Affected energy metabolism under manganese stress governs cellular toxicity

Excessive manganese exposure is toxic, but a comprehensive biochemical picture of this assault is poorly understood. Whether oxidative stress or reduced energy metabolism under manganese exposure causes toxicity is still a debate. To address this, we chose Δmnt P Escherichia coli, a highly manganese-sensitive strain, in this study. Combining microarray, proteomics, and biochemical analyses, we show that the chronic manganese exposure rewires diverse regulatory and metabolic pathways. Manganese stress affects protein and other macromolecular stability, and envelope biogenesis. Most importantly, manganese exposure disrupts both iron-sulfur cluster and heme-enzyme biogenesis by depleting cellular iron level. Therefore, the compromised function of the iron-dependent enzymes in the tricarboxylic acid cycle, and electron transport chain impede ATP synthesis, leading to severe energy deficiency. Manganese stress also evokes reactive oxygen species, inducing oxidative stress. However, suppressing oxidative stress does not improve oxidative phosphorylation and cell growth. On the contrary, iron supplementation resumed cell growth stimulating oxidative phosphorylation. Therefore, we hypothesize that affected energy metabolism is the primal cause of manganese toxicity.

Low melting agarose was poured to remove any bubble between polyacrylamide and pH strip. The 2 nd dimension electrophoresis was performed at 100 volts till the dye front reaches at the bottom. 2D gels were stained with coomassie brilliant blue R (CBB-R) and documented, as represented in the Fig. 1B. The protein spots which are labeled in the gel ( Fig. 1B) were excised and trypsinized accordingly 59 .

In-gel trypsinization of 2D spots
Each spot was excised with sterile blade, collected in a microfuge tube and chopped into smaller pieces. All samples (spots) were washed with sterile water and resuspended in 200-500µl destaining solution (100mM ammonium bicarbonate and acetonitrile (ACN) mix (1:1)), for 30 minutes at room temperature. Tubes were centrifuged for 5000rpm for 1 minute and supernatant was discarded. The gel pieces were resuspended in 100-200µl ACN for another 30 minutes at room temparature. The centrifugation step was repeated, liquid discarded and gel pieces dried in speed vacuum centrifuge for 30 min at 30C. The protein samples were dissolved in 30-50µl of Trypsin-HCl buffer (13ng/µl trypsin) (Promega) prepared in 50mM ammonium bicarbonate and 1mM HCl and kept for 2 hours at 37C.
The tubes were centrifuged briefly and 5-10 µl of sample elutes were taken for MS/MS.

Tandem MS/MS analysis
All experiments of MS and tandem MS were performed on an AB SCIEX MALDI TOF/TOF 5800 mass spectrometer with delayed extraction time 450 nanosecond to obtain maximum resolution. Matrix was prepared from -cyano-4-hydroxyl cinnamic acid in 50% acetonitrile plus 0.1% TFA. Trypsinized samples and matrix were mixed in 1:1 ratio. MS data was acquired at a laser repetition rate of 400Hz with total 1600 laser shots/spectrum.
Tandem MS data was acquired at 1000Hz in 1KV MS/MS mode with 2000 laser 6 shots/spectrum. Experiments were performed without DynamicEXIT algorithm and interpretation method was set for 40 strong precursors. TOF/TOF Explorer Series software was used for data acquisition and processing. The MS/MS files were analyzed by Mascot program (http://www.matrixscience.com/) to identify the proteins.

ICP-MS to determine intracellular metal levels
To determine the cellular metal contents, we grew the mntP strain of E. coli in the presence or absence of manganese. After three hours, the cell pellets were harvested and washed twice in milliQ water containing 1mM EDTA to remove cell surface bound metal ions. The cell pellets were dissolved in 2 ml of clean double distilled water and protein content was measured from a portion of the suspension. The remaining amount of cell suspensions were digested with concentrated nitric acid and 30% H2O2. The digested samples were filtered to get rid of any particles and metal contents were determined by ICP-MS facility provided by Punjab Biotechnology Incubator, Mohali, India. The metal concentration in the cell was determined by normalizing them to a total intracellular protein concentration of 300mg/ml, as described previously 5 .

Aggregated protein isolation from the cell extract
Cell pellets from 50 ml mntP strain (Mn-fed or unfed) were harvested, suspended in 250 l of buffer A (10mM potassium phosphate buffer, pH 6.5, 1mM EDTA, 20% sucrose, 1mg/ml lysozyme) with protease inhibitor cocktail (Roche), and incubated for 30 minutes. 500 l of buffer B (10mM Potassium phosphate buffer, pH 6.5, 1mM EDTA) was added thoroughly; samples were flash frozen in liquid nitrogen, and sonicated briefly.
The cell lysates were centrifuged at 2000g for 10 minutes to remove cell debris and intact cells. The protein concentrations in the supernatants were estimated by Bradford reagent 7 (BioRad). 200 µg of total proteins from each samples were taken out and mixed with the loading dye to visualize in the SDS-PAGE. The lysed cells containing equal amount of proteins were centrifuged at 15000g for 45 minutes at 4C. The pellets with inclusion bodies or aggregated proteins were resuspended in 500 l buffer B and centrifuged again at 15000g for 45 minutes at 4C. Pellets were collected and washed with 10% NP-40 and centrifuged at 15000g to remove membrane components. Pellets were finally resuspended in 30 l buffer B for SDS-PAGE analysis.

