Microbial competition reduces interaction distances to the low µm-range

Metabolic interactions between cells affect microbial community compositions and hence their function in ecosystems. It is well-known that under competition for the exchanged metabolite, concentration gradients constrain the distances over which interactions can occur. However, interaction distances are typically quantified in two-dimensional systems or without accounting for competition or other metabolite-removal, conditions which may not very often match natural ecosystems. We here analyze the impact of cell-to-cell distance on unidirectional cross-feeding in a three-dimensional system with competition for the exchanged metabolite. Effective interaction distances were computed with a reaction-diffusion model and experimentally verified by growing a synthetic consortium of 1 µm-sized metabolite producer, receiver and competitor cells in different spatial structures. We show that receivers cannot interact with producers ∼15 µm away from them, as product concentration gradients flatten close to producer cells. We developed an aggregation protocol and created variants of the receiver cells’ import system, to show that within producer-receiver aggregates even low affinity receiver cells could interact with producers. These results show that competition or other metabolite-removal of a public good in a three-dimensional system reduces the interaction distance to the low micrometer-range, highlighting the importance of concentration gradients as physical constraint for cellular interactions.


Plasmids Description Reference
pSEUDO::P usp45 -gfp Ery r , integration vector, pSEUDO::P usp45sfgfp(Bs) derivative, carrying the gene coding for the green fluorescent protein (Dasher-GFP). [21] chloride. Receiver cells were resuspended in 2 mL 0.225% sodium chloride, producer cells 115 were resuspended in 0.9% sodium chloride and diluted to an OD 600  Louis, MO, USA) and cells, and it was prepared as follows. Pre-cultures were washed with 145 phosphate buffered saline (PBS) and the OD 600 was measured to determine the cell 146 concentration (assuming OD 1 = 10 9 cells/mL). The total cell concentration in the aggregate 147 suspension was determined using flow cytometry (Accuri C6). The producer cell or aggregate 148 concentration in CDM with agarose was set to 2.7·10 6 /mL, the receiver cell concentration to 149 8.9·10 7 cells/mL. 150 300 µL water phase and 700 µL oil phase were mixed using a T10 basic ULTRA TURRAX 151 homogenizer with an S10N-5G dispersing element at 8000 rpm for 5 minutes. Emulsions 152 were subsequently placed on ice for at least 20 minutes, to solidify the agarose beads. After 153 solidification cells could not move and growth therefore resulted in micro-colony formation 154 To compare concentration gradients in two-and three-dimensional systems we made 191 reaction-diffusion models in COMSOL Multiphysics (supplementary information, section 192 4.2). The concentration gradients around a producer cell were calculated either in cube to 193 mimic a three-dimensional system, or in a thin plate to mimic a two-dimensional system 194 (plate thickness of 1.1 µm, roughly matching the producer cell diameter of 1 µm). In both 195 cases the total volume was 1 nL (10 6 cells/mL). The model predicted that in the thin plate the 196 maximal concentration is halved at 24 µm from the producer cell, while in the cube this 197 distance is 0.6 µm ( Figure S5). This indicated that in three-dimensional systems the distances 198 at which cells can interact are significantly shorter than in two-dimensional systems. 199

