The ubiquitous second messenger c-di-GMP controls growth and behavior of many bacteria. We have developed a novel Capture Compound Mass Spectrometry based technology to biochemically identify and characterize c-di-GMP binding proteins in virtually any bacterial species.
Considerable progress has been made during the last decade towards the identification and characterization of enzymes involved in the synthesis (diguanylate cyclases) and degradation (phosphodiesterases) of the second messenger c-di-GMP. In contrast, little information is available regarding the molecular mechanisms and cellular components through which this signaling molecule regulates a diverse range of cellular processes. Most of the known effector proteins belong to the PilZ family or are degenerated diguanylate cyclases or phosphodiesterases that have given up on catalysis and have adopted effector function. Thus, to better define the cellular c-di-GMP network in a wide range of bacteria experimental methods are required to identify and validate novel effectors for which reliable in silico predictions fail.
We have recently developed a novel Capture Compound Mass Spectrometry (CCMS) based technology as a powerful tool to biochemically identify and characterize c-di-GMP binding proteins. This technique has previously been reported to be applicable to a wide range of organisms1. Here we give a detailed description of the protocol that we utilize to probe such signaling components. As an example, we use Pseudomonas aeruginosa, an opportunistic pathogen in which c-di-GMP plays a critical role in virulence and biofilm control. CCMS identified 74% (38/51) of the known or predicted components of the c-di-GMP network. This study explains the CCMS procedure in detail, and establishes it as a powerful and versatile tool to identify novel components involved in small molecule signaling.
c-di-GMP is a key second messenger used by most bacteria to control various aspects of their growth and behavior. For instance, c-di-GMP regulates cell cycle progression, motility and the expression of exopolysaccharides and surface adhesins2-4. Through the coordination of such processes c-di-GMP promotes biofilm formation, a process which is associated with chronic infections of a range of pathogenic bacteria5. c-di-GMP is synthetized by enzymes called diguanylate cyclases (DGCs) that harbor a catalytic GGDEF domain4. Some DGCs possess an inhibitory site that down regulates the cyclase activity upon c-di-GMP binding. The degradation of c-di-GMP is catalyzed by two distinct classes of phosphodiesterases (PDEs) harboring either a catalytic EAL or HD-GYP domain6,7.
The majority of the known effector proteins that directly bind c-di-GMP belong to one of only three classes of proteins: catalytically inactive GGDEF or EAL domains and PilZ domains, small molecular switches that undergo conformational changes upon c-di-GMP binding8. DGCs, PDEs and PilZ proteins are well characterized and their domains can be predicted in silico relatively safely. A particular interest is now focused on the identification of new classes of c-di-GMP effectors. Some c-di-GMP effectors with different binding motifs were described recently such as the CRP/FNR protein family Bcam1349 in Burkholderia cenocepacia or the transcriptional regulator FleQ in P. aeruginosa9,10. In addition, c-di-GMP-specific riboswitches were recently identified and shown to control gene expression in a c-di-GMP-dependent manner11. The c-di-GMP binding motifs of different effectors are only poorly conserved making bioinformatic predictions of such proteins difficult. To address this issue, we developed a biochemical method, which is based on the use of a c-di-GMP specific Capture Compound combined with mass spectrometry 1,12,13.
We have recently engineered a novel trivalent c-di-GMP Capture Compound (cdG-CC, Figure 1)1. This chemical scaffold is composed of: 1) a c-di-GMP moiety used as bait to capture c-di-GMP binding proteins, 2) a UV-photoactivatable reactive group used to cross link the cdG-CC to the bound proteins and 3) a biotin to isolate the captured proteins using streptavidin-coated magnetic beads. The cdG-CC can be used to directly and specifically capture c-di-GMP effectors from complex mixture of macromolecules as cell lysates. Capture Compound based and chemical proteomics based approaches have previously been reported to be applicable to a wide range of organisms, e.g. Caulobacter crescentus, Salmonella enterica serovar typhimurium and P. aeruginosa1,14.
In this methodological paper, we provide an in depth description of the CCMS procedure using extracts of P. aeruginosa as an example. This study establishes CCMS as a powerful and versatile tool to biochemically identify novel components involved in small molecule signaling.
