The Cacti Microbiome: Interplay between Habitat-Filtering and Host-Specificity

Cactaceae represents one of the most species-rich families of succulent plants native to arid and semi-arid ecosystems, yet the associations Cacti establish with microorganisms and the rules governing microbial community assembly remain poorly understood. We analyzed the composition, diversity, and factors influencing above- and below-ground bacterial, archaeal, and fungal communities associated with two native and sympatric Cacti species: Myrtillocactus geometrizans and Opuntia robusta. Phylogenetic profiling showed that the composition and assembly of microbial communities associated with Cacti were primarily influenced by the plant compartment; plant species, site, and season played only a minor role. Remarkably, bacterial, and archaeal diversity was higher in the phyllosphere than in the rhizosphere of Cacti, while the opposite was true for fungi. Semi-arid soils exhibited the highest levels of microbial diversity whereas the stem endosphere the lowest. Despite their taxonomic distance, M. geometrizans and O. robusta shared most microbial taxa in all analyzed compartments. Influence of the plant host did only play a larger role in the fungal communities of the stem endosphere. These results suggest that fungi establish specific interactions with their host plant inside the stem, whereas microbial communities in the other plant compartments may play similar functional roles in these two species. Biochemical and molecular characterization of seed-borne bacteria of Cacti supports the idea that these microbial symbionts may be vertically inherited and could promote plant growth and drought tolerance for the fitness of the Cacti holobiont. We envision this knowledge will help improve and sustain agriculture in arid and semi-arid regions of the world.

