Optimized APEX2 peroxidase-mediated proximity labeling in fast- and slow-growing mycobacteria

Abstract

Proximity labeling is a technology for tagging proteins and other biomolecules in living cells. These methods use enzymes that generate reactive species whose properties afford high spatial resolution for the localization of proteins to subcellular compartments and the identification of endogenous interaction partners. Here we present the adaptation of the engineered peroxidase APEX2 to proximity labeling in mycobacteria, including the human pathogen Mycobacterium tuberculosis. APEX2 is uniquely suited for general use in bacteria because unlike other proximity labeling enzymes, it does not depend on metabolites like ATP that are found in the cytoplasm, but are absent from the bacterial periplasm. Importantly, we found that in slow-growing mycobacteria like M. tuberculosis, codon usage optimization is required for APEX2 export into the periplasm via fusion to an N-terminal secretion signal. APEX2 expressed from codon-optimized genes affords robust, compartment-specific protein labeling in the cytoplasm and the periplasm of both fast- and slow-growing species. Here we detail these updated constructs and provide an optimized protocol for APEX2-mediated protein labeling in mycobacteria. We expect this approach to be broadly useful for determining the localization of specific proteins, cataloguing subcellular proteomes, and identifying interaction partners of ‘bait’ proteins expressed as fusions to APEX2.

Introduction

Proximity labeling methods have afforded unprecedented spatial resolution of protein localization and the detection of protein-protein and protein-nucleic acid interactions in live cells. These methods take advantage of promiscuous enzymes that generate reactive species, which then modify targets of interest such as protein sidechains and RNA nucleobases. Biotin ligases create activated esters that modify exposed primary amines (i.e., lysine sidechains) and peroxidases generate radicals that label exposed aromatic groups (e.g., tyrosine sidechains). The short lifetimes and limited membrane permeation of these reactive species affords resolution on the order of nanometers and specificity to membrane-bound compartments such as organelles. We recommend this recent overview of proximity labeling approaches and their applications (Qin et al., 2021). Here we detail the use of the engineered peroxidase APEX2 to label proteins in mycobacteria and report modifications necessary for proximity labeling in the periplasm of slow-growing species, including the human pathogen Mycobacterium tuberculosis.

We and others have previously reported proximity labeling of proteins in the fast-growing non-pathogenic species Mycobacterium smegmatis. Veyron-Churlet et al. used biotin ligase-mediated labeling (BioID) to identify proteins proximal to the heparin-binding hemaglutinin protein HbhA (Veyron-Churlet et al., 2021). In our work we favor the engineered APEX2 peroxidase for proximity labeling primarily because APEX2 is broadly applicable to all bacterial compartments. Mycobacteria, although phylogenetically classified as gram-positive organisms, are surrounded by an additional lipid layer known as the mycomembrane and thus possess a periplasm. Unlike biotin ligases, peroxidases do not require metabolites such as biotin and ATP that are believed to be absent from the periplasm (Wulfing & Pluckthun, 1994). In Ganapathy et al., we showed compartment-specific labeling by APEX2 in the cytoplasm or periplasm of M. smegmatis (Ganapathy et al., 2018). Tagging of periplasmic proteins was achieved by APEX2 fused to the N-terminal secretion signal sequence from M. tuberculosis Mpt63 or BlaC, two validated substrates of the Sec and Tat protein secretion systems, respectively (McCann et al., 2007; McDonough et al., 2005). All constructs yielded robust protein labeling with biotin-phenol and a distinct pattern of labeling for Sec-APEX2 and Tat-APEX2, consistent with modification of periplasmic proteins (Figure 1A). In addition, we showed that comparing labeling by cytoplasmic and periplasmic APEX2 determines the localization of specific targets: Following enrichment for the label, one can assay for a protein of interest via a native antibody (Figure 1B) or an epitope when the target is co-expressed with APEX2 as a tagged fusion (Figure 1C). APEX labeling thus provides an accurate alternative to fractionation or extraction methods that are commonly used to localize proteins within mycobacteria.

Figure 1. Compartment-specific biotinylation of proteins in M. smegmatis by the engineered peroxidase APEX2. M. smegmatis expressing APEX2, Sec-APEX2 or Tat-APEX2 was grown without or with theophylline and subjected to the labeling protocol with biotin-phenol.

