are indeed differentially distributed, researchers could use this optimized cell separation method to enrich for certain organisms of interest.

REFERENCES 1. Chaparro JM, Sheflin AM, Manter DK, Vivanco JM. Manipulating the soil microbiome to increase soil health and plant fertility. Biol Fertil Soils 2012; 48: 489–499. 2. Long PE, Williams KH, Hubbard SS, Banfield JF. Microbial Metagenomics Reveals Climate-Relevant Subsurface Biogeochemical Processes. Trends Microbiol 2016; 24: 600–610. 3. Fernández-Luqueño F, Valenzuela-Encinas C, Marsch R, Martínez-Suárez C, Vázquez-Núñez E, Dendooven L. Microbial communities to mitigate contamination of PAHs in soil—possibilities and challenges: a review. Environ Sci Pollut Res 2010; 18: 12–30. 4. Tiedje JM, Asuming-Brempong S, Nüsslein K, Marsh TL, Flynn SJ. Opening the black box of soil microbial diversity. Appl Soil Ecol 1999; 13: 109–122. 5. Sameshima-Saito R, Chiba K, Hirayama J, Itakura M, Mitsui H, Eda S, et al. Symbiotic Bradyrhizobium japonicum Reduces N2O Surrounding the Soybean Root System via Nitrous Oxide Reductase. Appl Environ Microbiol 2006; 72: 2526–2532. 6. Schulz S, Brankatschk R, Dümig A, Kögel-Knabner I, Schloter M, Zeyer J. The role of microorganisms at different stages of ecosystem development for soil formation. Biogeosciences 2013; 10: 3983–3996. 7. Jobbagy E, Jackson R. The Vertical Distribution of Soil Organic Carbon and Its Relation to Climate and Vegetation. Ecol Appl 2000; 10: 423–436. 8. Fierer N, Schimel J, Holden P. Variations in microbial community composition through two soil depth profiles. Soil Biol Biochem 2003; 35: 167–176. 9. Ajwa H, Rice C, Sotomayor D. Carbon and Nitrogen Mineralization in Tallgrass Prairie and Agricultural Soil Profiles. Soil Sci Soc Am J 1998; 62: 942–951. 10. Jobbagy E, Jackson R. The Distribution of Soil Nutrients with Depth: Global Patterns and the Imprint of Plants. Biogeochemistry 2001; 53: 51–77. 11. Balesdent J, Basile-Doelsch I, Chadoeuf J, Cornu S, Derrien D, Fekiacova Z, et al. Atmosphere–soil carbon transfer as a function of soil depth. Nature 2018; 559: 599– 602. 12. Hillel D. Fundamentals of Soil Physics. 2013. 13. Brewer TE, Aronson EL, Arogyaswamy K, Billings SA, Botthoff JK, Campbell AN, et al. Ecological and Genomic Attributes of Novel Bacterial Taxa That Thrive in Subsurface Soil Horizons. mBio 2019; 10: e01318-19. 14. Bartelme RP, Custer JM, Dupont CL, Espinoza JL, Torralba M, Khalili B, et al. Influence of Substrate Concentration on the Culturability of Heterotrophic Soil Microbes

Isolated by High-Throughput Dilution-to-Extinction Cultivation. mSphere 2020; 5: e00024-20. 15. Bardgett RD, van der Putten WH. Belowground biodiversity and ecosystem functioning. Nature 2014; 515: 505–511. 16. Delgado-Baquerizo M, Oliverio AM, Brewer TE, Benavent-González A, Eldridge DJ, Bardgett RD, et al. A global atlas of the dominant bacteria found in soil. Science 2018; 359: 320–325. 17. Lasken RS, McLean JS. Recent advances in genomic DNA sequencing of microbial species from single cells. Nat Rev Genet 2014; 15: 577–584. 18. Steen AD, Crits-Christoph A, Carini P, DeAngelis KM, Fierer N, Lloyd KG, et al. High proportions of bacteria and archaea across most biomes remain uncultured. ISME J 2019; 13: 3126–3130. 19. Lloyd KG, Steen AD, Ladau J, Yin J, Crosby L. Phylogenetically Novel Uncultured Microbial Cells Dominate Earth Microbiomes. mSystems 2018; 3: e00055- 18. 20. Giovannoni S, Stingl U. The importance of culturing bacterioplankton in the ‘omics’ age. Nat Rev Microbiol 2007; 5: 820–826. 21. Carini P. A “Cultural” Renaissance: Genomics Breathes New Life into an Old Craft. mSystems 2019; 4: 1–5. 22. Overmann J, Abt B, Sikorski J. Present and Future of Culturing Bacteria. Annu Rev Microbiol 2017; 71: 711–730. 23. Thrash JC. Culturing the Uncultured: Risk versus Reward. mSystems 2019; 4: e00130-19. 24. Gao Y, Wu M. Free-living Bacterial Communities Are Mostly Dominated by Oligotrophs. bioRxiv 2018. 25. Carini P, Steindler L, Beszteri S, Giovannoni SJ. Nutrient requirements for growth of the extreme oligotroph ‘Candidatus Pelagibacter ubique’ HTCC1062 on a defined medium. ISME J 2013; 7: 592–602. 26. Poindexter JS. Oligotrophy: Fast and Famine Existence. Advances in Microbial Ecology. 1981. Plenum Press, pp 63–89. 27. Koch AL. Oligotrophs versus copiotrophs. BioEssays. 2001. pp 657–661. 28. Strange RE, Dark FA. ‘Substrate-Accelerated Death’ of Aerobacter aerogenes. Microbiol Soc 1965; 39: 215–228.

29. Staley J, Konopka A. Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats. Annual Review of Microbiology. 1985. pp 321–346. 30. Morita RY. Bacteria in Oligotrophic Environments. 1997. 31. Leff JW, Jones SE, Prober SM, Barberan A, Borer ET, Firn JL, et al. Consistent responses of soil microbial communities to elevated nutrient inputs in grasslands across the globe. Proc Natl Acad Sci U S A 2015; 112: 10967–10972. 32. Kuznetsov SI, Dubinina GA, Lapteva NA. BIOLOGY OF OLIGOTROPHIC BACTERIA. Annu Rev Microbiol 1979; 33: 377–387. 33. Stingl U, Tripp HJ, Giovannoni SJ. Improvements of high-throughput culturing yielded novel SAR11 strains and other abundant marine bacteria from the Oregon coast and the Bermuda Atlantic Time Series study site. ISME J 2007; 1: 361–371. 34. Schut F, de Vries EJ, Gottschal JC, Robertson BR, Harder W, Prins RA, et al. Isolation of Typical Marine Bacteria by Dilution Culture: Growth, Maintenance, and Characteristics of Isolates under Laboratory Conditions. Appl Environ Microbiol 1993; 59: 2150–2160. 35. Button DK, Schutl F, Quang P, Martin R, Robertson BR. Viability and Isolation of Marine Bacteria by Dilution Culture: Theory, Procedures, and Initial Results. Appl Environ Microbiol 1993; 59: 881–891. 36. Connon SA, Giovannoni SJ. High-throughput methods for culturing microorganisms in very-low-nutrient media yield diverse new marine isolates. Appl Environ Microbiol 2002; 68: 3878–3885. 37. Gomez EV, Garland JL, Roberts MS. Microbial structural diversity estimated by dilution–extinction of phenotypic traits and T-RFLP analysis along a land-use intensiﬁcation gradient. FEMS Microbiol Ecol 2004; 49: 253–259. 38. Aakra A, Utaker JB, Nes IF, Bakken LR. An evaluated improvement of the extinction dilution method for isolation of ammonia-oxidizing bacteria. J Microbiol Methods 1999; 39: 23–31. 39. Lindahl V, Bakken LR. Evaluation of methods for extraction of bacteria from soil. FEMS Microbiol Ecol 1995; 16: 135–142. 40. Khalili B, Weihe C, Kimball S, Schmidt KT, Martiny JBH. Optimization of a Method To Quantify Soil Bacterial Abundance by Flow Cytometry. mSphere 2019; 4: e00435-19. 41. Lane D. 16S/23S rRNA sequencing. Nucleic acid techniques in bacterial systematics. 1991. John Wiley & Sons, pp 115–175. 42. Loutit M. Problems in Studying Microbial Ecology. N Z Ecol Soc 1972; 19: 43–45.

43. Portillo MC, Leff JW, Lauber CL, Fierer N. Cell Size Distributions of Soil Bacterial and Archaeal Taxa. Appl Env Microbiol 2013; 79: 7610–7617. 44. Rickwood D, Ford T, Graham J. Nycodenz: A new nonionic iodinated gradient medium - ScienceDirect. Anal Biochem 1982; 123: 23–31. 45. Poté J, Bravo AG, Mavingui P, Ariztegui D, Wildi W. Evaluation of quantitative recovery of bacterial cells and DNA from different lake sediments by Nycodenz density gradient centrifugation. Ecol Indic 2010; 10: 234–240. 46. Holmsgaard PN, Norman A, Hede SC, Poulsen PHB, Al-Soud WA, Hansen LH, et al. Bias in bacterial diversity as a result of Nycodenz extraction from bulk soil - ScienceDirect. Soil Biology and Biochemistry. 47. Caporaso G, Ackermann G, Apprill A, Bauer M, Berg-Lyons D, Betley J, et al. EMP 16S Illumina Amplicon Protocol. 2018. 48. Ul-Hasan S, Bowers R, Figueroa-Montiel A, Licea-Navarro A, Beman J, Woyke T, et al. Community ecology across bacteria, archaea and microbial eukaryotes in the sediment and seawater of coastal Puerto Nuevo, Baja California. PLoS ONE 2019; 14. 49. Maron P-A, Schimann H, Ranjard L, Brothier E, Domenach A-M, Lensi R, et al. Evaluation of quantitative and qualitative recovery of bacterial communities from different soil types by density gradient centrifugation. Eur J Soil Biol 2006; 42: 65–73. 50. Courtois S, Frostegard A, Goransson P, Depret G, Jeannin P, Simonet P. Quantification of bacterial subgroups in soil: comparison of DNA extracted directly from soil or from cells previously released by density gradient centrifugation. Environ Microbiol 2001; 3: 431–439. 51. Liu J, Li J, Feng L, Cao H, Cui Z. An Improved Method for Extracting Bacteria from Soil for High Molecular Weight DNA Recovery and BAC Library Construction. J Microbiol 2010; 48: 728–733. 52. Davis KER, Joseph SJ, Janssen PH. Effects of Growth Medium, Inoculum Size, and Incubation Time on Culturability and Isolation of Soil Bacteria. Appl Environ Microbiol 2005; 71: 826. 53. Joseph SJ, Hugenholtz P, Sangwan P, Osborne CA, Janssen PH. Laboratory Cultivation of Widespread and Previously Uncultured Soil Bacteria. Appl Environ Microbiol 2003; 69: 7210. 54. Rappé MS, Giovannoni SJ. The Uncultured Microbial Majority. Annu Rev Microbiol 2003; 57: 369–394. 55. Rappe MS, Connon SA, Vergin KL, Giovannoni SJ. Cultivation of the ubiquitous SAR11 marine bacterioplankton clade. Nat Lond 2002; 418: 630–3.

56. Sait M, Hugenholtz P, Janssen P. Cultivation of globally distributed soil bacteria from phylogenetic lineages previously only detected in cultivation-independent surveys. Environ Microbiol 2002; 4: 654–666. 57. Janssen P, Yates P, Grinton B, Taylor P, Sait M. Improved culturability of soil bacteria and isolation in pure culture of novel members of the divisions Acidobacteria, Actinobacteria, Proteobacteria, and Verrucomicrobia. Appl Environ Microbiol 2002; 68: 2391–2396. 58. Brown CT, Hug LA, Thomas BC, Sharon I, Castelle CJ, Singh A, et al. Unusual biology across a group comprising more than 15% of domain Bacteria. Nature 2015; 523: 208–211. 59. Danczak RE, Johnston MD, Kenah C, Slattery M, Wrighton KC, Wilkins MJ. Members of the Candidate Phyla Radiation are functionally differentiated by carbon- and nitrogen-cycling capabilities. Microbiome 2017; 5. 60. Castelle CJ, Banfield JF. Major New Microbial Groups Expand Diversity and Alter our Understanding of the Tree of Life. Cell 2018; 172: 1181–1197. 61. Hug LA, Baker BJ, Anantharaman K, Brown CT, Probst AJ, Castelle CJ, et al. A new view of the tree of life. Nat Microbiol 2016; 1: 16048. 62. Kim S, Kang I, Seo J-H, Cho J-C. Culturing the ubiquitous freshwater actinobacterial acI lineage by supplying a biochemical ‘helper’ catalase. ISME J 2019; 13: 2252–2263. 63. Salcher MM, Schaefle D, Kaspar M, Neuenschwander SM, Ghai R. Evolution in action: habitat transition from sediment to the pelagial leads to genome streamlining in Methylophilaceae. ISME J 2019; 13: 2764–2777.

APPENDIX A: NycodenzTM Buoyant Density Cell Separation

1.1 Reagents

1) TSP-tween: 250 mM tetrasodium pyrophosphate (TSP) and 1.3% Tween 80

2) TSP-50: 50 mM TSP, filter sterilized through 0.2 µm filter

3) NycodenzTM: 0.974 mM NycodenzTM in TSP-50

4) Sterile saline: 0.154 mM NaCl in water, filter sterilized through 0.2 µm filter

5) Tris-EDTA buffer (TE): Add 1 mL of 1M Tris-Cl and 0.2 mL 0.5M EDTA to 98.8

mL of distilled water. Filter sterilize using 0.2-µm-filtered.