after 7 days of incubation at room temperature (25°C). After surface disinfection, stem tissues were divided in three sections (basal, medium and apical) for their independent dissection. From each section, stem fragments of ~1 cm 3 were cut and placed in a 20% glycerol solution and stored at -80 °C. Root tissues were also dissected in small fragments of 1 cm length under sterile conditions and stored with 20% glycerol at -80 °C.
DNA extraction. DNA extractions were performed using different commercial kits and traditional protocols depending on the nature of each sample. DNA extractions from thawed bulk soil, root-zone soil and rhizosphere samples were done with the MoBioTM PowerSoil DNA Isolation Kit (MoBio Laboratories, Carlsbad, USA) because of its ability to remove humic acids and other compounds that inhibit the PCR reaction. For the phyllosphere, thawed samples were processed using the MoBioTM PowerWater DNA Isolation Kit. For the root endosphere samples, 100 mg of each frozen sample was grounded to a powder in liquid nitrogen with a sterile mortar and pestle and was then prepped following a CTAB genomic DNA extraction protocol by Edwards, (2001) with the following lysis solution (100mM Tris pH8, 20mM EDTA, 1.4 M NaCl, 2% Hexadecyltrimethylammonium bromide, 0.2% Beta mercaptoethanol). For the stem endosphere samples, 100 mg each of basal, medium and basal frozen samples were grounded to a powder in liquid nitrogen with a sterile mortar and pestle and DNA extraction was performed following the protocol by Lopes et al., (1995). Equimolar amounts of DNA obtained from the apical, medium and basal stem sections were pooled in a composite stem endosphere sample for the next steps. Finally, composite samples of each treatment (144 samples) were prepared by mixing equal amounts of DNA from each biological replicate to yield 48 pooled samples. DNA quantity and quality were assessed along the process with a Nanodrop ND-1000 (Thermo scientific, Wilmington, USA) spectrophotometer.
PCR amplification and sequencing process. The PCR amplification and sequencing processes were performed at the Joint Genome Institute. The microbial molecular markers employed in this study were the 16S rRNA V4 for bacterial/archaeal communities and the ITS2 for fungal communities. We employed the 515F (5'-GTGCCAGCMGCCGCGGTAA-3') and 816R (5'-GGACTACHVGGGTWTCTAAT-3') primer sets for the 16S rRNA V4 amplification. For ITS2 amplification, we used the ITS9F (5'-GAACGCAGCRAAIIGYGA-3') and ITS4R (5'-TCCTCCGCTTATTGATATGC-3') primer sets. In each PCR amplification using primers 515F-816R, we used PNA clamps to reduce chloroplast and mitochondrial contamination and increase the number of bacterial/archaeal reads from the root endosphere, stem endosphere and phyllosphere samples as described in Lundberg et al., (2013). The amplification reactions were performed in triplicate using ~10 ng of pool DNA template per reaction and were conducted in 96-well plate format. In order to determine the OTU measurability thresholds, we employed four negative controls and several technical replicates in each 96-well plate as described in Coleman-Derr et al., (2016). The PCR conditions used were 94 °C for 3 min, followed by 30 cycles of 94 °C for 45s, 78 °C for 10s, 50 °C for 60s, and 72°C for 90s, and final extension to 72 °C for 10 min. Triplicate reactions from each sample were pooled and quantified with the Qubit High Sensitivity Assay kit (Life Technologies) on a Turner Biosystems fluorescence plate reader (Promega, Madison, WI, USA). Sets of 96 barcoded PCR products pooled in equimolar ratios and cleaned up using the AMPureXP magnetic beads (Beckman-Coulter, Indianapolis, IN, USA). Paired-end 2 x 250bp sequencing of the barcoded amplicons was performed on a MiSeq machine running v2 chemistry (Illumina Inc, San Diego CA, USA). Samples from Cacti were processed together with the samples derived from Agave species reported by Coleman-Derr et al., (2016) .
Data processing. The raw Fastq reads generated from Cacti samples were processed using a custom pipeline developed at the Joint Genome Institute. First, Illumina adapter sequences and PhiX sequencing control was removed from raw reads. Primer sequences were trimmed from the 5' end and low-quality bases were trimmed from the 3' end of reads prior to assembly of read1 and read2 with either FLASH (Magoč and Salzberg, 2011) for 16S data or Pandaseq (Masella et al., 2012) for ITS2 data. After that, 1,311,071 bacterial/archaeal and 2,752,860 fungal high-quality merged reads were obtained ( Figure  S3). These reads were clustered using the UPARSE pipeline (Edgar, 2013) to yield 40,759 bacterial/archaeal and 25,871 fungal OTUs at 97 and 95% identity, respectively. While 97% cut-off is generally sufficient for properly identifying fungal species using either ITS1 or ITS2, some species are split into two or more groups at this level of clustering (Blaalid et al., 2013). The JGI uses 95% for all its fungal ITS2 analyses, which will keep more fungal species from being split into two or more units, but at the expense of some species clustering together with other species.
Taxonomic lineage were assigned to each OTU using the RDP Naïve Bayesian Classifier (Wang et al., 2007) with custom reference databases. For the 16S V4 data, this database was compiled from the May 2013 version of the GreenGenes 16S database (DeSantis et al., 2006), the Silva 16S database (Quast et al., 2013) and additional manually-curated 16S sequences, trimmed to the V4 region. For the ITS2 data, this database was built from the UNITE database (Koljalg, 2013). After that the taxonomies were assigned to each OTU, we discarded some OTUs considering the following criteria: (1) OTUs that were not assigned a Kingdom level RDP classification score of at least 0.5, (2) OTUs that were not assigned to Kingdom Bacteria or Archaea for the 16S datasets, and (3) all OTUs that were not assigned to Kingdom Fungi for ITS2 datasets. Using several technical replicates included on each sequenced 96-well plate, we calculated a threshold for determining technical reproducibility for these OTUs using the progressive drop-out analysis described in Lundberg et al., (2013) (Coleman-Derr et al., 2016 Figure S2). This threshold determines the minimum cumulative read count across all plates for an OTU to be included in the analysis, and was determined to be at least 7 reads in at least 5 samples for the 16S dataset and at least 2 reads in at least 5 samples for the ITS2 dataset. These criteria yielded 4,012 and 3,541 measurable OTUs for bacteria/archaea and fungi, respectively for most downstream analyses (these OTUs are considered as the measurable OTUs and were deposited under Accession Numbers KU536055 -KU539595). Only for diversity analyses and in order to account for differences in sequencing read depth across samples, samples were randomly subsampled (rarefied) to 275 and 375 reads per sample in the bacterial/archaeal and fungal datasets, respectively. OTU measurable tables and metadata can be found as Supplementary Material (Data Sheet 1). All Illumina sequences related to this project were deposited under SRA accession: SRP068631.

