Heterologous expression, purification, and characterization of type II polyketide acyl carrier proteins

Summary/Abstract

The enzymes that comprise type II polyketide synthases (PKSs) are powerful biocatalysts that, once well-understood and strategically applied, could enable cost-effective and sustainable access to a range of pharmaceutically relevant molecules. Progress towards this goal hinges on gaining ample access to materials for in vitro characterizations and structural analyses of the components of these synthases. A central component of PKSs is the acyl carrier protein (ACP), which serves as a hub during the biosynthesis of type II polyketides. Herein, we share methods for accessing type II PKS ACPs via heterologous expression in E. coli. We also share how the installation of reactive and site-specific spectroscopic probes can be leveraged to study the conformational dynamics and interactions of type II PKS ACPs.

1. Introduction

Recent technical advances in bioinformatics and computational biology have led to the widespread discovery of intriguing biosynthetic gene clusters (BGCs) in a large swath of organisms [1,2]. BGCs encode suites of enzymes that cooperate to manufacture molecules thought to mediate an organism’s ecological interactions. These molecules, known as specialized metabolites (and commonly referred to as secondary metabolites or, more broadly, natural products), play a central role in medicine both as therapeutics and as inspiration for new pharmaceutically relevant compounds [3]. Because biosynthesis mediated by these enzyme assemblies requires only earth abundant, natural starting materials and uses water as a solvent, mixing-and-matching enzymes encoded by different BGCs could provide cost-effective and sustainable access to a host of pharmaceutically relevant “unnatural natural products”. Further biochemical and structural analyses of enzymes encoded by BGCs are necessary to make these goals a reality. However, the chemical innovation coordinated by the enzymes encoded in these BGCs is often difficult to assess, as the native organisms harboring BGCs of interest are often resistant to culture under laboratory conditions. Heterologous expression of the proteins encoded by BGCs can bypass these issues, opening the door for the investigation of previously out-of-reach products [4]. This approach can also centralize the production of target protein products from disparate sources into one strain, thereby saving both time and resources. Therefore, identifying a heterologous expression system to produce the enzymes encoded within BGCs is an important step toward harnessing biosynthetic protein assembly lines for access to new chemical diversity. Several genetically engineered strains of Escherichia coli have emerged as popular heterologous host species for such applications due to their straightforward culture conditions, rapid growth, well-studied genetics, and flexibility for protein expression [5-7].

Heterologous expression in E. coli is an effective strategy for the study of polyketide synthases (PKSs), multi-enzyme assembly lines encoded by BGCs that manufacture an array of pharmaceutically relevant chemicals. The products of type II PKSs are particularly lucrative targets for heterologous expression in E. coli because of the chemical diversity of this class of molecules and their strong track record for serving as widely used therapeutic agents [3]. Molecules manufactured by type II PKSs include important antibiotics such as tetracycline and doxorubicin, as well as anticancer agents such as doxycycline [8-10].

A central component of type II PKSs is the acyl carrier protein (ACP), which serves as a hub of activity during the manufacturing of polyketides. The ACP is a small (~9 kDa), helical, labile protein that is post-translationally modified at a conserved serine residue at the N-terminus of helix II with the addition of a 4'-phosphopantetheine (Ppant) “arm”. The terminal thiol of the 18Å Ppant arm is the point of attachment for malonyl-CoA-based molecular building blocks and β-keto intermediates throughout the assembly of the polyketide. The malonyl-ACP interacts iteratively with the acylated ketosynthase-chain length factor heterodimer (KS-CLF; sometimes referred to as the αKS and βKS) through repeated decarboxylative Claisen-like condensation reactions to extend the polyketide chain by two carbon units (Figure 1). Together, the ACP and KS-CLF are known as the “minimal type II PKS” and perform the essential task of building the backbone of the polyketide chain. This chain is modified by tailoring enzymes to create the final polyaromatic product.

Figure 1.