1.2 Soil Source Information

Table 1.2.1: Soil source general information Site Latitude/Longitude Depth Date Collected Fractionated Mt. Bigelow 32.42 °N, -110.72 °W 50 cm 5/5/2018 5/21/2019 Biosphere 2 32.33 °N, -110.50 °W 50 cm 5/23/2018 5/28/2019 Oracle Ridge 32.27 °N, -110.44 °W 50 cm 5/24/2018 5/23/2019 Marshall Gulch 32.25 °N, -110.46 °W 50 cm 5/25/2018 5/30/2019

Table 1.2.2: Soil source properties Site Moisture content (%) pH Elec. Cond. % Organic Matter (mmhos/cm) Mt. Bigelow 14.94 6.3 0.02 1.0 Biosphere 2 4.17 7.4 0.05 1.4 Oracle Ridge 4.17 6.26 0.02 0.5 Marshall Gulch 11.11 5.88 0.03 1.8

Table 1.2.3: Soil source nutrient content (in ppm) Site NO3-N P K Zn Fe Mn Cu Mt Bigelow 0.7 2.17 32.8 0.05 9.26 0.14 0.87 Biosphere 2 0.8 5.39 13 0.02 2.05 0.15 0.6 Oracle Ridge 0.6 3.35 11.8 0.01 10.9 0.28 0.28 Marshall Gulch 0.6 2.06 18.1 0.06 28.1 0.29 0.92

APPENDIX B 41 RESEARCH ARTICLE Applied and Environmental Science

Inﬂuence of Substrate Concentration on the Culturability of Heterotrophic Soil Microbes Isolated by High-Throughput Dilution-to-Extinction Cultivation Downloaded from Ryan P. Bartelme,a Joy M. Custer,a Christopher L. Dupont,b Josh L. Espinoza,b Manolito Torralba,b Banafshe Khalili,c Paul Carinia aDepartment of Environmental Science, University of Arizona, Tucson, Arizona, USA bDepartment of Environment and Sustainability, J. Craig Venter Institute, La Jolla, California, USA cDepartment of Ecology and Evolutionary Biology, University of California, Irvine, California, USA

ABSTRACT The vast majority of microbes inhabiting oligotrophic shallow sub- http://msphere.asm.org/ surface soil environments have not been isolated or studied under controlled laboratory conditions. In part, the challenges associated with isolating shallow subsurface microbes may persist because microbes in deeper soils are adapted to low nutrient availability or quality. Here, we use high-throughput dilution-to- extinction culturing to isolate shallow subsurface microbes from a conifer forest in Arizona, USA. We hypothesized that the concentration of heterotrophic substrates in microbiological growth medium would affect which microbial taxa were culturable from these soils. To test this, we diluted cells extracted from soil into one of two custom-designed defined growth media that differed by 100- fold in the concentration of amino acids and organic carbon. Across the two media, we isolated a total of 133 pure cultures, all of which were classified as Acti- on March 15, 2020 by guest nobacteria or Alphaproteobacteria. The substrate availability dictated which actinobacterial phylotypes were culturable but had no significant effect on the culturability of Alphaproteobacteria. We isolated cultures that were representative of the most abundant phylotype in the soil microbial community (Bradyrhizobium spp.) and representatives of five of the top 10 most abundant Actinobacteria phylotypes, including Nocardioides spp., Mycobacterium spp., and several other phylogenetically divergent lineages. Flow cytometry of nucleic acid-stained cells Citation Bartelme RP, Custer JM, Dupont CL, showed that cultures isolated on low-substrate medium had significantly lower Espinoza JL, Torralba M, Khalili B, Carini P. 2020. nucleic acid fluorescence than those isolated on high-substrate medium. These Influence of substrate concentration on the culturability of heterotrophic soil microbes results show that dilution-to-extinction is an effective method to isolate abun- isolated by high-throughput dilution-to- dant soil microbes and that the concentration of substrates in culture medium extinction cultivation. mSphere 5:e00024-20. influences the culturability of specific microbial lineages. https://doi.org/10.1128/mSphere.00024-20. Editor Susannah Green Tringe, U.S. IMPORTANCE Isolating environmental microbes and studying their physiology Department of Energy Joint Genome Institute under controlled conditions are essential aspects of understanding their ecology. Copyright © 2020 Bartelme et al. This is an Subsurface ecosystems are typically nutrient-poor environments that harbor di- open-access article distributed under the terms of the Creative Commons Attribution 4.0 verse microbial communities—the majority of which are thus far uncultured. In International license. this study, we use modified high-throughput cultivation methods to isolate sub- Address correspondence to Paul Carini, surface soil microbes. We show that a component of whether a microbe is cul- [email protected]. turable from subsurface soils is the concentration of growth substrates in the The influence of substrate concentration culture medium. Our results offer new insight into technical approaches and on the culturability of heterotrophic soil microbes isolated by high-throughput growth medium design that can be used to access the uncultured diversity of dilution-to-extinction cultivation. @Paul_Carini soil microbes. @MicrobialBart Received 10 January 2020 Accepted 14 January 2020 KEYWORDS genome streamlining, microbial cultivation, oligotrophy, soil microbial Published 29 January 2020 ecology

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 1 42 Bartelme et al.

oil microbial communities are tremendously diverse and mediate crucial aspects of Splant fertility, biogeochemistry, pollutant mitigation, and carbon sequestration (1–4). While the diversity and community composition of surface soils have been relatively well described, we know far less about the microbes inhabiting deeper soils (defined here as Ͼ10 cm below the surface), despite their key roles in soil formation and mineralization of key plant nutrients. In contrast to surface soils that are typically rich in plant-derived compounds, subsurface soils are often characterized by smaller amounts of mineralizable nitrogen, phosphorus, and organic carbon—much of which has a long residence time and is relatively recalcitrant to microbial degradation (5–9). The temperature and soil moisture of subsurface soils are also less variable than those of shallower soils that are exposed to seasonal changes in temperature and precipita- Downloaded from tion (10). These relatively stable and low-nutrient conditions found at depth constrain both the amount of microbial biomass present in the subsurface and the structure of these microbial communities (11–14). Many of the microbial taxa that are abundant in these subsurface environments are underrepresented in microbial culture and genome databases (11). Thus, there are large knowledge gaps in our understanding of the biology of a major fraction of subsurface soil microbes. Because subsurface soils are low-nutrient habitats, part of the challenge associated with culturing and studying the microbes that live belowground may be that they require low nutrient concentrations in order to be isolated or propagated in the http://msphere.asm.org/ laboratory (15). These microbes—often referred to as “oligotrophs”—are capable of growing under conditions where the supply or quality of nutrition is poor. Although oligotrophs dominate most free-living microbial ecosystems (16), the concept of oligotrophy itself is enigmatic. There is no coherent definition of what constitutes oligotrophic metabolism aside from their ability to grow at “low” nutrient concentrations—a definition that itself is arbitrary (17, 18). Kuznetsov et al. (19) identified three groups of cultivatable oligotrophs: (i) microbes that can be isolated on nutrient-poor medium but cannot be subsequently propagated, (ii) microbes that can be isolated on nutrient-poor medium but can be subsequently propagated on nutrient-rich medium, and (iii)

microbes that require special nutrient-poor medium for both isolation and propaga- on March 15, 2020 by guest tion. Although the molecular and genetic mechanisms that distinguish these three categories are poorly understood, several traits of oligotrophs have emerged from the study of microbes that numerically dominate oligotrophic ecosystems. For example, oligotrophs are typically small, slowly growing cells (20–23). The genome sizes of numerous lineages of microbes that dominate oligotrophic marine ecosystems tend to be highly reduced—an indication that microbial oligotrophy may be tied to reduction of genome size (24–26). These “streamlined” genomes often code for fewer copies of the rRNA gene operon and transcriptional regulator genes than microbes with larger genomes, suggesting that oligotrophs lack the ability to sense and rapidly respond to variable environmental conditions (12, 16, 92). Instead, genomic inventories of marine oligotrophs suggest a reliance on broad-specificity, high-affinity transporters that are constitutively expressed (22, 26–28). While the activities of abundant and ubiquitous microbes that inhabit oligotrophic marine environments have been extensively investigated in recent years (24, 29, 30), far fewer studies have focused on the activities of microbes that dominate oligotrophic soil environments. Several soil studies used low-throughput techniques to show that reduced-nutrient solid media facilitated the isolation of important soil microbes that were previously uncultured (31–34). While several agar-based high-throughput approaches have been used to isolate diverse microbes (35, 36), these approaches may not be appropriate to isolate microbes that thrive at micromolar amounts of growth substrate and do not form detectable colonies on solid media. Here, we adapt existing high-throughput dilution-to-extinction protocols, originally developed for isolating abundant aquatic oligotrophic bacteria, to facilitate the isolation of soil microbes. We hypothesized that the concentration of heterotrophic growth substrates in a growth medium would constrain which taxa were able to be isolated on a custom-designed defined artificial medium. We tested this by extracting cells from oligotrophic subsur-

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 2 43 Dilution-to-Extinction Cultivation from Soil face soils using Nycodenz buoyant density centrifugation (37) and inoculating high- throughput dilution-to-extinction experiments in two defined media that contained a 100-fold difference in the amounts of heterotrophic growth substrates. We isolated several bacteria that were representative of abundant phylotypes found in the original soil microbial community and two lineages representative of uncultured groups of microbes. In these experiments, the substrate concentration significantly influenced which actinobacterial genera were culturable in the laboratory but had no effect on which alphaproteobacterial lineages were culturable. Moreover, we show that cells isolated on low-nutrient medium had significantly lower SYBR green I nucleic acid fluorescence, suggesting that microbes isolated on low-nutrient medium may contain reduced nucleic acid content relative to those isolated on higher-nutrient medium. Downloaded from

RESULTS We collected shallow subsurface soil (55 cm) from the Oracle Ridge field site in a midelevation conifer forest that is part of the Santa Catalina Mountains Critical Zone Observatory in Arizona, USA. These soils contained very small amounts of total organic carbon (0.095%) and N-NO3 (0.3 ppm), indicating they were highly oligotrophic (see Fig. S1 in the supplemental material). We adapted existing high-throughput dilution- to-extinction approaches designed for aquatic microbes (38, 39) to culture soil microbes from these samples (Fig. 1). The primary modification to existing protocols was http://msphere.asm.org/ to add a buoyant density centrifugation cell separation step to detach inoculum cells from mineral soils prior to diluting cells into growth medium. To do this, we vortexed soil in a cell extraction buffer containing a nonionic surfactant and a dispersing agent. We layered this soil-buffer slurry over 80% Nycodenz and centrifuged it. During centrifugation, the mineral components of soil migrated through the Nycodenz, while cells “floated” on the surface of the Nycodenz. We extracted cells located at the Nycodenz interface, stained them with SYBR green I, and counted them on a flow cytometer. This extraction yielded 1.28 ϫ 105 cells mlϪ1 from 0.5 g wet soil. We diluted the extracted cells to an average of 5 cells wellϪ1 in deep-well polytetrafluoroethylene

96-well plates containing a custom-designed and defined growth medium that we on March 15, 2020 by guest named artificial subterranean medium (ASM), with low or high concentrations of heterotrophic growth substrates (ASM-low and ASM-high, respectively) (Fig. 1). The ASM-low and ASM-high media contained identical inorganic mineral and vitamin amendments but a 100-fold difference in the concentration of organic carbon and amino acids (Table S1). We designed these media to facilitate the growth of diverse chemoheterotrophic microbes by including an array of simple carbon compounds, polymeric carbon substrates, and individual amino acids (Table S1). We prepared triplicate 96-well plates for each growth medium formulation. These dilution-to- extinction experiments were screened for growth with flow cytometry after 4 weeks of incubation and again after 11 weeks of incubation (Fig. 1). Wells displaying growth (defined as those wells displaying 1.0 ϫ 104 cells mlϪ1) were subcultured into larger volumes and subsequently cryopreserved and identified by 16S rRNA gene sequencing (Fig. 1). Across the two medium types, a total of 214 wells (119 for ASM-low and 95 for ASM-high) displayed growth after 11 weeks of incubation. We successfully propagated 182 (85%) of the cultures from microtiter plates to polycarbonate flasks containing fresh medium. Of the cultures that successfully propagated, we confirmed that 73% (133 cultures) were pure cultures by amplifying and sequencing full-length 16S rRNA gene sequences from genomic DNA extractions. The remaining 49 cultures were mixed (forward and reverse 16S rRNA sequence reads did not assemble due to base ambi- guities) or, in rare instances, did not amplify under several amplification conditions. We defined microbial culturability using Button’s definition of microbial “viability” as determined in dilution-to-extinction experiments (40). Here, “culturability” is defined as the ratio of cells that grew into detectable cultures to the total number of cells initially diluted into a cultivation chamber (40). The culturability metric described here is informative to evaluate the suitability of a growth medium to isolate microbes and can

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 3 44 Bartelme et al. Downloaded from http://msphere.asm.org/ on March 15, 2020 by guest

FIG 1 Dilution-to-extinction workflow. Soils were collected and brought to the lab, where they were homogenized in cell extraction buffer, layered over a Nycodenz solution, and centrifuged (A). The cell layer was extracted from the Nycodenz solution and counted with flow cytometry (B). Counted cells were diluted into growth medium in 96-well microtiter plates to an average density of 5 cells wellϪ1 (C). After incubation, the 96-well microtiter plates were screened for growth with flow cytometry, and wells displaying growth were subcultured into larger volumes (D). After incubating the subcultures, flasks displaying growth were identified by 16S rRNA gene sequencing and molecular phylogeny (E). Aliquots of these identified subcultures were cryopreserved at Ϫ80°C (F). be applied across medium formulations and experiments so that different experiments can be directly compared. By the end of the experiment, we approached ϳ20% culturability across the two medium formulations (Fig. 2). In general, microbial culturability was higher for ASM-low than for ASM-high, but this effect was significant only after 4 weeks of incubation (Fig. 2; Wilcoxon rank sum test, P Յ 0.05 at 4 weeks). We assigned taxonomy to each 16S rRNA gene sequence using the SILVA database. All pure cultures isolated on ASM-low and ASM-high belonged to one of two bacterial phyla: Actinobacteria (110 cultures; 83% of the pure cultures) or Proteobacteria (23 cultures, all Alphaproteobacteria; 17% of the pure cultures) (Fig. 3). Across all experiments, the genera assigned to bacteria isolated on ASM-low

