Research

Genome-wide Transcription Dynamics

Tracking the levels of the global regulators of gene expression of E. coli

https://doi.org/10.1128/msystems.00065-24

Escherichia coli have evolved hundreds of transcription factors to tune the expression of thousands of genes. Interestingly, a few transcription factors control almost half of all genes and are thus named global regulators (GRs). Tracking the numbers of these GRs over time is essential to understand phenotypic modifications, e.g., under stress conditions. We have engineered a library of strains to track GR levels. Each strain has a single-copy plasmid coding for a fast-maturing green fluorescent protein whose transcription is controlled by a copy of the natural promoter of the GR. This library should become useful in scientific research, and future applications in therapeutics and the bioindustries. For more information, please visit: 10.1128/msystems.00065-24.

DNA  supercoiling  as  a  trigger  of  short-term, cold shock repressed genes of E. coli 

Adapting to cold is a key skill of gut bacteria of warm-blooded animals. We hypothesized that cold affects DNA supercoiling, which is regulated by gyrase. We performed two RNA-seqs, one after cold-shock, the other after adding Novobiocin, an antibiotic that represses gyrase. We found that many cold-shock repressed genes are similarly sensitive to Novobiocin. Next, we observed similar changes in the protein numbers of these genes to both perturbations. Moreover, under cold shock, nucleoid density increases, and  gyrases  and  nucleoid  become  colocalized. Also, the cell energy decreases, which hinders the resolution of positive supercoils. We suggest that responsiveness to low negative supercoiling is a core feature of E. coli’s short-term, cold shock transcriptional response program, and could be used to regulate the temperature sensitivity of synthetic circuits.

The transcription factor network of E. coli steers global responses to shifts in RNAP concentration

The role of transcription factor networks (TFNs) on the robustness, sensitivity, plasticity, and overall adaptability of microorganisms to environmental stresses remains largely unexplored. We studied how the TFN of E. coli responds to genome-wide perturbations caused by quick changes in medium richness, known to alter RNA polymerase (RNAP) concentrations. 

We found that, at the gene cohort level, the average magnitude of the mid-term transcriptional responses can be explained by the average absolute difference between the numbers of activating and repressing input TFs of a gene. Interestingly, this difference is strongly positively correlated with the average number of input TFs of a gene. Our result suggests that the global topological traits of TFNs influence their global response dynamics to genome-wide stresses. This may be a general principle by which TFNs regulate genome-wide information propagation. 

Analytical kinetic model of native tandem promoters in E. coli

We studied how tandem promoters interfere with each other, depending on their strengths and on the nucleotide distances between them., taking into consideration the nucleotide length occupied by an RNAP when bound to the transcription start site region. 

Illustration of tandem promoters, when and when not overlapping.

 Events leading to transcriptional interference between tandem promoters. (A) Sequence of events in transcription in isolated promoters. A similar set of events occurs in tandem promoters, if only one RNAP interacts with them at any given time. (B / C) Interference due to the occlusion of the downstream / upstream promoter by a bound RNAP, which will impede the incoming RNAP from binding to the TSS. (D) Interference of the activity of the RNAP from the upstream promoter by the RNAP on the downstream promoter. One of these RNAPs will be dislodged by the collision. Created with BioRender.com.

Workflow. (I) We identified genes controlled by tandem promoters in RegulonDB. (II) We measured their single-cell protein levels and RNA fold changes over time. (III) We used the data to tune the model. (IV) We used the model to explore the state space of protein expression. 

This is our first study where we combine RNA-seq, flow-cytometry of natural promoters, info from RegulonDB on transcription factor interactions and promoter sequences, strain libraries, and stochastic and analytical models. 

We proposed an analytical model of gene expression with RNAP-promoter occupancy times and distance between promoters, dTSS as the key regulators of transcription interference.

 This model can assist to predict the dynamics of new pairs of tandem promoters and, thus, contribute to the expansion of synthetic genetic libraries.

Single-gene Transcription Dynamics

Selected Publication:

CSD Palma, V Kandavalli, MNM Bahrudeen, M Minoia, V Chauhan and AS Ribeiro (2020) Dissecting the in vivo dynamics of transcription locking due to positive supercoiling buildup. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms 1863(5) 194515. doi: 10.1016/j.bbagrm.2020.194515.

We proposed a new strategy to dissect the contribution of transcription initiation locking due to PSB on the kinetics of RNA production. We started by validating a stochastic model of transcription (A) that accounts for positive supercoiling accumulation (reaction 4 and 5 in (A)) and removal (reaction 6 in (A)) due to Gyrase intervention. We then derived an analytical solution for the inverse of the mean rate of RNA production as a function of Gyrase levels. From qPCR and microscopy data at different concentrations of Gyrase, we applied a Lineweaver-Burk plot to dissect the in vivo transcription rate in absence of PSB (D). We validated this using the same gene when single-copy plasmid-borne (shown to be impervious to supercoiling). We then estimated the time in locked states (τLocked) and the number of transcription events prior to locking (N), which we validated by measurements after adding Novobiocin. Finally, we inferred a range of transcription initiation locking kinetics in a chromosomal location, obtainable by tuning the basal transcription rate, and validate it by measurements at different induction levels. This strategy for dissecting the transcription initiation locking kinetics due to PSB may contribute to resolving transcriptional programs of E. coli and in the engineering of synthetic genetic circuits.