1. Lysate Preparation
2. Removal of Free c-di-GMP and Other Nucleotides (Soluble Fraction Only)
3. Pellet Resuspension and Solubilization (Membrane Fraction Only)
4. Protein Concentration Measurement
5. Capture
6. Washing Steps
NOTE: (Magnet: see Materials List). Start with a capture of the magnetic beads in the PCR strip lid, with the magnet. Then replace the PCR strip by a new one containing the next washing solution. Remove the magnet and resuspend the beads, and incubate 2 min. Spin down and replace the lid by a fresh lid.
7. MS Sample Preparation
8. LC-MS/MS Analysis
9. Database Search
10. Label-free Quantification
To identify novel c-di-GMP effectors in P. aeruginosa we systematically used CCMS to analyze the soluble and membrane fractions of P. aeruginosa strain PAO1 from a log phase culture (OD600 = 0.5). Here we summarize and discuss representative results of this fishing expedition. Four independent biological replicas were used. For each experiment two different cdG-CC concentrations were used (5 µM and 10 µM). To probe for specificity, experiments were carried out in the presence or absence of 1 mM c-di-GMP as competitor and, finally, with a bead control (i.e. without cdG-CC) (Table 1).
When following the method described in detail above (Figure 2) we produced a list of captured proteins in a Scaffold format. Likely contaminants were removed. This included ribosomal proteins, streptavidin, trypsin, serum albumin, keratin and other human proteins. The protein identification false discovery rate (FDR) was set to 1% by using the Scaffold software, and the data exported to excel. The accession number provided by Scaffold can be converted into locus numbers using the VLOOKUP function in Excel linked to a list of the P. aeruginosa locus numbers. At this stage, the hit list comprises 768 proteins for the soluble fraction and 433 proteins for the membrane fraction. However, most proteins are not significantly enriched in the capture experiment. Thus, proteins that are likely captured non-specifically (positive in the bead control or only in the presence of c-di-GMP competitor) were removed. We calculated a spectral count ratio between the capture experiment and the competition control and only considered proteins with a ratio larger than two. In addition we employed a paired t-test on spectral counts to provide a significance measure between the capture experiment and the competition control, and set a permissive threshold of 0.1. Lastly, we considered only robust hits with at least four peptides identified in the four experiments for the 2 capture compound concentrations taken altogether. These criteria should be adjusted according to the needs and using the verified and predicted c-di-GMP binding proteins as standard to set the threshold. After sorting, the list was decreased to 76 hits for the soluble fraction, and 133 proteins for the membrane fraction. This included 13 soluble and 21 membrane proteins from P. aeruginosa that are known or predicted to bind c-di-GMP (Table 2). The other 63 soluble and 112 membrane proteins are new putative c-di-GMP binding proteins that do not contain one of the known c-di-GMP binding domains. These hits have now to be validated by testing their specific binding to c-di-GMP.
In a previous screen we fished GlyA2 (PA2444), GlyA3 (PA4602) and Gsp69 (PA1127)1. These 3 proteins were cloned, overexpressed, and purified from E. coli and could be validated to bind c-di-GMP in UV-cross linking experiments using 33P labeled c-di-GMP15. The Kds were determined to 1.0, 2.0 and 6.9 µM respectively, indicating that indeed novel effectors can be identified by using CCMS.
In addition to this representative example, we used CCMS with extracts of cells harvested from different growth conditions and with various intracellular c-di-GMP concentrations (n = 24). Overall we captured 74% (38/51) of the known or predicted P. aeruginosa PAO1 c-di-GMP signaling components (24/32 soluble proteins, 14/19 membrane proteins). Given that at least nine of these genes were shown to be transcribed under specific conditions (oxidative stress, quorum sensing, biofilms)16 and that some may not bind c-di-GMP at all, this degree of coverage might be close to saturation. This together with the observation that most of these components were captured with high specificity (Table 2) strongly argues that this technique is effective and powerful.
Figure 1: Chemical structure of the c-di-GMP Capture Compound.
Figure 2: CCMS workflow summary. After mechanical lysis the free nucleotides are removed using a PD10 exclusion column. Proteins from the soluble or membrane fractions are incubated with the cdG-CC and the mixture is exposed to UV irradiation to cross-link captured proteins. Steps of harsh washing are carried with compounds bound to streptavidin coated magnetic beads. On-bead tryptic digestion provides peptides, which are then separated from the beads and protonated for their mass spectrometry identification. Please click here to view a larger version of the figure.
