SARS-CoV-2 burden on the floor was associated with COVID-19 cases and outbreaks in two acute care hospitals: a prospective cohort study.

BACKGROUND: Recent work demonstrated that detection of SARS-CoV-2 on the floor of long-term care facilities is associated with impending COVID-19 outbreaks. It is unknown if similar results will be observed in hospitals. METHODS: Floor swabs were prospectively collected weekly from healthcare worker-only areas (eg, staff locker rooms) at two hospitals in Ontario, Canada for 39 weeks. Floor swabs were processed for SARS-CoV-2 using quantitative reverse-transcriptase polymerase chain reaction. Results were reported as percentage of positive floor swabs and viral copy number. Grouped fivefold cross-validation was used to evaluate model outbreak discrimination. RESULTS: SARS-CoV-2 RNA was detected on 537 of 760 floor swabs (71%). At Hospital A, overall positivity was 90% (95% CI: 85%-93%; N = 280); at Hospital B, overall positivity was 60% (95% CI: 55%-64%; N = 480). There were four COVID-19 outbreaks at Hospital A and seven at Hospital B during the study period. The outbreaks consisted of primarily patient cases (ie, 140 patient cases and 4 staff cases). For every 10-fold increase in viral copies, there was a 22-fold higher odds of a COVID-19 outbreak (OR = 22.0, 95% CI 7.3, 91.8). The cross-validated area under the receiver operating curve for SARS-CoV-2 viral copies for predicting a contemporaneous outbreak was 0.86 (95% CI 0.82-0.90). CONCLUSION: Viral burden of SARS-CoV-2 on floors, even in healthcare worker-only areas, was strongly associated with COVID-19 outbreaks in those hospital wards. Built environment sampling may support hospital COVID-19 outbreak identification, fill gaps in traditional surveillance, and guide infection prevention and control measures.

Unpredictability of the Fitness Effects of Antimicrobial Resistance Mutations Across Environments in Escherichia coli.

The evolution of antimicrobial resistance (AMR) in bacteria is a major public health concern, and antibiotic restriction is often implemented to reduce the spread of resistance. These measures rely on the existence of deleterious fitness effects (i.e. costs) imposed by AMR mutations during growth in the absence of antibiotics. According to this assumption, resistant strains will be outcompeted by susceptible strains that do not pay the cost during the period of restriction. The fitness effects of AMR mutations are generally studied in laboratory reference strains grown in standard growth environments; however, the genetic and environmental context can influence the magnitude and direction of a mutation's fitness effects. In this study, we measure how three sources of variation impact the fitness effects of Escherichia coli AMR mutations: the type of resistance mutation, the genetic background of the host, and the growth environment. We demonstrate that while AMR mutations are generally costly in antibiotic-free environments, their fitness effects vary widely and depend on complex interactions between the mutation, genetic background, and environment. We test the ability of the Rough Mount Fuji fitness landscape model to reproduce the empirical data in simulation. We identify model parameters that reasonably capture the variation in fitness effects due to genetic variation. However, the model fails to accommodate the observed variation when considering multiple growth environments. Overall, this study reveals a wealth of variation in the fitness effects of resistance mutations owing to genetic background and environmental conditions, which will ultimately impact their persistence in natural populations.

Detection of SARS-CoV-2 in schools using built environment testing in Ottawa, Canada: A multi-facility prospective surveillance study.

Classroom and staffroom floor swabs across six elementary schools in Ottawa, Canada were tested for SARS-CoV-2. Environmental test positivity did not correlate with student grade groups, school-level absenteeism, pediatric COVID-19-related hospitalizations, or community SARS-CoV-2 wastewater levels. Schools in neighbourhoods with historically elevated COVID-19 burden showed a negative but non-significant association with lower swab positivity.

Locust gut epithelia do not become more permeable to fluorescent dextran and bacteria in the cold.

The insect gut, which plays a role in ion and water balance, has been shown to leak solutes in the cold. Cold stress can also activate insect immune systems, but it is unknown whether the leak of the gut microbiome is a possible immune trigger in the cold. We developed a novel feeding protocol to load the gut of locusts (Locusta migratoria) with fluorescent bacteria before exposing them to -2°C for up to 48 h. No bacteria were recovered from the hemolymph of cold-exposed locusts, regardless of exposure duration. To examine this further, we used an ex vivo gut sac preparation to re-test cold-induced fluorescent FITC-dextran leak across the gut and found no increased rate of leak. These results question not only the validity of FITC-dextran as a marker of paracellular barrier permeability in the gut, but also to what extent the insect gut becomes leaky in the cold.

