Analysis of Gram-negative Bacteria Peptidoglycan by Ultra-performance Liquid Chromatography

[Abstract] Bacteria are surrounded by a protective peptidoglycan cell wall. Provided that this structure and the enzymes involved are the preferred target for our most successful antibiotics, determining its structural and chemical complexity is of the highest interest. Traditionally, high-performance liquid chromatography (HPLC) analyses have been performed, but these methods are very time consuming in terms of sample preparation and chromatographic separation. Here we describe an optimized method for preparation of Gram-negative bacteria peptidoglycan and its subsequent analysis by ultra-performance liquid chromatography (UPLC). The use of UPLC in peptidoglycan analyses provides a dramatic reduction of the sample volume and hands-on time required and, furthermore, permits in-line mass spectrometry (MS) of the UPLC resolved muropeptides, thus facilitating their identification. This method improves our capability to perform high throughput analysis to better understand the cell-wall biology.

Cava and de Pedro, 2014; Yadav et al., 2018). Our method allows us to study the effect of genetic perturbations on cell-wall composition in a high-throughput manner. Furthermore, combining highthroughput PG analysis with genetic perturbations, we can explore the functions of different cell-wallmodifying enzymes. In fact, we have recently studied how class-A Penicillin-Binding Proteins (aPBPs) contribute to cell-wall integrity in Escherichia coli (Vigouroux et al., 2020). By combining PG analysis, morphology analysis and single-molecule tracking, we have demonstrated that the major aPBP PBP1b contributes to cell-wall integrity by repairing cell-wall defects, while the Rod complex governs rod-like cell shape.
Here we describe in detail the procedure for Gram-negative bacteria sacculi isolation, muramidase digestion and muropeptide separation by liquid chromatography. We also provide some general instructions on compositional and structural analysis. The isolation of PG relies on the insolubility of the sacculi in SDS, which allows a relatively easy method to obtain high amounts of PG, and on the availability of specific enzymes (muramidases or lysozymes) that split the MurNAc-(β1-4)-GlcNAc glycosidic bonds which hold the structure together and disassemble it into its individual subunits. The next step requires the use of sensitive and reliable methods that permit the resolution, identification and quantification of the different PG subunits. Since the 80's, HPLC has been traditionally used for this purpose (Glauner et al., 1988). While this technology was a revolution at the time and revealed an unexpected complexity in PG structure and composition, it has been used essentially unchanged for more than 30 years. However, there are three critical limitations: i) the requirement for inorganic buffers, incompatible with mass spectrometry in-line analysis that would facilitate identification of subunits; ii) the very low sample through-put, which requires a few days for sample preparation and several hours of HPLC run time per sample; and iii) the requirement for large sample volumes (injection volumes of 100-500 µl), due to the relatively low sensitivity of the chromatographic systems. The introduction of UPLC replacing its predecessor HPLC permits to dramatically reduce sample size (100x), while increasing the speed (20x), without compromising the quality of the data. The new systems allow the use of new and improved materials for reverse phase chromatography, including stationary phases with a very small particle size (in the range of 2 µm) that withstand very high pressures, increasing resolution, speed and sensitivity, which are essential requisites for high throughput analysis.
We have developed a new transformative approach that circumvents the limitations of the already available methods (e.g., Desmarais et al., 2013;Kühner et al., 2014). Our protocol i) can be easily adapted for the more frequently available UPLC machines by anyone with a basic knowledge of UPLC techniques; ii) has cut down sample preparation and run times dramatically, allowing for the processing and analysis of large numbers of PG samples per day; iii) is MS-compatible, samples can be collected and subjected to MS analysis without prior desalting steps or can even be analyzed in MS-systems coupled to the UPLC for a faster muropeptide identification, thus revealing a much higher PG chemical complexity than previously anticipated (i.e., identification of minor muropeptides); iv) and requires much less of sample, which is particularly important for in vivo samples (low sample size amount). 4 www.bio-protocol.org/e3780

Procedure
A. Cell lysate preparation 1. Grow cultures to the desired optical density in the appropriate culture medium for the bacteria.
Record the culture optical density for normalization. For optimal sample preparation, aim for 10 10 bacteria or higher (e.g., 10 ml at 1 unit OD600, 50 ml at 0.2 units OD600). Samples can be scaled up but require more reagents and are more time-consuming.
For statistical analysis, prepare and process samples in triplicates.    Figures 1C and 1D and measure absorbance at 204 nm. 6. Run a blank injection to monitor and subtract the baseline. 5. Contamination with other components or short re-equilibration time between runs also contribute to the appearance of ghost peaks.

Perform the LC run using the gradient described in
6. When sample concentration is low, the baseline drift and noise become more evident. Run a blank injection with water and subtract the baseline during the data processing.

Data analysis
Prepare and run samples in triplicates for statistical analysis.
For optimal comparison conditions, use the same volumes, washes and treatments for all samples.
A. Peptidoglycan profile 1. Extract the raw data: retention time (min) and absorbance at 204 nm (arbitrary units).
2. For baseline correction, subtract the data of a blank injection of water (Figure 2A).
3. Define the chromatographic processing regions removing not useful data, typically the injection front and the wash at the end of the run (datapoints ≤ 2-3 min, datapoints ≥ 12-13 min) ( Figure  9 www.bio-protocol.org/e3780   8. Represent the results as a PG feature table ( Figure 2D).
A real example of PG analysis is shown in Figure 3.  a. This reaction is highly exothermic, safety measures must be taken.
b. The NaOH solution is relatively unstable and its concentration can change over time, so for better accuracy, it is better prepared fresh.