200
Design of a synthetic consortium and three-dimensional spatial structure for growth 201 To study how concentration gradients constrain interactions between micro-organisms in a 202 three-dimensional environment, we extended the cubic model to contain producer and 203 receiver cells, and analyzed the impact of cell-to-cell distance on the interaction 204 (supplementary information, section 4.3). To experimentally validate the model results we 205 constructed synthetic consortia using four L. lactis strains. 1) A "producer" that takes up 206 lactose and hydrolyzes it intracellularly to glucose and galactose. It was engineered to not 207 metabolize glucose, which was therefore secreted while the cells grew on galactose. 2) A 208 GFP-expressing "receiver" that can take up and grow on glucose, but not lactose. 3) A "non-209 producer" that takes up lactose. It uses both the glucose and galactose moiety for growth, and 210 therefore does not secrete glucose. 4) A "competing glucose-consumer" ( Figure 1A). To co-211 culture these cells in a three-dimensional system, glucose-producers and -receivers (the 212 unidirectional cross-feeders) were encapsulated in solidified agarose beads with an average 213 diameter of ~40 µm. For negative controls, glucose-producers were replaced by glucose-214 "non-producers". Cells were embedded in the beads either as separate cells (~15 µm between 215 cells) or as aggregates (0 µm between cells) ( Figure 1B). During incubation agarose beads 216 were separated either by oil or by CDM ( Figure 1C). Separation by oil prevented diffusion of 217 glucose from beads, enabling us to validate that cells can grow and interact in agarose beads. 218 Separation by CDM resulted in glucose diffusion from beads, enabling us to study 219 unidirectional cross-feeding in presence of a concentration gradient in a three-dimensional 220 system. To investigate the effect of metabolite-removal on the interaction distances, 221 interactions were analyzed in presence and absence of competing glucose-consumers outside 222 the beads ( Figure 1C). 223

226
The four L. lactis strains that were used to make synthetic consortia: 1) "producers" which take up 227 lactose and secrete glucose, 2) "receivers" which take up glucose and express GFP, 3) "non-228 producers" which take up lactose but do not secrete glucose and 4) "competing glucose-consumers" To analyze if we could detect growth in agarose beads we cultured producers and receivers in 238 beads surrounded by oil (no glucose diffusion from beads) and analyzed the beads with flow 239 cytometry before and after incubation. Beads inoculated with mono-cultures of producers 240 (lactose as carbon source) or receivers (glucose as carbon source) showed an increased 241 forward scatter after incubation, indicating that cells could grow inside beads. In agreement 242 with our expectation the fluorescence/scatter ratio of beads with growth was low for beads 243 with producers and high for beads with GFP-expressing receivers ( Figure 2B). 244 To validate that cells could also interact within beads, we made beads such that ~21% of 245 them contained both a producer and receivers and ~79% contained receivers only. Because 246 the metabolic interaction is unidirectional, we expected that in presence of lactose the 247 producers would always grow, while the receivers would only grow when glucose, secreted 248 by producers, was available to them ( Figure 2A). After incubation 14±1% of the agarose 249 beads showed an increased forward scatter, and these beads had a high fluorescence/scatter 250 ratio ( Figure 2B, Table 2). This is close to the expected 21% of beads with producer and 251 receivers, indicating that receivers could only grow in beads with producers. 252 Together this setup forms a synthetic consortium where spatial interactions can be 253 manipulated in a three-dimensional environment, and which allows the detection of growth 254 and interactions using flow cytometry. 255 256 Under glucose competition receivers cannot interact with producers ~15 µm away 257 In the example above glucose could not diffuse from beads and each agarose bead acted as an 258 individual compartment. In contrast, when glucose can diffuse from agarose beads the model 259 predicted that the glucose concentration flattens close to the producer. In that case receivers 260 at a distance of 15 µm from a producer in the same bead are exposed to similar glucose 261 concentrations as receivers in beads without a producer ( Figure 2C). If this prediction was 262 correct, we expected that most receivers can grow when the global glucose concentration 263 builds up, while in case of glucose competition the glucose concentration stays low and even 264 receivers 15 µm away from a producer in the same bead should not be able to grow. We 265 therefore incubated agarose beads in CDM, which allows glucose diffusion from beads. 266 Without competing glucose-consumers in the CDM outside the beads 75±7% of the beads 267 showed growth and these beads had a high fluorescence/scatter ratio ( Figure 2D, Table 2). 268 This indicates growth of both receivers with and receivers without a producer in their bead. 269 When we took the same beads but added competing glucose-consumers outside the beads, 270 only 15±3% of the beads contained growth. In the beads where growth was observed the 271 fluorescence/scatter ratio was low, indicating that only producers grew ( Figure 2D, Table 2). 272 These results are consistent with the model predictions and show that under glucose 273 competition receivers cannot interact with producers even if they are only ~15 µm away. 274 although the model predicted that all receivers could grow ( Figure 2C and 2D). These beads 277 could be false negatives caused by our conservative gating strategy, or by empty beads with 278 single fluorescent cells attached to their outside. Conversely, for the beads gated as "growth", 279 we observed an increased fluorescence/scatter ratio compared to the single receiver controls 280 ( Figure 2D). It is known that fluorescence of individual cells increases with decreasing 281 growth rate [32,33], suggesting that in co-cultures the higher fluorescence/scatter ratio could 282 be caused by glucose limited and therefore slower growth of the receivers in the beads. 283 Together this data shows that competition for glucose in a three-dimensional environment 284 prevents interactions at ~15 µm distance, because the presence of competing public good-285 consumers leads to steep concentration gradients. 286 287