The spatial and temporal distribution of SARS-CoV-2 from the built environment of COVID-19 patient rooms: A multicentre prospective study.

BACKGROUND: SARS-CoV-2 can be detected from the built environment (e.g., floors), but it is unknown how the viral burden surrounding an infected patient changes over space and time. Characterizing these data can help advance our understanding and interpretation of surface swabs from the built environment. METHODS: We conducted a prospective study at two hospitals in Ontario, Canada between January 19, 2022 and February 11, 2022. We performed serial floor sampling for SARS-CoV-2 in rooms of patients newly hospitalized with COVID-19 in the past 48 hours. We sampled the floor twice daily until the occupant moved to another room, was discharged, or 96 hours had elapsed. Floor sampling locations included 1 metre (m) from the hospital bed, 2 m from the hospital bed, and at the room's threshold to the hallway (typically 3 to 5 m from the hospital bed). The samples were analyzed for the presence of SARS-CoV-2 using quantitative reverse transcriptase polymerase chain reaction (RT-qPCR). We calculated the sensitivity of detecting SARS-CoV-2 in a patient with COVID-19, and we evaluated how the percentage of positive swabs and the cycle threshold of the swabs changed over time. We also compared the cycle threshold between the two hospitals. RESULTS: Over the 6-week study period we collected 164 floor swabs from the rooms of 13 patients. The overall percentage of swabs positive for SARS-CoV-2 was 93% and the median cycle threshold was 33.4 (interquartile range [IQR]: 30.8, 37.2). On day 0 of swabbing the percentage of swabs positive for SARS-CoV-2 was 88% and the median cycle threshold was 33.6 (IQR: 31.8, 38.2) compared to swabs performed on day 2 or later where the percentage of swabs positive for SARS-CoV-2 was 98% and the cycle threshold was 33.2 (IQR: 30.6, 35.6). We found that viral detection did not change with increasing time (since the first sample collection) over the sampling period, Odds Ratio (OR) 1.65 per day (95% CI 0.68, 4.02; p = 0.27). Similarly, viral detection did not change with increasing distance from the patient's bed (1 m, 2 m, or 3 m), OR 0.85 per metre (95% CI 0.38, 1.88; p = 0.69). The cycle threshold was lower (i.e., more virus) in The Ottawa Hospital (median quantification cycle [Cq] 30.8) where floors were cleaned once daily compared to the Toronto hospital (median Cq 37.2) where floors were cleaned twice daily. CONCLUSIONS: We were able to detect SARS-CoV-2 on the floors in rooms of patients with COVID-19. The viral burden did not vary over time or by distance from the patient's bed. These results suggest floor swabbing for the detection of SARS-CoV-2 in a built environment such as a hospital room is both accurate and robust to variation in sampling location and duration of occupancy.

Detection of Covid-19 Outbreaks Using Built Environment Testing for SARS-CoV-2.

BACKGROUND: Environmental surveillance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) through wastewater has become a useful tool for population-level surveillance. Built environment sampling may provide a more spatially refined approach for surveillance in congregate living settings. METHODS: We conducted a prospective study in 10 long-term care homes (LTCHs) between September 2021 and November 2022. Floor surfaces were sampled weekly at multiple locations within each building and analyzed for the presence of SARS-CoV-2 using quantitative reverse transcriptase polymerase chain reaction. The primary outcome was the presence of a coronavirus disease 2019 (Covid-19) outbreak in the week that floor sampling was performed. RESULTS: Over the 14-month study period, we collected 4895 swabs at 10 LTCHs. During the study period, 23 Covid-19 outbreaks occurred with 119 cumulative weeks under outbreak. During outbreak periods, the proportion of floor swabs that were positive for SARS-CoV-2 was 54.3% (95% confidence interval [CI], 52 to 56.6), and during non-outbreak periods it was 22.3% (95% CI, 20.9 to 23.8). Using the proportion of floor swabs positive for SARS-CoV-2 to predict Covid-19 outbreak status in a given week, the area under the receiver-operating characteristic curve was 0.84 (95% CI, 0.78 to 0.9). Among 10 LTCHs with an outbreak and swabs performed in the prior week, eight had positive floor swabs exceeding 10% at least 5 days before outbreak identification. For seven of these eight LTCHs, positivity of floor swabs exceeded 10% more than 10 days before the outbreak was identified. CONCLUSIONS: Detection of SARS-CoV-2 on floors is strongly associated with Covid-19 outbreaks in LTCHs. These data suggest a potential role for floor sampling in improving early outbreak identification.

