This protocol describes how to perform cell-type-specific protein labeling with azidonorleucine (ANL) using a mouse line expressing a mutant L274G-Methionine tRNA synthetase (MetRS*) and the necessary steps for labeled cell-type-specific proteins isolation. We outline two possible ANL administration routes in live mice by (1) drinking water and (2) intraperitoneal injections.
Understanding protein homeostasis in vivo is key to knowing how the cells work in both physiological and disease conditions. The present protocol describes in vivo labeling and subsequent purification of newly synthesized proteins using an engineered mouse line to direct protein labeling to specific cellular populations. It is an inducible line by Cre recombinase expression of L274G-Methionine tRNA synthetase (MetRS*), enabling azidonorleucine (ANL) incorporation to the proteins, which otherwise will not occur. Using the method described here, it is possible to purify cell-type-specific proteomes labeled in vivo and detect subtle changes in protein content due to sample complexity reduction.
Aberrant protein homeostasis is caused by an imbalance in protein synthesis and degradation. Several diseases are related to alterations in protein homeostasis. The hallmark of some diseases is the presence of aggregates in different subcellular locations and brain areas. Protein homeostasis is not only important in disease but also plays a crucial role in normal organ and cellular function1. For example, protein synthesis is necessary for many forms of neuronal plasticity2,3, as determined by the use of chemical inhibitors that block protein synthesis4. However, it is neither clear in which cell-types the proteome is altered to support learning and memory, nor is it understood which specific proteins in each cell-type increase or decrease in their synthesis or degradation. Thus, a comprehensive study of protein homeostasis requires the capability to differentiate proteomes coming from specific cell-types. Indeed, the identification of cell-type-specific proteomes to study cellular processes occurring in a multicellular environment has been an important hurdle in proteomics. For this reason, we developed a technique using MetRS* expression combined with bio-orthogonal methods that has proven to be an effective way to identify and purify cell-type specific proteomes, filling this gap5,6,7.
The expression of a mutant MetRS* (MetRS L274G) allows for the loading of the non-canonical methionine analog ANL into the corresponding tRNA8,9 and its subsequent incorporation into proteins. When MetRS* expression is regulated by a cell-type-specific promoter, the non-canonical amino acid will be incorporated into the proteins in a cell-selective manner. Once ANL is incorporated in the proteins, it can be selectively functionalized by click-chemistry and subsequently either visualized by imaging (FUNCAT) or by Western Immunoblot (BONCAT). Alternatively, proteins can be selectively purified and identified by mass spectrometry (MS). Using this technology, we created a mouse line expressing the MetRS* protein under the control of the Cre recombinase. Considering the increasing number of available Cre-mouse lines, the MetRS* system can be used in any field to study any cell-type from any tissue for which there is an existing Cre-line. Protein labeling with ANL is possible in vitro or in vivo, and does not alter mouse behavior or protein integrity6. Labeling timespan can be adapted to the scientific question of each researcher, labeling newly synthesized proteins (shorter labeling times) or entire proteomes (longer labeling times). The use of this technique is limited by the number of cells of the type that the researcher is willing to study; hence protein isolation from cell-types with low numbers or low metabolic rates is not possible by this method. The goal of the presented method is to identify cell-type-specific proteins/proteomes labeled in vivo. In this protocol, we describe how to label cell-type-specific proteomes with ANL in live mice and purify the labeled proteins. After purification, proteins can be identified by routine mass spectrometry protocols5,10. The reduction of sample complexity achieved in this method by the selective purification of proteins from specific cellular populations allows the experimenter to detect subtle changes in proteomes, for example, in response to environmental changes. Purification of the labeled proteins can be achieved in ~10 days, not including the MS analysis or the labeling period. Here, we describe two methods for ANL administration to MetRS* expressing mice, namely (1) adding the amino acid in the drinking water, and (2) introducing ANL by intraperitoneal injections. Regardless of the method chosen for ANL administration, the isolation and purification steps are the same (from step 2 on).
All experiments with animals were performed with permission from the local government offices in Germany (RP Darmstadt; protocols: V54-19c20/15-F122/14, V54-19c20/15-F126/1012) or Spain (Committee of Animal Experiments at the UCM and Environmental Counselling of the Comunidad de Madrid, protocol number: PROEX 005.0/21) and are compliant with the Max Planck Society rules and Spanish regulations and follow the EU guidelines for animal welfare.
