We work across the boundaries of the traditional scientific disciplines to address fundamental and challenging biological questions related to gene expression, chromosome structure, and the bacterial cell cycle. In particular, we strive to test quantitative predictions and to identify inconsistencies in current biological models. This agenda has made it necessary to sharpen the experimental tools both in terms of throughput, sensitivity, and spatiotemporal resolution. A few areas where we have explored how the biological cell has solved physical challenges are outlined in the following sections.
The search problem
We have explored how molecules search for information in the genome in several different projects (Hammar et al., Science 2012; Hammar et al., Nat. Gen. 2017, Jones et al., Science 2017, Li et al. Nature Physics 2009). We conclude that the transcription factor (TF) lacI, which act as a repressor of the lactose digestion machinery in the cell, searches for its binding site using a combination of 3D diffusion and 1D sliding on the chromosomal DNA. The TF slides around 45 bases before detaching after 1 ms and although it searches the entire genome, it takes only a few minutes for the protein to find and bind the operator.
In the ongoing work on LacI, we are measuring how the TF interrogates DNA at the microsecond time scale. In contrast to LacI that searches the DNA sequence through interactions with the DNA grooves, Cas9 actually needs to unwind the double helix and interrogate the DNA sequence to discriminate right from wrong. Although Cas9’s search problem is somewhat reduced by the fact the potential targets are defined by the presence of a PAM sequence, it still presents a daunting task. We could measure that it takes on average 6 hours for a dCas9 molecule to find its target sequence. Compared to the search time of the transcription factors it’s an eternity, but it is the price that the CRISPR/Cas9 system has to pay in order to be reprogrammable to target any sequence.
Our ongoing work on the search problem addresses the question of how a broken chromosome can find its homologous sister after a double-stranded break.
Noise in gene expression
We have demonstrated how the binding strength of transcription factors affects noise in gene expression (Grönlund et al., Nat. Comm. 2013; Grönlund et al. PNAS 2010). Following a biological cue, it can take several minutes before a TF finds and binds its target sequence. As a result, negative feedback (where the expression of a TF is inhibited by the TF itself) cannot at the same time be fast and strong. If the TF binds strongly to the repressor site, most proteins are necessarily produced in a burst immediately after gene replication and if the binding is very weak, is practically unregulated. To this end, we predict the existence of an optimal TF binding strength, where the time the binding site is free is not dependent on TF concentration. To produce biological evidence of this optimum is an ongoing effort.
Cell cycle control
As a result of our investigations of the coordination of replication and cell division (Walldén et al. Cell, 2016), we have found that genetically identical bacterial cells display a large variability in sizes and generation times also in a constant environment. Given this variation, it is not obvious how the cells manage to coordinate DNA replication and cell division, to ensure that there are always two chromosome copies ready for each division. The problem is especially hard considering that rapidly growing bacteria have multiple replications of the same chromosome ongoing at the same time. We suggest that E. coli solves this challenging problem of replication-division coordination by initiating replication at a fixed cell volume per chromosome independent of growth rate and division volume that. Following initiation, division occurs after a time interval that is determined by the cell’s growth rate. Currently, we are investigating the mechanistic details of how the cell can sense its size or possibly an increase in size.
New technologies
Many of our biological questions has driven technological innovation. The development of customized microfluidics for single molecule experiments in living bacterial cells has been essential for the transcription factor and cell cycle work (Hammar et al., Science 2012; Hammar et al., Nat. Gen. 2014, Jones et al., Science 2017, Wallden et al. Cell 2016)
New approaches to using single molecule tracking have been critical to determine intracellular kinetics rates (English et al., PNAS 2011; Sanamrad et al. PNAS 2014; Balzarotti et al. Science 2017, Volkov et al., Nat. Chem.Biol. 2018). These studies have also involved important improvements in the analysis and software (Persson et al., Nat. Meth. 2012; Lindén et al., Bioinformatics 2016).
In an effort to bridge the current gap between advanced imaging and high-throughput genetic engineering, we have developed DuMPLING (Lawson et al., Mol. Syst. Biol. 2017). This optical technology can be used to phenotype a large strain library and subsequently identify the genetic identity of the strains by in situ genotyping. Combined with visionary cloning strategies, this method promise to deliver new insights on previously inaccessible scientific problems.
We also aim to further increase the resolution with which it is possible to investigate biological systems in living cells. The Poltrack system (Marklund et al. BioRxiv 2018) is an extension of the MINFLUX microscopy (Balzarotti et al. Science 2016) that uses ultrafast hardware-based feedback laser scanning and nanosecond time tagging of individual photons to determine the absolute orientation of the fluorophore dipole moment in the sub-millisecond time regime. This technique closes an important gap in studying dynamic systems in the life sciences. Antibiotic resistance
We also work with different aspects of antibiotic resistance, where we contribute a genome-wide screening approach to identify different resistance mechanisms, as well as more applied approaches such as the development of a rapid resistance determination test for clinical samples (Baltekin et al., PNAS 2017).