Golden Gate Assembly of Aerobic and Anaerobic Microbial Bioreporters.

Microbial bioreporters provide direct insight into cellular processes by producing a quantifiable signal dictated by reporter gene expression. The core of a bioreporter is a genetic circuit in which a reporter gene (or operon) is fused to promoter and regulatory sequences that govern its expression. In this study, we develop a system for constructing novel Escherichia coli bioreporters based on Golden Gate assembly, a synthetic biology approach for the rapid and seamless fusion of DNA fragments. Gene circuits are generated by fusing promoter and reporter sequences encoding yellow fluorescent protein, mCherry, bacterial luciferase, and an anaerobically active flavin-based fluorescent protein. We address a barrier to the implementation of Golden Gate assembly by designing a series of compatible destination vectors that can accommodate the assemblies. We validate the approach by measuring the activity of constitutive bioreporters and mercury and arsenic biosensors in quantitative exposure assays. We also demonstrate anaerobic quantification of mercury and arsenic in biosensors that produce flavin-based fluorescent protein, highlighting the expanding range of redox conditions that can be examined by microbial bioreporters. IMPORTANCE Microbial bioreporters are versatile genetic tools with wide-ranging applications, particularly in the field of environmental toxicology. For example, biosensors that produce a signal output in the presence of a specific analyte offer less costly alternatives to analytical methods for the detection of environmental toxins such as mercury and arsenic. Biosensors of specific toxins can also be used to test hypotheses regarding mechanisms of uptake, toxicity, and biotransformation. In this study, we develop an assembly platform that uses a synthetic biology technique to streamline construction of novel Escherichia coli bioreporters that produce fluorescent or luminescent signals either constitutively or in response to mercury and arsenic exposure. Beyond the synthesis of novel biosensors, our assembly platform can be adapted for numerous applications, including labeling bacteria for fluorescence microscopy, developing gene expression systems, and modifying bacterial genomes.

Identifying the drivers of computationally detected correlated evolution among sites under antibiotic selection.

The ultimate causes of correlated evolution among sites in a genome remain difficult to tease apart. To address this problem directly, we performed a high-throughput search for correlated evolution among sites associated with resistance to a fluoroquinolone antibiotic using whole-genome data from clinical strains of Pseudomonas aeruginosa, before validating our computational predictions experimentally. We show that for at least two sites, this correlation is underlain by epistasis. Our analysis also revealed eight additional pairs of synonymous substitutions displaying correlated evolution underlain by physical linkage, rather than selection associated with antibiotic resistance. Our results provide direct evidence that both epistasis and physical linkage among sites can drive the correlated evolution identified by high-throughput computational tools. In other words, the observation of correlated evolution is not by itself sufficient evidence to guarantee that the sites in question are epistatic; such a claim requires additional evidence, ideally coming from direct estimates of epistasis, based on experimental evidence.

An sRNA Screen for Reversal of Quinolone Resistance in Escherichia coli.

In light of the rising prevalence of antimicrobial resistance (AMR) and the slow pace of new antimicrobial development, there has been increasing interest in the development of adjuvants that improve or restore the effectiveness of existing drugs. Here, we use a novel small RNA (sRNA) screening approach to identify genes whose knockdown increases ciprofloxacin (CIP) sensitivity in a resistant strain of Escherichia coli 5000 sRNA constructs were initially screened on a gyrA S83L background, ultimately leading to 30 validated genes whose disruption reduces CIP resistance. This set includes genes involved in DNA replication, repair, recombination, efflux, and other regulatory systems. Our findings increase understanding of the functional interactions of DNA Gyrase, and may aid in the development of new therapeutic approaches for combating AMR.

The distribution of fitness effects among synonymous mutations in a gene under directional selection.