An overview of the type II polyketide biosynthesis of doxorubicin, a well-known type II polyketide, highlighting the crucial role that the acyl carrier protein (ACP) plays in the biosynthetic process. The ACP is primed by malonyl transferase to add a reactive malonyl moiety to the free thiol of the 4'-phosphopantetheine (Ppant) arm, which is denoted by a squiggly line. (A) The primed ACP interacts with an acylated ketosynthase-chain length factor (KS-CLF) in a Claisen-like decarboxylative condensation reaction, resulting in a diketide tethered to the ACP as a thioester. (B) The thiolate of the active site serine residue of the KS-CLF acts as the nucleophile in a nucleophilic acyl substitution reaction, shifting the growing acyl chain back to the KS-CLF. (C and D) These steps repeat as the ACP and KS-CLF interact to iteratively produce the polyketide backbone of the final product. (E) A cohort of tailoring enzymes, including cyclase (CYC), methyltransferase (MT), oxygenase (OXY), methyl esterase (ME), and ketoreductase (KR) modify the completed polyketide backbone to finish the synthesis of the secondary metabolite (F).

Understanding ACP-enzyme interactions may give researchers the ability to engineer hybrid PKSs. These PKSs would consist of enzymes encoded by a variety of evolutionarily diverged BGCs and could provide sustainable access to “unnatural natural products” with novel structures and biologically relevant activities. Type II PKSs share a common ancestor with fatty acid synthases (FASs), which suggests the possible compatibility between the components of type II PKSs and FASs for the future construction of combinatorial synthases and the production of novel, hybrid products with promising applications [11-13].

Type II PKSs are found across a variety of species [11]. The most well-studied type II PKSs are those found in Actinomycetes, an order of the Actinobacteria, a phylum of bacteria that are notoriously onerous to work with in the lab [14,15]. Heterologous expression of type II PKSs could improve the efficiency of characterizing these systems; however, challenges in expressing soluble Actinobacterial KS-CLF heterodimers in heterologous hosts–including E. coli–has slowed the application of these techniques. In 2015, a bioinformatics study led by Hillenmeyer and Charkoudian suggested that the BGCs responsible for the transcription and translation of the minimal type II PKS are present in many strains anciently diverged from the Actinobacteria [11]. The great majority of these recently discovered pathways remain uncharacterized; neither their biosynthetic machinery nor their molecular products have been well-studied. Fortunately, many of these putative type II PKS BGCs are found in phyla such as the Firmicutes and Proteobacteria, and hence are attractive candidates for heterologous expression in tractable, non-Actinobacterial heterologous hosts such as the well-known proteobacterium E. coli. In just the past year, researchers have shown that genes encoding type II PKS KS-CLFs from non-Actinobacterial species can indeed be expressed in E. coli, thereby opening a new era for the in vitro characterization of type II PKSs [14-17].

With newfound access to type II PKS KS-CLFs for in vitro characterization, it is important to streamline the production of a range of ACPs and develop methods to study their interactions with KS-CLFs. These advances will enable the molecular-level characterization of protein-protein and protein-substrate interactions that facilitate the production of type II polyketides and inform bioengineering studies to produce hybrid synthases capable of manufacturing “unnatural natural products”. Herein we present our optimized methods for the cloning, E. coli heterologous expression, and purification of type II PKS ACPs. We also share methods to determine the molecular weight (MW) of ACPs using a single quadrupole liquid chromatography-mass spectrometry (LC-MS), assess the solvation environment of the ACP Ppant arm and its molecular cargo using site-specific vibrational spectroscopy, and visualize ACP-KS interactions using a tandem mechanistic crosslinking/colorimetric assay.