Figure 3: Volcanoplots of P. aeruginosa proteins significantly enriched by CCMS. Following LC-MS/MS analysis and label-free quantification, proteins were sorted as described in the text. Log2-intensity ratio of detected peptide between the capture and competition experiments were calculated and plotted versus values derived from significance analysis (modified t-statistic, empirical Bayes method17). Proteins within the significance thresholds for p-values <0.05 and intensity ratios >1.5-fold are indicated in a grey box. The 4 replicates for the soluble fraction (A) and the membrane fraction (B) were performed in the presence of 10 μM c-di-GMP-CC, and the competition experiment with 1 mM c-di-GMP. The circled dots correspond to known c-di-GMP binding proteins. Please click here to view a larger version of the figure.
Capture: | Buffer | Chemical | Source | Concentration |
Bacterial Lysis Buffer 10x | MES | Sigma | 67 mM | |
pH 7.5 | HEPES | Sigma | 67 mM | |
NaCl | Merck | 2 M | ||
Na Acetate | Merck | 67 mM | ||
DTT | Fluka | 10 mM | ||
DNaseI | Roche | 20 U/ml | ||
Complete Protease Inhibitor Cocktail | Roche | 1 tab / 10 ml | ||
Capture Buffer 5x | HEPES | Sigma | 100 mM | |
KAc | Sigma | 250 mM | ||
MgAc (anhydrous) | Sigma | 50 mM | ||
Glycerol | Sigma | 50% (V/V) | ||
GDP, GTP, ATP, CTP | sigma | 1 mM each | ||
Wash buffer 5x | Tris-HCl | Merck | 1 M | |
(for the soluble fraction) | EDTA | Sigma | 0.5 M | |
pH 7.5 | NaCl | Sigma | 5 M | |
n-octyl-β-D-glucopyranoside | Anagrade (Affymetrix) | 42.5 µM | ||
Wash buffer 5x | Tris-HCl | Merck | 1 M | |
(for the membrane fraction) | EDTA | Sigma | 0.5 M | |
pH 7.5 | NaCl | Sigma | 5 M | |
Other chemicals: | Ammonium Bicarbonate (ABC) | Fluka | ||
Urea | Applichem | |||
MS sample preparation: | Buffer | Chemical | Source | Concentration |
C18 Buffer A | TFA | Pierce | 0.1% (V/V) | |
H2O HPLC grade | 99.9% (V/V) | |||
C18 Buffer B | TFA | Pierce | 0.1% (V/V) | |
Acetonitrile | Biosolve | 49.9% (V/V) | ||
H2O HPLC grade | 50% (V/V) | |||
C18 Buffer C | TFA | Pierce | 0.1% (V/V) | |
Acetonitrile | Biosolve | 5% (V/V) | ||
H2O HPLC grade | 94.9% (V/V) | |||
LC Buffer A | Formic acid | Sigma | 0.15% (V/V) | |
Acetonitrile | Biosolve | 2% | ||
H2O HPLC grade | 97.85% | |||
Other chemicals: | tris(2-carboxyethyl)phosphine (TCEP) | Sigma | ||
iodoacetamide (IAA) | Sigma | |||
N-acetyl-cysteine | Sigma | |||
Endoproteinase Lys-C | Wako | |||
Trypsin | Promega |
Table 1: Buffers composition. Summary of the buffers composition, chemicals and suppliers. The c-di-GMP-capture compound, c-di-GMP (for the competition control), streptavidin coated magnetic beads, capture buffer, and washing buffer are included in the caproKit.
Bead control | Capture experiment | Competition control | |
Protein extract (10 mg/ml) | 30 µl | 30 µl | 30 µl |
c-di-GMP (10 mM) | 0 µl | 0 µl | 10 µl |
Nucleotides (10 mM of each nucleotide) | 10 µl | 10 µl | 10 µl |
Capture buffer 5x | 20 µl | 20 µl | 20 µl |
H2O | 42 µl | 32 µl | 22 µl |
30 min incubation | |||
c-di-GMP~CC (stock 100 µM) | 0 µl | 5-10 µl | 5-10 µl |
Final concentrations: | Bead control | Capture experiment | Competition control |
c-di-GMP~CC (µM) | 0 µM | 5-10 µM | 5-10 µM |
Competitor c-di-GMP (µM) | 0 µM | 0 µM | 1,000 µM |
Table 2: Capture reaction mix. Summary of the reaction mix for the bead control (i.e. without capture compound), the capture experiment (with the c-di-GMP-CC), and the competition control which contains a large excess of c-di-GMP. The c-di-GMP-CC final concentration can be adjusted, and is typically set between 5 and 10 µM.