Western Blotting
mntP cells were grown to log phase in the presence and absence of 1mM manganese for 2 hours. Subsequently, 20µg/ml chloramphenicol and 10 µg/ml tetracycline were added to the culture and cells were collected at 0 15, 30 min time points and washed with saline before sonication in lysis buffer (20 mM Tris, pH 8, 200 mM NaCl and protease inhibitor cocktail buffer). Protein concentration was estimated by Bradford and equal amount of lysate were loaded for control and treated samples in 12% SDS-PAGE. Semidry transfer system was used to transfer proteins onto nitrocellulose membrane for 1hr at constant volt (10V). The blot was blocked in 5% skimmed milk in TBST (Tris-NaCl buffer with 0.01% Tween-20) buffer, followed by incubation with primary antibody (rabbit anti-TF) (Sigma) and secondary antibody (Goat anti-rabbit HRP conjugate) for 1 hour at room temperature followed by three washes with TBST after each incubation. The blot was developed with WesternBright TM ECL western blotting detection kit (Advansta) and observed in BioRad Imager. Scion image software was used to measure anti-TF band intensities and normalizations were done against band intensity at 0 minutes. 8
The OD was recorded at 340nm. Absorbance is directly proportional to the cellular aconitase activity. Relative aconitase activity was calculated against per milligrams of cellular proteins.

Catalase activity assay
Catalase assay was performed by foam forming ability of the mntP strain carrying mutated plasmids in the presence of H2O2, as described 33 . Briefly, manganese-fed and unfed cells pellets were harvested from 10 ml log phase cultures and resuspended in 100l normal saline. 25 l was used to estimate the proteins and remaining portions were placed in the narrow glass test tubes. Addition of concentrated H2O2 produced foam in the tube.
The heights of the foam correspond to the activity of the cellular catalases. Relative catalase activity was calculated against per milligrams of cellular proteins.

NDH-1 and SDH activity assays
NDH-1 and SDH assays were performed as mentioned 57 . Briefly, untreated, manganese-fed and iron-supplemented cells were collected by centrifugation. The cell pellets were suspended in 50 mM MES buffer (pH 6.0) with 10% glycerol and lysozyme 50µg/ml, and disrupted by sonication. Cell debris were separated by low speed centrigugation. One-halves of the cell extracts were directly used to test NDH-1 activity.

9
The other-halves of the cell extracts were used to prepare membrane fraction centrifuging at 45,000 g for 2 hours. For NDH-1 activity assay, a specific substrate, deamino-NADH has been used. The oxidation of deamino-NADH was monitored using a spectrophotometer at 340 nm (εNADH = 6.22 mM -1 . cm -1 ), in a reaction mixture containing 50 mM MES (pH 6.0), 10 % glycerol, 200 μM deamino-NADH. Relative NDH-1 activities were determined against per milligrams of cellular proteins. For SDH activity assay, the membrane faction was suspended in the same buffer, and dichlorophenol indophenol (DCPIP) reduction was monitored in 50 mM Tris-HCl (pH 7.5), 4 mM succinate, 1 mM KCN for 30 min at 30°C.

Glutamate synthase (GS) assay
GS assays were performed as mentioned 58 . Briefly, growing cells were collected by centrifuging at 12000 g for 10 min. The cell pellets were sonicated in the buffer containing 0·1 M KCl and 0·5% (v/v) 2-mercaptoethanol, adjusted to pH 7·6, and centrifuged to remove cell drbris. GS activity was determined using a spectrophotometer by following the glutamine-dependent oxidation of NADPH at 340 nm. The reaction mixture contained 50 mM HEPES buffer pH 8·5, 1% (v/v) 2-mercaptoethanol, 3·65 mM glutamine, 3 mM 2-oxoglutarate, 0·2 mM NADPH and 0·1 ml cell-free extract in a final volume of 1 ml. Relative GS activities were normalized per milligram of cellular proteins.