2. Materials

2.1. General Equipment

1. Thermal cycler

2. Pipettes and tips

3. Agarose gel electrophoresis materials

4. 0.2 mL microcentrifuge tubes

5. 1.5 mL microcentrifuge tubes

6. Nuclease-free H₂O

7. Thermo Scientific™ NanoDrop 2000 or other UV-visible (UV-vis) spectrophotometer

8. LB broth supplemented with 50 μg/mL kanamycin

9. Orbital shaker with heating/cooling capabilities

10. LB agar plates supplemented with 50 μg/mL kanamycin

11. Inoculating loop/sterile toothpick

12. Sterile glass beads

13. Incubator

14. Benchtop Centrifuge

15. Hot water bath

2.2. Amplification of ACP Genes

1. DNA polymerase (such as Q5® High-Fidelity DNA Polymerase)

2. Zymo DNA Clean & Concentrator™ Kit

3. (Optional) Oligonucleotide PCR primers, designed in Section 3.1

2.3. Gibson Assembly®

1. pET-28a (+) vector

2. NdeI and EcoRI restriction endonucleases

3. Zymoclean™ Gel DNA Recovery Kit

4. Gibson Assembly® Master Mix

2.4. Preparation and Transformation of Plasmids

1. Chemically competent E. coli cells optimized for molecular cloning (e.g., DH5ɑ or XL1 Blue)

2. QIAprep® Spin Miniprep Kit

3. Microcentrifuge

4. Chemically competent E. coli cells optimized for protein production (e.g., BL21 or BAP1 [18])

5. Parafilm

2.5. Expression of ACPs in E. coli

1. 2 L Baffled culture flask with sponge top or other loose-fitting top to allow for aeration

2. Autoclave

3. 1M Isopropyl β-D-1-thiogalactopyranoside (IPTG)

4. High-speed floor centrifuge

5. Large (~500 mL) centrifuge tubes compatible with floor centrifuge

6. 50 mL conical tubes

7. Spatula

2.6. Purification and On-Column Phosphopantetheinylation of ACPs

1. Lysis Buffer: 50 mM sodium phosphate, pH 7.6, 300 mM NaCl, 10 mM imidazole, and 10% (v/v) glycerol

2. Ultrasonic homogenizer with tip probe

3. Ni-NTA agarose affinity resin

4. Fritted gravity column with stopcock and cap

5. Wash Buffer: 50 mM sodium phosphate, pH 7.6, 300 mM NaCl, and 30 mM imidazole

6. 50 mM sodium phosphate buffer, pH 7.6

7. 50 mM Dithiothreitol (DTT)

8. 250 mM MgCl₂

9. 50 mM Coenzyme A or relevant Coenzyme A derivative

10. 150 μM R4-4 Sfp enzyme[19,20]

11. Elution Buffer: 50 mM sodium phosphate, pH 7.6, 100 mM NaCl, and 300 mM imidazole)

12. Dialysis materials. We use Thermo Scientific™ Slide-A-Lyzer™ Dialysis Cassette (2K MWCO) with corresponding cassette buoys and syringes.

13. Bucket, 2 L or larger

14. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) materials

2.7. Cyanylation of holo-ACP Ppant arm

1. Protein quantification assay materials. We use Pierce™ Coomassie (Bradford) Protein Assay Kit

2. Ellman’s Reagent (5,5'-dithiobis-(2-nitrobenzoic acid) or DTNB)

3. PD-10 desalting column

4. Sodium cyanide (NaCN), reagent grade

5. Double distilled water (ddH₂O)

6. Centrifugal concentration tubes, 3 kDa MWCO

7. Liquid nitrogen

8. Cryogenic dewar

9. Cryogenic tongs

10. Cryogenic gloves

2.8. Collecting and Analyzing a Vibrational Spectrum

1. Bruker Optics Vertex 70 FT-IR with photovoltaic MCT detector

2. Constant N₂ purge source: either a dry N₂ tank or, more ideally and sustainably, the output from a Parker-Balston IR purge gas generator

3. Harrick 13 mm diameter demountable liquid cell

4. General purpose wipes (i.e.,Kimwipes™)

5. OPUS software package

6. Origin software package

2.9. Colorimetric ACP-KS Mechanistic Crosslinking Assay

1. Aliquot of desired KS (or KSCLF)

2. Ellman’s Reagent (5,5'-dithiobis-(2-nitrobenzoic acid) or DTNB)

2.10. Tracking Ppant Sequestration Activity Using Raman Spectroscopy

1. 275 μM ACP in 50 mM sodium phosphate buffer, pH 7.6

2. 2.5 mM DTT

3. 23 mM MgCl₂

4. 18 mM ATP, pH adjusted to 7.6

5. 0.8 μM Acyl-acyl carrier protein synthetase (AasS) in Tris buffer, pH 7.6

6. 4.6 mM carboxylic acid probe (4-pentynoic acid (C5), 7-octynoic acid (C8), or 12-tridecynoic acid (C13)), stock in isopropanol