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 4 45 Dilution-to-Extinction Cultivation from Soil

10 ** 8

6 Downloaded from percent culturability (v) 4

0 http://msphere.asm.org/

ASM-low ASM-high ASM-low ASM-high 4 weeks 11 weeks

FIG 2 Microbial culturability (v) was greater on ASM-low than on ASM-high. Bar heights are the mean percent culturability Ϯ standard deviation in 96-well microtiter plates (n ϭ 3) as calculated from the initial cell inoculum and the proportion of wells positive for growth (40). Double asterisks indicate Wilcoxon rank sum test P values of Յ0.05. were signiﬁcantly distinct from those isolated on ASM-high (Kruskal-Wallis rank sum, ␹2 ϭ 19.05, P ϭ 1.28 ϫ 10Ϫ5). However, these differences were largely driven by

significant differences in culturability across medium types for the Actinobacteria but on March 15, 2020 by guest not for the Alphaproteobacteria (Dunn test, P Յ 0.000 for Actinobacteria and P ϭ 0.128 for Alphaproteobacteria). Just over half of the Alphaproteobacteria (57%) were isolated on ASM-low medium, and the remaining 43% were isolated on ASM-high (Fig. 3 and Fig. S2). Cultures that were classified as Bradyrhizobium spp. were the most frequent alphaproteobacterial isolates (13 isolates), seven of which were isolated on ASM-low medium. Cultures classified as Reyranella spp. and Nordella spp. were also isolated on both ASM-high and ASM-low medium. Of the remaining five Proteobacteria cultures, three were isolated on ASM-low (Afipia [1 culture], Rhizobiales [1 culture], and Bauldia [1 culture]), and two were isolated on ASM-high (Pseudolabrys [1 culture] and a Xanthobacteraceae sp. [1 culture]). The actinobacterial cultures belonged to three classes: Actinobacteria (107 cultures), Thermoleophilia (2 cultures), and Acidimicrobiia (1 culture). Of these Actino- bacteria, 65 (59%) were isolated on ASM-low, and 45 (41%) were isolated on ASM-high. The cultures were numerically dominated by two genera that were differentially isolated on ASM-low and ASM-high: Nocardioides and Mycobacterium. Nocardioides spp. (46 cultures) were exclusively isolated on ASM-low medium (Fig. 3 and Fig. S3). Other cultures that were isolated on ASM-low included those classified as Arthrobacter (3 cultures), Marmoricola (2 cultures), Nakamurella (2 cultures), Aeromicrobium (1 culture), Blastococcus (1 culture), and Patulibacter (1 culture) (Fig. 3 and Fig. S3). While the majority of cultures classified as Mycobacterium sp. were isolated on ASM-high (38 cultures), we isolated seven mycobacterial cultures on ASM-low medium—five of which form a phylogenetically distinct cluster from those isolated on ASM-high (Fig. 3 and Fig. S3). Other actinobacterial cultures isolated on ASM-high included Jatrophihabitans (4 cultures), Conexibacter (1 culture), and Amycolatopsis (1 culture). Interestingly, we isolated what are likely the first members of two novel actinobacterial lineages on ASM-low. The first such culture—Microtrichales sp. strain

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 5 46 Bartelme et al. Downloaded from http://msphere.asm.org/

FIG 3 ASM-low and ASM-high cultured distinct Alphaproteobacteria (a) and Actinobacteria (b). Bar heights are the number of cultures obtained for each taxon and are colored by the medium type on which they were isolated. The genera assigned to bacteria isolated on ASM-low were distinct from those isolated on ASM-high (Kruskal-Wallis rank sum, ␹2 ϭ 19.05, P ϭ 1.28 ϫ 10Ϫ5). These differences were driven by differences in culturability across medium types for actinobacterial genera but not for alphaproteobacterial genera (Dunn test, P Յ 0.000 for Actinobacteria and P ϭ 0.128 for Alphaproteobacteria). on March 15, 2020 by guest

AZCC_0197—belongs to the Microtrichales order of the Acidimicrobiia class. The best 16S rRNA gene sequence match to an existing isolate is 93.4% identity to Aquihabitans daechungensis strain G128. However, strain AZCC_0197 more closely matched numerous 16S rRNA gene sequences from environmental clones of uncultured Acidimicrobiia. The second lineage—Frankiales sp. strains AZCC_0102 and AZCC_0072—were classi- ﬁed as members of the Frankiales order of the Actinobacteria class with best matches of Ͻ97% nucleotide identity to existing Frankiales isolates (41). Several of the microbes we isolated were representative of abundant members of the subsurface soil microbial community at the Oracle Ridge site. We matched the 16S rRNA gene sequences from our cultures to the phylotypes derived from the 55 cm Oracle Ridge soil sample. The 16S rRNA gene sequences from our cultures matched 13 phylotypes (at 97% identity [Fig. 4]) that account for 11.0% Ϯ 1.6% (mean Ϯ standard deviation [SD], n ϭ 3) of the total ampliﬁable microbial community. For example, the 16S rRNA gene sequences from our Bradyrhizobium isolates match a single Bradyrhi- zobium phylotype that was the most abundant phylotype at 55 cm (relative abundance of 5.7% Ϯ 0.3% [mean Ϯ SD, n ϭ 3] [Fig. 4]). Additionally, we isolated representatives of abundant Actinobacteria (Fig. 4), including two Mycobacterium phylotypes (the 11th and 17th most abundant phylotypes overall), Nocardioides (the 13th most abundant phylotype overall), and two Arthrobacter phylotypes (16th and 1,271st most abundant phylotypes overall). The other Actinobacteria cultured in these experiments represent rarer phylotypes in bulk soils. The 16S rRNA gene sequences from several of our pure cultures did not match any of the phylotypes derived from these soils at Ն97% identity, including Nakamurella (2 cultures), Nocardioides (5 cultures), Mycobacterium (1 culture), Jatrophihabitans (1 culture), Patulibacter (1 culture), Conexibacter (1 culture), Rhizobiales sp. (1 culture), Reyranella (1 culture), and Microtrichales sp. strain AZCC_0197.

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 6 47 Dilution-to-Extinction Cultivation from Soil

Relative abundance (percent) 5.0 6.0 7.0 0 1.0 00.5 0 0.05 0.1 0

40 Depth (cm) 60 Bradyrhizobium Mycobacterium (OTU 62) Frankiales Aeromicrobium ave. phylotype rank: 1 ave. phylotype rank: 11 ave. phylotype rank: 26 ave. phylotype rank: 311 0 Downloaded from

60 Nocardioides Jatrophihabitans Nordella ave. phylotype rank: 13 ave. phylotype rank: 46 ave. phylotype rank: 334 0 http://msphere.asm.org/

40 Depth (cm)

60 Arthrobacter (OTU 162) Reyranella Amycolatopsis ave. phylotype rank: 16 ave. phylotype rank: 114 ave. phylotype rank: 550 0

20 on March 15, 2020 by guest 40

60 Mycobacterium (OTU 376) Pseudolabrys Arthrobacter (OTU 2028) ave. phylotype rank: 17 ave. phylotype rank: 197 ave. phylotype rank: 1271

FIG 4 The cultures isolated in this study were representative of several abundant soil lineages that show dynamic depth distributions in Oracle Ridge soils. Points are the mean relative abundances Ϯ standard deviation (n ϭ 3) of 16S rRNA gene sequence phylotypes that matched the 16S rRNA gene sequences obtained from cultured isolates at Ն97% identity. Error bars that are not visible are located behind the symbol. Assigned genus names and the average (n ϭ 3) relative rank of each phylotype at 55 cm are shown. Cultures classiﬁed at the genus level as Mycobacterium and Arthrobacter cultures matched more than one phylotype in the cultivation-independent surveys. The best-matching OTU number is shown in parentheses. Triangles are Alphaproteobacteria. Circles are Actinobacteria.

In many environments, relative nucleic acid content can be estimated with flow cytometry analysis of cells stained with nucleic acid-staining dyes (42–44). Given that many microbes inhabiting low-nutrient environments exhibit reduced genome sizes, we sought to determine whether the nucleic acid fluorescence measured for our cultures partitioned by the growth medium on which they were isolated. For each culture, we identified the closest match to an available genome sequence and found that the genome length of these “best hit” matches was significantly correlated with the average fluorescence of SYBR green I stained cells (Spearman’s rho ϭ 0.3, P ϭ 0.0012), indicating that the nucleic acid fluorescence we quantified by flow cytometry may be indicative of genome size differences. We found that cultures isolated on ASM-low exhibited significantly lower mean nucleic acid fluorescence than those isolated on ASM-high (Fig. 5) (Kruskal-Wallis rank sum ␹2 ϭ 24.8, P ϭ 6.27 ϫ 10Ϫ7). The overall mean fluorescence was not significantly different across the phyla assigned to each isolate (Kruskal-Wallis rank sum ␹2 ϭ 0.210, P ϭ 0.647) but was significant across

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 7 48 Bartelme et al. Downloaded from http://msphere.asm.org/

FIG 5 The mean nucleic acid fluorescence of taxa isolated on ASM-low was significantly lower than for those microbes isolated on ASM-high. Points are the mean natural logarithm (ln) of the quantified nucleic acid fluorescence (in arbitrary units [AU]) of fixed and SYBR green I stained stationary-phase cultures. The mean fluorescence value was obtained from manually gated histogram plots of fluorescence within the Guava EasyCyte software. Only those cultures that were defined as pure cultures are plotted. The horizontal line in each plot is the mean fluorescence value, and the box surrounding the mean is a 95% confidence interval. Shading illustrates the relative distribution of fluorescence values within each medium type.

individual genus assignments (Kruskal-Wallis rank sum ␹2 ϭ 62.4, P ϭ 2.98 ϫ 10Ϫ6). on March 15, 2020 by guest Moreover, the mean nucleic acid fluorescence values within a given genus were similar (Fig. S4). For example, Mycobacterium isolates had relatively high nucleic acid fluorescence, regardless of which medium they were isolated on (Fig. S4). In contrast, Nocardioides (ASM-low) and Jatrophihabitans (ASM-high) displayed relatively low nucleic acid fluorescence. Interestingly, we observed a clear nucleic acid fluorescence dichotomy across the Bradyrhizobium isolates isolated on ASM-high (Fig. S4).

DISCUSSION We designed a proof-of-concept workflow to determine the feasibility of high- throughput dilution-to-extinction cultivation for the isolation of soil microbes. The method was based on a workflow for isolating microbes from oligotrophic marine environments (38). However, unlike aquatic samples, microbial cells in soils are heter- ogeneously dispersed within, or attached to, a complex matrix comprised of noncellular organic matter and minerals. The complexity of this soil matrix complicates accurate enumeration of viable cells because mineral and organic matter can interfere with flow cytometry. To circumvent these issues, we separated cells by gently shaking soils in a cell extraction buffer containing a dispersing agent and a nonionic surfactant. Cells were separated from this slurry by buoyant density centrifugation (Fig. 1). This procedure allowed cells to be floated on top of a dense solution of Nycodenz while allowing minerals to migrate through the Nycodenz solution (45). We estimated the expected culturability and calculated the actual culturability using the statistical framework of dilution culture growth outcomes described by Button et al. (40). The expected number of pure cultures (û) was estimated across all experiments using the formula û ϭϪn(1 Ϫ p) ϫ ln(1 Ϫ p), where p is the proportion of wells displaying growth (214 growth chambers displaying growth/576 chambers inoculated ϭ 0.37) and n is the number of inoculated chambers (576 chambers in total).