Protein name | Locus ID | Domain architecture | capture experiment / competition experiment 1 | |||||||
a) Soluble fraction | cdG-CC = 5 µM | cdG-CC = 10 µM | ||||||||
– | PA4843 | REC-REC-GGEEF* | 14/0 | 14/0 | 13/0 | 11/0 | 14/0 | 12/0 | 14/0 | 14/0 |
WspR | PA3702 | REC-GGEEF* | 9/0 | 9/0 | 10/0 | 9/0 | 11/0 | 10/0 | 11/0 | 11/0 |
– | PA2567 | GAF-SPTRF-EAL | 8/0 | 4/0 | 9/0 | 0/0 | 7/0 | 3/0 | 8/0 | 8/0 |
– | PA3353 | PilZ | 11/0 | 12/0 | 13/0 | 12/0 | 12/0 | 10/0 | 11/0 | 12/0 |
– | PA0290 | PAS-GGDEF | 5/0 | 3/0 | 6/0 | 5/0 | 8/0 | 5/0 | 6/0 | 6/0 |
– | PA5295 | GDDEF-EAL | 3/0 | 3/0 | 3/0 | 1/0 | 6/0 | 6/0 | 5/0 | 4/0 |
FimX | PA4959 | PAS-GDSIF-EVL | 23/1 | 21/0 | 21/0 | 11/0 | 24/3 | 23/2 | 22/0 | 20/0 |
– | PA4608 | PilZ | 3/0 | 3/0 | 3/0 | 0/0 | 3/0 | 2/0 | 3/0 | 3/0 |
– | PA0012 | PilZ | 3/0 | 2/0 | 2/0 | 2/0 | 2/0 | 2/0 | 4/0 | 2/0 |
– | PA2989 | PilZ | 1/0 | 1/0 | 2/0 | 1/0 | 1/0 | 2/0 | 3/0 | 3/0 |
– | PA4324 | PilZ | 2/0 | 2/0 | 1/0 | 1/0 | 2/0 | 2/0 | 1/0 | 2/0 |
– | PA3177 | GGEEF | 2/0 | 1/0 | 3/0 | 0/0 | 1/0 | 1/0 | 3/0 | 1/0 |
– | PA4396 | REC-DEQHF | 0/0 | 1/0 | 4/0 | 0/0 | 1/0 | 0/0 | 5/0 | 1/0 |
– | PA0169 | GGEEF* | 3/0 | 2/0 | 6/0 | 7/0 | 7/1 | 6/2 | 9/1 | 7/1 |
– | PA2799 | PilZ | 1/0 | 0/0 | 2/0 | 0/0 | 0/0 | 0/0 | 3/0 | 1/0 |
– | PA5017 | PAS-GAF-PAS-ASNEF-EAL | 1/0 | 2/0 | 1/0 | 0/0 | 1/0 | 3/2 | 0/0 | 0/0 |
– | PA5487 | GGEEF* | 0/0 | 0/0 | 0/0 | 1/0 | 1/0 | 0/0 | 1/0 | 1/0 |
b) Membrane fraction | cdG-CC = 5 µM | cdG-CC = 10 µM | ||||||||
– | PA2072 | CHASE4-TM-PAS-GGDEF-EAL | 13/1 | 25/0 | 27/0 | 19/0 | 36/0 | 36/0 | 31/0 | 23/0 |
– | PA0861 | TM-PAS-GGDEF-ELL | 6/1 | 14/0 | 13/0 | 10/0 | 17/0 | 18/0 | 13/0 | 8/0 |
– | PA3353 | PilZ | 6/0 | 10/0 | 9/0 | 7/0 | 10/0 | 10/0 | 6/0 | 5/0 |
– | PA3343 | 5TM-GGDEF | 3/0 | 7/0 | 7/0 | 4/0 | 12/0 | 10/0 | 7/0 | 7/0 |
– | PA1181 | MASE1-PAS-PAS-PAS-PAS-GGDEF-ELL | 3/0 | 6/0 | 9/0 | 3/0 | 12/0 | 12/0 | 5/0 | 2/0 |
– | PA0847 | TM-CHASE4-HAMP-PAS-GGDEF | 0/0 | 4/0 | 4/0 | 1/0 | 15/0 | 13/0 | 8/0 | 6/0 |
– | PA0575 | PBPb-TM-PAS-PAS-PAS-PAS-GGDEF-EAL | 1/0 | 7/0 | 6/0 | 3/0 | 10/0 | 10/0 | 6/0 | 2/0 |
yfiN | PA1120 | 2TM-HAMP-GGDEF | 2/0 | 4/0 | 3/0 | 3/0 | 5/0 | 4/0 | 3/0 | 1/0 |
– | PA0290 | PAS-GGDEF | 1/0 | 4/0 | 3/0 | 2/0 | 1/0 | 5/0 | 1/0 | 2/0 |
– | PA4929 | 7TMR:DISMED2-7TMR:DISMED2-GGDEF | 2/0 | 4/0 | 2/0 | 2/0 | 2/0 | 2/0 | 2/0 | 1/0 |