The fitness effects of synonymous mutations, nucleotide changes that do not alter the encoded amino acid, have often been assumed to be neutral, but a growing body of evidence suggests otherwise. We used site-directed mutagenesis coupled with direct measures of competitive fitness to estimate the distribution of fitness effects among synonymous mutations for a gene under directional selection and capable of adapting via synonymous nucleotide changes. Synonymous mutations had highly variable fitness effects, both deleterious and beneficial, resembling those of nonsynonymous mutations in the same gene. This variation in fitness was underlain by changes in transcription linked to the creation of internal promoter sites. A positive correlation between fitness and the presence of synonymous substitutions across a phylogeny of related Pseudomonads suggests these mutations may be common in nature. Taken together, our results provide the most compelling evidence to date that synonymous mutations with non-neutral fitness effects may in fact be commonplace.

Insights Into Arsenite and Arsenate Uptake Pathways Using a Whole Cell Biosensor.

Despite its high toxicity and widespread occurrence in many parts of the world, arsenic (As) concentrations in decentralized water supplies such as domestic wells remain often unquantified. One limitation to effective monitoring is the high cost and lack of portability of current arsenic speciation techniques. Here, we present an arsenic biosensor assay capable of quantifying and determining the bioavailable fraction of arsenic species at environmentally relevant concentrations. First, we found that inorganic phosphate, a buffering agent and nutrient commonly found in most bioassay exposure media, was in fact limiting As(V) uptake, possibly explaining the variability in As(V) detection reported so far. Second, we show that the nature of the carbon source used in the bioassay differentially affects the response of the biosensor to As(III). Finally, our data support the existence of non-specific reduction pathways (non-ars encoded) that are responsible for the reduction of As(V) to As(III), allowing its detection by the biosensor. To validate our laboratory approach using field samples, we performed As(III) and As(V) standard additions on natural water samples collected from 17 lakes surrounding Giant Mine in Yellowknife (NWT), Canada. We found that legacy arsenic contamination in these lake water samples was accurately quantified by the biosensor. Interestingly, bioavailability of freshly added standards showed signs of matrix interference, indicative of dynamic interactions between As(III), As(V) and environmental constituents that have yet to be identified. Our results point toward dissolved organic carbon as possibly controlling these interactions, thus altering As bioavailability.

Pseudomonas aeruginosa Biofilm Antibiotic Resistance Gene ndvB Expression Requires the RpoS Stationary-Phase Sigma Factor.

Chronic, biofilm-based bacterial infections are exceptionally difficult to eradicate due to the high degree of antibiotic recalcitrance exhibited by cells in biofilm communities. In the opportunistic pathogen Pseudomonas aeruginosa, biofilm recalcitrance is multifactorial and arises in part from the preferential expression of resistance genes in biofilms compared to exponential-phase planktonic cells. One such mechanism involves ndvB, which we have previously shown to be expressed specifically in biofilms. In this study, we investigated the regulatory basis of this lifestyle-specific expression by developing an unstable green fluorescent protein (GFP) transcriptional reporter to observe the expression pattern of ndvB We found that in addition to its expression in biofilms, ndvB was upregulated in planktonic cells as they enter stationary phase. The transcription of ndvB in both growth phases was shown to be dependent on the stationary-phase sigma factor RpoS, and mutation of a putative RpoS binding site in the ndvB promoter abolished the activity of the promoter in stationary-phase cells. Overall, we have expanded our understanding of the temporal expression of ndvB in P. aeruginosa and have uncovered a regulatory basis for its growth phase-dependent expression.IMPORTANCE Bacterial biofilms are more resistant to antibiotics than free-living planktonic cells, and understanding the mechanistic basis of this resistance can inform treatments of biofilm-based infections. In addition to chemical and structural barriers that can inhibit antibiotic entry, the upregulation of specific genes in biofilms contributes to the resistance. We investigated this biofilm-specific gene induction by examining expression patterns of ndvB, a gene involved in biofilm resistance of the opportunistic pathogen Pseudomonas aeruginosa We characterized ndvB expression in planktonic and biofilm growth conditions with an unstable green fluorescent protein (GFP) reporter and found that the expression of ndvB in biofilms is dependent on the stationary-phase sigma factor RpoS. Overall, our results support the physiological similarity between biofilms and stationary-phase cells and suggest that the induction of some stationary-phase genes in biofilms may contribute to their increased antibiotic resistance.