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 8 49 Dilution-to-Extinction Cultivation from Soil

Based on this equation, we expected û ϭ 168 pure cultures across all experiments. The number of pure cultures we obtained (133 cultures) was within 21% of this value. However, this result is conservative because it does not account for the cultures that were initially scored as positive for growth but could not be successfully subcultured. Some of these cultures may be oligotrophic taxa that were initially cultivatable but failed to successfully propagate, as described by Kuznetsov et al. (19). Alternatively, cultures that failed to propagate from microtiter plates to larger volumes might have been false positives, where flow cytometer instrument noise or well-to-well carryover was mistaken for a low-density culture. The mean culturability we observed for a given experiment (1.4% to 11% [Fig. 2]) was comparable to dilution-to-extinction cultivation studies of marine microbes, which report 0.5% to 14.3% culturability (39). Similar to Downloaded from previous observations for soil microbes (33), we observed increased culturability with longer incubation times (Fig. 2). We speculate that the culturability was higher on ASM-low than on ASM-high because the cells were extracted from an extremely oligotrophic soil sample (see Fig. S1 in the supplemental material), and perhaps the results would have been different if the original inoculum originated from more productive soils. The concentration of heterotrophic growth substrates in the isolation medium significantly influenced our ability to isolate certain actinobacterial lineages (Fig. 3). In particular, Nocardioides were exclusively isolated on ASM-low and most Mycobacterium http://msphere.asm.org/ isolates were isolated on ASM-high. We designed our media to include a defined but diverse range of carbon substrates that have been successfully used to isolate chemoheterotrophic microbes from soil or oligotrophic taxa from other environments (33, 46). Organic carbon type and availability are crucial for heterotrophic soil microbes because carbon acts as both a source of electrons for respiration and a nutrient for growth. To accommodate this requirement, many common microbial growth medium formulations for heterotrophic microbes supply diverse growth substrates (yeast extract or casein digests, for example), usually at concentrations much higher than are normally available in situ. Two key assumptions made with these common medium formulations

are that (i) microbes will use only the relevant constituents and any remaining com- on March 15, 2020 by guest pounds will have minor or no effect on microbial growth and (ii) microbes grow optimally in the laboratory when nutrient availability is much greater than their half-saturation constant (47). While many commonly studied microbes have the capac- ity to grow on complex, high-nutrient formulations, environmental nucleic acid data inform us that the vast majority of Earth’s microbes remain uncultured (46, 48). Our results indicate that the concentration of nutrients in a growth medium may be as important as the constituents of the growth medium for cultivation of uncultured environmental microbes. Numerous studies have demonstrated that dilute growth medium is superior to substrate-rich growth medium for the isolation of novel soil microbes (33, 34, 49, 50). However, the physiological explanation of why low-nutrient medium facilitates the growth of diverse microbes, or high nutrient concentrations inhibit the growth of some taxa, remains unclear. One possible explanation for these concentration-dependent effects may be that growth medium formulations applied at high concentrations contain large amounts of inhibitory substances—substances that are reduced to non- inhibitory levels in dilute medium formulations. For example, a key amino acid transporter in Chlamydia trachomatis can be blocked by nonessential amino acids, prevent- ing the transport of required amino acids, resulting in growth inhibition (51). A similar phenomenon was demonstrated in the extreme marine oligotroph “Candidatus Pe- lagibacter ubique,” where alanine was conditionally required for cell division but abolished growth at higher concentrations (52). Furthermore, reactive oxygen species can be produced during the autoclaving of nutrient-rich medium that either directly inhibit growth or combine with organics in the medium to form inhibitory compounds (53, 54). Finally, growth inhibition may be the result of misbalanced regulation of growth or accumulation of nutrient storage structures (poly-␤-hydroxybutyrate, for example), ultimately leading to cell lysis (55). A better understanding of the mecha-

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 9 50 Bartelme et al. nism(s) that enables growth on low-nutrient medium—or prevents growth on high- nutrient medium—may help us design better strategies for isolating uncultivated lineages. Critically, the collection of cultures we describe here, which were isolated on medium with identical constituents applied at different concentrations, is a first step toward an experimental method capable of addressing these questions. While several of the taxa we isolated were abundant microbial members of the shallow subsurface microbial community (Fig. 4), other isolates were rare or not identified in the cultivation-independent soil microbial community. The cultivation of additional microbial phylotypes that were not observed in molecular analyses of the same samples has been observed (56, 57). The dilution-to-extinction approach we used here favors the cultivation of abundant microbes in a given sample (40), such that the Downloaded from isolation of rarer taxa or taxa that were not observed in the original sample was unexpected. There are several possible explanations for this observation. First, the buoyant density separation protocol that we used to separate cells may have intro- duced biases. Nycodenz cell separation approaches do not recover all microbial cells from soil, and the separable fraction can be compositionally distinct from the nonsepa- rated soils (58). These effects could potentially skew the proportions of microbes that were diluted into microtiter plate wells. Alternatively, the absence of a particular taxon in a soil microbiome analysis may be the result of insufficient sequencing depth (56). Moreover, the “universal” primers used in the soil microbiome analysis (59) may not http://msphere.asm.org/ have primed DNA from some of the divergent lineages we cultured as efficiently as other phylotypes in the soil microbiome, resulting in either underrepresentation of these phylotypes in the original community or no amplification at all. We do not have sufficient evidence indicating which of these scenarios may explain our ability to culture cells that were not apparent in the microbiome analysis. Finally, as is true in any microbial cultivation experiment, there were many taxa that we did not isolate. In particular, these soils contained high relative abundances of Verrucomicrobia related to “Candidatus Udaeobacter copiosus” (60) and Acidobacteria (subgroup 6), which belong to highly sought-after lineages of uncultured microbes (61). The reasons for not

isolating these (and other) lineages are numerous but may be the result of inappro- on March 15, 2020 by guest priate medium composition (62, 63), toxic compounds in the cell separation constituents, long doubling times (Ͼϳ6 days), or dormancy (reviewed in reference 64). We provide evidence that microbes cultured from oligotrophic soils on low-nutrient medium may have reduced nucleic acid content relative to microbes isolated on richer medium (Fig. 5). Depending on the taxa in question, and their effective population size, microbial genome reduction can be driven by either genetic drift or “streamlining” selection (reviewed in reference 65). Genome streamlining is strongly linked with microbial oligotrophy in free-living aquatic microbes as a mechanism to reduce the overhead cost of replication in periodically nutrient-limited environments (reviewed in reference 21). However, direct evidence for genome streamlining in terrestrial microbes has been elusive. For example, metagenome-assembled genomes of abundant and ubiquitous uncultured Verrucomicrobia suggest that some lineages may contain reduced genome sizes (60). A more recent study showed that fire-affected warm soils selected for groups of microbes with significantly smaller genomes than cooler soils (66). Yet, there are few definitive ways to identify the growth preferences of taxa with reduced genomes short of culturing them and studying their growth dynamics under controlled conditions. The appearance of reduced nucleic acid content in cultures isolated on ASM-low may be an indication that genome reduction may be a successful life strategy for soil oligotrophs. Alternative explanations for the apparent differences in nucleic acid content in microbes cultured in ASM-low may be that (i) the ploidy of stationary-phase cells grown in ASM-low may be lower than those isolated in ASM-high, (ii) unknown cellular constituents of cells grown in ASM-low may quench SYBR green I fluorescence in the assay conditions we used, or (iii) cells isolated in ASM-high may form small microaggregates that are not completely dispersed prior to flow cytometry. The development of cultivation techniques emphasizing the high-throughput and sensitive detection of microbial growth on low-nutrient medium revolutionized the

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 10 51 Dilution-to-Extinction Cultivation from Soil

ﬁeld of aquatic microbial ecology by culturing microbes that were previously “uncul- turable” using standard techniques (38, 39, 67–69). Here, we show that similar cultivation principles facilitate the cultivation of abundant soil microbes. We demonstrate that, in addition to scrutinizing the nutritional composition of a given growth medium, the concentration of growth substrates in the growth medium must also be considered. Although we do not yet understand the mechanism of substrate-induced growth inhibition, there is evidence that this phenomenon is widespread and may impede laboratory cultivation efforts. Future studies to deduce the molecular mechanisms of substrate-induced growth inhibition will likely lead to new cultivation approaches that will allow us to isolate abundant free-living oligotrophic microbes.

MATERIALS AND METHODS Downloaded from Soil source and nutrient analysis. Fresh soil samples were collected from a soil pit on 16 August 2017 at the Oracle Ridge site in the Catalina Jemez Critical Zone Observatory (coordinates: 32.45 N, Ϫ110.74 W, elevation 2,103 m). The mean annual precipitation at Oracle Ridge is 87 cm yearϪ1, and the mean annual surface temperature is 12°C (70). The soils were Typic Ustorthent (70). The dominant vegetation at the site is ponderosa pine (Pinus ponderosa), with sparse Douglas ﬁr (Pseudotsuga menziesii). We collected ϳ300-g subsamples from 0 cm, 10 cm, 20 cm, 30 cm, 40 cm, and 55 cm depths. Soils were kept cool with ice packs for Ͻ4 h while in transit to the laboratory. At the laboratory, the soils were sieved to 2 mm and kept at 4°C for 50 days, at which point cells were separated from mineral soil. Standard soil chemical analyses were performed at the Colorado State University Soil Water and Plant

Testing Laboratory using their standard protocols. We analyzed the microbial community composition at http://msphere.asm.org/ each depth (see “Soil microbial community analysis” below) and conducted cultivation experiments from the 55 cm soil sample. Soil microbial community analysis. We extracted DNA from 1.0 g subsamples (n ϭ 3) using MoBio PowerSoil DNA extraction kits according to the manufacturer’s instructions. We amplified 16S rRNA gene fragments using 515F-Y (5 -TATGGTAATTGTGTGYCAGCMGCCGCGGTAA-3 ) and 926R (5 -AGTCAGTCAG = = = GGCCGYCAATTCMTTTRAGT-3 )(71). PCR products were purified using the QIAquick PCR purification kit = (Qiagen, Germantown, MD) per manufacturer’s specifications. Cleaned products were quantified using Tecan fluorometric methods (Tecan Group, Mannedorf, Switzerland), normalized, and pooled for Illumina MiSeq sequencing using custom sequencing primers and the MiSeq Reagent v2 500 cycle kit (Illumina, San Diego, CA) according to the manufacturer’s protocols. We identified phylotypes based on the generation of de novo operational taxonomic units (OTUs) from raw Illumina sequence reads using the UPARSE pipeline at a stringency of 97% identity (72). Paired-end reads were trimmed of adapter sequences, barcodes, and primers prior to assembly. We discarded low-quality and singleton sequences on March 15, 2020 by guest and dereplicated the remaining sequences before calculating relative abundances. Chimera filtering of the sequences was completed during clustering, while taxonomy was assigned to the OTUs with mothur (73) using version 123 of the SILVA 16S rRNA database (74) as the reference. We generated OTU and taxonomy assignment tables for subsequent analyses. Cell separation. Cells were separated from sieved soils using buoyant density centrifugation with Nycodenz modified from reference 37 to isolate viable cells. Briefly, we added 0.5 g wet soil to 44.8 ml of cell extraction buffer (137.5 mM NaCl, 26.78 mM tetrasodium pyrophosphate, and 0.27% [vol/vol] Tween 80). The soil-buffer slurry was vortexed for 30 s and shaken horizontally on a platform shaker for 2 h at 4°C. We layered 15-ml aliquots of this soil-buffer slurry over 10.0 ml of 80% (wt/vol) Nycodenz solution in 50 mM tetrasodium pyrophosphate. We used 50 ml Nalgene Oak Ridge high-speed polycarbonate centrifuge tubes for buoyant density centrifugations. Tubes containing the soil-buffer solution with Nycodenz were centrifuged at 17,000 ϫ g for 30 min at 16°C. We extracted three 0.5-ml aliquots from the resulting buoyant density preparation at a location of ϳ25 mm above the bottom of the tube (coincident with the approximate level of the top of the Nycodenz solution) to sterile microcentrifuge tubes containing 1.0 ml 137.5 mM NaCl. The microcentrifuge tubes containing Nycodenz/NaCl were vortexed and centrifuged for 20 min at 17,000 ϫ g. The resulting cell pellets were resuspended in 137.5 mM NaCl, pooled, and stored at 4°C. Medium design rationale. The ASM media were custom designed to facilitate the growth of a broad range of soil chemoheterotrophic microbes (see Table S1 in the supplemental material). Both ASM-high and ASM-low were buffered with phosphate. To this, we added minerals at concentrations derived from an “artificial rainwater” recipe (75), trace elements as described in trace element solution SL-10 with the addition of LaCl3, and vitamins as described elsewhere (52). We added heterotrophic growth substrates that included 21 amino acids and a diverse range of simple carbon substrates including 2-C substrates (glycerol and acetate), 3-C substrates (pyruvate), 4-C substrates (succinate, butyrate, and isobutyrate), 5-C substrates (ribose and valerate), a 6-C substrate (glucose), an 8-C substrate (N-acetylglucosamine), and a 10-C substrate (decanoic acid) (Table S1). We also added several polymeric growth substrates including pectin, methylcellulose, alginate, starch, and xylan (Table S1). We calculated the added carbon amount to be ϳ200 mg C literϪ1 for ASM-high and ϳ2 mg C literϪ1 for ASM-low. Dilution-to-extinction. An aliquot of cells extracted from the buoyant density separation was fixed with 1.75% (final [vol/vol]) formaldehyde and stained with SYBR green I (final stain concentration was a 1:4,000 dilution of commercial stock) for 3.5 h at room temperature in the dark. Cells were enumerated using a Millipore Guava flow cytometer, as described elsewhere (52). We diluted cells into artificial subterranean medium (ASM)-high or ASM-low nutrient medium (Table S1) to a density of 5 cells mlϪ1

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 11 52 Bartelme et al.