morA | PA4601 | TM-TM-PAS-PAS-PAS-PAS-GGDEF-EAL | 3/0 | 7/0 | 7/3 | 5/0 | 9/0 | 10/0 | 5/0 | 4/0 |
– | PA1851 | 5TM-GGDEF | 1/0 | 2/0 | 1/0 | 2/0 | 4/0 | 3/0 | 1/0 | 1/0 |
– | PA2870 | TM-GGDEF | 0/0 | 0/0 | 1/0 | 0/0 | 4/0 | 4/0 | 4/0 | 2/0 |
– | PA3311 | TM-MHYT-MHYT-MHYT-AGDEF-EAL | 1/1 | 5/0 | 7/1 | 4/0 | 8/0 | 8/0 | 5/1 | 3/0 |
bifA | PA4367 | TM-GGDQF-EAL | 1/0 | 2/0 | 1/0 | 2/0 | 1/0 | 2/0 | 2/1 | 1/0 |
– | PA4608 | PilZ | 0/0 | 1/0 | 1/0 | 0/0 | 3/0 | 3/0 | 2/0 | 2/0 |
– | PA4332 | 5TM-GGEEF | 1/0 | 3/0 | 2/0 | 1/0 | 1/0 | 1/0 | 2/0 | 0/0 |
– | PA0012 | PilZ | 1/0 | 1/0 | 1/0 | 1/0 | 1/0 | 1/0 | 1/0 | 0/0 |
– | PA2989 | PilZ | 4/0 | 8/0 | 7/0 | 7/0 | 5/2 | 11/2 | 7/2 | 8/3 |
– | PA1433 | HAMP-RGGEF-KVL | 0/0 | 0/0 | 0/0 | 0/0 | 1/0 | 1/0 | 1/0 | 1/0 |
– | PA4843 | REC-REC-GGEEF | 0/0 | 0/0 | 1/1 | 1/0 | 0/0 | 1/0 | 1/0 | 0/0 |
* = GGDEF domain containing an I site | ||||||||||
1 number of spectral counts of identified peptides |
Table 3: P. aeruginosa known c-di-GMP signaling components specifically captured. Identified proteins were first sorted as described in the text. Proteins are identified with to their name and locus number, and we indicate their architecture predicted with the NCBI Conserved Domain Database online tool (n = 4, cdG-CC = 5 µM or 10 µM) in order to show the capture specificity and the reproducibility of the method.
Special care should be taken at several steps of the protocol. The protein concentration is a critical parameter with a concentration of 10 mg/ml being difficult to reach when cells are grown under specific growth conditions (e.g. biofilms or small colony variants). Thus, the pellet resuspension should be performed in a low volume of lysis buffer. Protein concentrations can be decreased to 8 mg/ml. Compared to the method published by Nesper et al.1, we added various nucleotides to the capture reaction to minimize non-specific capturing of nucleotide binding proteins. Although the addition of nucleotides improved the specificity, it may at the same time prevent the capture of proteins that bind different nucleotides at the same site. For example the effector FleQ, which was recently shown to bind ATP and c-di-GMP18 was fished specifically in the absence of ATP in our previous experiment1, but not anymore in the data presented here in the presence of an excess of ATP.