1. In vivo metabolic labeling with ANL
2. Tissue harvesting, lysis, and protein extraction
3. Protein purification
Following the described protocol (summarized in Figure 1), ANL was administered to mice either by daily intraperitoneal injections (400 mM ANL 10 mL/kg, Nex-Cre::MetRS*) for 7 days, or via drinking water (0.7% Maltose, 1% ANL, CamkII-Cre::MetRS*) for 21 days. After labeling, the corresponding brain areas were dissected, lysed, alkylated, and clicked. Click reactions were analyzed by SDS-PAGE and Western Immunoblot. Representative images of the experiments are shown in Figure 2 for ANL administration by IP injection and in Figure 3 for ANL administration via drinking water. Note that the aim of the figures is to show that both labeling protocols work, not to compare them. Comparison is not possible because different neuronal populations are labeled in the two experiments shown (excitatory neurons of the hippocampus and cortex, and the Purkinje neurons of the cerebellum).
A further experiment provides an example for the determination of the optimal DST-cleavable alkyne concentration (Figure 4). In this example, the fold change between labeled sample and control is highest at an alkyne concentration of 14 μM. This alkyne concentration is then applied to all samples. After verification of the labeling of each sample in the experiment with the chosen alkyne amount, all samples were subjected to the purification of ANL-labeled/biotin-clicked proteins by affinity binding using Neutravidin beads (Figure 5). In this step, proteins are bound to the beads, washed, and subsequently eluted by reducing the disulfide bridge present in the DST-alkyne. After this last step, the biotin and part of the alkyne remain bound to the beads. The ANL-containing proteins (bound to the rest of the alkyne) are recovered in the elution buffer. To evaluate the efficiency of the elution step, one-third of the eluate volume is loaded onto an SDS-PAGE gel and visualized using a sensitive total protein stain method. At least a 3-fold intensity difference of total protein stain between negative controls and ANL-labeled samples must be observed to achieve interpretable results with mass spectrometry. Completion of all steps described in this protocol will be followed by MS sample preparation, acquisition, and analysis5. Although there are no special requirements for MS (each lab may use its routine MS protocol5,10), it should be kept in mind that the amount of purified proteins will be generally low (in the order of nanograms). Figure 6 provides an example of MS results with a clear enrichment in the ANL-labeled sample as compared to the control (Figure 6A). This difference was already visible in the total protein stain shown in Figure 5. Besides changes in peptide intensity, there are also unique proteins found in both samples (Figure 6B).
Figure 1: Work pipeline for cell-type-specific protein purification by BONCAT. Following protein labeling with ANL, the tissue of interest is dissected and lysed, tagged with biotin by click chemistry, and the amount of labeling for each sample is evaluated by BONCAT. Outliers (that failed to incorporate ANL) and representative samples can be differentiated in this step. One of the representative samples is clicked again to find the optimized cleavable alkyne dosage for subsequent protein purification. The alkyne dosage achieving the best signal-to-noise ratio is applied to every biological replicate. Cell-type-specific proteins are obtained by affinity purification. Purified samples are studied by mass spectrometry, and proteins are identified. This figure has been modified from Alvarez-Castelao et al. (2017)6. Please click here to view a larger version of this figure.
Figure 2: ANL administration by intraperitoneal injections (IP). BONCAT was performed to evaluate protein labeling in the hippocampus (HP) and cortex (CX) following metabolic labeling of proteins with ANL by daily IP injections for 7 days. Wild-type mice samples as negative control (wt) were clicked in parallel to the labeled samples (MetRS*). Labeled proteins are from the excitatory neurons using a Nex-Cre::MetRS* line16. Please click here to view a larger version of this figure.
Figure 3: ANL administration by drinking water. BONCAT was performed to evaluate protein labeling in the Purkinje neurons of the cerebellum after administration of ANL to GAD-Cre::MetRS* mice via drinking water for 21 days. The figure shows a negative control (wt) and a labeled sample (MetRS*) lysed and clicked in parallel. This figure has been modified from Alvarez-Castelao et al. (2017)6. Please click here to view a larger version of this figure.
Figure 4: DST-alkyne dosage titration. One biological replicate per experiment is chosen as a representative sample to titrate the optimal alkyne concentration. Three concentrations (14, 28, and 56 μM) of the alkyne were tested here in the click reaction for BONCAT. The labeling ratio between the ANL-labelled sample (MetRS*) and the negative control (wt) determines the signal-to-noise ratio (shown in the graph). 14 μM is the best alkyne concentration obtained in this experiment. Cortex tissue from the ANL-labeled Nex::MetRS* mouse line was used for this experiment. Please click here to view a larger version of this figure.
Figure 5: Purified proteins. Labeled proteins from Purkinje neurons were purified and stained using SYPRO Ruby. This figure has been modified from Alvarez-Castelao et al. (2017)6. Please click here to view a larger version of this figure.