and aliquoted 1.0 ml of the dilute cell suspension into the wells of 2 ml polytetrafluoroethylene 96-well microtiter plates (Cowie Technology, New Castle, DE) so that on average each well contained 5 cells. Plates were covered with plastic lids that allowed air circulation and incubated at 16°C in the dark under aerobic conditions. We screened the dilution-to-extinction plates for growth by fixing (1.75% formaldehyde) and staining (1:4,000 dilution of commercial SYBR green I stock) aliquots for 18 h in the dark at room temperature and counting by flow cytometry (EMD-Millipore Guava EasyCyte), as described previously (52). We screened plates for growth at 4 and 11 weeks after inoculation. Positive cultures were defined as cultures that exceeded 1.0 ϫ 104 cells mlϪ1. Actual and theoretical culturability estimates. Culturability estimates were determined by the equation V ϭϪln(1 Ϫ p)/X, where V is the estimated culturability, p is the proportion of inoculated cultivation chambers that displayed measurable growth (number of chambers positive for growth/total number of chambers inoculated), and X is the number of cells added to each cultivation chamber as estimated from dilutions (40). The number of pure cultures (û) was estimated as follows: û ϭϪn(1 Ϫ p) ϫ

ln(1 Ϫ p), where n is the number of inoculated growth chambers and p is the proportion of inoculated Downloaded from wells displaying growth (40). Culture transfer and storage. We subcultured positive growth chambers into 25 ml of the respective growth medium (ASM-high or ASM-low) in acid-washed, sterile polycarbonate flasks and incubated them at 16°C. At the time of transfer, we assigned cultures a unique internal identification number for our Arizona Culture Collection (AZCC). Flasks were monitored for growth every other week for 2 months. Flasks displaying growth within 2 months were cryopreserved in 10% glycerol and stored at Ϫ80°C. If no growth appeared within 2 months, the cultures were discarded and the assigned AZCC number was retired. Mean fluorescence calculations. We calculated the mean fluorescence of each culture from the subcultures grown in 25 ml volumes at 12 to 15 weeks after inoculating. Culture aliquots were fixed and http://msphere.asm.org/ stained for 15 to 18 h as described above under “Dilution-to-extinction.” We manually gated histograms of the intensity of SYBR green I fluorescence (in arbitrary units) and extracted the mean fluorescence of the gated peak for each culture using the GuavaSoft software package. “Best hit” genomes were determined by subjecting the full-length 16S rRNA gene sequence of our isolates to a BLAST search against the NCBI Microbial Genomes database using web-blast (76). We extracted the total genome length from each best-hit genome. Culture identification. Cultures were identified by full-length 16S rRNA gene sequencing. Briefly, we filtered 5 to 10 ml of cell biomass from 25 ml cultures onto 0.2 ␮m pore size Supor filters and extracted DNA using a Qiagen PowerSoil DNA extraction kit according to the manufacturer’s instructions. We amplified full-length 16S rRNA genes from the resulting DNA using the 27F-1492R primer set (27F, 5 -AGAGTTTGATCMTGGCTCAG-3 ; 1492R, 5 -ACCTTGTTACGACTT-3 [77]). The reaction mix consisted of = = = = Promega’s GoTaq HotStart 2ϫ PCR master mix with final concentrations of 0.4 ␮M 27F and 0.4 ␮M 1492R primers, and 1 to 11.5 ␮l of template DNA, in a total reaction volume of 25 ␮l. The thermocycling profile was once at 94°C for 10 min followed by 36 cycles of 94°C for 45 s, 50°C for 90 s, and 72°C for 90 s, and on March 15, 2020 by guest a single 72°C extension for 10 min. The resulting amplicons were cleaned and Sanger sequenced from both the 27F and 1492R primers by Eurofins Genomics (Louisville, KY, USA) using their standard techniques. Sequences were curated using 4Peaks (78) and Geneious Prime v2019.0.1 (79). Reads were trimmed and assembled using the moderate setting in Geneious. Forward and reverse Sanger PCR reads that failed to build a full-length 16S rRNA gene with these metrics were considered “mixed” cultures and not analyzed further. Culture taxonomy and determination of taxonomic differences across growth medium formulations. High-quality full-length 16S rRNA gene sequences from the cultures were used to assign taxonomy and reconstruct phylogenetic relationships. We assigned taxonomy to all assembled 16S rRNA gene sequences using the SILVA database SINA aligner v128 (80). A Shapiro-Wilk test of normality was conducted in base R (81) on the distribution of SILVA genus assignments from both medium types. After concluding the data were nonparametric, we performed a Kruskal-Wallis test in R (assigned genus ϳ medium type). We performed a post hoc analysis (Dunn test, in R) to determine whether culturability within a phylum varied by growth medium type. Taxonomic selection for phylogenetic reconstruction. To reconstruct a phylogeny of full-length 16S rRNA genes, our culture sequences were compared to NCBI’s Microbial Genomes and environmental sequence databases using web-blast (76). The top five hits for each sequence from each NCBI database were chosen based on the highest percent coverage and lowest E value score and included in the reconstruction. Escherichia coli K-12 was used as the outgroup of the alphaproteobacterial phylogeny, and Bacillus subtilis was used as the outgroup for the actinobacterial tree. These sequences aligned with MAFFT (82) with turn checking enabled. The alignment was then trimmed using trimAl (83) with the “automated1” setting to optimize sequence trimming for maximum-likelihood (ML) phylogenetic analyses. We reconstructed phylogenetic relationships from this trimmed alignment in the CIPRES Gateway (84). Maximum-likelihood (ML) trees were constructed using IQ-TREE with 10,000 ultrafast bootstrap trees and Bayesian Information Criterion to select the best-fit nucleic acid substitution model (85, 86). For Actinobacteria, we used the SYMϩR10 model, and for Alphaproteobacteria, we used the GTRϩFϩIϩG4 model. After an initial round of ML trees, sequence alignments were heuristically curated with IQ-TREE to eliminate sequences that appeared in the tree more than once. Finalized ML trees were then imported into the ARB environment (87), where any duplicate sequences from our AZCC cultures were added to the ML trees through ARB’s quick add parsimony function. Final trees were visualized with FigTree (88). Environmental contextualization of AZCC isolates. We matched the AZCC isolate full-length 16S rRNA gene sequences against a database of the clustered OTUs from the shallow soil depth profile

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 12 53 Dilution-to-Extinction Cultivation from Soil

samples (see “Soil microbial community analysis” above) using the usearch_global command (89)ata stringency of Ն97% identity, in both strand orientations, with maxaccepts ϭ 1 and maxrejects ϭ 0. Data availability. Full-length Sanger-sequenced 16S rRNA gene sequences are available on NCBI GenBank under accession numbers MK875836 to MK875967. Illumina data from the 55-cm Oracle Ridge community are available on the NCBI SRA under accession numbers SRR9172130 to SRR9172198.

SUPPLEMENTAL MATERIAL Supplemental material is available online only. FIG S1, EPS file, 1.3 MB. FIG S2, PDF file, 0.4 MB. FIG S3, PDF file, 0.6 MB. FIG S4, EPS file, 1.5 MB.

TABLE S1, XLSX ﬁle, 0.01 MB. Downloaded from

ACKNOWLEDGMENTS We thank Nathan Abramson, Jasper Bloodsworth, Brenna Bourque, Amanda Howe, Bridget Taylor, and H. James Tripp for assisting with sample collection, culture maintenance, and DNA extractions relevant to this work. Funding from this work came from startup funds provided to P.C. from the Univer- sity of Arizona’s Technology and Research Initiative Fund (the Water, Environmental,

and Energy Solutions initiative) and seed grants from the Center for Environmentally http://msphere.asm.org/ Sustainable Mining and The University of Arizona College of Agriculture and Life Sciences. Work at JCVI was supported by P01AI118687.

REFERENCES 1. Chaparro JM, Sheﬂin AM, Manter DK, Vivanco JM. 2012. Manipulating the structure of bacterial and archaeal communities in soil. Soil Biol Biochem soil microbiome to increase soil health and plant fertility. Biol Fertil Soils 50:58–65. https://doi.org/10.1016/j.soilbio.2012.03.011. 48:489–499. https://doi.org/10.1007/s00374-012-0691-4. 13. Blume E, Bischoff M, Reichert JM, Moorman T, Konopka A, Turco RF. 2. Long PE, Williams KH, Hubbard SS, Banﬁeld JF. 2016. Microbial metag- 2002. Surface and subsurface microbial biomass, community structure enomics reveals climate-relevant subsurface biogeochemical processes. and metabolic activity as a function of soil depth and season. Appl Soil Trends Microbiol 24:600–610. https://doi.org/10.1016/j.tim.2016.04.006. Ecol 20:171–181. https://doi.org/10.1016/S0929-1393(02)00025-2.

3. Fernández-Luqueño F, Valenzuela-Encinas C, Marsch R, Martínez-Suárez 14. Spohn M, Klaus K, Wanek W, Richter A. 2016. Microbial carbon use on March 15, 2020 by guest C, Vázquez-Núñez E, Dendooven L. 2011. Microbial communities to efficiency and biomass turnover times depending on soil depth— mitigate contamination of PAHs in soil—possibilities and challenges: a implications for carbon cycling. Soil Biol Biochem 96:74–81. https://doi review. Environ Sci Pollut Res Int 18:12–30. https://doi.org/10.1007/ .org/10.1016/j.soilbio.2016.01.016. s11356-010-0371-6. 15. Overmann J, Abt B, Sikorski J. 2017. Present and future of culturing bacteria. 4. Tiedje JM, Asuming-Brempong S, Nüsslein K, Marsh TL, Flynn SJ. 1999. Annu Rev Microbiol 71:711–730. https://doi.org/10.1146/annurev-micro Opening the black box of soil microbial diversity. Appl Soil Ecol 13: -090816-093449. 109–122. https://doi.org/10.1016/S0929-1393(99)00026-8. 16. Gao Y, Wu M. 2018. Free-living bacterial communities are mostly dom- 5. Jobbagy EG, Jackson RB. 2000. The vertical distribution of soil organic inated by oligotrophs. bioRxiv 350348. https://www.biorxiv.org/content/ carbon and its relation to climate and vegetation. Ecol Appl 10:423–436. 10.1101/350348v1. https://doi.org/10.1890/1051-0761(2000)010[0423:TVDOSO]2.0.CO;2. 17. Morita RY. 1997. Bacteria in oligotrophic environments: starvation- 6. Fierer N, Schimel JP, Holden PA. 2003. Variations in microbial community survival lifestyle, 1st ed. Chapman and Hall, New York, NY. composition through two soil depth profiles. Soil Biol Biochem 35: 18. Schut F, Prins RA, Gottschal JC. 1997. Oligotrophy and pelagic marine 167–176. https://doi.org/10.1016/S0038-0717(02)00251-1. bacteria: facts and fiction. Aquat Microb Ecol 12:177–202. https://doi 7. Ajwa HA, Rice CW, Sotomayor D. 1998. Carbon and nitrogen mineralization .org/10.3354/ame012177. in tallgrass prairie and agricultural soil profiles. Soil Sci Soc Am J 62:942–951. 19. Kuznetsov SI, Dubinina GA, Lapteva NA. 1979. Biology of oligotrophic https://doi.org/10.2136/sssaj1998.03615995006200040014x. bacteria. Annu Rev Microbiol 33:377–387. https://doi.org/10.1146/annurev 8. Jobbágy EG, Jackson RB. 2001. The distribution of soil nutrients with .mi.33.100179.002113. depth: global patterns and the imprint of plants. Biogeochemistry 53: 20. Cho J-C, Giovannoni SJ. 2004. Cultivation and growth characteristics of a 51–77. https://doi.org/10.1023/A:1010760720215. diverse group of oligotrophic marine gammaproteobacteria. Appl Envi- 9. Balesdent J, Basile-Doelsch I, Chadoeuf J, Cornu S, Derrien D, Fekiacova ron Microbiol 70:432–440. https://doi.org/10.1128/aem.70.1.432-440 Z, Hatté C. 2018. Atmosphere-soil carbon transfer as a function of soil .2004. depth. Nature 559:599–602. https://doi.org/10.1038/s41586-018-0328-3. 21. Giovannoni SJ, Cameron Thrash J, Temperton B. 2014. Implications of 10. Hillel D. 1980. Fundamentals of soil physics. Academic Press, New York, streamlining theory for microbial ecology. ISME J 8:1553–1565. https:// NY. doi.org/10.1038/ismej.2014.60. 11. Brewer TE, Aronson EL, Arogyaswamy K, Billings SA, Botthoff JK, Camp- 22. Lauro FM, McDougald D, Thomas T, Williams TJ, Egan S, Rice S, DeMaere bell AN, Dove NC, Fairbanks D, Gallery RE, Hart SC, Kaye J, King G, Logan MZ, Ting L, Ertan H, Johnson J, Ferriera S, Lapidus A, Anderson I, Kyrpides G, Lohse KA, Maltz MR, Mayorga E, O’Neill C, Owens SM, Packman A, N, Munk AC, Detter C, Han CS, Brown MV, Robb FT, Kjelleberg S, Pett-Ridge J, Plante AF, Richter DD, Silver WL, Yang WH, Fierer N. 2019. Cavicchioli R. 2009. The genomic basis of trophic strategy in marine Ecological and genomic attributes of novel bacterial taxa that thrive in bacteria. Proc Natl Acad SciUSA106:15527–15533. https://doi.org/10 subsurface soil horizons. mBio 10:e01318-19. https://doi.org/10.1128/ .1073/pnas.0903507106. mBio.01318-19. 23. Poindexter JS. 1981. Oligotrophy: fast and famine existence. Adv Microb 12. Eilers KG, Debenport S, Anderson S, Fierer N. 2012. Digging deeper to Ecol 5:63–90. https://doi.org/10.1007/978-1-4615-8306-6_2. find unique microbial communities: the strong effect of depth on the 24. Giovannoni SJ. 2017. SAR11 bacteria: the most abundant plankton in the

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 13 54 Bartelme et al.