Selected Publication:

One topic of our lab is transcription dynamics, which affects cell-to-cell variability. Having E. coli cells to naturally change their RNA polymerase concentrations, we compared RNA production rates of a target gene and inferred the time RNAP spends prior and after commitment to transcription. For the Plac/ara-1 promoter, under full induction, the closed complex took ∼788 s while subsequent steps took ∼193 s, on average. Also, the promoter intermittently switches to an inactive state that, on average, last ∼87 s, due to the repression by LacI. This method can be used to resolve the rate-limiting steps of the in vivo dynamics of initiation of prokaryotic promoters.

Selected Publication

Small Synthetic Genetic Circuits

Bacteria process information to bind past events to future actions, so as to adapt to environment changes. This is made possible by a timely organized execution of multiple tasks such as counting time, sensing the environment, and decision making, which are performed in parallel, by semi-independent, tuned small genetic circuits. These circuits differ in structure, which defines the function. Meanwhile, the properties of their internal components, e.g. the kinetics of transcription initiation of the promoters, define the efficiency with how functions are performed.

We constructed a single-copy repressilator (SCR) by implementing the original repressilator circuit of Elowitz et al on a single-copy F-plasmid. We studied its behaviour as a function of temperature and compared to the original low-copy-number repressilator (LCR). In optimal temperature, the dynamics of the two systems differ, but respond similarly to temperature changes. Interestingly, the SCR is more robust to lower temperatures and perturbations than the original LCR, which loses functionality as temperature increases beyond 30 C, due to the loss of functionality of one of its proteins, CI.

Selected Publications:

Inner Cellular Biophysics

Cells have a complex internal spatial organization, ruled by biophysical laws. While E. coli is likely one of the simplest case-studies, there are still many questions about the mechanisms that regulate their internal organization. We have so far focused on three issues, namely, the temporal spatial organization of: protein aggregates, Z-rings, and Serine Chemoreceptors.

Protein aggregates and cellular aging 

The accumulation of non-functional protein aggregates is believed to be a cause for cellular aging. It has been found that cell lineages are able to distribute these aggregates unevenly by the cells of each generation, causing rejuvenation in many cells and accelerated aging in others. 

Using tracking of individual fluorescent proteins and protein aggregates in live cells, we studied their short- and long-term dynamics and spatial distribution in the cells' cytoplasm. One of our finds is that the nucleoid excludes aggregates from midcell, forcing them to mostly remain at the poles. Following cell divisions, this leads to accumulation of aggregates in only some individuals of a lineage at each generation, which will exhibit slower division rates i.e. aging. 

Example microscopy images prior and after segmentation. (A) DAPI-stained nucleoids in cells, (B) cells with visible cytoplasm (filled with MS2-GFP proteins) along with MS2-GFP tagged RNA molecules (synthetic aggregates), visible as bright white “spots”, and (C) segmentation of the images in (A) and (B) merged into one image. Dark grey areas show segmented cells while segmented nucleoids are shown in lighter grey and synthetic aggregates are shown as small white spots. 
This figure shows the anisotropy in the protein aggregates motions, as a function of the distance from the cell center,. Note the positive peak at 0.6, which means that the aggregates preferential move away from midcell. This peak occurs exactly where the nucleoid 'physical boundary' is. Meanwhile, another peak (negative) occurs at the cell borders, forcing the aggregates to remain inside the cell, as expected. 

We also studied this process in cells subject to low temperatures. We show that the process of segregation of aggregates to the poles is hampered, due to increased cytoplasm viscosity.

Selected Publications:

Z-rings and Cell Division

Cell division in E coli is morphologically symmetric due to, among other, their ability to place the Z-ring at midcell. At sub-optimal temperatures, this symmetry decreases at the single-cell level. Using fluorescence microscopy, we observed FtsZ-GFP and DAPI-stained nucleoids to assess the robustness of the symmetry of Z-ring formation and positioning in cells at sub-optimal and critical temperatures. We found that the Z-ring formation and positioning is robust at sub-optimal temperatures, as the Z-ring ’ s mean width, density and displacement from midcell maintain correlation to one another. However, at critical temperatures, the Z-ring displacement from midcell is greatly increased. This is due to enhanced distance between the replicated nucleoids and, thus, reduced Z-ring density, which explains the weaker precision in setting a morphologically symmetric division site. This also occurs in rich media and is cumulative, i.e. combining richer media and critically high temperatures enhances the asymmetries in division, which is evidence that the causes are biophysical. To support this, we showed that the effects are reversible, i.e. shifting cells from optimal to critical, and then to optimal again, reduces and then enhances the symmetry in Z-ring positioning, respectively. Overall, we found that Z-ring positioning in E. coli is a robust biophysical process under sub-optimal temperatures, and that critical temperatures cause significant asymmetries in division.

