Using computer simulations to model bacterial microcompartment assembly

Bacterial microcompartments are protein shells found in bacteria that surround enzymes and other chemicals required for certain biological reactions.  For example, the carboxysome is a type of microcompartment that enables bacteria to convert the products of photosynthesis into sugars (thus taking carbon out of the atmosphere).  During the formation of a microcompartment, the outer protein shell assembles around hundreds of enzymes and chemicals required for the reaction.  Because the intermediates in this assembly process are small and short-lived, it is hard to study in detail using experiments. It is therefore useful to develop computational models that can help explain how proteins collect the necessary cargo, and then assemble into an ordered shell with the cargo on the inside.  The videos in this post show some examples of computer simulations of a model for bacterial microcompartment assembly, with each video corresponding to a different set of parameters that control the strengths of interactions among the proteins and cargo.

The study is described in the research article “Many-molecule encapsulation by an icosahedral shell” by Jason Perlmutter, Farzaneh Mohajerani, and Michael Hagan in eLife (eLife 2016;10.7554/eLife.14078).

Video 1: Multistep assembly of a microcompartment encapsulating hundreds of molecules (I) video1
Video 2: Multistep assembly of a microcompartment encapsulating hundreds of molecules (II)  video2
Video 3: Assembly of a microcompartment and encapsulation of hundreds of diffuse cargo molecules  video3

Deciding the fate of a stalled RNA polymerase

Ever wondered what happens when the transcription machinery runs into a DNA lesion or a protein roadblock? Alexandra M. Deaconescu, corresponding author and research associate in the Grigorieff laboratory together with HHMI Investigator and Biochemistry Professor Dr. Nikolaus Grigorieff and Dr. Irina Artsimovitch (Ohio State University) address this question in a new review “Interplay of DNA repair with transcription: from structures to mechanisms” featured in the latest issue of Trends in Biochemical Sciences. The review describes emerging mechanisms of transcription-coupled DNA repair with emphasis on the bacterial system.

Protein Flexing: A New Look at Transcription-Coupled DNA Repair

Alexandra M. Deaconescu, a research associate in the Rosenstiel Basic Medical Sciences Research Center and 2008-2010 Fellow of the Damon Runyon Cancer Research Foundation, together with Professor of Biochemistry and HHMI Investigator Nikolaus Grigorieff and collaborators in the Laboratory of Dr. Irina Artsimovitch at Ohio State University have just published a new study in PNAS, which delineates novel mechanistic details of transcription-coupled DNA repair.

In any cell, there is intense interplay between various DNA-based transactions, such as replication, transcription and DNA repair. More than twenty years ago, it was discovered that DNA lesions that cause stalling of RNA polymerase molecules elicit a form of preferential nucleotide excision repair (NER) that exists in both eubacteria and eukaryotes, and specifically targets the transcribed DNA strand. Termed transcription-coupled DNA repair (TCR), the process is found to be carried out in bacteria by an ATPase called Mfd or TRCF (see Figure, right). In TCR, TRCF performs two functions: 1) it recognizes a damage-stalled RNA polymerase (RNAP), then dissociates it off the DNA using energy derived from ATP hydrolysis and 2) it recruits DNA repair enzymes via binding to the UvrA subunit of the Uvr(A)BC NER machinery [1].   The Uvr(A)BC machinery is one of the main players in bacterial DNA repair, and distinguishes itself from other DNA repair proteins by its ability to repair a remarkably diverse repertoire of lesions by utilizing a “cut and patch” mechanism, whereby an oligonucleotide containing the damage is excised and the gap later filled.

The cellular role of TRCFs extends beyond TCR. Because of their ability to forward translocate and dissociate stalled RNAPs (or  “backtracked” RNAPs that have slid backwards on the template) [2], TRCFs are also involved in transcription elongation regulation [3, 4], resolution of head-on collisions of the transcription apparatus with the DNA replication machinery [5], and antibiotic resistance [6, 7]. In humans, the effects of impaired TCR are systemic and complex. Mutations in the transcription-repair coupling factor CSB lead to Cockayne Syndrome [8], a progeroid (accelerated-aging) disease characterized by severe developmental abnormalities and neurodegeneration, and whose etiology is currently poorly understood.

To elucidate the mechanism underpinning UvrA recruitment by TRCF, Deaconescu crystallized and solved the X-ray structure of a core UvrA-TRCF complex (Figure, left) demonstrating that UvrA binding involves unmasking of a conserved intramolecular surface within TRCF via a gating motion of the C-terminal domain (red in Figure above). Despite significant effort so far, Deaconescu is still trying to coax nucleotide-bound TRCF to form crystals suitable for X-ray diffraction. These would be highly informative because ATP is required for DNA binding, and its hydrolysis leads to TRCF translocation on dsDNA and ultimately release of RNAP off the damaged template.  Because diffracting crystals eluded her, and to further find out how ADP/ATP modulate the structure of TRCF, Deaconescu learned small-angle X-ray scattering techniques suitable for probing TRCF in solution in the absence and presence of nucleotides, thus circumventing the need for highly-ordered crystals. Then, the Brandeis team and their collaborators at Ohio State employed domain-locking disulfide engineering in conjunction with functional assays to gain a deeper understanding of what TRCF looks like during its catalytic cycle and upon binding to UvrA.  They find that the two main functions of TRCF (RNAP release and UvrA binding) can be uncoupled, suggesting that UvrA recruitment may only occur during/post RNAP release, and not upon RNAP binding as had been proposed earlier in the literature [9]. Furthermore, they show that the ternary elongation complex (consisting of RNAP, template and nascent RNA), but not naked DNA, significantly stimulates ATP hydrolysis by TRCF. Thus, bacterial TRCF operates in a manner reminiscent of that utilized by eukaryotic chromatin remodeling factors, and are preferentially stimulated by nucleosomes over naked DNA substrates.