Ionic Strength Differentially Affects the Bioavailability of Neutral and Negatively Charged Inorganic Hg Complexes.

Mercury (Hg) bioavailability to bacteria in marine systems is the first step toward its bioamplification in food webs. These systems exhibit high salinity and ionic strength that will both alter Hg speciation and properties of the bacteria cell walls. The role of Hg speciation on Hg bioavailability in marine systems has not been teased apart from that of ionic strength on cell wall properties, however. We developed and optimized a whole-cell Hg bioreporter capable of functioning under aerobic and anaerobic conditions and exhibiting no physiological limitations of signal production to changes in ionic strength. We show that ionic strength controls the bioavailability of Hg species, regardless of their charge, possibly by altering properties of the bacterial cell wall. The unexpected anaerobic bioavailability of negatively charged halocomplexes may help explain Hg methylation in marine systems such as the oxygen-deficient zone in the oceanic water column, sea ice or polar snow.

Evolution of Cost-Free Resistance under Fluctuating Drug Selection in Pseudomonas aeruginosa.

Antibiotic resistance evolves rapidly in response to drug selection, but it can also persist at appreciable levels even after the removal of the antibiotic. This suggests that many resistant strains can both be resistant and have high fitness in the absence of antibiotics. To explore the conditions under which high-fitness, resistant strains evolve and the genetic changes responsible, we used a combination of experimental evolution and whole-genome sequencing to track the acquisition of ciprofloxacin resistance in the opportunistic pathogen Pseudomonas aeruginosa under conditions of constant and fluctuating antibiotic delivery patterns. We found that high-fitness, resistant strains evolved readily under fluctuating but not constant antibiotic conditions and that their evolution was underlain by a trade-off between resistance and fitness. Whole-genome sequencing of evolved isolates revealed that resistance was gained through mutations in known resistance genes and that second-site mutations generally compensated for costs associated with resistance in the fluctuating treatment, leading to the evolution of cost-free resistance. Our results suggest that current therapies involving intermittent administration of antibiotics are contributing to the maintenance of antibiotic resistance at high levels in clinical settings. IMPORTANCE Antibiotic resistance is a global problem that greatly impacts human health. How resistance persists, even in the absence of antibiotic treatment, is thus a public health problem of utmost importance. In this study, we explored the antibiotic treatment conditions under which cost-free resistance arises, using experimental evolution of the bacterium Pseudomonas aeruginosa and the quinolone antibiotic ciprofloxacin. We found that intermittent antibiotic treatment led to the evolution of cost-free resistance and demonstrate that compensatory evolution is the mechanism responsible for cost-free resistance. Our results suggest that discontinuous administration of antibiotic may be contributing to the high levels of antibiotic resistance currently found worldwide.

Adaptive synonymous mutations in an experimentally evolved Pseudomonas fluorescens population.

Conventional wisdom holds that synonymous mutations, nucleotide changes that do not alter the encoded amino acid, have no detectable effect on phenotype or fitness. However, a growing body of evidence from both comparative and experimental studies suggests otherwise. Synonymous mutations have been shown to impact gene expression, protein folding and fitness, however, direct evidence that they can be positively selected, and so contribute to adaptation, is lacking. Here we report the recovery of two beneficial synonymous single base pair changes that arose spontaneously and independently in an experimentally evolved population of Pseudomonas fluorescens. We show experimentally that these mutations increase fitness by an amount comparable to non-synonymous mutations and that the fitness increases stem from increased gene expression. These results provide unequivocal evidence that synonymous mutations can drive adaptive evolution and suggest that this class of mutation may be underappreciated as a cause of adaptation and evolutionary dynamics.

The biofilm-specific antibiotic resistance gene ndvB is important for expression of ethanol oxidation genes in Pseudomonas aeruginosa biofilms.

