Chronic exposure of soybean plants to nanomolar cadmium reveals specific additional high-affinity targets of Cd toxicity.

Solving the global environmental and agricultural problem of chronic low-level Cd exposure requires better mechanistic understanding. Here, soybean (Glycine max) plants were exposed to Cd concentrations ranging from 0.5 nM (background concentration, control) to 3 µM. Plants were cultivated hydroponically under non-nodulating conditions for 10 weeks. Toxicity symptoms, net photosynthetic oxygen production, photosynthesis biophysics (Chl fluorescence: Kautsky and OJIP), Cd binding to proteins (metalloproteomics by HPLC-ICPMS) and Cd ligands in LHCII (XANES), accumulation of elements, Chl and metabolites were monitored.

The following Supporting Information is available for this manuscript: In this file: Fig. S1: Nutrient utilization efficiency and utilization index of soybean plants exposed to various Cd concentrations for 10 weeks. : Acetone extracts from soybean leaves exposed to low, medium and high but sublethal Cd concentrations for 10 weeks with focus on the Chl molecules.   Protocol S1: Identification of isolated proteins from selected fractions of membrane proteins from the leaves of soybean plants exposed to 50 nM Cd.
Protocol S2: Metabolomics. Isolation and identification of (hydrophilic) metabolites from leaves of soybean plants exposed to various Cd concentrations.
Protocol S3: Lipidomics. Isolation and identification of lipophilic compounds from soybean leaves of soybean plants exposed to various Cd concentrations.

In separate Excel files:
Table S1 (Excel file): List of metabolites detected in leaves and roots of soybean plants exposed to various Cd concentrations for 10 weeks.  Table S4 (Excel file): Identification of membrane proteins and peptides isolated from leaves of soybean plants exposed to 50 nM Cd. The fractions containing LHC II trimers, as well as a LHC II neighbouring fraction eluting together with Cd were subjected to MALDI-TOF analyses.
In a separate pdf file: Table S5: Reports from all statistical tests that were used for the description of results in the manuscript. Figure S1: Nutrient utilization efficiency and utilization index of soybean plants exposed to various Cd concentrations for 10 weeks.

Figure S2
Acetone extracts from soybean leaves exposed to low, medium and high but sublethal Cd concentrations for 10 weeks with focus on the Chl molecules.  All mass spectra were calibrated internally using peptides from the auto digestion of trypsin.
The analysis by MALDI-TOF/TOF mass spectrometry produces peptide mass fingerprints and the peptides observed with a signal to noise greater than 12 can be collated and represented as a list of monoisotopic molecular weights.
Proteins ambiguously identified by peptide mass fingerprints were subjected to MS/MS sequencing analyses using the 4800 Proteomics Analyzer (Applied Biosystems, Framingham, MA).

Suitable precursors from the MS spectra were selected for MS/MS analyses by Colision
Induced Disociation (CID) using atmospheric gas and 1 Kv ion reflector operating mode. The precursor mass windows isolation was +/-4 Da. The plate model and default calibration were optimized for the MS-MS spectra processing. The parameters for the combined search (Peak list from Peptide mass fingerprint and MS-MS spectra) were the same as described above. In all protein identification the probability scores were greater than the score fixed by Mascot as significant with a p-value minor than 0.05.
Proteins that were not identified by Mascot database searching were subsequently subjected to de novo sequencing analyses, based on the fragmentation spectra of peptides, using DeNovo tool software (Applied Biosystems), tentative sequences were manually checked and validated.

Sample extraction
A slightly modified version of the protocol described by James et. al. (DOI: 10.1007/s11306-016-0956-2) was used to extract the metabolites. In short, the extraction consists of a beadbeating step and liquid-liqiud extraction using chloroform, ultrapure water and methanol.
The aqueous and organic phases are collected and dried. The dried organic phase has been used for the lipidomics analysis (see below) and the dried aqueous phase for all other analysis (after reconstituting in 200 μL ultra-pure water).

Quality control samples
For quality control, a mixed pooled sample (QC sample) was created by taking a small aliquot from each sample. This sample was analysed with regular intervals throughout the sequence.
Matrix effects have been tested for quantified compounds by spiking the QC sample in a minimum of two levels.

Target LC
For the detection of NADPH, AMP, ADP, ATP, Acetyl-CoA and Glyceraldehyde-3-phosphate a targeted LC-method were applied. The method is using a zwitterionic HILIC column and is based on the method described by West et al. (DOI 10.1007/s11306-016-0956-2). Besides from the compounds mentioned above also a range of sugars were extracted from the data obtained from this analysis.

Sample preparation
Samples reconstituted samples were diluted 5 times in eluent A prior to analysis.

Data processing
The data was analysed using both a targeted and an untargeted approach. The targeted approach was used to extract the response of compounds included in the standard list, which covered the 142 compounds listed below. For the untargeted approach feature extraction was conducted using mzMine. A feature is a peak characterized by a mass and a retention time. Since many compounds gives rise to a signal in more than one mass trace (e.g. naturally occurring C13 isotopes, adducts, and fragments) a compound will almost always be represented by more than one feature with the same retention time but different masses.
Features from mzMine will be identified with Xiiii (

GC-metabolites
Gas Chromatography -Mass Spectrometry (GC-MS) is a widely applied analytical tool in metabolomics. Due to its high separation power, its capacity for reliable identification of hundreds of metabolites and its low cost, GC-MS is often the first choice for metabolite analysis. However, GC-MS systems is limited to detect volatile compounds and, consequently, chemical derivatization of non-volatile compounds is required. The GC-metabolites method converts amino and non-amino organic acids into volatile carbamates and esters.

Sample preparation
Samples (see above) have been derivatised using MCF (methyl chloroformate).

Data processing
The large amount of raw GC-MS data is processed by software developed by MS-Omics and collaborators. The software uses the powerful PARAFAC2 model and can extract more compounds and cleaner MS spectra than most other GC-MS software (see their homepage www.msomics.com for more information).

Protocol S3: Lipidomics
Metabolites were extracted as above. The aqueous and organic phases were collected and dried.
The dried organic phase has been is used for the lipidomics analysis and the dried aqueous phase for all other analysis. Samples were reconstituted in an isopropanol / acetontril / watermixture.
Data were processed using Compound Discoverer 3.0 (ThermoFisher Scientific). First, features were extracted from the raw data. One compound often give rise to a signal in more than one mass trace (due to e.g. naturally occurring C13 isotopes, adducts, and/or fragments), a compound will therefore almost always be represented by more than one feature with the same retention time but different masses. The feature detection was followed by grouping of features belonging to the same compound. This additional information (e.g. isotope pattern) was then used together with the accurate mass to determine the molecular formula.
Lipids clearly not belonging to the plant kingdom were omitted from the analyses.
The total information collected for each compound were then used in the following identification step. For these data there are generally three levels of annotation: Level 1: Annotations on this level are the most secure identifications. They are based on three pieces of information: accurate mass, MSMS spectra and known retention time obtained from standards analysed on the same system.
Leve 2: Annotations on this level is based on two pieces of information; either accurate mass and MSMS spectra or accurate mass and known retention time as obtained from standards analysed on the same system.
Level 3: Annotations on this level is based on library searches using the accurate mass and elemental composition alone. Be aware that annotations on this level should be used with care, as more than one elemental composition could be matched with the same accurate mass, even with the high accuracy on the instruments we use, and it is impossible to distinguish between isomers on this annotation level.
For unidentified compounds the elemental composition is determined, if a proper match is found between the accurate mass obtained and the isotope pattern.