oceans. Annu Rev Mar Sci 9:231–255. https://doi.org/10.1146/annurev nucleic acid-content bacteria across diverse freshwater ecosystems. -marine-010814-015934. ISME J 12:1344–1359. https://doi.org/10.1038/s41396-018-0070-8. 25. Swan BK, Tupper B, Sczyrba A, Lauro FM, Martinez-Garcia M, Gonzalez 44. Liu J, Hao Z, Ma L, Ji Y, Bartlam M, Wang Y. 2016. Spatio-temporal JM, Luo H, Wright JJ, Landry ZC, Hanson NW, Thompson BP, Poulton NJ, variations of high and low nucleic acid content bacteria in an exorheic Schwientek P, Acinas SG, Giovannoni SJ, Moran MA, Hallam SJ, Cavic- river. PLoS One 11:e0153678. https://doi.org/10.1371/journal.pone chioli R, Woyke T, Stepanauskas R. 2013. Prevalent genome streamlining .0153678. and latitudinal divergence of planktonic bacteria in the surface ocean. 45. Liu J, Li JQ, Feng L, Cao H, Cui Z. 2010. An improved method for Proc Natl Acad SciUSA110:11463–11468. https://doi.org/10.1073/pnas extracting bacteria from soil for high molecular weight DNA recovery .1304246110. and BAC library construction. J Microbiol 48:728–733. https://doi.org/10 26. Santoro AE, Dupont CL, Richter RA, Craig MT, Carini P, McIlvin MR, Yang .1007/s12275-010-0139-1. Y, Orsi WD, Moran DM, Saito MA. 2015. Genomic and proteomic char- 46. Steen AD, Crits-Christoph A, Carini P, DeAngelis KM, Fierer N, Lloyd KG, acterization of “Candidatus Nitrosopelagicus brevis”: an ammonia- Cameron Thrash J. 2019. High proportions of bacteria and archaea oxidizing archaeon from the open ocean. Proc Natl Acad SciUSA across most biomes remain uncultured. ISME J 13:3126–3130. https:// 112:1173–1178. https://doi.org/10.1073/pnas.1416223112. doi.org/10.1038/s41396-019-0484-y. 27. Noell SE, Giovannoni SJ. 2019. SAR11 bacteria have a high afﬁnity and 47. Monod J. 1949. The growth of bacterial cultures. Annu Rev Microbiol Downloaded from multifunctional glycine betaine transporter. Environ Microbiol 21: 3:371–394. https://doi.org/10.1146/annurev.mi.03.100149.002103. 2559–2575. https://doi.org/10.1111/1462-2920.14649. 48. Lloyd KG, Steen AD, Ladau J, Yin J, Crosby L. 2018. Phylogenetically novel 28. Carini P, Dupont CL, Santoro AE. 2018. Patterns of thaumarchaeal gene uncultured microbial cells dominate earth microbiomes. mSystems expression in culture and diverse marine environments. Environ Micro- 3:e00055-18. https://doi.org/10.1128/mSystems.00055-18. biol 20:2112–2124. https://doi.org/10.1111/1462-2920.14107. 49. Sait M, Hugenholtz P, Janssen PH. 2002. Cultivation of globally distrib- 29. Könneke M, Bernhard AE, de la Torre JR, Walker CB, Waterbury JB, Stahl uted soil bacteria from phylogenetic lineages previously only detected DA. 2005. Isolation of an autotrophic ammonia-oxidizing marine ar- in cultivation-independent surveys. Environ Microbiol 4:654–666. chaeon. Nature 437:543–546. https://doi.org/10.1038/nature03911. https://doi.org/10.1046/j.1462-2920.2002.00352.x. 30. Santoro AE, Casciotti KL. 2011. Enrichment and characterization of 50. Janssen PH, Yates PS, Grinton BE, Taylor PM, Sait M. 2002. Improved ammonia-oxidizing archaea from the open ocean: phylogeny, physiol- culturability of soil bacteria and isolation in pure culture of novel http://msphere.asm.org/ ogy and stable isotope fractionation. ISME J 5:1796–1808. https://doi members of the divisions Acidobacteria, Actinobacteria, Proteobacteria, .org/10.1038/ismej.2011.58. and Verrucomicrobia. Appl Environ Microbiol 68:2391–2396. https://doi 31. Hu SJ, Van Bruggen AHC, Grünwald NJ. 1999. Dynamics of bacterial .org/10.1128/aem.68.5.2391-2396.2002. populations in relation to carbon availability in a residue-amended soil. 51. Braun PR, Al-Younes H, Gussmann J, Klein J, Schneider E, Meyer TF. 2008. Appl Soil Ecol 13:21–30. https://doi.org/10.1016/S0929-1393(99)00015-3. Competitive inhibition of amino acid uptake suppresses chlamydial 32. Saito A, Mitsui H, Hattori R, Minamisawa K, Hattori T. 1998. Slow-growing growth: involvement of the chlamydial amino acid transporter BrnQ. J and oligotrophic soil bacteria phylogenetically close to Bradyrhizobium Bacteriol 190:1822–1830. https://doi.org/10.1128/JB.01240-07. japonicum. FEMS Microbiol Ecol 25:277–286. https://doi.org/10.1111/j 52. Carini P, Steindler L, Beszteri S, Giovannoni SJ. 2013. Nutrient require- .1574-6941.1998.tb00480.x. ments for growth of the extreme oligotroph “Candidatus Pelagibacter 33. Davis KER, Joseph SJ, Janssen PH. 2005. Effects of growth medium, ubique” HTCC1062 on a deﬁned medium. ISME J 7:592–602. https://doi inoculum size, and incubation time on culturability and isolation of soil .org/10.1038/ismej.2012.122. bacteria. Appl Environ Microbiol 71:826–834. https://doi.org/10.1128/ 53. Tanaka T, Kawasaki K, Daimon S, Kitagawa W, Yamamoto K, Tamaki H, AEM.71.2.826-834.2005. Tanaka M, Nakatsu CH, Kamagata Y. 2014. A hidden pitfall in the 34. Joseph SJ, Hugenholtz P, Sangwan P, Osborne CA, Janssen PH. 2003.

preparation of agar media undermines microorganism cultivability. Appl on March 15, 2020 by guest Laboratory cultivation of widespread and previously uncultured soil Environ Microbiol 80:7659–7666. https://doi.org/10.1128/AEM.02741-14. bacteria. Appl Environ Microbiol 69:7210–7215. https://doi.org/10.1128/ 54. Carlsson J, Nyberg G, Wrethén J. 1978. Hydrogen peroxide and super- aem.69.12.7210-7215.2003. oxide radical formation in anaerobic broth media exposed to atmo- 35. Zengler K. 2017. Protocols for high-throughput isolation and cultivation, spheric oxygen. Appl Environ Microbiol 36:223–229. https://doi.org/10 p 27–35. In McGenity TJ, Timmis KN, Balbina N (ed), Hydrocarbon and lipid microbiology protocols: isolation and cultivation. Springer, Berlin, .1128/AEM.36.2.223-229.1978. Germany. 55. Baxter M, Sieburth JM. 1984. Metabolic and ultrastructural response to 36. Jiang CY, Dong L, Zhao JK, Hu X, Shen C, Qiao Y, Zhang X, Wang Y, glucose of two eurytrophic bacteria isolated from seawater at different Ismagilov RF, Liu SJ, Du W. 2016. High-throughput single-cell cultivation enriching concentrations. Appl Environ Microbiol 47:31–38. https://doi on microfluidic streak plates. Appl Environ Microbiol 82:2210–2218. .org/10.1128/AEM.47.1.31-38.1984. https://doi.org/10.1128/AEM.03588-15. 56. Lau JT, Whelan FJ, Herath I, Lee CH, Collins SM, Bercik P, Surette MG. 37. Khalili B, Weihe C, Kimball S, Schmidt KT, Martiny J. 2019. Optimization 2016. Capturing the diversity of the human gut microbiota through of a method to quantify soil bacterial abundance by flow cytometry. culture-enriched molecular profiling. Genome Med 8:72. https://doi.org/ mSphere 4:e00435-19. https://doi.org/10.1128/mSphere.00435-19. 10.1186/s13073-016-0327-7. 38. Stingl U, Tripp HJ, Giovannoni SJ. 2007. Improvements of high- 57. Diakite A, Dubourg G, Dione N, Afouda P, Bellali S, Ngom II, Valles C, throughput culturing yielded novel SAR11 strains and other abundant Million M, Levasseur A, Cadoret F, Lagier JC, Raoult D. 2019. Extensive marine bacteria from the Oregon coast and the Bermuda Atlantic Time culturomics of 8 healthy samples enhances metagenomics efficiency. Series study site. ISME J 1:361–371. https://doi.org/10.1038/ismej.2007 PLoS One 14:e0223543. https://doi.org/10.1371/journal.pone.0223543. .49. 58. Portillo MC, Leff JW, Lauber CL, Fierer N. 2013. Cell size distributions of 39. Connon SA, Giovannoni SJ. 2002. High-throughput methods for cultur- soil bacterial and archaeal taxa. Appl Environ Microbiol 79:7610–7617. ing microorganisms in very-low-nutrient media. Appl Environ Microbiol https://doi.org/10.1128/AEM.02710-13. 68:3878–3885. https://doi.org/10.1128/aem.68.8.3878-3885.2002. 59. Walters W, Hyde ER, Berg-Lyons D, Ackermann G, Humphrey G, Parada 40. Button DK, Schut F, Quang P, Martin R, Robertson BR. 1993. Viability and A, Gilbert JA, Jansson JK, Caporaso JG, Fuhrman JA, Apprill A, Knight R. isolation of marine bacteria by dilution culture: theory, procedures, and 2015. Improved bacterial 16S rRNA gene (V4 and V4-5) and fungal initial results. Appl Environ Microbiol 59:881–891. https://doi.org/10 internal transcribed spacer marker gene primers. mSystems 1:e00009-15. .1128/AEM.59.3.881-891.1993. https://doi.org/10.1128/mSystems.00009-15. 41. Tahon G, Willems A. 2017. Isolation and characterization of aerobic 60. Brewer TE, Handley KM, Carini P, Gilbert JA, Fierer N. 2016. Genome anoxygenic phototrophs from exposed soils from the Sør Rondane reduction in an abundant and ubiquitous soil bacterium “Candidatus Mountains, East Antarctica. Syst Appl Microbiol 40:357–369. https://doi Udaeobacter copiosus.” Nat Microbiol 2:16198. https://doi.org/10.1038/ .org/10.1016/j.syapm.2017.05.007. nmicrobiol.2016.198. 42. Wang Y, Hammes F, Boon N, Chami M, Egli T. 2009. Isolation and 61. Carini P. 2019. A “cultural” renaissance: genomics breathes new life into characterization of low nucleic acid (LNA)-content bacteria. ISME J an old craft. mSystems 4:e00092-19. https://doi.org/10.1128/mSystems 3:889–902. https://doi.org/10.1038/ismej.2009.46. .00092-19. 43. Proctor CR, Besmer MD, Langenegger T, Beck K, Walser JC, Ackermann 62. Olsen RA, Bakken LR. 1987. Viability of soil bacteria: optimization of M, Bürgmann H, Hammes F. 2018. Phylogenetic clustering of small low plate-counting technique and comparison between total counts and

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 14 55 Dilution-to-Extinction Cultivation from Soil

plate counts within different size groups. Microb Ecol 13:59–74. https:// 77. Lane DJ. 1991. 16S/23S rRNA sequencing, p 115–175. In Stackebrandt E, doi.org/10.1007/BF02014963. Goodfellow M (ed), Nucleic acid techniques in bacterial systematics. 63. Aagot N, Nybroe O, Nielsen P, Johnsen K. 2001. An altered pseudomonas John Wiley and Sons, Chichester, United Kingdom. diversity is recovered from soil by using nutrient-poor pseudomonas- 78. Griekspoor A, Groothuis T. 2004. 4Peaks, 1.8. Nucleobytes, Aalsmeer, selective soil extract media. Appl Environ Microbiol 67:5233–5239. Netherlands. https://doi.org/10.1128/AEM.67.11.5233-5239.2001. 79. Biomatters, Ltd. 2019. Geneious prime, v2019.0.1. Biomatters, Ltd., Auck- 64. Lennon JT, Jones SE. 2011. Microbial seed banks: the ecological and land, New Zealand. evolutionary implications of dormancy. Nat Rev Microbiol 9:119–130. 80. Pruesse E, Peplies J, Glöckner FO. 2012. SINA: accurate high-throughput https://doi.org/10.1038/nrmicro2504. multiple sequence alignment of ribosomal RNA genes. Bioinformatics 65. Batut B, Knibbe C, Marais G, Daubin V. 2014. Reductive genome evolu- 28:1823–1829. https://doi.org/10.1093/bioinformatics/bts252. tion at both ends of the bacterial population size spectrum. Nat Rev 81. R Core Team. 2014. R: a language and environment for statistical com- Microbiol 12:841–850. https://doi.org/10.1038/nrmicro3331. puting, 3.1.1. R Foundation for Statistical Computing, Vienna, Austria. 66. Sorensen JW, Dunivin TK, Tobin TC, Shade A. 2019. Ecological selection 82. Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment soft- for small microbial genomes along a temperate-to-thermal soil gradient. ware version 7: improvements in performance and usability. Mol Biol

Nat Microbiol 4:55–61. https://doi.org/10.1038/s41564-018-0276-6. Evol 30:772–780. https://doi.org/10.1093/molbev/mst010. Downloaded from 67. Rappe MS, Connon SA, Vergin KL, Giovannoni SJ. 2002. Cultivation of the 83. Capella-Gutiérrez S, Silla-Martínez JM, Gabaldón T. 2009. trimAl: a tool ubiquitous SAR11 marine bacterioplankton clade. Nature 418:630–633. for automated alignment trimming in large-scale phylogenetic analyses. https://doi.org/10.1038/nature00917. Bioinformatics 25:1972–1973. https://doi.org/10.1093/bioinformatics/ 68. Kim S, Kang I, Seo JH, Cho JC. 2019. Culturing the ubiquitous freshwater btp348. actinobacterial acI lineage by supplying a biochemical ‘helper’ catalase. 84. Miller MA, Pfeiffer W, Schwartz T. 2010. Creating the CIPRES Science ISME J 13:2252–2263. https://doi.org/10.1038/s41396-019-0432-x. Gateway for inference of large phylogenetic trees. In 2010 Gateway 69. Salcher MM, Schaeﬂe D, Kaspar M, Neuenschwander SM, Ghai R. 2019. Computing Environments Workshop. IEEE, New York, NY. Evolution in action: habitat-transition leads to genome-streamlining in 85. Minh BQ, Nguyen MAT, Von Haeseler A. 2013. Ultrafast approximation Methylophilaceae. ISME J 13:2764–2777. https://doi.org/10.1038/s41396 for phylogenetic bootstrap. Mol Biol Evol 30:1188–1195. https://doi.org/ -019-0471-3. 10.1093/molbev/mst024. 70. Lybrand RA, Rasmussen C. 2015. Quantifying climate and landscape http://msphere.asm.org/ 86. Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ. 2015. IQ-TREE: a fast position controls on soil development in semiarid ecosystems. Soil Sci and effective stochastic algorithm for estimating maximum-likelihood Soc Am J 79:104–116. https://doi.org/10.2136/sssaj2014.06.0242. phylogenies. Mol Biol Evol 32:268–274. https://doi.org/10.1093/molbev/ 71. Parada AE, Needham DM, Fuhrman JA. 2016. Every base matters: assess- msu300. ing small subunit rRNA primers for marine microbiomes with mock 87. Ludwig W, Strunk O, Westram R, Richter L, Meier H, Yadhukumar, communities, time series and global ﬁeld samples. Environ Microbiol 18:1403–1414. https://doi.org/10.1111/1462-2920.13023. Buchner A, Lai T, Steppi S, Jobb G, Förster W, Brettske I, Gerber S, Ginhart 72. Edgar RC. 2013. UPARSE: highly accurate OTU sequences from microbial AW, Gross O, Grumann S, Hermann S, Jost R, König A, Liss T, Lüßmann amplicon reads. Nat Methods 10:996–998. https://doi.org/10.1038/ R, May M, Nonhoff B, Reichel B, Strehlow R, Stamatakis A, Stuckmann N, nmeth.2604. Vilbig A, Lenke M, Ludwig T, Bode A, Schleifer K. 2004. ARB: a software 73. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, environment for sequence data. Nucleic Acids Res 32:1363–1371. Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, Sahl JW, Stres B, https://doi.org/10.1093/nar/gkh293. Thallinger GG, Van Horn DJ, Weber CF. 2009. Introducing mothur: open- 88. Rambaut A, Drummond A. 2016. FigTree 1.4.3. source, platform-independent, community-supported software for de- 89. Edgar RC. 2010. Search and clustering orders of magnitude faster