(A) Streptavidin blot analysis of total lysates was used to detect protein biotinylation. Asterisks and arrowheads indicate examples of APEX2 expression-independent and -dependent bands; respectively. Data are representative of >3 independent experiments. (B) Biotinylated proteins were enriched by avidin affinity purification and analyzed by immunoblot with antibody against the M. tuberculosis antigen 85 complex, a group of three homologous proteins expressed in the cell wall. Purified M. tuberculosis Ag85A (34 kDa) was included as a positive control for the antibody. All data are from the same image; intervening lanes were removed for clarity. Data are representative of 2 independent experiments. (C) M. smegmatis expressing APEX2 or Sec-APEX2 from a multi-copy episomal plasmid and the cell wall protein-epitope fusion LprG-3XFLAG (27 kDa) or NA-LprG-3XFLAG (which lacks the N-terminal secretion signal (Drage et al., 2010) and thus accumulates in the cytosol; 24 kDa) from an integrated chromosomal copy were grown without or with theophylline and subjected to the labeling protocol with biotin-phenol. Biotinylated proteins were enriched by avidin affinity purification. Anti-FLAG immunoblot analysis with chemiluminescence detection was used to assess expression and enrichment of LprG-3XFLAG and NA-LprG-3XFLAG. Data are representative of 3 independent experiments. Lane labels I and O indicate input and output for avidin enrichment. Figure adapted and used with permission (Ganapathy et al., 2018).

Biotin-phenol is the substrate most widely used with APEX2 and results in biotinylated proteins that are conveniently detected and enriched via binding to streptavidin reagents. However, mycobacteria natively biotinylate some proteins, complicating the identification of APEX2-dependent labeling. Therefore, we also reported tyramide azide and tyramide alkyne as alternate substrates that enable Cu(I)-mediated azide-alkyne cycloaddition (CuAAC) to a wide variety of commercially available reagents for detection and enrichment (Ganapathy et al., 2018). Based on a qualitative assessment of the relative degree of labeling by the azide and alkyne reagents, we recommend the use of tyramide azide in mycobacteria.

In more recent work we found that labeling by APEX2 expressed in the cytosol was also robust in slow-growing M. bovis BCG and M. tuberculosis (Mtb). However, neither Sec-APEX2 nor Tat-APEX2 yielded the distinct profile of labeling expected for selective modification of periplasmic proteins. We hypothesized that the signal sequence fusions were not efficiently secreted in slow-growing species. In an attempt to improve recognition and processing by secretion systems, we tested signal sequences from Mpt64 and Cfp21, which are secreted proteins and predicted Sec substrates (Weldingh et al., 1998; Wiker et al., 2000), and from PepA, which is a confirmed Tat substrate (Marrichi et al., 2008). We also extended the linker between the signal peptide and APEX2 from two to ten amino acids. Only fusions to signal sequences from Mpt64 and Cfp21 yielded detectable protein labeling in Mtb H37Rv, but the pattern was similar to cytoplasmic labeling, suggesting that these constructs were expressed but not exported.

We then hypothesized that inefficient translation of the eukaryotic APEX2 gene was preventing successful secretion and therefore processing, folding, heme insertion, and activity of Sec-APEX2 in the periplasm. Indeed, Sec-APEX2 expressed from an Mtb codon usage-optimized gene (APEX2m) showed robust expression and peroxidase activity and a pattern of protein labeling distinct from that of cytoplasmic APEX2 in all mycobacterial strains tested thus far (M. smegmatis, M. tuberculosis H37Rv, and the attenuated strain M. tuberculosis mc²6020 ΔpanCD ΔlysA (Sambandamurthy et al., 2005); Figure 2). We therefore recommend the use of APEX2m codon-optimized constructs for all APEX2-mediated proximity labeling applications in mycobacteria. To facilitate use by the broader research community, we have generated theophylline riboswitch-inducible multi-copy episomal and single copy integrating constructs containing the APEX2m gene (AddGene: 176842–176845); maps for Sec-APEX2 vectors are shown in Figure 3). These plasmids are also appropriate for generating APEX2 fusions to ‘bait’ proteins, towards labeling and detecting their interaction partners.

Figure 2. Compartment specific labeling of proteins in M. tuberculosis by APEX2 expressed from a gene optimized for M. tuberculosis H37Rv codon usage.

