Introduction
In previous studies it has been shown that multicopy plasmids (MCP) may confer many evolutionary advantages to the bacteria. Such as evolutionary trade-offs \citep{Rodriguez_Beltran_2018} <,more> and evolution to antibiotic resistance \cite{San_Millan_2016}. In the case of antibiotic resistance MCP increases bacteria evolvability by increasing the gene dosage and thus, increassing the chance to acquire a beneficial mutation. Moreover, if a mutation occurs in the plasmid replication control mechanism it could increases the copy number and for those plasmids that replicate through the RNAI-RNAII system it doubles the mean plasmid copy number.
When a mutation occurs within a plasmid gene, at the beginning it is only in one plasmid of one cell, then the mutation spreads in the population by replication and segregation. This phenomena has been described in detail in \cite{Rodriguez-Beltran2018} and has been called plasmid mediated heterozygosis (PMH). It is a natural occurring transient state, and it is believed/understood that heterozygosis disappears upon selection in favor of one allele.
In this work we achieve to explore the limits of PMH. A previous result demonstrated that, at population level, heterozygosis is maintained under fluctuating environments even under strong selection. But there is still much to know in the matter, like what are the plasmids dynamics upon selection? Is the heterozygosis maintained within the cell or though homozygotic subpopulations? Having allelic variants could be useful to ensure the survival of a population against a sudden change of environment, but what are the necessary conditions to maintain heterozygosis? Are fluctuating environments enough to maintain it? Is PMH maintained under constant environments?
In order to address this questions we need to use a single cell approach. We used a complex microfluidic system to perform experiments and acquire information of individual cells submitted to different periods of environments fluctuations . We used the plasmid heterozygotic strain (HT) described in \cite{Rodriguez-Beltran2018} to perform experiments. HT is an Escherichia coli MG1655 bearring two plasmids, pBGT1 and pBRT12. Both plasmids share the same replication origin, the pBGT1 plasmid carries the blaTEM1 gene which confers resistance to Ampicillin (AMP) and a GFP as a reporter, the pBRT12 plasmid carries a blaTEM12 gene that confers resistance to Ceftazidime (CAZ) and a RFP as reporter.
We also developed an stochastic agent based model to make simulations of our experiments and to explore the plasmids dynamics under a large set of selective pressures and periods of fluctuations.
Methods
Microfluidics
Our microfluidic system consists in three coupled parts: 1) The microfluidic device. It is a PDMS chip with two inputs for media one shunt*, two exits for waste and an entry for the cells, it has 48 chambers of 30x40x.9 μm in which we the trapped bacteria form a monolayer of between 500 to 1000 bacteria. This chip architecture has been previously published in [REF]. 2) The Dial-A-Wave (DAW) system. A computer controlled motor series that control <sliders> on rails to control the height of the medium. With this system we can dynamically control the flows inside the chip. The motors [ref] are controlled by Stepper Phidget boards [ref] and the last ones by an open source software [ref]. The slider/rails we bought on [ref]. 3) The Image Acquisition system. A Nikon eclipse Ti-E epiflourescence microscope was used. It counts with an incubator <need?>, a perfect focus system, and a motorized platine. The microscope was controlled by NIS Elements program.
Experiments methodology
The HT cells were cultivated in LB media with cloraphenicol (Chl) 15 mg/l and CAZ 0.5 mg/l to balance the portion of plasmids, concentrations were proven by growth in solid plate, data not shown. Then this culture is used to inoculate 200 ml of fresh LB and cultivated at 30 C until it reaches 0.3 OD. Then it is centrifuged and cells are resuspended in 5 ml of LB microfluidic media and this dense culture is used to populate the chip. The media used for the microfluidic experiments is LB with arabinose at 0.5%, tween20 at 0.075% and the experimeriment indicated antibiotic concentration. Before introduccion of antibiotics during the experiments cells were observed for a variable amount of time growing under LB or LB with amp 50 mg/l to as behavior reference to compare the response to antibiotics.