Estimation of intracellular NAD and NADH
Abcam colorimetric assay kit was used to estimate NAD/NADH levels. 10 ml log phase culture of E. coli was grown in the presence or absence of 1mM MnCl2 or MnCl2+0.5mM FeCl2 for 2 hours and washed with cold PBS. Extraction of NADH/NAD were done by adding 400 µL of extraction Buffer followed by 4 freeze-thaw cycles on liquid nitrogen and by sonicating briefly. The lysates were centrifuge for 5 minutes at 4°C at 15000 rpm. The supernatants were collected and passed through 10kD spin column to remove NAD/NADH decomposing before performing the assay. Half of the filtrates were used directly with assay reagents to measure total NAD levels (NADt=NAD+NADH).
Remaining halves were incubated at 60C for 30 minutes to decompose NAD, and assay was performed to measure NADH levels. Relative NAD or NADH levels were calculated against per milligrams of cellular proteins.

Determination of intracellular ATP level
Relative ATP measurement was done using ATP Bioluminescence Assay Kit CLS II, Roche. Briefly, cell pellets were collected and washed in 1X PBS. The cell pellets were resuspended in pre-boiled ATP extraction buffer (100 mM Tris, pH 7.75 and 4mM EDTA, pH 8.0) and one small portion was taken out to measure protein concentration. The remaining portions were incubated for 2 minutes at 100°C. The samples were centrifuged for 5 minutes at 1000g and the supernatant was transferred to the fresh microfuge tubes. 50µl of sample and 50 µl of luciferase reagent were added in Black 96 well microplate and luminescence were recorded using a luminometer. The relative light unit (RLU) values were recorded. Standard curve was generated using known ATP concentrations. Finally, ATP levels were normalized against per milligrams of proteins.

Probing the intracellular and extracellular reactive oxygen species (ROS)
ROS species were detected by H2DCFDA and DHR123 fluorescent dyes (Thermofisher). To measure the intracellular ROS using flow cytometry, approximately equal number of cells were taken (different conditions) and pellet was done. Cell pellet was washed with 1X PBS and distributed in two parts. In one part 1X PBS was added and in another part equal volume of 10µM H2DCFDA dye was added. Staining was done for one hour and the data was acquired using FACS acuri (BD) at FL1 laser for 0.1 million cells.

Measuring relative intracellular pH changes
To measure the relative intracellular pH changes, fluorometric intracellular pH assay kit (Sigma-MAK150) was used. The protocol was little modified because kit is standardized for eukaryotic cells. Around 0.1 million control, manganese-fed and iron supplemented cells were harvested and washed in 1X PBS. After washing the cells were suspended in HHBS buffer supplied with Kit and distributed in two halves. In one half HHBS buffer was added and in another half dye loading solution was added (composition as described in kit). Staining was done for one hour and immediately the data was acquired using FACS acuri (BD) for one lakh cells using FL1 laser.

Estimation of intracellular pyruvate level
We used pyruvate colorimetric assay kit (Abcam). Cells were resuspended in pyruvate assay buffer and sonicated in ice for 2-4 min. Portions of lysates were taken out to measure the protein concentrations by Bradford method. The lysate was de-proteinised as recommended in the protocol. The samples were centrifuged at 13000 rpm for 15 min in cold before the assay was performed. 50µl of lysate was used for reaction with probe and enzyme mix provided in the kit following the mentioned protocol and blanks were set with pyruvate buffer as suggested. The OD was taken at 570 nm using a spectrophotometer.
Standard curve was prepared using known concentration of pyruvates. Pyruvate level was normalized against per milligrams of cellular proteins.

DNA damage studies by Confocal and FACS
To study DNA damage in presence of manganese and array of other chemicals (ROS scavengers and Spermidine) in combination with manganese confocal microscopy and flow cytometery was used. The cells were grown for 4 hours in the presence of 1 mM MnCl2 and pelleted. Pellet was washed with 1X PBS and the cells were fixed in 4% formaldehyde at 37C for 10 minutes. The cells were again pelleted, washed and finally dissolved in 1X PBS. 10µl sample was used to prepare slides and rest of the sample was used for flow cytometry study. FACS data was acquired using FACS acuri (BD) for 0.1 million cells using FL1 laser and imaging was done using Nikon confocal microscope using 488 laser. The relative cell lengths were measured by confocal images using Image J.

Measurement of survivability
Since ribosome exhibits chaperone function 62,63 , an elevated expression of ribosomal genes (Supplementary Table S1) might also be linked to the protein folding. Repressed profile of fimbrial genes (Supplementary Table S1) suggests why manganese stress reduces the adhesive properties of the cells. 1b and 1c). We reason that activation of proteolysis could limit the cellular concentration of these proteins.
To test the activated protein degradation, we focused on the TF chaperone levels in

Fig S1. qRT-PCR data validates microarray experiment
Bar diagrams represents manganese-mediated differential expression (fold changes) of the genes, as observed by qRT-PCR experiments. Corresponding differential expression under microarray has been shown in the parenthesis.   Table S4.