M. tuberculosis mc²6020 expressing APEX2 or Sec-APEX2 from the codon usage-optimized APEX2m gene were grown with or without theophylline and subjected to the labeling protocol for (A) biotin-phenol or (B) tyramide azide followed by CuAAC coupling to fluorescein-alkyne. Total lysates were analyzed by (A) streptavidin blot or (B) fluorescence imaging. In (B) cultures of two independent clones were analyzed. Asterisks indicate examples of APEX2 expression-independent bands. Data are represented of >3 independent experiments.

Figure 3. Plasmid maps for pRibo-Sec-APEX2m and pRiboI-Sec-APEX2m.

The restriction sites most relevant to cloning are indicated. Plasmids and partial sequences are available from Addgene (ID: 176844, 176845). Full maps and sequences are available from the authors upon request. Figure created with SnapGene.

We provide below an updated protocol for APEX2 proximity labeling in mycobacteria (Figure 4). To aid troubleshooting we include instructions for a whole-cell colorimetric activity assay based on the peroxidase substrate guaiacol: both APEX2 and Sec-APEX2 should give a robust result by eye if expression and activity are as expected. Additional adaptations to our previously published methods include (1) affinity enrichment conditions conducive to subsequent proteomic analysis and (2) confirmation of APEX2 expression by direct detection with a commercial native antibody, which obviates the need for an epitope tag on APEX2.

Figure 4. Experimental flow for APEX2 expression and protein labeling in mycobacteria.

The procotol specifies labeling with tyramide azide followed by coupling to fluorescein (FAM-alkyne), but is compatible with other alkyne reagents for other modes of detection or downstream applications, including enrichment as shown for proteins labeled with biotin-phenol. Figure created with BioRender.com.

BEFORE YOU BEGIN

Timing: 1–2 days

1. Strains:

   1. This protocol assumes the use of M. smegmatis mc²155 or M. tuberculosis (H37Rv, mc²6020, or other strain of interest) transformed with a plasmid encoding APEX2m under control of a theophylline-inducible promoter.
   2. For information on transformation, selection, and culturing of mycobacteria, see (Goude & Parish, 2008; Parish, 2021).
2. Stock solutions for preparation of growth media

   1. Sterilize by 0.2 μM membrane filtration (“filter sterile”) unless otherwise noted.
   2. 80 mg/mL lysine in ultrapure water
   3. 24 mg/mL pantothenate in ultrapure water
   4. 20% v/v Tween-80 in ultrapure water
   5. 10% v/v Tyloxapol in ultrapure water
   6. 50% w/v glucose in ultrapure water
   7. 50% v/v glycerol in ultrapure water (can also be sterilized by autoclaving)
   8. 25 mg/mL kanamycin in ultrapure water (1000X stock)
3. Growth media

   1. M. smegmatis: 7H9 Cas medium (4.7 g Middlebrook 7H9 base, 10 g casamino acids, 4 mL 50% v/v glycerol, 4 mL 50% w/v glucose, 2.5 mL 20% v/v Tween-80 to 1 L with ultrapure water)
   2. M. tuberculosis: 7H9 OADC medium (4.7 g Middlebrook 7H9 base, 10 mL 50% v/v glycerol, 2.5 mL 10% v/v Tyloxapol, 100 mL oleic acid-albumin-dextrose-catalase (OADC) supplement to 1 L with ultrapure water)
   3. M. tuberculosis mc²6020: 7H9 OADC-pan-lys medium

      1. Prepare 1000X stocks, filter sterile.
      2. Per 1L 7H9 OADC medium, add 2 g casamino acids, 1 mL lysine stock, 1 mL pantothenate stock.
   4. For inducing APEX2 expression: 20 mM theophylline in growth medium (see Note below).
   5. Filter sterilize media. Alternatively, autoclave prior to addition of glucose, Tween-80, Tyloxapol, theophylline and/or ADC or OADC supplement. Store at 4 °C.
4. Stock solutions