The cdG-CC should be carefully protected from light. Although ambient light contains only a small fraction of UV, it is recommended to keep the capture compound stock wrapped in aluminum foil, as well as the capture mix prior to activation by UV irradiation. The washing steps that follow can be very stringent to increase the specificity, as the captured proteins are covalently bound to the cdG-CC. Regarding the LC-MS/MS analysis, the experiments should be carried in a clean keratin free environment. Moreover, HPLC compatible buffers should be used, especially after the washing steps. The candidate list typically comprises between 300 and 800 proteins (for a quadruplicate), with low variations between the replicates (see Table 2 as an example).
Some parameters like the protein and the cdG-CC concentration might need to be optimized depending on the organisms analyzed. Since low abundant proteins or proteins expressed only under specific conditions can easily be missed, care should be taken regarding the culture conditions employed. This problem can be overcome by comparing the hit list with a global protein ATLAS collected for the same culture conditions. Finally, the optimization of the detergent can be challenging, as it needs to be optimized with respect to its ability to solubilize membrane proteins and also needs to be MS compatible.
One needs to keep in mind that the c-di-GMP molecule of the CC is chemically modified, as it is linked via the 2’OH group of one ribose to the rest of the scaffold. This modification could alter its ability to bind to some effectors thereby providing false-negatives. In this context it is noteworthy that we never captured proteins that harbor an EAL domain but lack a GGDEF domain in P. aeruginosa, although EAL proteins were captured in other species, like Caulobacter crescentus. This could be due to a poor access or a low affinity of the cdG-CC to the binding site, or to the degradation of the cdG-CC by EAL proteins. In contrast, many nucleotide-binding proteins were captured with a relatively low specificity and, to a large degree, are probably false positives. Further validation of the specific binding to c-di-GMP using techniques such as DRaCALA20, UV cross-linking15, differential scanning fluorimetry (DSF)21, Microscale thermophoresis (MST)22, isothermal calorimetry (ITC)23–24… is thus necessary.
It is also possible that a fraction of the c-di-GMP moiety of the capture compound is degraded by phosphodiesterases from the cell lysate. This is one of the reasons why the procedure has to be carried out at 4 °C, therefore limiting the phosphodiesterases activity before cross-linking.
The procedure can be adapted to many bacterial species, and has been successfully used for 3 different bacterial species with very minor modifications1. Capture Compound based technology can reduce false-positives by using thorough washing (e.g. 1 M salt, high detergent concentration, 2 M urea in case of membrane proteins, 80% acetonitrile), as compared to other techniques that do not rely on covalent binding. Given that validation of candidates can be a tedious and time-consuming process, this is a major advantage to alternative methods like chemical proteomics based approaches14.
This illustrated video method establishes CCMS as a powerful and versatile tool to identify and characterize novel components involved in small molecule signaling. In the future, similar Capture Compounds harboring other selectivity groups could be used to capture proteins involved in small molecule signaling, such as the novel c-di-AMP effectors.
The authors have nothing to disclose.
We thank Alberto Reinders for his work in optimizing the CCMS conditions for P. aeruginosa. We also thank Pablo Manfredifor the annotation of the P. aeruginosa proteins. This work was supported by the Swiss National Science Foundation (SNF) Sinergia grant CRSII3_127433.
caproBox | caprotec bioanalytics | 1-5010-001 (220 V) | UV lamps coupled to a cooling 96-plate cooling block, for the photoactivation |
caproMag | caprotec bioanalytics | included in the CCMS Starter Kit | For easy handling of magnetic particles without pipetting |
c-di-GMP caproKit | caprotec bioanalytics | upon request | The kit contains the c-di-GMP-capture compound, c-di-GMP (for the competition control), streptavidin coated magnetic beads, capture buffer, and washing buffer |
Disposable PD-10 Desalting Columns | GE Healthcare | 17-0851-01 | |
12-tube PCR strips | Thermo Scientific | AB-1114 | |
UIS250v sonicator with VialTweeter | Hielscher ultrasound technology | UIS250v and VialTweeter | |
Miniature French Pressure Cell | Thermo Electron Corporation | FA-003 |