Bacterial response to spatial gradients of algal-derived nutrients in a porous microplate

Acrylic and polydimethylsiloxane (PDMS) molds preparation

The incubating device for the porous microplate was designed using a CAD software (Solidworks, Dassault Systèmes) and the exported drawing files were used to laser cut 1/4” and 1/8” acrylic sheet (Universal Laser Systems; Supplementary Fig. S2). After washing the cut acrylic parts with deionized water, they were attached by acrylic (Weld-On) and epoxy (3 M) adhesives that were followed by a curing process for ~18 h. Polydimethylsiloxane (PDMS) (Sylgard 184, Dow Corning) was cast onto the acrylic mold and cured at 80 °C for at least 3 h. The PDMS mold was carefully detached from the acrylic surface by dispensing isopropyl alcohol (VWR) into the area between the PDMS and the acrylic molds (Fig. 2a).

Porous microplate preparation

Synthesis of copolymer HEMA–EDMA was based on previously described protocols [30, 31] and details are given as follows. Prepolymer solution HEMA − EDMA was prepared by mixing 2-hydroxyethyl methacrylate (HEMA; monomer, 24 wt.%, Sigma-Aldrich), ethylene glycol dimethacrylate (EDMA; crosslinker, 16 wt.%, Sigma-Aldrich), 1-decanol (porogen, 12 wt.%, Sigma-Aldrich), cyclohexanol (porogen, 48 wt.%, Sigma-Aldrich) and 2,2-dimethoxy-2-phenylacetophenone (DMPAP; photoinitiator, 1 wt.%). The solution was stored at room temperature without light exposure until further use. Glass slides (75 × 50 mm², VWR) were chemically cleaned by sequentially soaking in 1 M hydrochloric acid and 1 M sodium hydroxide for one hour, followed by rinsing with deionized water and air drying. The prepolymer solution was cast onto the PDMS mold and a glass slide was placed on the mold. The solution was then polymerized under ultraviolet light with a wavelength 365 nm for 15 min by using a commercial UV lamp (VWR). The photopolymerized device was detached from the PDMS mold and stored in a jar containing methanol (VWR) until further use (Fig. 2a). The jar was refilled with new methanol twice in order to remove the remaining porogen and uncrosslinked monomers from the hydrogel.

Upon each incubation experiment with the porous microplate, each device was decontaminated by replacing the solvent with 70% alcohol (VWR) and storing it for 24 h. They were immersed in a pre-autoclaved jar for two weeks with f/2 medium with omitted silicate, where the jar was refilled once with a new sterile medium to adjust its pH for the algal culture and remove any solvent remaining in the hydrogel. Before inoculating microbial cells, each microplate was taken out from the jar and the media remaining on the top surface was removed by absorbing it with a pre-sterilized wipe to minimize the chance for cross-contamination between wells (Fig. 2b).

Scanning electron microscopy

Photopolymerized HEMA − EDMA was removed from methanol and dried in air for at least one week to evaporate the excess solvent. A ~5 × 5 mm² specimen was collected from the dried copolymer and attached to a pin stub. The stub was loaded on a scanning electron microscope (SEM; MERLIN, Carl Zeiss), and the specimen was characterized with imaging software (SmartSEM, Carl Zeiss) with 16,270X magnification and an operating voltage of 1 kV. The SEM imaging was performed at the Electron Microscopy Facility in the MIT Materials Research Science and Engineering Centers (MRSEC; Fig. 2c).

Strains and culturing conditions

Axenic P. tricornutum CCMP 2561 was acquired from the National Center for Marine Algae and Microbiota (NCMA) and shown to be axenic via epifluorescence microscopy and sequencing of the 16 S rRNA gene [11]. P. tricornutum was maintained in f/2 medium with 20 g L⁻¹ commercially available sea salts (Instant Ocean, Blacksburg) and with omitted silicate, which we will refer to as f/2-Si [11, 16]. Batch cultures were grown at 20 °C with a 12 h light/12 h dark diurnal cycle and a light intensity of 200 μmol photons m⁻² s⁻¹ (Exlenvce). Every 2–3 weeks, axenic cultures were monitored for bacterial contamination by streaking culture samples on marine broth agar [33], that tests for contamination by bacteria that can grow on agar media and is not definitive. Every 6–12 months, every axenic and bacterial co-culture of P. tricornutum was inspected for the absence/presence of bacteria by staining the cellular DNA with 0.1% v/v SYTO BC Green Fluorescent Acid Stain (Thermofisher, Supplementary Fig. S1).