   1. PBS (phosphate-buffered saline, pH 7.4)
   2. PBST80 (PBS with 0.05% v/v Tween-80)
   3. PBST (PBS with 0.1% v/v Tween-20)
   4. 50 mM biotin-phenol (biotin tyramide) in dimethyl sulfoxide (DMSO)
   5. 50 mM tyramide-azide in DMSO
   6. Click reaction lysis buffer (20 mM Tris, 150 mM NaCl, 1% v/v Triton X-100, 1% v/v SDS, pH 7.4)
   7. 1 mM fluorescein-alkyne (FAM alkyne, 5-isomer) in ultrapure water
   8. 6 mM TBTA [tris(benzyltriazolylmethyl)amine] in DMSO
   9. 10 mM CuSO₄ in ultrapure water
   10. 50 mM EDTA (ethylenediaminetetraacetic acid) in ultrapure water
   11. 20% w/v DDM (n-dodecyl-β-D-maltoside) in ultrapure water
   12. 100X protease inhibitor cocktail (17 g/L phenylmethanesulfonyl fluoride, 33 g/L benzamidine hydrochloride, 0.137 g/L pepstatin A, 0.03 g/L leupeptin, 0.2 g/L chymostatin dissolved in ethanol)
   13. Western blot blocking buffer (5% w/v bovine serum albumin in PBST)

Alternatives:

APEX2m labeling has also been validated for mycobacteria grown in the following media:

1. M. smegmatis: 7H9 ADC medium (Middlebrook 7H9 with 0.5% v/v glycerol, 0.05% Tween-80, 10% albumin-dextrose-catalase (ADC) supplement)

2. M. smegmatis and M. tuberculosis: Roisin-Fe medium (see Note)

   1. To prevent precipitation, add all ingredients in order.
   2. Prepare 1000X Trace Elements solution (per 1L): 80 mg ZnCl₂, 22 g FeCl3-6H₂O, 20 mg CuSO₄, 20 mg MnCl₂-4H2O, 20 mg Na₂B₄O₇-10H₂O, 20 mg (NH₄)6Mo₇O₂₄-4H₂O.
   3. Prepare separate 1 M CaCl₂ and 1 M MgCl₂ solutions.
   4. For 1 L, add 1 g KH₂PO₄, 2.5 g Na₂HPO₄, 5.9 g NH₄Cl, 2 g K₂SO₄, 0.5 mg biotin
   5. Add 1 mL Trace Element solution, 0.5 mL 1M CaCl₂, 0.5 mL 1 M MgCl₂.
   6. Add 10 mL 50% v/v glycerol, 2.5 mL 10% v/v Tyloxapol.
   7. Adjust to pH 6.6. Medium can be filter sterilized or autoclaved prior to addition of Tyloxapol.

Note:

• To be consistent with the proximity labeling literature, we refer to the commercial reagent biotin tyramide as biotin-phenol throughout.

• Theophylline is close to saturation at 20 mM and may require sonication for solubilization. Solutions stored at 4 °C may precipitate, but at 20–22 °C are stable for at least a year. We therefore recommend that 20 mM theophylline stock solutions be prepared and stored at 20–22 °C in medium. Alternatively, theophylline can be prepared in medium at lower concentrations (e.g., a working concentration of 2 mM) and stored at 4 °C.

• Roisin-Fe is a modified Roisin’s medium (Beste et al., 2005).

  • We found that iron is limiting in standard Roisin’s medium with respect to supporting APEX activity: M. smegmatis expressing APEX2 and grown in standard Roisin medium yields less robust protein labeling than the same strain grown in 7H9-Cas medium. The recipe reported here has increased iron content to match the concentration of [Fe³⁺] in Middlebrook 7H9 broth base (82 μM).
  • Because NH₄Cl is the sole nitrogen source in Roisin’s medium, this medium can be used for stable (nitrogen) isotope labeling of amino acids in culture (SILAC) in mycobacteria (Li et al., 2021; Touchette et al., 2017). We found that M. smegmatis growth is also sustained at a lower concentration of 3 g/L NH₄Cl. For SILAC applications, using this lower NH₄Cl concentration conserves expensive isotopically pure reagents (i.e., ¹⁵NH₄Cl) without compromising growth. Note that this has not yet been validated in slow-growing mycobacterial species.

EQUIPMENT

• Bead beater (Omni International Bead Ruptor 12)

• Fluorescence scanner for infrared detection (LI-COR Odyssey Clx)

• Fluorescence scanner for visible wavelength detection (Azure Sapphire Biomolecular Imager)

• Shaking incubator (New Brunswick Scientific Classic Series C24 incubator shaker)

• Centrifuge (Sorvall Legend Mach 1.6R)

• Microcentrifuge (Eppendorf Centrifuge 5424 R)

• UV-visible spectrophotometer (Amersham Biosciences)

• Mini-PROTEAN electrophoresis cell (Bio-Rad)

• Semi-dry transfer apparatus (Bio-Rad)

MATERIALS

• 500 mM H₂O₂ (19.6-fold dilution of a 30% commercial solution (9.8 M)).