Bacteria growing in biofilms are responsible for a large number of persistent infections and are often more resistant to antibiotics than are free-floating bacteria. In a previous study, we identified a Pseudomonas aeruginosa gene, ndvB, which is important for the formation of periplasmic glucans. We established that these glucans function in biofilm-specific antibiotic resistance by sequestering antibiotic molecules away from their cellular targets. In this study, we investigate another function of ndvB in biofilm-specific antibiotic resistance. DNA microarray analysis identified 24 genes that were responsive to the presence of ndvB. A subset of 20 genes, including 8 ethanol oxidation genes (ercS', erbR, exaA, exaB, eraR, pqqB, pqqC, and pqqE), was highly expressed in wild-type biofilm cells but not in ΔndvB biofilms, while 4 genes displayed the reciprocal expression pattern. Using quantitative real-time PCR, we confirmed the ndvB-dependent expression of the ethanol oxidation genes and additionally demonstrated that these genes were more highly expressed in biofilms than in planktonic cultures. Expression of erbR in ΔndvB biofilms was restored after the treatment of the biofilm with periplasmic extracts derived from wild-type biofilm cells. Inactivation of ethanol oxidation genes increased the sensitivity of biofilms to tobramycin. Together, these results reveal that ndvB affects the expression of multiple genes in biofilms and that ethanol oxidation genes are linked to biofilm-specific antibiotic resistance.

Pseudomonas aeruginosa tssC1 links type VI secretion and biofilm-specific antibiotic resistance.

Biofilm-specific antibiotic resistance is influenced by multiple factors. We demonstrated that Pseudomonas aeruginosa tssC1, a gene implicated in type VI secretion (T6S), is important for resistance of biofilms to a subset of antibiotics. We showed that tssC1 expression is induced in biofilms and confirmed that tssC1 is required for T6S.

Membrane proteases and aminoglycoside antibiotic resistance.

We present genetic studies that help define the functional network underlying intrinsic aminoglycoside resistance in Pseudomonas aeruginosa. Our analysis shows that proteolysis, particularly that controlled by the membrane protease FtsH, is a major determinant of resistance. First, we examined the consequences of inactivating genes controlled by AmgRS, a two-component regulator required for intrinsic tobramycin resistance. Three of the gene products account for resistance: a modulator of FtsH protease (YccA), a membrane protease (HtpX), and a membrane protein of unknown function (PA5528). Second, we screened mutations inactivating 66 predicted proteases and related functions. Insertions inactivating two FtsH protease accessory factors (HflK and HflC) and a cytoplasmic protease (HslUV) increased tobramycin sensitivity. Finally, we generated an ftsH deletion mutation. The mutation dramatically increased aminoglycoside sensitivity. Many of the functions whose inactivation increased sensitivity appeared to act independently, since multiple mutations led to additive or synergistic effects. Up to 500-fold increases in tobramycin sensitivity were observed. Most of the mutations also were highly pleiotropic, increasing sensitivity to a membrane protein hybrid, several classes of antibiotics, alkaline pH, NaCl, and other compounds. We propose that the network of proteases provides robust protection from aminoglycosides and other substances through the elimination of membrane-disruptive mistranslation products.

Mutational analysis of Burkholderia thailandensis quorum sensing and self-aggregation.

Acyl-homoserine lactone (acyl-HSL) quorum-sensing signaling is common to many Proteobacteria. Acyl-HSLs are synthesized by the LuxI family of synthases, and the signal response is mediated by members of the LuxR family of transcriptional regulators. Burkholderia thailandensis is a member of a closely related cluster of three species, including the animal pathogens Burkholderia mallei and Burkholderia pseudomallei. Members of this group have similar luxI and luxR homologs, and these genes contribute to B. pseudomallei and B. mallei virulence. B. thailandensis possesses three pairs of luxI-luxR homologs. One of these pairs, BtaI2-BtaR2, has been shown to produce and respond to 3OHC(10)-HSL and to control the synthesis of an antibiotic. By using a markerless-exhange method, we constructed an assortment of B. thailandensis quorum-sensing mutants, and we used these mutants to show that BtaI1 is responsible for C(8)-HSL production and BtaI3 is responsible for 3OHC(8)-HSL production. We also show that a strain incapable of acyl-HSL production is capable of growth on the same assortment of carbon and nitrogen sources as the wild type. Furthermore, this mutant shows no loss of virulence compared to the wild type in mice. However, the wild type self-aggregates in minimal medium, whereas the quorum-sensing mutant does not. The wild-type aggregation phenotype is recovered by addition of the BtaI1-R1 HSL signal C(8)-HSL. We propose that the key function of the BtaR1-BtaI1 quorum-sensing system is to cause cells to gather into aggregates once a sufficient population has been established.