Deaconescu previously “looked” at TCR using X-rays – as a graduate student she solved the first structure of an intact transcription-repair coupling factor from any organism using X-ray crystallography [10]. She now hopes to reconstitute the larger intermediates that form during TCR and bridge low- with high-resolution information using hybrid structural methods, particularly electron cryo-microscopy, and ultimately formulate a cogent model of how TRCFs operate in cells.

Bacteria have RNAs that sense fluoride, and channels that tranport it

Fluoride: unless you’re a synthetic chemist or a dentist, you probably don’t worry about this ion very often.  But, according to a new paper published in Science, bacteria do, and have done for a very long time.

The work, spearheaded by Ron Breaker’s group at Yale University, identified a novel RNA motif that selectively binds fluoride ion.  In response to Fbinding, this motif, called a riboswitch, undergoes a structural change that leads to increased transcription of downstream genes.  These genes encode crucial metabolic enzymes that are strongly inhibited by fluoride ion, like enolase and pyrophosphatase, as well as members of a family of chloride transport proteins, the CLC’s.  The CLC’s that are associated with F riboswitches are clustered together in a phylogenetic clade distant from well-characterized CLC’s.  Could these “chloride” channel proteins actually assist with fluoride export?  Randy Stockbridge, a Brandeis postdoc working in Chris Miller’s lab, contributed to the findings by showing that this subset of riboswitch-associated CLC’s do, in fact, transport F, whereas “conventional” CLC’s strictly exclude F.   The F riboswitches, and the F CLC’s, are found among a huge variety of bacteria and archaea, from plant and human pathogens to benign soil and seawater-dwelling bugs, leading to the inference that F toxicity has been a consistent evolutionary pressure.

You’re probably wondering just how much fluoride there is in the environment.  Fluoridated municipal drinking water contains about 80 micromolar F, and natural F- concentrations in the environment can be  higher and lower than that number.   In acidic environments especially, F might accumulate to much higher levels in bacteria.  With a pKa of 3.4, a small amount of F is present as HF at low pH, and the uncharged HF can diffuse cross the cell membrane into the cell.  Once in the cytoplasm, where the pH is around 7, HF dissociates, and F can’t diffuse across the membrane back into the environment.  Unless, of course, evolution has provided that bacterium a system to transport F out of the cell…

see also

Detecting Mutations the Easy Way

Recent Brandeis Ph.D graduate, Tracey Seier (Molecular and Cell Biology Program), Professor Sue Lovett, Research Assistant Vincent Sutera, together with former Brandeis undergraduates Noor Toha, Dana Padgett and Gal Zilberberg have developed a set of bacterial strains that can be used as “mutational reporters”.  Students in the Fall 2009 BIOL155a, Project Laboratory in Genetics and Genomics, course also assisted in the development of this resource. This work has recently been published in the journal Genetics.

These Escherichia coli strains carry mutations in the lacZ (β-galactosidase) gene that regain the ability to metabolize lactose by one, and only one, specific type of mutation. This set allows environmental compounds to be screened for effects on a broad set of potential mutations, establishing mutagen status and the mutational specificity in one easy step.

This strain set is improved over previous ones in the inclusion of reporters that are specific for certain types of mutations associated with mutational hotspots in gene. Mutations at these sites occur much more frequently than average and involve DNA strand misalignments at repeated DNA sequences rather than DNA polymerase errors. Such mutations are associated with human diseases, including cancer progression, and have been under-investigated because of the lack of specific assays. Using this strain set, Seier et al. also identified a mutagen, hydroxyurea, used in the treatment of leukemia and sickle cell disease, which affects only the “hotspot” class of mutations. This strain set, which will be deposited in the E. coli Genetic Stock Center,  will facilitate the screening of potential mutagens, environmental conditions or genetic loci for effects on a wide spectrum of mutational events.



Left: E. coli colonies showing lacZ mutant revertants (blue pimples) arising on a white colony on growth medium containing the beta-galactosidase indicator dye,  X-gal


Bacterial phenomics

As a self-confessed prokaryotic chauvinist, I’m always on the lookout for new interesting papers aimed at understanding bacterial metabolism and regulation. A recent paper in Cell, entitled “Phenotypic Landscape of a Bacterial Cell” by a group of authors including Biology Professor Susan Lovett demonstrates the application of high-throughput screening to finding new bacterial phenotypes. Approximately 4000 E.coli mutant strains, representing deletions of individual non-essential genes, were plated on 324 different media representing a total of over 100 different stress conditions, and the growth followed by image analysis. Approximately half of the genes screened had one or more identifiable phenotypic repsonses. This approach allows the identification of genes that are conditionally essential, genes that are involved in multiple resistance, etc. This represents a new automated method for identifying phenotypes (hence “phenomics’) and understanding the roles of genes of as yet unidentified function in bacteria. The data set is publicly available at

Protected by Akismet
Blog with WordPress

Welcome Guest | Login (Brandeis Members Only)