In vitro biochemical characterization of seed-borne bacteria associated with M. geometrizans and O. robusta
In vitro tolerance to water stress Growth capacity under water reduction. Bacterial cells were grown overnight or up to one day in Tryptone Soya Broth (TSB) at 28 °C and 150 rpm. Optical density of cultures was then measured at 600 nm (OD 600nm ) and adjusted to 1.0 in all cases. Subsequently, 10 µl of serial dilutions of 10 2 , 10 4 , 10 6 and 10 8 of adjusted cultures were poured in 10% TSA plates containing 405 g/l, 202.5 g/l, 101.25 g/l and 0 g/l (control) of sorbitol. Finally, the plates were incubated at 28 °C for a period of five days and the growth was monitored every 24 hrs (Kavamura et al., 2013).

Exopolysaccharide staining with alcian blue.
A loop of cells growing on the exopolysaccharides selective media was taken and suspended in 1 ml of a solution of Alcian Blue (alcian blue and acetic acid, 1:8 (v/v)). The suspension was mixed and incubated one hour at room temperature. Centrifugation followed at 12 krpm for 5 min; the supernatant discarded and the pellet washed with 200 µl of sterile distilled water. A second round of centrifugation took place at 10 krpm for 5 min; the supernatant removed and the pellet resuspended in 20 µl of sterile distilled water. 10 µl of the staining was observed under a light microscope at a magnification of 100X (Leica DM 750, Leica Microsystems).
The alcian blue dye binds to the carboxyl groups of the polysaccharides without binding to bacterial cell walls, allowing a clear differentiation (Paulo et al., 2012).

Plant growth promotion traits by direct mechanisms
Nitrogen fixation capacity in solid and semi-solid media. 10 µl of serial dilutions of 10 2 , 10 4 , 10 6 and 10 8 of grown and adjusted cultures (OD 600nm = 1.0) were inoculated in Winogradsky media at pH 7 (100 ml of Winodradsky salts (5 g of K 2 HPO 4 , 2.5 g of MgSO 4 ·7H 2 O, 2.5 g of NaCl, 0.05 g of Fe(SO 4 ) 3 and 0.05 g of MnS0 4 for gauging one liter), 0.5 g of CaCO3, 0.5% of bromothymol blue (as an indicator of pH), 10 g of carbon source and 8 g of agar to one liter of medium) with different carbon sources (glucose, sucrose and sucrose/mannitol). We used growth on TSA as control for each strain. Nitrogen fixation is characterized by the formation of halo with a shift from blue-green coloration to yellow, this due to acidification of the culture medium. The size and coloration was measured and recorded every 24 hrs for a period of 5 days (Hardy et al., 1968). Semi-solid media is exactly as above reported, but agar concentration was lowered to 2 g per liter instead of 8. Evaluation in semi-solid media allowed the determination of the oxygen requirements of evaluated strains. As in solid medium, nitrogen fixation is characterized by the formation of a color shift from blue-green to yellow. Coloration and position of bacterial growth (superficial, middle or bottom of the tube) were monitored every 24 hrs for a period of three days (Hardy et al., 1968).
Indole acetic acid (IAA) production. 100 µl of grown and adjusted bacterial cultures (OD 600nm = 1.0) were inoculated into 10 ml of 10% TSB media supplemented with 5 mM Ltryptophan and incubated in dark at 28 °C for a period of 48 hrs. Then, the cultures were centrifuged at 10 krpm for 5 min and 750 µl of the supernatant were treated with 750 µl of Salkowski reagent (50 ml of perchloric acid (35%) and 1 ml of FeCl 3 (0.5 M)) for 30 min under protection from light. As negative control, 750 µl of culture without bacterial inoculation was used. Finally, the optical density was measured in a spectrophotometer at 530 nm (Beckman DU640). The production of a pink to red color indicates the production of IAA (Gordon and Weber, 1951). We developed a calibration curve with commercial IAA using concentrations of 3-50 ng/ml. Phosphate solubilizing capacity. We employed the colorimetric method Nautiyal (1999) with some modifications in order to detect available phosphate. 100 µl of grown and adjusted bacterial cultures (OD 600nm = 1.0) were incubated in 10 ml of NBRIP (National Botanical Research Institute's Phosphate) media (1% glucose, 0.5% Ca 3 (PO 4 ) 2 , 0.5% MgCl 2 , 0.02% KCl, 0.025% MgSO 4 ·7H 2 O and 0.01% (NH 4 ) 2 SO 4 ) at 28 °C, 180 rpm for 15 days. Afterwards, 1 ml of each sample was taken, transferred to 1.5 ml tubes and centrifuged at 10 krpm for 5 min. Then, 145 µl of supernatant was collected and addition of 570 µl of sterile distillated water and 285 µl of molybdate-vanadate (5% ammonium molybdate, 0.25% ammonium vanadate, 1:1 (v/v)) followed. The mixture was incubated for 10 min at room temperature. As negative control, we used uninoculated NBRIP media. Finally, the optical density was measured in a spectrophotometer at 420 nm (Beckman DU640). We developed a calibration curve with KH 2 PO 4 using concentrations of 50-350 ng/ml. Siderophores detection. 10 µl of grown and adjusted bacterial cultures (OD 600nm = 1.0) were inoculated into CAS media and incubated at 28 °C until growth could be observed. The CAS medium uses a ferrichrome complex which changes color as a result of iron loss. Siderophores that have more affinity for iron chromogen can capture the ferry ironchromogen complex, turning dye color from blue to yellow-orange (De los Santos-Villalobos et al., 2012).