Aggregated producers and receivers interact even under glucose competition 288
In the presence of steep concentration gradients microbial interactions might be facilitated by 289 bringing producers and receivers in close proximity. Consistently, the model predicted that 290 cell aggregation would allow receivers to grow under glucose competition ( Figure 2E). We 291 developed a protocol to make producer-receiver aggregates. Defined aggregates were formed 292 by adding positively charged producers to an excess of negatively charged receivers, ensuring 293 that producers were directly surrounded by receivers. In this way we obtained a mixture of 294 single receivers and aggregates of one producer and approximately eight receivers ( Figure  295 1B). We aimed to add an aggregate to ~21% of the beads, but we could only roughly estimate 296 the aggregate concentration in the mixture based on the added amount of positively charged 297 cells. However, underestimating this percentage would not affect the results, as we only 298 analyze agarose beads with growth after incubation (supplementary information, section 3). 299 after incubation we saw an increased scatter in 3±2% of the beads ( Figure 2F, Table 2),  301 indicating only growth in beads with both producers and receivers. The fluorescence/scatter 302 ratio of beads with growth was increased compared to the producer mono-culture ( Figure 2F), 303 indicating growth of both producers and receivers. Beads with grown aggregates of non-304 producers and receivers showed a fluorescence/scatter ratio similar to the producer mono-305 culture ( Figure 2F), indicating that the aggregation protocol or the data analysis procedure did 306 not influence the readout. 307 Therefore, the results show that cell aggregation facilitates microbial interactions, even in a 308 three-dimensional system with competition for the public good. 309 310  Research had no role in the study design, data collection and analysis, decision to publish, or 507 preparation of the manuscript. 508

Section 1: Agarose bead size-and volume-distributions 511
We prepared agarose beads surrounded by oil and made pictures with a microscope (9 per 512 emulsion, Figure S1A shows an example). Pictures were subsequently analyzed with ImageJ 513 to identify the beads (Figure S1B), and to measure size-and volume distributions ( Figure  514 S1C and S1D). Small droplets were not always identified, but as they contain only little 515 volume this only marginally affects the analysis. Beads on the edge of the picture were 516 excluded from the analysis. Formed emulsions were polydisperse but distributions of 517 replicates were reproducible, with mean volume ± SEM of 26±2 pL (diameter of 37 µm). To establish if addition of 10 9 L. lactis MG1363 cells per mL outside agarose beads 544 prevented cross-talk between beads, we mixed beads with producers and beads with receivers 545 and incubated them in presence of lactose in different spatial structures ( Figure S2). After 546 incubation surrounded by oil only producers were grown, which was expected as glucose 547 could not diffuse from beads. When glucose could diffuse from beads and no L. lactis 548 MG1363 cells were added outside the beads, both producers and receivers grew. However, in 549 presence of 10 9 L. lactis MG1363 cells per mL outside the agarose beads only producers 550 grew, suggesting that glucose leaving beads with producers was mainly consumed by 551 L. lactis MG1363 cells outside the beads and did not reach receivers in neighboring beads. 552 The glucose concentration outside the beads probably did not exceed the low micro-molar 553 range, as the K m for glucose of the highest affinity transporter in L. lactis MG1363 is 13 µM 554 Reaction. The net glucose rate r s (mol/m 3 /s) results as the difference between production and 667 consumption at a certain position in space, r s = q p C x -q s C x . The specific glucose production 668 rate (q p ) of L. lactis NZ9000 Glc-Lac+ is the same as its specific lactose uptake rate, as each 669 lactose molecule contains one glucose molecule. The q p was therefore set to a constant value 670 of 1 molP/CmolX/h [47] and applied within the producer cells. Simulations which did include 671 the lactose concentration and Monod kinetics for lactose consumption yielded similar results 672 as simulations with a constant q p , therefore we adopted the simpler constant rate. For 673 receivers the glucose uptake was assumed with a saturation (Monod) kinetics, 674 q s = q s max · C s /(K s + C s ). We used the K s of the highest affinity glucose transporter of L. lactis