• 2X Quencher (20 mM sodium ascorbate, 20 mM sodium azide, and 10 mM Trolox in PBST)

  • Trolox must be solubilized in ethanol prior to addition.
  • E.g., for 40 mL 2X Quencher, dissolve 0.1 g Trolox in 1 mL ethanol and add to all other ingredients dissolved in 39 mL PBST80.

• 1X Quencher (1:1 dilution of 2X Quencher with PBST80)

• 167 mM sodium ascorbate in ultrapure water

STEP-BY-STEP METHOD DETAILS

Mycobacterium smegmatis growth and APEX2 expression

Timing: 1 day

1. Prepare 7H9 Cas medium with 25 μg/mL kanamycin.

2. Inoculate with M. smegmatis and incubate to a final OD₆₀₀ = 0.8 – 1 (as determined by measuring the absorbance at 600 nm). Unless noted otherwise, all incubations with M. smegmatis are at 37 °C with vigorous shaking (250 rpm).

3. Subculture to OD₆₀₀ = 0.25 in 15 mL 7H9 Cas medium with or without 2 mM theophylline.

4. Incubate cultures for 6 h (~2 doubling times for M. smegmatis).

Mycobacterium tuberculosis growth and APEX2 expression

Timing: 5–7 days

1. Prepare 7H9 OADC medium with 25 μg/mL kanamycin

2. For M. tuberculosis mc26020, use 7H9 OADC-pan-lys medium.

3. Inoculate with M. tuberculosis and incubate to a final OD600 = 0.8 – 1. Unless noted otherwise, all incubations with M. tuberculosis are at 37 °C with moderate shaking (110 rpm).

4. Subculture to OD600 = 0.25 in 15 mL growth medium with or without 2 mM theophylline.

5. Incubate cultures for 48 h (~2 doubling times for M. tuberculosis).

Note:

• Based on our previous work with reporters such as GFP, induction of the riboswitch with 2 mM theophylline leads to maximal expression after ~2 doubling times for both M. smegmatis and M. tuberculosis (Seeliger et al., 2012).

• Under the specified conditions bacteria are in mid- to late-log phase during induction. Expression and activity of APEX2 in other phases of growth has not been tested.

• Mycobacterial growth is slower in Roisin’s medium than in supplemented 7H9 media: Doubling times are ~5 h for M. smegmatis and ~36 h for M. tuberculosis. Growth and induction times should be adjusted accordingly.

Guaiacol activity assay

Timing: 5 – 10 minutes

1. Transfer 500 µl culture into a 1.5-mL centrifuge tube for each growth condition (with or without theophylline).

2. Add 5 µl guaiacol and 5 µl 30% w/v hydrogen peroxide to the inside of the caps of each tube.

   1. Adding reagents to the cap enables multiples reactions to be synchronized more easily for short incubation times and side-by-side comparison.

3. Close and invert all tubes at the same time to initiate the reaction.

4. Record the formation of red coloration (Figure 5).

   1. The red pigment formed by guaiacol oxidation can be transient. We recommend recording the reaction as quickly as possible.
   2. If the coloration is weak, briefly centrifuging the tubes (1 min, 10,000 x g) after initiating the reaction pellets the cells and aids detection of any coloration.

Figure 5. A colorimetric assay enables rapid assessment of APEX2 activity.

in whole cells. M. smegmatis was grown with or without theophylline. Guaiacol substrate and hydrogen peroxide were added to an aliquot of culture to initiate the peroxidase reaction.

APEX2-mediated protein labeling

Timing: 2 – 4 h

• 1Steps 1–5 below are performed at 20–22 °C.
• 2
  Harvest the cells by centrifugation at 2,500 x g for 10 min. Decant and resuspend cells in 1 mL growth medium and transfer to a 2-mL centrifuge tube.

