The goal of Biological Physics is to reach an understanding of biological processes that is based on a mathematical description that is quantitative and predictive. In Einstein’s words, we would like to describe biological systems with equations that are “as simple as possible, but not simpler”.
The key biological process at the heart of Biology is the process of self-replication. The amazing process that allows a cell to make an (almost) perfect copy of itself, is what defines Biology.
Self-replication can occur within 20 minutes (E.coli- Movie 1), within a day (mammalian cells- Movie 3) or be halted for thousands of years (spores). Our goal is to reach a deep understanding of the way physical, biochemical and evolutionary processes shape these time-scales. We develop tools and approaches to study the time distribution of self-replication is various systems, from bacteria to mammalian cells. We have found that the distribution of self-replication times has surprising features with direct clinical consequences. For example, bacteria that halt the process of self-replication for long periods can be detected in quasi all bacterial populations, and those bacteria are responsible for the failure of many antibiotic treatments. Similar behavior in cancer cells may underlie the failure of anti-cancer treatments.
Self-replication arrest as a bet-hedging strategy underlies bacterial persistence
In general, growth and self-replication expose bacteria to many dangers. To grow bacteria must import nutrients from the environment, increase the cell wall and replicate their DNA, complex and vital processes that can be deregulated by antibiotics. Using micro-fluidic devices to trap single bacteria under the microscope, we demonstrated that a small fraction of bacteria that are in a dormant state, prior to the antibiotic exposure, may underlie the persistence to antibiotics, as reduced growth or growth arrest significantly diminishes the potency of most antibiotics (Balaban et al., Science 2004). Digging further into the molecular mechanism responsible for dormancy, we found out that threshold amplification of noise in a toxin-antitoxin module can underlie bacterial persistence (Rotem et al., PNAS 2010).
Evolution of a “bacterial timer” that halts self-replication for a defined time and allows bacteria to evade antibiotic killing
The ability of microorganisms to overcome antibiotic treatments is one of the top concerns of modern medicine. The effectiveness of many antibiotics has been reduced by bacteria's ability to rapidly evolve and develop strategies to resist antibiotics. Bacteria achieve this by specific mechanisms that are tailored to the molecular structure or function of a particular antibiotic. For example, bacteria would typically develop drug resistance by evolving a mutation that breaks down the drug. We found out that bacterial population may follow a different evolutionary process. Using the quantitative approach of physicists, we developed experimental tools to measure precisely the bacterial response to antibiotics and developed a mathematical model of the process. The model predicted that a daily three-hour dose would enable the bacteria to predict delivery of the drug, and go dormant for that period in order to survive.
To test the predictions, we delivered antibiotics to bacterial populations in the lab for precisely three hours each day. After only ten days we were able to observe the bacteria using a new survival tactic. When exposed to these repeated cycles of antibiotic treatments, the bacteria evolved an adaptation to the duration of the antibiotic stress by remaining dormant for the treatment period. The results demonstrated that bacteria can evolve an anti-drug response within days. Most significantly, it showed for the first time that bacteria can develop a biological clock timer to survive under antibiotic exposure. When additional population were exposed to antibiotics for different periods, lasting 3, 5, or 8 hours, each of the populations adapted by prolonging their dormant stage to match the exposure duration (Fridman et al., Nature 2014).
With this new understanding of how bacterial populations evolve survival strategies against antibiotics, scientists could develop new approaches for slowing the evolution of antibiotic resistance. In the future, it may help devising different treatment schedules.
How bacteria survive long starvation periods
In his thesis, Jacques Monod pioneered the precise and quantitative analysis of bacterial cultures. He focused on mathematically defining and measuring the growth rate of exponentially
Several species of bacteria are known to have elaborated differentiation programs leading to protective structures called “spores” that allow long term survival in harsh and nutrient depleted environments. However, bacteria that do not sporulate, such as E.coli, are also able to survive extended starvation period. We showed that in order to maintain long-term viability under carbon starvation, E.coli shuts down its metabolism by 90% upon gradual nutrient depletion. Then, this reduced metabolism is maintained constant for several days, despite the absence of growth. We call this reduced but constant metabolism CASP (Constant Activity Stationary Phase) (Gefen O. et al., PNAS 2014).
Tools to study growth arrested bacteria
growing bacterial populations. These conceptual and experimental tools laid the foundation for quantitative studies of growing bacteria, which could be compared and reproduced. In contrast to these advances in the characterization of growing bacteria, growth arrest, which is the prevalent state of bacteria in nature, is far from being characterized in a similarly satisfactory manner. In Monod’s growth curve, two growth arrested phases are described: the “stationary phase”, where growth arrest is due to nutrients depletion, and the “lag phase”, during which bacteria adapt to the supply of fresh nutrients. We showed that single growth arrested bacteria can be visualized in microfluidic devices at the lag phase (Gefen O. et al., PNAS 2008) and at stationary phase (Gefen O. et al., PNAS 2014).
The lag phase duration, namely the time until division resumes when exposed to fresh nutrients, can vary greatly from cell to cell, even in clonal populations. We have developed an automated scanning systems, based of office documents scanners, that extract the single cell lag time distribution (Levin-Reisman et al., Nat Methods (2010), Levin-Reisman et al., Jove (2014))
Self-replication variability in mammalian cells
We have extended our analysis of single cell division times to study mammalian cells. Ongoing work reveals surprising features in the inheritance of the cell-cycle duration along a lineage of cells.