Figure 6: Protein identification and quantification. The Purkinje cell proteome was obtained by mass spectrometry using cerebellum from a GAD-Cre::MetRS* mouse line as starting material. (A) shows increased abundances (peptide intensity) of the identified proteins in ANL labeled samples compared to wild-type (wt) mice. (B) The Purkinje proteome was obtained by pooling proteins enriched in the ANL-labeled samples in A (defined by >3-fold intensities in MetRS* mice compared to wt), and unique proteins found in the MetRS*-expressing mice. This figure has been modified from Alvarez-Castelao et al. (2017)6. Please click here to view a larger version of this figure.
The critical aspects of the protocol are; the inclusion of negative controls, having enough biological replicates, ANL administration route, amount, and duration, alkylation of the samples, alkyne concentration, and β-mercaptoethanol elimination when using DST-alkyne.
It is key to include negative control samples proceeding from ANL-labeled animals without Cre driver and therefore no MetRS* expression. These samples must be subjected to every step described in the protocol in parallel to the samples from ANL-labeled Cre-induced MetRS* mice. Non-clicked samples are not valid controls, as any results achieved using only this control may be due to non-specific click. We have not observed incorporation of ANL in cells not expressing the MetRS* gene, nor MetRS* expression in cells with no Cre; thus, wt animals labeled with ANL can be used as a negative control.
Given that the protocol hereby described consists of many steps where samples can be lost, it is recommended to calculate biological replicates bearing technical failure in mind. For brain samples, there is approximately 30% replicate loss.
We describe here two ANL administration routes and durations. This does not exclude that other administration routes (e.g., adding ANL to the food), work equally well. With respect to the administered ANL amount, the experiments shown here were performed using relatively high amounts of ANL. Labeling with lower amounts is also possible depending on the experimental set-up. The shortest time span of ANL labeling reported in this protocol is 1 week, and the longest 21 days. Shorter or longer labeling periods could be applied, but we have not determined it. Researchers should determine the ANL administration route, ANL dosage, and ANL labeling period adequate for the specific experimental question under study by considering the metabolic properties of the tissue, the cell-type of interest, and the experimental question under study.
Sample alkylation, also known as capping, is an important step to avoid background click17. When alkylation is properly done, the non-specific click as monitored by the control samples is reduced, and the difference between the negative control and ANL-labelled samples is larger. Elimination of free IAA after alkylation is also a key step. The presence of small amounts of IAA during the click reaction will limit its efficiency. Occasionally, the desalting step which also eliminates IAA (step 2.6), has to be repeated to ensure proper removal of IAA.
There are several biotin-alkynes and -DBCO reagents commercially available from a variety of companies. The main differences among them regard the polyethyleneglycol (PEG) linker chain length and the absence, presence, and type of cleavable groups. Regardless of the type used, excess amounts of alkyne lead to non-specific click to proteins bearing only natural amino acids. This can be easily prevented by precise and careful titration of the alkyne amount. As described, the best way to achieve this is using the actual lysed and alkylated samples to be used in the subsequent protein purification steps (step 2.6) and mass spectrometry analysis.
When DST-alkynes are used in the click reaction, β-mercaptoethanol reduces the disulfide bridge and dissociates the labeled proteins from the beads in the elution step (step 3.3). β-mercaptoethanol must be removed from the samples to allow the enzymatic digestion needed for MS. There are several ways to do this (e.g., lyophilization18, excision from a gel19, or cleaning using S-Trap columns20), and the choice should be made by the MS laboratory processing the samples.
Modifications and troubleshooting of the method should be directed to optimize the critical steps mentioned in the protocol. For instance, if there is too much variability, add more biological replicates; if the labeling is too low, increase ANL concentration, use longer labeling time spans, or test other ANL administration routes that could be more efficient for the studied tissue. For protein alkylation, a previous reducing step to the addition of IAA could be implemented.
The main limitation of the method is the purification of proteomes arising from small cell populations even if a small tissue area is dissected. Proteomes of more numerous cell populations that are, however, spread out over larger tissue areas are also difficult to access. Nevertheless, the in vivo labeling technique described here can still be applied to assess the referred cell-types by imaging (e.g., with FUNCAT or FUNCAT-PLA21).
This method is currently the only method available for in vivo study of cell-type-specific proteomes based on full-length proteins and for in situ visualization of cell-type-specific complete proteins. Other non-canonical amino acids such as azidohomoalanine (AHA) can also be used for in vivo protein labeling and purification of proteomes, but lack cell cell-type specificity22,23. Other approaches such as puromycin are suitable for providing a fixed picture of protein synthesis24,25. Nevertheless, using non-canonical amino acids for longer periods of labeling is possible, also reflecting protein degradation and showing a more accurate cellular proteome. BioID based methods are used to identify proteins in specific subcellular regions, regardless of their moment/place of synthesis26.