Predicted concentration gradients in two-and three-dimensional reaction-diffusion 693
systems 694 To analyze the difference in concentration gradients in two-and three-dimensional systems 695 we modelled production by one producer cell in two different geometries, as represented in 696 Figure S5. In the two-dimensional system the model predicts that the product concentration is  Figure S6 shows the predicted glucose concentration ( Figure S6A) and glucose production 725 rate ( Figure S6B) over a plane crossing the producer cell and four of the eight receiver cells. 726 Profiles for aggregated cells and for cells ~15 µm away from each other are shown. Because 727 it is difficult to know the actual q p and q s max inside agarose beads, we also performed a 728 sensitivity analysis ( Figure S7). A 5-fold change in q p and q s max resulted in similar 729 concentration gradients and it did not affect our hypotheses. 730 NZ9000_GFP_glcU the q s max was reduced with a factor four, which reflects the differences in 786 V max of the transporters [22]. We calculated the glucose uptake for the different mutants after 787 5 hours in presence of competing glucose-consumers, without considering growth of the cells 788 ( Figure S8). The effective diffusion coefficient (D eff,s ) varies from 10-70%, depending on the 789 density of the micro-colony [29,31]. Figure S7 shows the glucose uptake when D eff,s is 10%, 790 30% and 70%. We included a sensitivity analysis for five-fold changes in q p and q s max values, 791 which all showed similar trends as the reference (1x q p and 1x q s Strains were incubated in CDM + 0.2 wt% glucose in a 96-well plate. The OD 600 was 817 measured every six minutes for 24 hours using a SPECTRAmax 384 plus plate reader 818 (Molecular Devices, San Jose, CA, USA). OD 600 measurements were background corrected, 819 ln-transformed and the slope of the region with exponential growth was calculated as the 820 growth rate (Table S2). 821 822 Table S2. Growth rates of receivers with different glucose transporters (n=22).

affinities and V max 826
We constitutively expressed GFP in three previously constructed L. lactis NZ9000 mutants 827 with a single glucose transporter [27], and analyzed their growth in different spatial 828 structures. This experiment focused on beads incubated in CDM (allowing glucose diffusion 829 from beads), as we expected that under these conditions the transporter characteristics of 830 receivers would be important. Figure S9A shows the experimental results when receivers 831 were ~15 µm from a producer within the same bead and incubated in CDM, whereas in 832 Figure S9B the beads were incubated in medium with 10 9 glucose-consumers per mL. Figure  833 S9C shows the experimental results of producer-receiver aggregates, incubated in CDM with 834 10 9 glucose-consumers per mL. Without a competing glucose-consumers we observed 835 growth of both receivers with and receivers without a producer in their bead ( Figure S9A), 836 while with competing glucose-consumers only producers could grow ( Figure S9B). In 837 producer-receiver aggregates receivers were able to grow, despite the presence of competing

931
Aggregates of producers and receivers, incubated in beads surrounded by CDM with 10 9 glucose-932 consumers per mL.