Bacterial community samples (referred to as “phycosphere enrichments”) were obtained from mesocosms of P. tricornutum and maintained as previously described [11, 16]. Briefly, an outdoor P. tricornutum mesocosm sample in natural seawater was collected in Corpus Christi, TX and filtered with 0.6–1 µm pores to remove larger algal cells. The bacterial filtrates were inoculated to an axenic algal culture, maintained in f/2-Si media for ~3 months, and washed with a sterile medium to enrich for phycosphere-associated bacteria. These enriched communities were subsequently co-cultured with P. tricornutum in f/2-Si media for ~4 years prior to the start of the experiments.

Two bacterial strains, Marinobacter sp. 3-2 and Algoriphagus sp. ARW1R1, were isolated from the phycosphere enrichment samples (Supplementary Table S1). The isolates were either maintained by growing on marine broth agar plates at 30 °C or by co-culturing with P. tricornutum through inoculation of a single colony into the axenic culture.

P. tricornutum culture in porous microplate

Three baseline experiments were designed to study how the alga P. tricornutum interacts with its associated bacteria in the porous microplate (Fig. 1). For experiments assessing the algal growth in the microplate, axenic P. tricornutum was acclimated to a copolymer environment in advance by inoculating a stationary phase-culture to a separate microplate. After acclimation for 4 days, the culture was diluted to ~1 × 10⁶ cells ml⁻¹ and transferred to the experimental microplate. Three replicated microplates were placed in a single transparent covered container (128 × 85 × 10 mm³, VWR) which was filled with ~25 ml f/2-Si medium to keep the microplate hydrated throughout the incubation period of 20 days with an initial culture volume of 75 µl (Fig. 1a). The procedures were conducted under a biosafety cabinet to prevent any biological contamination. The cells were incubated under the same conditions as described above for the batch cultures (temperature, light intensity, diurnal cycle).

Growth of P. tricornutum was measured by counting cells using a hemocytometer (Electron Microscopy Sciences) or flow cytometry (described later). Specific growth rates were calculated from the natural log of the cell densities in triplicate sampled during an exponential growth phase (day 3 for the batch culture, day 5 for the porous microplate system; Fig. 3a).

Algal-bacterial isolates experiment

Marinobacter sp. 3-2 and Algoriphagus sp. ARW1R1 were prepared by inoculating a single colony into marine broth and growing overnight (30 °C, 150 r.p.m.). The cells were washed with f/2-Si twice by centrifuging at 2258 rcf for 4 min and diluted to a density of ~4 × 10⁶ cells ml⁻¹ (Marinobacter sp. 3-2) and ~4 × 10⁵ cells ml⁻¹ (Algoriphagus sp. ARW1R1). The cell densities were initially set to resemble in situ conditions with P. tricornutum-associated bacterial communities where Marinobacter displayed relative abundances several fold higher than Algoriphagus (Supplementary Note S1) [16]. The diluted bacterial cultures were transferred to the surrounding wells of the experimental microplate with pre-acclimated P. tricornutum with a density of ~1 × 10⁷ cells ml⁻¹ in the center well, as described above (Fig. 1c).

Every 2–3 days, 5 µl samples were collected from wells and transferred to a 96-well plate (Corning). Before collecting the samples, the culture was thoroughly mixed using a micropipette to avoid sampling bias resulting from any spatial heterogeneity. Cells were then diluted 35 times with 1 x phosphate-buffered saline solution (PBS) (Mallinckrodt), followed by the addition of 16% formaldehyde (Thermo Scientific) to a final concentration of 2% adjusting to a final volume of 200 µl. Fixed samples were stored at 4 °C for no more than 4 weeks. After collecting the cell samples for 20 days, the remaining cultures were taken out from the device and were streaked on marine broth agar plates to check if any of the culture wells had been cross-contaminated, based on different colonies morphologies (Supplementary Fig. S3). We detected cross-contamination in 5 out of 108 bacterial samples, and they were excluded from the analysis. No less than three replicates were retained after excluding the cross-contaminated samples.

Flow cytometry

Fluorescent counting beads for flow cytometry (Alignflow^TM, Thermofisher) were diluted 20 times with PBS and 50 µl was added to each well of the 96-well plate for calibration. SYBR Green I nucleic acid stain (Thermofisher) was added to the samples with a final concentration of 0.1% v/v, and allowed to sit at room temperature at least for 30 min without light exposure.

Flow cytometry analysis was conducted on a BD FACS Canto II HTS to quantify the number of algal and bacterial cells with parameter setting as follows: (voltage) FSC = 580, SSC = 370, GFP = 400, PE = 330, PerCP = 647, PE-Cy7 = 677, Alexa Fluor 680 = 290, APC-Cy7 = 410, Pacific Blue = 440, AmCyan = 539. Populations were plotted with GFP-A and APC-Cy7-A intensities, allowing a clear distinction between counting beads, stained algal, and bacterial cells. The data were exported to.csv files using FlowJo (BD) and converted into cell density data using MATLAB (Mathworks).