Plant growth promotion traits by indirect mechanisms
Ammonia production. 100 µl of grown and adjusted bacterial cultures (OD 600nm = 1.0) were inoculated in 10 ml peptone media and incubated for 48 hrs at 28 °C. Then, from each culture 1 ml was taken and transferred to a 1.5 ml microcentrifuge tube and 50 µl of reagent Nessler (10% HgI 2 , 7% KI and 50% NaOH (32%)) were added. Mixtures developing a light yellow color indicate a small amount of ammonia production, whereas a strong yellow to brown color indicate a high production of ammonia together with a precipitate resulting from the reaction of ammonia with mercuric iodide (Cappuccino and Sherman, 1992).

Production of hydrogen cyanide (HCN).
The strains were streaked on TSA media supplemented with 10% glycine (4.4 g/l) and incubated at 28 °C. After 24 hrs of incubation and under sterile conditions, a sterile plate-sized filter paper soaked with a solution of 0.5% picric acid and 2% Na 2 CO 3 was placed in each petri dish. The plates were sealed and incubated with the filter paper up to 48 hrs. Hydrogen cyanide production is considered positive if paper filter turns orange to red-brown. Color and cell growth were monitored every 24 hrs (Bakker and Schippers, 1987).
Capacity cellulose hydrolysis. 20 µl of grown and adjusted bacterial cultures (OD 600nm = 1.0) were adsorbed on filter paper discs placed in Mandels media (2.0 g of KH 2 PO 4 , 0.4 g of CaCl 2 , 5.0 mg of FeSO 4 ·7H 2 O, 1.4 g of (NH 4 ) 2 SO 4 , 0.3 g urea, 0.3 g of MgSO 4 ·7H 2 O, 1.6 mg of MnSO 4 ·H 2 O, 1.4 g of ZnSO 4 ·7H 2 O, 20 mg of CoCl 2 , 1.0 g Tween 80, 25.0 g of peptone, 1.0 g of carboxymethylcellulose and 12g of agar). Then, the plates were incubated at 28 °C until radial growth could be observed (Teather and Wood, 1982). Congo red staining was performed at 1% (w/v) trying to cover the entire plate. After 15 min, excess dye was removed and 1M NaCl was added for 15 min. Finally, the presence of halos of hydrolysis was analyzed and recorded. Figure S1. Examples of stems, roots and soils recovered from M. geometrizans and O. robusta in the field.

Suplementary Figures
. d. OTUs with 98% homology with respect to isolated strains from the two Cacti species and in the process of genome sequencing.