  1. Biotin-phenol labeling: Add 20 µl of 50 mM biotin-phenol (1 mM final concentration) and incubate for 30 min at 37 °C.
  2. Tyramide-azide labeling: Add 20 µl of 50 mM tyramide-azide (1 mM final concentration) and incubate for 15 min at 37 °C.
• 3Add 2 µl of freshly prepared 500 mM H₂O₂ to the caps of each centrifuge tube.
• 4Close and invert the tubes at the same time to initiate the reaction.
• 5After one minute, add 1 mL 2X Quencher to each tube. Unless noted otherwise, all centrifugation steps for washing labeled cells are in a 2 mL microcentrifuge tube at 10,000 x g for 10 minutes.
• 6
  Wash the cells follows:

  1. 2 mL 1X Quencher
  2. 2 mL PBST80
  3. 1.5 mL 1X Quencher
  4. 1.5 mL PBST80

Pause Point: Washed cells can be pelleted and stored at −80 °C for at least 3 months and thawed on ice prior to proceeding with the protocol.

• 7Steps 8–11 below are performed at 4 °C or on ice with pre-cooled buffers.
• 8Resuspend the cells in 1.5 mL PBS and transfer to 2-mL screw-cap tube containing ~0.5 mL 0.1 mm zirconia beads.
• 9Lyse the cells with the bead beater in 5 cycles of 30s each at 6 m/s with 5-min incubations on ice between each cycle.
• 10Pellet beads and cell debris by centrifugation for 10 min at 12,000 x g at 4 °C.
• 11Transfer supernatant to a fresh 1.5-mL centrifuge tube.

Pause Point: Clarified lysates can be flash frozen with liquid nitrogen and stored at −80 °C for at least 3 months and thawed on ice prior to proceeding with the protocol.

Coupling to azide-labeled proteins via copper-mediated azide-alkyne cycloaddition

Timing: 1–2 hours

• 1Normalize lysates to 1.12 mg/mL by BCA assay according to manufacturer instructions.
• 2Transfer 89.2 μL (100 μg) each lysate into a 1.5-mL centrifuge tube.
• 3
  Add to each reaction in the following order:

  1. 3 µl 167 mM sodium ascorbate
  2. 2.5 µl 1 mM fluoroscein-alkyne
  3. 3.3 µl 6 mM TBTA
  4. 2 µL 10 mM CuSO₄
• 4Incubate with gentle mixing (e.g., on a tube or platform rotator) and protected from light for 1 h at 20–22 °C.
• 5Quench with 2 µL 50 mM EDTA per reaction. Add SDS-PAGE loading dye.

Pause point: SDS-PAGE samples can be stored at −20 or −80 °C until further analysis.

• 6Load 10 μg lysate per well and resolve on a 10% SDS-PAGE gel.
• 7Image SDS-PAGE gel using settings appropriate for fluorescein excitation/emission wavelengths.

Western blot analysis

Timing: 3–4 hours

1. Normalize lysates to 1 mg/mL by BCA assay using manufacturer provided guidelines.

2. Load 10 μg lysate per well for each growth condition and resolve a 10% SDS-PAGE gel.

3. Transfer SDS-PAGE gel to a nitrocellulose membrane using a semi-dry transfer apparatus at 25 V for 50 min.

4. Incubate membrane with blocking buffer for 1 h. This and all subsequent steps at 20–22 °C.

5. To detect APEX2 (Figure 6), incubate with α-APX2 primary (1:5000 in blocking buffer) for 1h (see Note)

   1. Wash three times with PBST for 5 min per wash.
   2. Incubate with 1:10,000 goat-α-rabbit IR Dye 800 LT for 1h
   3. Wash three times with PBST and then once with PBS for 5 min per wash.

6. To detect biotinylated proteins, incubate with 1:10,000 Streptavidin IR-Dye 680 LT for 1 h. Wash three times with PBST and once with PBS for 5 min per wash.

7. Image using settings appropriate for dye excitation/emission wavelengths (see Note).

Figure 6. Detection of APEX2 by anti-APX2 antibody.

M. smegmatis expressing APEX2 or Sec-APEX2 was growth with or without theophylline. Total lysates were analyzed by immunoblot using an anti-APX2 antibody raised against ascorbate peroxidase from Arabidopsis thaliana. The doublet for Sec-APEX2 may indicate protein before (31 kDa predicted molecular weight) and after (28 kDa) cleavage of the N-terminal secretion signal by signal peptidase in the periplasm. Asterisks indicate non-specific bands. Data are representative of >3 independent experiments.

Note:

• Incubation with primary antibody can also be done at 4 °C for 12–16 h.