In the mouse line that we established (available from Jackson Lab, stock no. 028071), the mutant methionine tRNA synthetase (MetRS*) that allows for the incorporation of ANL into the cells is expressed in a Cre dependent manner. With this mouse line as a tool, it is possible to exclusively express the MetRS* in cell-types for which Cre-driver lines are available. This arrangement endows the method with considerable versatility, making it useful in almost every area of biomedical research.
The authors have nothing to disclose.
B.A-C is funded by the Spanish Ministry of Science and Innovation (Ramón y Cajal-RYC2018-024435-I), by the Autonomous Community of Madrid (Atracción de Talento-2019T1/BMD-14057), and MICINN (PID2020-113270RA-I00) grants. R. A-P is funded by Autonomous Community of Madrid (Atracción de Talento-2019T1/BMD-14057). E.M.S. is funded by the Max Planck Society, an Advanced Investigator award from the European Research Council (grant 743216), DFG CRC 1080: Molecular and Cellular Mechanisms of Neural Homeostasis, and DFG CRC 902: Molecular Principles of RNA-based Regulation. We thank D.C Dieterich and P. Landgraf for their technical advise and the synthesis of the DST-Alkyne. We thank E. Northrup, S. Zeissler, S. Gil Mast, and the animal facility of the MPI for Brain Research for their excellent support. We thank Sandra Goebbels for sharing the Nex-Cre mouse line. We thank Antonio G. Carroggio for his help with English editing. B.A-C. designed, conducted, and analyzed experiments. B.N-A, D.O.C, R.A-P, C. E., and S. t. D. conducted and analyzed experiments. B.A-C and E.M.S. designed experiments, and supervised the project, B.A-C wrote the paper. All authors edited the paper.
12% Acrylamide gels | GenScript | SurePAGE, Bis-Tris, 10 x 8, 12% | |
β-Mercaptoethanol | Sigma | M6250 | Toxic; use a lab coat, gloves and a fume hood. |
Ammonium bicarbonate | Sigma | 9830 | Toxic; use a lab coat, gloves and a fume hood |
ANL | Synthesized as described previously for AHA (see references 5 and 11) | ||
ANL-HCl | IrishBotech | HAA1625.0500 | |
Benzonase | Sigma | E1014 | |
Biotin alkyne | Thermo, | B10185 | |
Chicken antibody anti-GFP | Aves | 1020 | |
Complete EDTA-free protease inhibitor | Roche | 4693132001 | Toxic; use a lab coat and gloves. |
Copper (I) bromide | Sigma | 254185 | 99.999% (wt/wt) |
Disulfide tag (DST)-alkyne | Synthesized as reported in reference number 15, in which it is referred to as probe 20 | ||
DMSO | Sigma | 276855 | |
Filters | Merk | SCGP00525 | |
Iodo acetamide (IAA) | Sigma | I1149 | |
IR anti chicken 800 | LI-COR | IRDye 800CW | Donkey anti-Chicken Secondary Antibody |
IR anti rabit 680 | LI-COR | IRDye 680RD | Goat anti-Rabbit IgG Secondary Antibody |
Maltose | Sigma | M9171 | |
Manual Mixer | BioSpec Products | 1083 | |
NaCl | Sigma | S9888 | |
N-ethylmaleimide | Sigma | 4259 | Toxic; use a lab coat, gloves and a fume hood. |
NeutrAvidin beads | Pierce | 29200 | |
Nitrocellulose membrane | Bio-rad | 1620112 | |
PBS 1X | Thermo | J62036.K2 | |
PBS 1X pH 7.8 | Preparation described in reference number 5 | ||
PD SpinTrap G-25 columns | GE Healthcare | Buffer exchange | |
Pierce BCA Protein Assay Kit | Thermo, | 23225 | Reagents in the Pierce BCA Protein Assay Kit are toxic to aquatic life. |
Polyclonal rabbit anti-biotin antibody | Cell Signaling | 5597 | |
PVDF membrane | Millipore | IPVH00010 | |
SDS 10% | Sigma | 71736 | |
SDS-PAGE Running buffer MOPS | GenScript | M00138 | |
SYPRO Ruby stain | Sigma | S4942 | |
Table automatic Vortexer | Eppendorf | Mixmate | |
Triazole ligand | Sigma | 678937 | |
Triton X-100 | Sigma | T9284 | |
Water | Sigma | W4502 | Molecular biology grade |