scribing and comparing microbial communities. Appl Environ Microbiol than BLAST. Bioinformatics 26:2460–2461. https://doi.org/10.1093/ on March 15, 2020 by guest 75:7537–7541. https://doi.org/10.1128/AEM.01541-09. bioinformatics/btq461. 74. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, Peplies J, 90. Carini P, Marsden PJ, Leff JW, Morgan EE, Strickland MS, Fierer N. 2016. Glöckner FO. 2013. The SILVA ribosomal RNA gene database project: Relic DNA is abundant in soil and obscures estimates of soil microbial improved data processing and web-based tools. Nucleic Acids Res 41: diversity. Nat Microbiol 2:16242. https://doi.org/10.1038/nmicrobiol D590–D596. https://doi.org/10.1093/nar/gks1219. .2016.242. 75. Kox MAR, Lüke C, Fritz C, van den Elzen E, van Alen T, Op den Camp HJM, 91. Delgado-Baquerizo M, Oliverio AM, Brewer TE, Benavent-González A, Lamers LPM, Jetten MSM, Ettwig KF. 2016. Effects of nitrogen fertilization Eldridge DJ, Bardgett RD, Maestre FT, Singh BK, Fierer N. 2018. A global on diazotrophic activity of microorganisms associated with Sphagnum atlas of the dominant bacteria found in soil. Science 359:320–325. magellanicum. Plant Soil 406:83–100. https://doi.org/10.1007/s11104 https://doi.org/10.1126/science.aap9516. -016-2851-z. 92. Eichorst SA, Kuske CR, Schmidt TM. 2011. Inﬂuence of plant polymers 76. Johnson M, Zaretskaya I, Raytselis Y, Merezhuk Y, McGinnis S, Madden on the distribution and cultivation of bacteria in the phylum Acido- TL. 2008. NCBI BLAST: a better web interface. Nucleic Acids Res 36: bacteria. Appl Environ Microbiol 77:586–596. https://doi.org/10.1128/ W5–W9. https://doi.org/10.1093/nar/gkn201. AEM.01080-10.

January/February 2020 Volume 5 Issue 1 e00024-20 msphere.asm.org 15 56

1 Supplementary Information 2 For 3 4 The concentration of heterotrophic growth substrates in culture medium is a crucial 5 determinant of the culturability of subsurface soil microbes 6 7 Ryan P. Bartelme1, Joy M. Custer1, Christopher Dupont2, Josh L. Espinoza2, Manolito 8 Torralba2, Banafshe Khalili3, and Paul Carini1 9 10 11 Corresponding author: [email protected] 12 13 14 Supplementary Table 1. Artificial Subterranean Growth Medium (ASM) -low and -high 15 composition including concentrations of carbon sources, amino acids, vitamins, trace metals, 16 amino acids, and inorganic nutrients. 17 18 Supplementary Figure 1: The soils used for cultivation were oligotrophic. (A) points are the 19 measured soil organic carbon content of 184 soil samples collected from shallow subsurface pits 20 located across the United States (from ref. (11)). (B) points are the measured NO3-N content of 21 121 soils (113 surface soils) collected across the United States from (91, 92). The organic carbon 22 percent and NO3-N measured in our 55 cm sample from Oracle Ridge is illustrated with a red 23 dashed line in both panels. Our sample falls on the low end of both organic carbon content and 24 NO3-N content across this broad selection of soils. 25 26 27 Supplementary Figure 2: Alphaproteobacterial IQ-TREE Maximum Likelihood Full 28 Length 16S rRNA gene sequence phylogeny. The tree was constructed using a GTR+F+I+G4 29 nucleotide substitution model iterated over 10,000 ultrafast bootstraps. The scale bar indicates 30 the FigTree proportionally transformed branch length. The 16S rRNA gene sequence from E. 31 coli K-12 was used as an outgroup. Clade bootstrap support is indicated by colored nodes as 32 indicated in the legend. This tree includes 16S rRNA gene sequences that amplified from all pure 33 alphaproteobacterial cultures isolated in this study (bolded black leaves denoted with ‘AZCC’), 34 as well as their best sequence matches to NCBI microbial genomes (dark blue leaves), NCBI 16S 35 rRNA gene sequences from cultured isolates (orange leaves), and NCBI 16S rRNA gene 36 sequences from environmental clones (green leaves). The medium concentration on which each 37 AZCC culture was isolated are indicated by colored bars on the right side of the phylogenetic 38 tree (red for ASM-high and blue for ASM-low). 39 40 Supplementary Figure 3: Actinobacterial IQ-TREE Maximum Likelihood Full Length 16S 41 rRNA gene sequence phylogeny. The tree was constructed using a SYM+R10 nucleotide 42 substitution model iterated over 10,000 ultrafast bootstraps. The scale bar indicates the FigTree 43 proportionally transformed branch length. The 16S rRNA gene sequence from B. subtilis DSM- 44 10 was used as an outgroup. Clade bootstrap support is indicated by colored nodes as indicated 45 in the legend. This tree includes 16S rRNA gene sequences that amplified from all pure

1 57

46 actinobacterial cultures isolated in this study (bolded black leaves denoted with ‘AZCC’), as well 47 as their best sequence matches to NCBI microbial genomes (dark blue leaves), NCBI 16S rRNA 48 gene sequences from cultured isolates (orange leaves), and NCBI 16S rRNA gene sequences 49 from environmental clones (green leaves). The medium concentration on which each AZCC 50 culture was isolated are indicated by colored bars on the right side of the phylogenetic tree (red 51 for ASM-high and blue for ASM-low). Actinobacterial classes are indicated by colored vertical 52 bars to the right of the tree figure: Thermoleophilia (purple), Acidimicrobiia (teal), and 53 Actinobacteria (Orange-Red). 54 55 56 Supplementary Figure 4: Distribution of nucleic acid fluorescence values by Phylum (a) 57 and Genus (b). Data are natural logarithm-transformed nucleic acid fluorescence values of fixed 58 and SYBR Green I-stained stationary phase cultures. Plots show all genera with n ≥ 3 cultures in 59 one or both media types. Shaded regions illustrate the distribution of fluorescence values within 60 culture media groups.

2 58

61 62

3 59

63 64

4 60

5 61

6 APPENDIX C 62 Bacterial DNA Isolation CTAB Protocol

Bacterial genomic DNA isolation using CTAB

Version Number: 3 Start Production Date: 8-25-04 Stop Production Date: (current) Authors: William S. and Helene Feil, A. Copeland Reviewed by: M. Haynes 11-12-12

Summary This scaled up CTAB method can be used to extract large quantities of large molecular weight DNA from bacteria and other microbes.

Materials & Reagents Materials/Reagents/Equipment Vendor Stock number

Disposables 1.5-mL microcentrifuge tube Eppendorf 22 36 320-4 50-mL Nalgene Oak Ridge polypropylene centrifuge tube VWR 21010-568 10-mL pipette Falcon 357551 1-mL pipette tips MBP 3781

Reagents CTAB (*see preparation notes at end) Sigma H-6269 NaCl Sigma S-3014 TE buffer (10mM Tris; 1 mM EDTA, pH 8.0). Ambion 9858 Lysozyme Sigma L-6876 Proteinase K Qiagen 19131 5M NaCl Ambion 9759 10% SDS Sigma L-4522 Chloroform Sigma C-2432 Isoamyl alcohol Sigma I-9392 Phenol Sigma P-4557 Isopropanol VWR PX-1835-14 Ethanol AAPER ------DNase-free RNAse I (100 mg/mL) Epicentre N6901K Molecular biology grade DNase-free water

Equipment Hot Plate 250 mL glass beaker Magnetic stir rod Thermometer Last Updated 10/23/12

2800 Mitchell Drive, Walnut Creek, CA 94598 www.jgi.doe.gov 63

Automatic pipette dispenser Sorval 500 Plus centrifuge (DuPont, Newtown, CT) 65°C water bath 37°C incubator/heat block 56°C heat block

Procedure Cell preparation and extraction techniques. (Modification of “CTAB method”, in Current Protocols in Molecular Biology)

Cell growth: To minimize gDNA sampling bias (e.g., excess coverage of sequences around the origin of replication) please take precautions NOT to proceed with DNA isolation while most of the cell population is in the stage of active DNA replication. We recommend collaborators to check the cell growth prior to DNA isolation. DNA should be prepared from cell culture that is either in late log phase or early stationary phase. If the cells are in the early log phase, the culture should be placed on ice or 4°C to slow down the growth and allow DNA replication to complete prior to cell lysis and DNA isolation.

If at all possible, please produce more DNA from a single isolation event than is strictly required for library creation and freeze aliquots of the extra DNA. Then, should more DNA be required for finishing it will be available. If extra cells are available instead, please consider storing extra aliquots in 15-40% glycerol at -80°C.

1.5ml 30ml 60ml 1. Grow cells (see above) in broth and pellet at 10,000 rpm for 5 min or scrape from plate. 2. Transfer bacterial suspension to the appropriate centrifuge tube. 3. Spin down cells in microfuge or centrifuge at 10,000 rpm for 5 minute. 4. Discard the supernatant. 5. Resuspend cells in TE. 6. Adjust to OD600 ! 1.0 with TE buffer (10mM Tris; 1 mM EDTA, pH 8.0) 7. Transfer given amount of cell suspension to a clean centrifuge tube. ------740!l 14.8ml 29.6ml 8. Add lysozyme (conc. 100mg/ml). Mix well. ------20!l 400!l 800!l This step is necessary for hard to lyse gram (+) and some gram (–) bacteria. 9. Incubate for 30 min. at 37°C. 10. Add 10% SDS. Mix well. ------40!l 800!l 1.6ml 11. Add Proteinase K (10mg/ml). Mix well. ------8!l 160!l 320!l 12. Incubate for 1-3 hr at 56°C. If cells are not lysed (as seen by cleared solution with increased viscosity) incubation can proceed overnight (16 hrs). 13. Add 5 M NaCl. Mix well. ------100!l 2ml 4ml

Last Updated 11/12/2012

2800 Mitchell Drive, Walnut Creek, CA 94598 www.jgi.doe.gov 64

14. Add CTAB/NaCl (heated to 65°C). Mix well. ------100!l 2ml 4ml 15. Incubate at 65°C for 10 min. 16. Add chloroform:isoamyl alcohol (24:1). Mix well. ------0.5ml 10ml 20ml 17. Spin at max speed for 10 min at room temperature. 18. Transfer aqueous phase to clean microcentrifuge tube (should not be viscous). 19. Add phenol:chloroform:isoamyl alcohol (25:24:1). Mix well. ------0.5ml 10ml 20ml 20. Spin at max speed for 10 min at room temperature. 21. Transfer aqueous phase to clean microcentrifuge tube. 22. Add chloroform:isoamyl alcohol (24:1). Mix well. ------0.5ml 10ml 20ml 23. Spin at max speed for 10 min at room temperature. 24. Transfer aqueous phase and add 0.6 vol isopropanol (-20°C). (e.g. if 400 µl of aqueous phase is transferred, add 240 µl of isopropanol. ---- Add 0.6 vol. ---- 25. Incubate at -20°C for 2 hrs to overnight. 26. Spin at max speed for 15 min at 4°C. 27. Wash pellet with cold 70% ethanol (directly from -20°C freezer), spin at max speed for 5 min. 28. Discard the supernatant and let pellet dry at room temp. This may take some time (20 min. to several hours, depending on humidity). 29. Resuspend in ~170 !l of DNase-free water. Proceed to RNAse treatment. 1.1 Set up the following reaction in a 1.5ml microcentrifuge tube (multiple reactions can be done in different tubes): Note: RNase I @ 10U/!l, one unit digests 100 ng of RNA per second

DNA (in H20) 170!l 10X RNase I buffer 20!l RNase I 10!l 200ul 1.2 Mix & Spin down. 1.3 Incubate tube at 37°C for 1 hr. Checkpoint: Check a small aliquot (5ul) on an agarose gel with a no treatment control. Run gel 10-15 min. If there is only a trace of RNA, proceed with next step, heat inactivation. If a large amount of RNA is still present, add another 10!l of RNase I and repeat the incubation.