Bacterial community experiment

To examine bacterial community structure changes induced by the presence of the diatom, ~1 × 10⁶ P. tricornutum cells ml⁻¹ were placed in the center well and the results were compared to control incubations where sterile media was placed in the center (Fig. 1d). Phycosphere enrichment samples [11, 16] (with the algal cells removed with a 0.8 μm filter) were used as a bacterial community inoculant where 100 µl samples were placed in the surrounding wells in the array. Layer 1 and Layer 2 refer to bacterial wells with a distance from the centre well of 8 mm and 16 mm, respectively. On the outmost well, 100 µl sterile f/2-Si medium was inoculated as a blank control for 16 S rRNA community analysis.

The devices were incubated in a container (GasPak™ EZ container systems, BD) which was filled with f/2-Si medium. All preparation steps were performed in a laminar flow hood to prevent any bacterial contamination. The container was incubated under the same condition as described above. After 7 days, each well sample was collected and filtered into 96-well filter plate (0.2 μm pore size, Pall AcroPrep) to remove liquid, and was stored at −20 °C until further analysis.

16 S ribosomal RNA gene amplification and sequencing

Twenty microliters of sterile DNA-free water were added to each well of the filter plates, and cells were lysed to release the DNA with heat (95 °C for 10 min). The 16 S rRNA gene was directly amplified without cleanup, using primers targeting the v4 region and modified to include Illumina platform adaptor sequence on the 5′ ends. The PCR contained 10 μl of 5 Prime MasterMix, 1 μl of 10 μM of forward 16 S primer (5′ – TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG[GTGYCAGCMGCCGCGGTAA] − 3′) [34], 1 µl of reverse 16 S primer (5′ – GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG[GGACTACNVGGGTWTCTAAT] -3′) [35], and 5 μl of DNA template (actual primer sequences highlighted with brackets). Cycling conditions were as follows: denaturation at 94 °C for 3 min, followed by 30 cycles of denaturation at 94 °C for 45 s, annealing at 51 °C for 30 s and extension at 72 °C for 1.0 min. The final extension was conducted at 72 °C for 10 min and the samples were held at 4 °C. The second round of amplification added Illumina Dual Nextera XT indexes and sequencing adaptors [36]. Each library was quantified with a Qubit broad range dsDNA assay and equimolar amounts of each library were pooled. The size and concentration of the final pool were verified by a D5000 High Sensitivity assay on the Agilent Tapestation. Six pM of the pooled libraries were combined with 15% phiX and paired-end sequenced on an Illumina MiSeq for 500 cycles.

Paired-end MiSeq reads were filtered to remove read pairs that contained Illumina adapter or barcode kmers with bbduk v38.22 using options k = 31 and hdist = 1 [37]. Read trimming and quality filtering was done with DADA2 v1.12.1 using options trimLeft = c(4, 14), truncLen = c(200,150), maxN = 0, maxEE = c(2, 2), truncQ = 2 [38]. Read pairs remaining after filtering were further processed into amplicon sequence variants (ASVs) retaining ASVs of length 274 nucleotides with at least 2 reads in at least 2 samples. The sequences were then aligned with MUSCLE v3.8.1551 [39] and clustered into an approximately-maximum-likelihood tree with Fasttree v2.1.10 [40]. Taxonomy was assigned using IDTAXA [41] against the Silva version 132 SSU database [42]. ASVs were analyzed using the R package phyloseq v1.30.0 [43]. Samples were quality-filtered by removing taxa not found at least 10 times in at least 3 samples and normalized by Cumulative Sum Scaling (CSS) method using the R package metagenomeSeq v1.28.2 [44]. In brief, a mean of 95,770 (from 11,535 to 241,656) and 1477 (from 61 to 5183) read pairs were sequenced per sample of bacterial wells and of quality-filtered negative controls, respectively (Supplementary Table S2). Quality-filtered negative samples have been selected from 18 total negative samples based on their read pairs of less than 10,000. The other five negative control samples with reading pairs of more than 10,000, were dropped from the analysis as they suggested that cross-contamination had occurred from the adjacent bacterial communities for these samples. Another 36 bacterial cultures, located in between layers 1 and 2, were not assigned with a layer number and excluded from the statistical analysis. R package phyloseq v1.30.0 was used to perform principal coordinate analysis (PCoA), vegan v2.5.6 to perform permutational multivariate analysis of variance (PERMANOVA) [45] based on weighted Unifrac [46], and R function (aov) was used to perform one-way and two-way analysis of variance (ANOVA) tests using packages tidyverse v1.3.0 [47], vegan v2.5.6 [48].

Source: Ecology - nature.com