• The specified antibodies are for infrared fluorescence detection. Any appropriate (secondary) antibody and associated detection method can be used, such as visible fluorescence or luminescence.

Enrichment of biotin-tyramide labeled proteins

Timing: 2–3 hours

• 1
  Scale preparation of APEX2-labeled lysates as follows (see Note):

  1. Follow “Growth and APEX2 expression” above for either M. smegmatis or M. tuberculosis as appropriate, except subculture into 50 mL or 250 mL growth medium for APEX2 or Sec-APEX2, respectively.
  2. Follow “APEX2-mediated protein labeling” as above except:

     1. Resuspend cell pellets in 5 mL or 25 mL growth medium for APEX2 or Sec-APEX2, respectively, prior to incubation with 50 mM biotin-phenol.
• 2Pellet the cells for 10 min at 2,500 x g.
• 3Add protease inhibitor cocktail to 1X final concentration in PBS.
• 4Decant supernatant and resuspend pellet in 2 mL (for APEX2) or 5 mL (for Sec-APEX2) PBS and transfer in 1-mL aliquots to 2 mL screw-cap tubes filled with 0.5 mL 0.1 mm zirconia beads.
• 5Lyse the cells with 5 cycles of bead beating for 30 s per cycle at 6 m/s. Between each cycle, cool samples on ice for 5 minutes.
• 6Pellet cell debris by centrifugation at 10,000 x g at 4 °C. Pool supernatants in fresh tubes.
• 7Add DDM to a final concentration of 1% w/v and incubate with gentle mixing (e.g., on a tube rotator or platform) at 4 °C for 1 h.
• 8Quantify the protein concentrations by BCA assay (see Note).
• 9Add 300 μL (APEX2) or 750 μL (Sec-APEX2) neutravidin beads per sample to a 5-mL centrifuge tube. Centrifuge beads for 5 min at 500 x g and decant supernatant. Wash 3 times with 1X volume PBS.
• 10Add lysate to washed beads and incubate with gentle mixing at 22 °C for 1 h.
• 11
  Pellet the beads by centrifugation for 5 min at 500 x g

  1. If desired, supernatants can be transferred to fresh tubes and used to check for efficiency of enrichment.
• 12Wash the resin 4X with PBS volume equivalent to bead volume per wash.
• 13
  For Western Blot analysis: Transfer 100 μL beads to fresh tube.

  • a
    Pellet beads by centrifugation for 5 min at 500 x g. Aspirate supernatant.
  • b
    Elute proteins with 100 μL 2X SDS-PAGE loading dye and flick the tubes to mix.
  • c
    Boil the suspension for 5 min at 100 °C.
  • d
    Pellet beads by centrifugation for 5 min at 500 x g.
  • e
    Transfer supernatant to a fresh tube.

Pause point: SDS-PAGE samples can be stored at −20 or −80 °C until further analysis.

• • fResulting eluate can be used in “Western Blot analysis” above to detect biotinylated proteins or to detect enrichment of a specific protein of interest using an appropriate antibody.

• 14For proteomics: Remaining beads are appropriate for on-bead digestion or other protocols for protein identification. Check with your core facility, vendor, or collaborator for a specific protocol for further sample preparation for MS/MS.

Note:

• This protocol is optimized for identification of enriched proteins by on-bead digestion and LC-MS/MS proteomics. The culture volumes for strains expressing APEX2 or Sec-APEX2 were empirically optimized based on the yield of labeled proteins (as determined qualitatively by silver stain) from the respective constructs in M. tuberculosis mc²6020. In general, for Sec-APEX2 labeling, a larger culture volume and higher total lysate protein concentration are required to obtain yields equivalent those obtained from APEX2 labeling.

• The recommended minimum total protein amounts per sample are 2–3 mg (APEX2) and 12–15 mg (Sec-APEX2) to obtain a sufficient yield of enriched protein for subsequent identification by unlabeled proteomics. These minimum amounts were empirically optimized using M. tuberculosis mc²6020.

QUANTIFICATION AND STATISTICAL ANALYSIS

• No quantitative analysis is explicitly required in these protocols except in the determination of protein concentration by the manufacturer’s instructions for the BCA assay.

• To determine the localization of proteins of interest by comparing labeling by APEX vs. Sec-APEX2 as reported in Ganapathy et al., we recommend analyzing Western blots using ImageJ or similar imaging software as provided by imaging instrument manufacturers (e.g., LI-COR Image Studio).