7. CW-Raman spectrometer

8. β-mercapto ethanol (BME)

2.11. Liquid Chromatography-Mass Spectroscopy (LC-MS) Analysis of ACPs

1. Liquid Chromatography-Mass Spectrometer (LC-MS) Equipped for Electrospray Ionization-Mass Spectroscopy

2. Waters XBridge Protein BEH C4 column

3. Solvent A: water + 0.1% formic acid

4. Solvent B: acetonitrile + 0.1% formic acid

3. Methods

3.1. Amplification of ACP Genes for Assembly into Expression Vectors

There are two popular options for the acquisition of the gene encoding ACPs of interest for Gibson Assembly® into an expression vector. The target gene can be ordered from a synthetic biology company, or the gene can be amplified via polymerase chain reaction (PCR) from genomic DNA (either purchased or extracted from the wild-type host). This latter option will be described here.

1. The algorithm that we use for primer design (written in Python), can be found in the Supporting Information, titled Gibson_assembly_primer_design.py. Simply execute the file (on a machine with Python3 installed) in a command line window (such as Terminal on Mac or Linux, or Command Prompt on Windows) and enter the coding sequence (5' to 3') of the gene of interest when prompted. The results in the command line represent primer sequences designed to insert the gene of interest into the NdeI/EcoRI endonuclease restriction sites of the pET-28a (+) expression vector, either via traditional restriction digestion/ligation assembly or Gibson Assembly® (see Notes 1-4) [21]. Other software programs, such as SnapGene and Serial Cloner, offer robust primer design algorithms as well.

2. Using an online calculator (such as this Thermo Fisher Scientific version: http://www.thermoscientificbio.com/webtools/tmc/), determine the melting temperatures of the primer pair. The difference between the melting temperatures of the primers should be roughly less than or equal to 5 °C (see Note 5). Synthesize the primers or order them from a synthetic biology company.

3. Following the protocol provided by the polymerase manufacturer, perform PCR using the designed primers and target genomic DNA (see Notes 6 and 7). Use an annealing temperature 3 °C warmer than the lower of the two melting temperatures. Purify the amplified gene block using a Zymo DNA Clean & Concentrator™ Kit (see Note 8).

3.2. Gibson Assembly® and Transformation of Plasmid into Competent Cells

1. Prepare pET-28a (+) vector by digesting it with both NdeI and EcoRI restriction enzymes, as directed by the supplier of the enzymes (see Note 9). This will linearize the circular plasmid by cutting it at the NdeI and EcoRI restriction sites, preparing the vector for compatibility with the previously designed Gibson Assembly® primers (see Note 10). After the restriction digestion, purify the linearized pET-28a (+) vector. We use a Zymoclean™ Gel DNA Recovery Kit for this step.

2. Once the vector has been prepared and the primers have been obtained, set up the reaction by combining the reagents described in Table 1 in a 0.2 mL microcentrifuge tube.

3. Place the 0.2 mL microcentrifuge tube with prepared Gibson Assembly® mixture in a thermal cycler at 50 °C for one hour.

4. After one hour, transfer the entire 20 μL reaction mixture to 100 μL of chemically competent E. coli cells in a 1.5 mL microcentrifuge tube. Incubate on ice for 30 minutes (see Note 13).

5. Transfer competent cells to a 42 °C water bath for exactly 90 seconds.

6. Remove cells from the water bath and incubate on ice for 2 minutes.

7. Add 200 μL of lysogeny broth (LB) without antibiotic to the microcentrifuge tube and shake for one hour at 200 RPM and 37 °C.

8. Remove the culture from the shaker and transfer the complete mixture onto a LB plate prepared with the appropriate antibiotic. We typically use pET-28a (+) for ACP expression, which encodes kanamycin resistance (50 μg/mL) (see Note 14). Spread the prepared cells around the plate using sterile glass beads or other sterile glass surface.

9. Leave the plate open in a sterile environment (either a sterile hood or near an open flame) for 10-15 minutes to allow the plate to dry.

10. Incubate the dry plate, with lid on and agar-side up, at 37 °C. Colonies harboring the assembled plasmid should begin to appear within 24-48 hours (see Note 15).

