Background

The peptidoglycan (PG) sacculus is a mesh-like, essential macromolecule that encases the cytoplasmic membrane in most bacteria. It is composed of glycan chains made of alternating β-1, 4-linked N-acetylglucosamine and N-acetylmuramic acid residues connected via short peptides and is required to maintain the shape and osmotic stability of a bacterial cell (Typas et al., 2011). Growing and dividing bacteria synthesize new PG and incorporate it into their PG sacculus, and these processes are targeted by important antibiotics such as the β-lactams and glycopeptides. To visualize the growth sites on the PG sacculus Miguel de Pedro developed an elegant method in which D-cysteine (D-Cys) is first homogeneously incorporated into PG of growing bacteria (by then unknown enzymes), followed by a chase period in the absence of D-Cys. PG sacculi were isolated, followed by the selective biotinylation of D-Cys residues and the visualization of biotin with nano-gold labeled antibodies by electron microscopy (de Pedro et al., 1997). PG growth sites were characterized by regions of reduced label (when D-Cys became 'diluted' upon incorporation of new, label free PG during the chase) or without label (a zone of exclusively new PG). Although this method was crucial to determine the modes of PG segregation in model species such as Escherichia coli (de Pedro et al., 1997) and Caulobacter crescentus (Aaron et al., 2007), it was relatively cumbersome and used mainly in specialist laboratories. Over the last years, Erkin Kuru, Michael VanNieuwenhze and Yves Brun developed an ever-increasing palette of fluorescent D-amino acid (FDAA) probes that covalently label PG in cells and can therefore be used to visualize PG by fluorescence microscopy (Kuru et al., 2012 and 2015; Hsu et al., 2017). New 'rotor probes' of this series, rFDAA, produce fluorescence only upon incorporation into PG, enabling real-time visualization of probe incorporation and making it unnecessary to wash cells after the labeling procedure (Hsu et al., 2019). Presumably FDAAs are incorporated into PG by PBPs and LDTs, as are other D-amino acids (Cava et al., 2011; Lupoli et al., 2011). FDAAs efficiently label sites of active PG synthesis and have therefore been successfully used to track PG synthesis sites in different bacterial species (Kuru et al., 2012; Radkov et al., 2018). They are easy to use and reliably label the PG in many bacteria and they have contributed to major discoveries in a variety of bacterial species (Radkov et al., 2018), for example, the discovery of PG in Chlamydia trachomatis and its visualization (Liechti et al., 2014 and 2016), the tracking of septal PG synthesis dynamics in B. subtilis (Bisson-Filho et al., 2017) and the visualization of preseptal PG synthesis in E. coli (Pazos et al., 2018).

We recently showed that copper(II) inhibits LD-TPases in vivo and in vitro (Peters et al., 2018). Within this study we performed long-pulse in situ labeling of E. coli cells with the FDAA 7-hydroxycoumarincarbonylamino-D-alanine (HADA), which was previously shown to uniformly label cells (Kuru et al., 2012 and 2015). Indeed, using the published procedure wild-type cells showed homogeneous label along the lateral wall but, unexpectedly, constricting cells showed a reduced signal at the septum. No signal could be detected in a strain lacking all six LD-TPases (Peters et al., 2018). During cell growth PG synthesis takes place at the division septum due to the activities of PG synthases (PBP1B and PBP3) (de Pedro et al., 1997; Bertsche et al., 2006). We reasoned that the HADA signal at mid-cell was low despite the enhanced PG synthesis because the incorporated HADA might be removed by PG hydrolases during the time the cells were harvested and washed at neutral pH, resulting in the loss of label at the division septum. Indeed, labeling of chlamydial PG with fluorescent dipeptide required the inactivation of DD-carboxypeptidases by amplicillin (Liechti et al., 2014). Here we optimized the long-pulse HADA labeling protocol to reduce the potential loss of HADA label by PG hydrolases. Specifically, we stopped cell growth and label incorporation by adding sodium citrate buffer pH 2.25 to the growing cells and washed the cells once with sodium citrate buffer at pH 3.0, followed by two washing steps with phosphate-buffered saline (PBS, pH 7.4), prior to processing the samples for fluorescence microscopy. With this procedure cells of wild-type and several mutants were homogeneously labeled at the side-wall and had enhanced signal at the septum of constricting cells. We conclude that the rapid incubation and washes of cells at acidic pH preserves the HADA label by preventing the removal of incorporated HADA by PG hydrolases (Amanuma and Strominger, 1980; Stefanova et al., 2002), resulting in improved signal detection at cell division sites.