1.4 Heat inactivate enzyme at 70°C for 15 min. 1.5 Place tube on ice to cool. 2. Ethanol Precipitation 2.1 Add 1/10 volume of 3M Sodium Acetate to your sample. 2.2 Add 2.5 volumes of 100% ethanol. Last Updated 11/12/2012

2800 Mitchell Drive, Walnut Creek, CA 94598 www.jgi.doe.gov 65

2.3 Mix and spin down sample. 2.4 Place at -80°C for 30 min (-20°C 2 hrs to overnight). 2.5 Spin sample at 4°C for 20 min to pellet DNA. 2.6 Carefully, pour off supernatant. 2.7 Wash pellet with 70% ethanol (cold). 2.8 Spin sample at 4°C for 3-5 min. 2.9 Pull off all ethanol with pipet tip. 2.10 Air dry pellet (or vacuum dry for 5-15 min using no heat). 2.11 Resuspend pellet with 100 !l of TE. 2.12 If multiple reactions, combine them. 2.13 Run 1 µl in a 1% agarose gel to check quality. 2.14 Store DNA @ -80°C or -20°C.

Measure DNA concentration with fluorometer dsDNA assay (Qubit or equivalent) or UV absorption (Nanodrop). The 260/280 ratio should be approximately 1.8. The 260/230 ratio should be 1.8 – 2.2 for pure DNA. Note that residual phenol absorbs strongly at 270 nm and will inflate the apparent DNA concentration. If using Nanodrop check whether the peak (which should be at about 258 nm) is shifted toward 270 nm. Note that the JGI requires submission of a Qubit/fluorometric measurement. Nanodrop readings are not acceptable QC measurements for the JGI.

Notes and precautions.

-In step 1, do not use too many bacterial cells (an OD600 of not more than 1.2 is recommended), or DNA does not separate well from the protein.

-Most of the time, inverting several times is sufficient to mix well. Shaking too hard will shear the DNA.

-Use any protocol for DNA precipitation, the one in this protocol works well.

Reagent/Stock Preparation

CTAB/NaCl (hexadecyltrimethyl ammonium bromide)

Dissolve 4.1 g NaCl in 80 ml of water and slowly add 10 g CTAB while heating ("65°C) and stirring. This takes more than 3 hrs to dissolve CTAB. Adjust final volume to 100 ml and sterilize by filter or autoclave.

Last Updated 11/12/2012

2800 Mitchell Drive, Walnut Creek, CA 94598 www.jgi.doe.gov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`(9CE::(#&( a::%;C( #C( b50;(W@[:(;1(-.c+.)(N+-&+5"?(M05&"d(&0(S0;0$"%0+'('+',"%'.0%C(( .C( H&(&S.'(,0.%&(20+(N#%(,.,"&&"(044(%"N"''#52(#;0+%&'(405('&0NU(405(450T"%('&0NU'C( ..C( R"--"&'(;#2(#-'0(Q"(450T"%(405(4+&+5"("d&5#N&.0%'C(*&05"(#&(@E:_>C( QC( >"%&5.4+$"(&+Q"'(#&(WD::(5,;(d(9:(;.%+&"'(#%)().'N#5)('+,"5%#&#%&C(4,)%K#Q&R%#*-$%# *-;%$(&)&()#2*#(,)#L",-.:>#*,0%)20%*#&#",('%$#L%()$26-'%#2*#$%S-2$%.<( NC( e".$S(,"--"&C(*&.%$(;#''('S0+-)(Q"(E:@9::(;$(K"&(K".$S&(60,&.;#-("d&5#N&.0%'(#&(fa:@I:(;$( 405(K"&(K".$S&7C(X4(20+(K#'S(.%(R3*(#%)(K".$S(.%(#(;.N50N"%&5.4+$"(&+Q"?(0,&.;#-(K".$S&'(;#2(Q"( E:@[:(;$(405(gK#'S")(K".$S&hC( .C( 1#5$"5(,"--"&'(;#2(Q"().M.)")(#%)("d&5#N&.0%'(,"5405;")(.%(,#5#--"-i("d&5#N&")(G=H(,(":# 6$,0#)+%#*&0%#*2('"%#L,",(2%*(;#2(Q"(N0;Q.%")(-#&"5C(( ..C( !"'+',"%)(,"--"&(.%([D:(O1(04(9j(R3*(.%(E(;1(;.N50N"%&5.4+$"(&+Q"'( EC( A"#&(.%#N&.M#&"LU.--(&S"(5"'+',"%)")(;2N0Q#N&"5.#(#&(]D_>(405(a:(;.%+&"'((( M<# >"%&5.4+$"(405(9:(;.%+&"'(#&(WD::(5,;(#%)().'N#5)('+,"5%#&#%&(&0(5";0M"(5"'.)+#-(R3*C(4,)%K#Q&:# $%S-2$%#&(#&..2)2,(&"#T5UA#02(-)%*#26#)+%#*-;%$(&)&()#2*#L",-.:<# WC( H))(W::(O1(04(-2'.'(Q+44"5(,-+'(9::(O1(04(E:J(*G*(#%)(;.d(Q2(,.,"&&.%$C( ( VC&$":W#Q%L+&(2L&"#32*$-;)2,(( DC( H))(9D:(;$(04(;#&5.d(;#&"5.#-(6$-#''(:C9@;;().#;"&"57(#%)(S0;0$"%.T"(#&([:::(5,;(405(W(N2N-"'(d([:( '"N0%)'("#NS(6b.'S"5(*N."%&.4.N(M05&"d"5(K.&S(<03.0(;.N50N"%&5.4+$"(&+Q"(S0-)"5(#)#,&"5i(#-&"5%#&.M"( S0;0$"%.T"5(04(R5"N"--2'(EW(S0;0$"%.T"5([(k([:('"N?(+,(&0(aI::(5,;?(V#'"5(E:9:?(!"4([E7C(4,)%K#X&)L+# )+%#)-/%*#L&$%6-"":#),#&Y,2.#"%&R&'%<( ( V8&)%W#P%""#1&""#.2'%*)2,(# aC( H))(D:(O1(04(-2'0T2;"(69::;$L;1('&0NUi(4.%#-(N0%N"%&5#&.0%(9:(;$L;17?(;.d(Q2(&+5%.%$("%)@0M"5@"%)(Q2( S#%)('"M"5#-(&.;"'?(#%)(.%N+Q#&"(#&([B_>(d(9(S0+5C((4,)%K#!""#*-/*%S-%()#Y,$)%F2('#*+,-".#/%#&Y,2.%.<# 4,)%K#Z"&**#/%&.*#&$%#*)2""#2(#*,"-)2,(I#)+-*>#)+,$,-'+#02F2('#2*#20;,$)&()<( BC( H))(W:(O1(P.#$"%(R50&".%#'"(V(6ECD(;$L;1i(4.%#-(N0%N"%&5#&.0%(04(E::(+$L;-7?(;.d(Q2(50&#&.%$("%)@0M"5@ "%)(Q2(S#%)?(#%)(.%N+Q#&"(#&([Bl>(405(]:(;.%+&"'(i("M"52([:(;.%+&"'?(;.d(Q2(50&#&.%$("%)@0M"5@"%)( '"M"5#-(&.;"'(6#4&"5([:(#%)(a:(;.%+&"'7C(( IC( >"%&5.4+$"(#&(WD::(5,;(d(9:(;.%+&"'(40--0K")(Q2(9W:::(5,;(d(E(;.%+&"'C(4,)%K#D+%#/%&.*#12""#(,)#;%""%)# ),#)+%#/,)),0#&(.#*,#/%#Y%$:#L&$%6-"#$%0,Y2('#6$,0#)+%#L%()$26-'%#),#(,)#.2*)-$/#)+%#/%&.*<((( ]C( m"52(N#5"4+--2?(K.&S0+&(#$.&#&.%$(&S"(Q"#)'?(,.,"&(044('+,"5%#&#%&(#%)(&5#%'4"5(&0(R1Y(&+Q"'C( ! 67 ! ! ! ( P+",$,6,$0K2*,&0:"#&"L,+,"#%F)$&L)2,(# 9:C(X;;").#&"-2(,5.05(&0(+'"?(,"--"&(RS#'"(10NU(Y"-(6R1Y7(#&(9E?:::(@(9a?:::(d($(405(E:(&0([:('"N0%)'C( 99C(H))(D::(O1(,S"%0-8NS-050405;8.'0#;2-(#-N0S0-(6ED8EW897(&0(&S"(R1Y(&+Q"'C( 9EC(R-#N"(R1Y(&+Q"'(0%(#(A+-#<.d"5(#&(E:(5,;(405(9:(;.%+&"'(&0(&S050+$S-2(;.d(05$#%.N(#%)(#c+"0+'(,S#'"'i( &S.'('S0+-)(405;(#(S0;0$"%0+'('+',"%'.0%<( 9[C(>"%&5.4+$"(#&(WD::(5,;(405(9:(;.%+&"'C#/S"(R1Y($"-(K.--(405;(#(Q#55."5(Q"&K""%(&S"(#c+"0+'(#%)(05$#%.N( ,S#'"'C( 9WC(>#5"4+--2(,.,"&(044(%+N-".N(#N.)@N0%&#.%.%$(#c+"0+'(+,,"5(,S#'"(&0(45"'S(E(;1(;.N50N"%&5.4+$"(&+Q"'C( 4,)%K#!S-%,-*#"&:%$#*+,-".#/%#&/,-)#[AA5\TA#-8<( (( ]*,;$,;&(,"#;$%L2;2)&)2,(#&(.#C"-)2,(# 9DC(H))(9L9:(M0-+;"(6E:@WD(+17(04(D(<(=#>-(40--0K")(Q2(9(M0-+;"(6E::@WD:(+17(04(.'0,50,#%0-C( 9aC(X%N+Q#&"(0M"5%.$S&(#&(500;(&";,"5#&+5"C( U@<#>"%&5.4+$"(#&(9W?:::(5,;(405([:(;.%(#&(E:@EEl>C(4,)%K#D%0;%$&)-$%#L,()$,"#2*#),#;$%Y%()#,Y%$+%&)2('<# 9IC(e#'S(K.&S(B::(O1(04(B:J("&S#%0-(#%)(N"%&5.4+$"(#&(9W?:::(5,;(405(9:(;.%+&"'C(4,)%K#^%%;#)+%#)-/%*# ;,*2)2,(%.#2(#)+%#*&0%#.2$%L)2,(#*,#)+&)#:,-#&"1&:*#R(,1#1+%$%#)+%#;%""%)#V*+,-".#/%W#&*#2)#2*#,6)%(# 2(Y2*2/"%_( 9]C(G0(#%(#)).&.0%(E@[(;05"(K#'S"'(6B::(O1(04(B:J(F/\A?(',.%(d(9W:::5,;(d(9:;.%(#%)().'N#5)( '+,"5%#&#%&7C(4,)%K#9,-#0&:#1&()#),#*&Y%#)+%#62$*)#*-;%$(&)&()#-()2"#:,-$#%F)$&L)2,(#2*#L,0;"%)%#),#%(*-$%# (,#",**#,6#34!<( E:C(1"&(&S"(5";#.%.%$("&S#%0-("M#,05#&"(Q2(-"#M.%$(#&(500;(&";,"5#&+5"(405(9D(;.%+&"'(K.&S(&+Q"(-.)'(N5#NU")C( 4,)%K#?,0%#1&)%$#0&:#$%0&2(<#D-/%*#L&(#/%#.-0;%.#-;*2.%#.,1(#,(),#^20#12;%*#),#+%";#'%)#$2.#,6# 1&)%$#26#:,-#.%*2$%#/-)#&(:#%)+&(,"#*+,-".#%Y&;,$&)%#,66#&6)%$#UT#02(-)%*<( E9C(H))(D:@9::(O-(04(K#5;")(F3(6&S#&(S#'(Q""%(S"#&")(&0(aDl>(d(9:(;.%+&"'7(( EEC(1"#M"(0M"5%.$S&(#&(500;(&";,"5#&+5"(0%(#(%+&#&05C( E[C(*&05"(#&(Wl>(405(;.%.;+;(04([@W()#2'(,5.05(&0(c+#-.&2(#''"'';"%&C( ( `-&"2):#&**%**0%()*# EWC(A"#&(G=H(&0(aDl>(d(9(S0+5C(P+#-.&2(#''"'';"%&'('S0+-)(Q"(N0;,-"&")(K.&S(=#%0)50,( *,"N&50,S0&0;"&52?(P+Q.&n(4-+050;"&52?(#%)($"-("-"N&50,S05"'.'C(e"(5"N0;;"%)(9a*@Q#'")(R>!(K.&S( *#%$"5('"c+"%N.%$(,5.05(&0(eY*(&0("%'+5"(%0(N0%&#;.%#&.0%C( #C( =#%0)50,8( .C( Y0#-8(Ea:LEI:(04(9Ca@9CI( ..C( Y0#-(Ea:LE[:(04(EC:C( QC( P+Q.&8(S.$S(N0%N"%&5#&.0%'LS.$S(c+#%&.&2C(( .C( R#N3.0($0#-8(.)"#-(&0&#-(G=H(.'(oD+$( ..C( <.%X\=($0#-8(oW::%$(K.&S(;.%.;+;(N0%N"%&5#&.0%(04(oD[(%$L+1( NC( Y"-8(:CaJ(#$#50'"($"-?(:C9J(F&35?(5+%(-0K(#%)('-0K(6W:m(d(9CD@E(S0+5'7C(( .C( Y0#-8(1#5$"(Q5.$S&(&0,@S"#M2(';"#5'LQ#%)'(0%($"-(&S#&(#5"(oWIUQ(0%("d&"%)(-#))"5C(( EDC(X4('#;,-"(c+#-.&2(.'('+Q0,&.;#-?(0,&.0%'(405(N-"#%@+,(.%N-+)"8(( )C( 10%$"5("-+&.0%(#&(Wl>C( "C( !","#&(.'0,50,#%0-(,5"N.,.&#&.0%( .C( 3"'&(&0(S#M"('&.%$(M0-+;"(04(9::O1(G=H(#&(-"#'&C(!"#N&.0%('S0+-)(Q"('N#-")(&0( 5"#N&.0%(M0-+;"(6'&(#&('&",(9D7C( 4C( *R!X@Q"#)(Q#'")(N-"#%@+,(