11. Once colonies have appeared, continue by preparing the amplified plasmid or store the plate, sealed with parafilm at 4 °C, until ready to proceed.

Table 1.

For typical assemblies, Gibson Assembly® has the highest rate of success when the molar ratio of the digested vector-to-DNA insert is 1:3 (see Notes 11 and 12). Determine the concentration of the two DNA segments (the DNA insert and the digested plasmid) in ng/μL. This can be done using NanoDrop UV-vis. Then use the following formula to calculate the molar concentration of each component:

$$\frac{pmol}{\mu \mathrm{L}}=\frac{(ng∕\mu \mathrm{L})\cdot 1000}{(\#bp)\cdot 650\text{Da}}$$

Where ng/μL is the measured concentration of DNA, # bp is the length of the DNA segment, 650 Da is an approximation of the average mass of a nucleotide base, and 1000 is a conversion factor. Add enough of each DNA segment to the mixture to achieve the desired concentration. The total volume of DNA (x+y) should not exceed 5 μL. If the volume x+y is less than 5 μL, add z = 5-(x+y) μL of nuclease-free water to make up the difference.

Reagent	Volume (μL)
DNA Insert	x
Linearized pET-28a (+) Vector	y
Nuclease-Free H2O	z
Gibson Assembly® Master Mix	15
Total	20

3.3. Preparation of Amplified Plasmid and Transformation into Expression Strain

1. Pick a single colony from the selective plate prepared previously and inoculate a ~5 mL culture in LB liquid medium containing the appropriate selective antibiotic (e.g., 50 μg/mL kanamycin). Incubate overnight (for 12-16 hours) at 37 °C, shaking at 200 RPM (see Notes 16 and 17).

2. Harvest the cells by centrifugation (5,400 x g) for 10 minutes at 4 °C. Remove all traces of supernatant by inverting the open centrifuge tube until all of the LB has been drained. If the cells have been pelleted correctly, they will remain securely attached to the bottom of the tube upon inversion.

3. Purify the plasmid from the cell pellets using a kit such as the QIAprep Spin Miniprep Kit and a microcentrifuge (see Note 18).

4. After purification, measure the concentration of DNA and assess purity using NanoDrop UV-vis. An A₂₆₀/A₂₈₀ ratio of ~1.7-2.0 indicates pure DNA. Additionally, confirm the sequence fidelity of the putative plasmid by sequencing (see Note 19).

5. Store purified plasmids at −20 °C (or −80 °C for long-term storage) and thaw before use. Once the plasmid sequence has been confirmed, transform into a production strain of E. coli (either BL21 or BAP1) as follows in Steps 6-11 (see Note 20).

6. Add 1 μL of plasmid to 100 μL of chemically competent cells.

7. Incubate on ice for 30 minutes.

8. Heat shock in a water bath at 42 °C for 90 seconds.

9. Remove from the water bath and incubate on ice for 2 minutes.

10. Add 200 μL of LB and incubate at 37 °C, shaking at 200 RPM, for 1 hour.

11. Plate 100 μL of culture onto LB semi-solid agar supplemented with kanamycin and incubate overnight at 37 °C.

3.4. Expression of ACPs in E. coli

1. Inoculate a seed culture of transformed cells containing the expression plasmid of choice in 10 mL of liquid LB treated with kanamycin (see Note 21).

2. Incubate overnight at 37 °C, shaking at 200 RPM.

3. Prepare 1 L of LB in a 2 L baffled culture flask and sterilize by autoclave.

4. Cool to room temperature and add 50 μg/mL kanamycin (for pET-28a (+)) (see Note 22).

5. In a sterile hood (or using aseptic technique), inoculate 1 L of prepared LB with the seed culture (which should now be cloudy due to overnight growth of cells). Incubate at 37 °C, shaking at 200 RPM.

6. Monitor the OD₆₀₀ of the production culture by measuring the absorbance at 600 nm. Continue incubating until the OD₆₀₀ reaches 0.4-0.6 (see Notes 23 and 24).

7. At appropriate OD₆₀₀, collect a 1 mL aliquot in a microcentrifuge tube and freeze. This sample will be used for subsequent SDS-PAGE analysis as a “pre-induction control”. Then add 250 μL of 1M isopropyl β-D-1-thiogalactopyranoside (IPTG) to each culture in a sterile hood and shake the cultures at 18 °C, 125 RPM overnight for 14-18 hours to induce the expression of the target ACP.

