Discovery in microbiology has often benefitted from the development of new tools and techniques for describing bacterial cell processes like cell division. To that end, the genes and proteins involved in division site selection have painstakingly been discovered and their roles characterized for the past several decades. While the reported molecular mechanisms explain in part how division site selection occurs in some bacteria, they are unable to explain why physically division takes place in one specific place. We adapted in Georg Fantner’s engineering lab at the École Polytechnique Fédérale de Lausanne (EPFL) an atomic force microscope (AFM) that “feels” and probes the surface of rod-shaped actinobacteria, like Mycobacterium smegmatis, a model organism used in John McKinney’s microbiology lab for studying tuberculosis.
Mycobacteria are of particular interest because they do not possess any known systems governing division site selection. Thus, we reasoned that a non-conventional approach was needed for studying division site selection in mycobacteria. Together with Adrian Nievergelt, Mélanie Hannebelle, Joelle Ven and Pascal Odermatt, we adapted AFM to conducting long-term time-lapse imaging of mycobacteria allowing us to continuously follow bacterial growth and division for tens of generations corresponding to over one week of elapsed time. Long-term time-lapse AFM imaging was achieved by non-specifically immobilizing mycobacteria on a hydrophobic coverslip surface and imaging the cell surface with a tiny mechanical probe, called a cantilever. Our AFM movies offered us the first dynamic view of the mycobacterial surface at a resolution that is 100 times better than optical fluorescence microscopy amenable to time-lapse imaging.
We were surprised to observe an undulating surface morphology along the length of mycobacterial cells, composed of “wave-troughs” and “wave-peaks”. The wave-troughs are roughly repeating waves ~1.8 µm long and ~50 nm high, which is too small to resolve with conventional microscopy. That these morphological features were never before described is due to their minuscule size and that these surface features could never before be visualized dynamically.
Time-lapse images revealed that cells always divided within a wave-trough nearest to midcell and that “daughter” cells inherited the “mother” cell’s off-center wave-troughs. Wave-troughs are formed at the cell poles as a function of cell elongation and can exist for up to three generations before they are selected as the division site, making these morphological features the earliest known landmark of bacterial division site selection.
Having previously constructed a dual atomic force and optical fluorescence microscopy setup by Pascal Odermatt, a doctoral student in Georg Fantner’s group, we were able to time the appearance of morphological events apparent at the cell surface by AFM with the appearance of biological markers of division. Curiously, in mutant cells that undergo asymmetric divisions and are impaired for symmetrical chromosome partitioning, divisions also occurred at an off-center wave-trough corresponding to a local DNA minimum. Our work culminates in a model by which wave-troughs act as “licensed” sites of cell division and chromosome partitioning determines within which of the wave-troughs to divide.
We describe a physical mechanism for bacterial division site selection. It is the first such study to directly visualize and describe why division may occur at a specific site in mycobacteria. The descriptive nature of the physical mechanism we describe is in principle no different than a “molecular” mechanism that a conventional biology lab would report. No known “molecular” mechanism exists governing division site selection in mycobacteria and there is no evidence suggesting that such a mechanism should exist. A “physical” mechanism might offer a unique opportunity to describe why a process occurs in bacteria. Our work offers us a novel perspective on how one might go about finding a “molecular” mechanism responsible for division site selection in mycobacteria. A focus of future research is to identify whether there exists a relationship between wave-trough formation and the recruitment of division machinery.
The paper in Nature Microbiology is here: http://go.nature.com/2tYcqfY