Example images of cells at 37 ° C. (A) A phase contrast image with cell segmentation. (B) DAPI-stained nucleoids visualized by epifluorescence microscopy. (C) Confocal images with FtsZ-GFP labelled rings with cell segmentation and Z-ring segmentation (light grey line) based on threshold selection.

Selected Publication:

Serine Chemoreceptors 

 We studied whether nucleoid exclusion contributes to the segregation and retention of Tsr chemoreceptor clusters at the cell poles. We used live time-lapse, single-cell microscopy measurements, mutant cells, etc. 

Overall, we found that the single-cell spatial distributions of Tsr clusters have heterogeneities and symmetries that are consistent with nucleoid exclusion, being one of the mechanisms by which they are positioned in the cells.

Spatial  distributions  of  fluorescence  intensity of Tsr-Venus proteins along the major cell axis. Black line is from 1195 control cells. Gray line is the same distribution averaged over 68 anucleate D mukB cells.

Stochastic Simulators

We have been developing detailed models of transcription and small genetic circuits. With these, we made predictions on how transcription and genetic circuits operate. This has provide us tangible hypotheses on how genetic circuits operate and how can they be modified to perform desired functions. Most of our studies on real cells have been based on these hypotheses and models. For this, we developed modelling strategies and simulators. 

Selected Publications on modelling strategies:

Selected Publications on simulators:

Software for image analysis of microscopy data

We have developed software for analyzing microscopy data: 

Selected Publications :

Signal Processing of Single-Cell, Single-Molecule Biology Data

Estimating RNA numbers in single cells by RNA fluorescent tagging and flow-cytometry 

We proposed a new methodology for quantifying mean, standard deviation, and skewness of single-cell distributions of RNA numbers, from flow cytometry data on cells expressing RNA tagged with MS2d-GFP. The quantification, based on scaling flow-cytometry data from microscopy single-cell data on integer-valued RNA numbers, produces precise, big data on in vivo single-cell distributions of RNA numbers and can assist in studies of transcription.

 A) Example microscopy image of cells carrying the reporter gene coding for MS2d-GFP, prior to the production of target RNAs. The cells are visible due to carrying a large amount of MS2d-GFP proteins; B) Example microscopy image of cells carrying the reporter gene coding for MS2d-GFP, after the production of target RNAs. The RNAs tagged with MS2d-GFP are visible as bright spots; (C) Mean total cell background fluorescence intensity and mean total fluorescence intensity of all RNA spots in individual cells (in arbitrary units), as measured by confocal microscopy (Methods, Section 2.5). Data from > 300 cells. The error bars are the standard error of mean. (D) Example image of a cell, along with the results of the segmentation of the cell border (blue line) and of the RNA spots within (red circles) using the tailored software ‘SCIP’ (Martins et al., 2018) (Methods, Section 2.5). (E) Left: example image of a cell along with a yellow line, manually introduced to obtain a fluorescence intensity profile using imageJ (Abramoff et al., 2004). Right: pixel intensity (in arbitrary units) along the yellow line shown on the left image. The peaks correspond to the regions where the two spots (tagged MS2d-GFP RNAs) are located. (F) Mean fluorescence intensity of individual tagged RNA molecules over time since first appearing. 10 tagged RNAs were tracked, all from cells with only one RNA. Also shown is the standard error of the mean (vertical bars). (G) Total fluorescence intensity (in arbitrary units) of cell populations over time, as measured by spectrophotometry, obtained from cells with target and reporter plasmids induced (brown line) and from cells with only the reporter plasmid induced (blue line).
(A) Mean, M, (B) Standard deviation, Sd, and (C) Skewness, S, of single-cell distributions of integer-valued RNA numbers obtained by microscopy, as a function of IPTG concentration (x-axis). The standard error of M, Sd and S of RNA numbers was estimated as described in Methods, section 2.7. (D) Mean, (E) Standard deviation, and (F) Skewness of the single-cell distribution of F/W values obtained by flow cytometry. The red error bars are standard errors of the statistics (Methods, Section 2.7). The blue error bars are the standard error of the statistics after adding variability estimated from eight technical replicates of cells carrying only the reporter gene (Supplementary Methods, Section 1.6).

Selected Publication:

VIII) Learning and Behavioral Changes using Mouse Paw-Preference as a case-study

We have being interested in Learning and Behavior. Paw preference is a very interesting case-study since the amount of learning (information) can be quantified from measurements, such as the one shown, where the choice of which paw the mouse uses at each reach, as he learns, can be observed, and the number of possible choices (left or right) can also be quantified. 

So far, some of our interesting results include:

Selected Publications:

The data is collected by our long-time collaborators Fred G. Biddle (picture below) and Brenda Eales.

Fred in his lab at the University of Calgary, Canada :-)