ADVANTAGES

• The APEX2m gene, which is optimized for M. tuberculosis H37Rv codon usage, expresses APEX2 at higher levels than the originally reported APEX2 gene in all compartments and all mycobacterial species thus far tested.

• The pRibo vectors provide expression via a dose-dependent theophylline riboswitch. Close to maximal expression is expected at 2 mM theophylline, as recommended in this protocol. However, APEX2 expression can be adjusted as desired by changing the concentration of theophylline. Note that theophylline concentrations ≥4 mM can cause growth defects over two doubling times and longer in all mycobacterial species thus far tested.

LIMITATIONS

• High expression levels of either APEX2 or a protein of interest may lead to protein labeling by APEX2 expressed in a different compartment. We speculate that this is due to the moderate diffusion of phenoxyl radicals across lipid membranes. To accurately localize proteins, we therefore recommend expressing proteins of interest at native levels or expressing APEX2 at the lowest level necessary to detect labeling of the target.

• APEX2 generates phenoxyl radicals that label aromatic groups: on proteins, primarily tyrosine sidechains. Therefore, proteins that lack exposed tyrosine sidechains may not be detected.

SAFETY CONSIDERATIONS AND STANDARDS

• Risk Level classification for non-pathogenic and attenuated organisms can vary between institutions. Ensure that your use of recombinant organisms and biosafety containment conforms to the guidelines and are approved by the appropriate bodies at your institution.

• All virulent strains of M. tuberculosis must be handled at Biosafety Level 3.

• Attenuated M. tuberculosis strains such as mc²6020 can be handled at Biosafety Level 2 per individual institutional approval.

• M. smegmatis is can be handled at Biosafety Level 1 or 2 per individual institutional approval.

KEY RESOURCES TABLE.

Note that not all areas will be used in every protocol

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-APX2	Abcam	ab222414
Goat-anti-rabbit IRDye 800CW Lt	LI-COR	926-32211
Streptavidin IRDye 680 Lt	LI-COR	926-68031
Bacterial and Virus Strains
M. smegmatis mc2155	ATCC	700084
M. tuberculosis mc26020 ΔpanCD ΔlysA	Gift of Dr. William Jacobs, AECOM
M. tuberculosis H37Rv	ATCC	27294
Chemicals, Peptides, and Recombinant Proteins
ADC supplement	Fisher	B12352
Benzamidine hydrochloride	Sigma	434760
Biotin	Fisher	PI29129
Biotin tyramide (biotin-phenol)	Iris Biotech	LS-3500
Bovine serum albumin	GoldBio	A-420
Casamino acids	VWR	VWRJ851-1KG
Chymostatin	Sigma	230790
FAM alkyne, 5-isomer	Lumiprobe	41B0
Guaiacol	Sigma	G5502
Hydrogen peroxide (30% w/v)	Fisher	H325-500
Leupeptin	Roche	11017101001
Lysine	Sigma	L5501
Middlebrook 7H9 broth base	HI MEDIA	M198
n-dodecyl-β-D-maltoside	GoldBio	DDM25
OADC supplement	Fisher	B12351
Pantothenate	Sigma	21210
Pepstatin A	Sigma	77170
Phenylmethanesulfonyl fluoride	Sigma	P7626
Sodium ascorbate	Sigma	11140
Sodium azide	Sigma	71289
Theophylline	Sigma	T1633
Tris(benzyltriazolylmethyl)amine (TBTA)	Alfa Aesar	H66485
Triton-X100	Sigma	T8787
Trolox	Fisher	AC218940050
Tween-20	VWR	VWRV0777-1L
Tween-80	Sigma	P1754
Tyloxapol	Sigma	T0307-50G
Tyramide-azide	(Ganapathy et al., 2018)
Zirconia beads, 0.1 mm	BioSpec Products	11079101z
Critical Commercial Assays
BCA assay	Fisher	PI23225
Neutravidin enrichment kit	Fisher	PI29201
Recombinant DNA
pRibo-APEX2m	This work	Addgene: 176842
pRiboI-APEX2m	This work	Addgene: 176843
pRibo-Sec-APEX2m	This work	Addgene: 176844
pRiboI-Sec-APEX2m	This work	Addgene: 176845