8. After shaking overnight, collect a 1 mL aliquot in a microcentrifuge tube and freeze for subsequent SDS-PAGE analysis as a “post-induction control”.

9. Pour into large (~500 mL) centrifuge tubes and balance. Balancing is very important when using a high-speed centrifuge.

10. Collect cells by centrifugation. (4,500 x g, 15 min, 4 °C) (see Note 25).

11. While spinning down, label a 50 mL conical tube for each protein and incubate on ice.

12. After centrifugation is complete, pour off supernatant. Scrape cells into the prepared conical tube using a clean spatula.

13. Either freeze the collected cells at −80 °C to be lysed at a later date or continue with the purification procedure as follows.

3.5. Purification and On-Column Phosphopantetheinylation of ACPs

Unless otherwise indicated, all materials should be kept at 4 °C (on ice at all times and in a cold room if possible).

1. With the sample on ice, gently pipette the cell pellet from ~1 L of ACP culture in 20 mL Lysis Buffer until it is evenly suspended. Do not vortex.

2. Sonicate using an ultrasonic probe at 12 W in ten pulses (30 seconds on, 30 seconds off) to lyse cells (see Note 26).

3. Centrifuge at 17,000 x g for 45 minutes to remove cell debris; the His₆-tagged target protein should be soluble and remain suspended in the supernatant (see Note 27).

4. Meanwhile, prepare 1.5 mL Ni-NTA agarose affinity resin. Resuspend resin slurry in the bottle by shaking gently and transfer ~3 mL to a conical tube. Centrifuge at 500 x g for 5 minutes (using a benchtop centrifuge) to settle the resin; remove the supernatant. Add 5 column volumes (CVs) of Lysis Buffer and invert gently to equilibrate resin. Centrifuge at 500 x g for 5 minutes to settle the resin; remove the supernatant while being careful to not discard the resin. Repeat equilibration with 5 CVs of Lysis Buffer, centrifuge, and remove the supernatant.

5. After the centrifugation in Step 3 has reached completion, collect 50 μL of the supernatant for future SDS-PAGE analysis. Transfer the remaining supernatant (containing the target protein) to the tube containing the equilibrated, settled resin.

6. Combine resin and protein by shaking on an orbital shaker at 200 RPM for 15 minutes.

7. Add mixed protein/resin to a fritted gravity column and wash with 5-10 CVs of Lysis Buffer.

8. Measure the A₂₈₀ of the Wash Buffer using a spectrophotometer to establish a baseline absorbance (likely between 0.04-0.02). Wash the resin with Wash Buffer until the A₂₈₀ returns to the baseline (i.e., until the absorbance of the eluent is similar to that of the Wash Buffer).

9. To the column, add 1-2 mL of 50 mM sodium phosphate buffer (pH 7.6) and allow the buffer to run through. Keep the resin wet, leaving about 100-200 μL of the buffer above the resin. Collect 50 μL of the flow through for future SDS-PAGE analysis.

10. Prepare 4 mL of the ‘Sfp reaction mixture’ by combining the following (see Note 28):

    1. 2.4 mL of 50 mM phosphate buffer, pH 7.6
    2. 1.2 mL of 50 mM dithiothreitol (DTT) (see Note 29)
    3. 160 μL of 250 mM MgCl₂
    4. 200 μL of 50 mM CoA or CoA derivative
    5. 80 μL of 150 μM R4-4 Sfp enzyme

11. Add the resulting mixture to the column, cap/seal, and leave overnight at room temperature. Gentle rocking overnight using a nutating rocker is advised, but not required.

12. The next day, transfer the column to a 4 °C cold room. Wash the column with Wash Buffer to remove excess CoA until the absorbance of the eluent returns to the baseline (CoA absorbs strongly at 260 nm) (see Note 30). Collect 50 μL of the wash for future SDS-PAGE analysis.

13. Elute the now holo-ACP using Elution Buffer. Collect 10-12 x 1 mL fractions and measure the absorbance at 280 nm. Evaluate the 260/280 ratio to determine protein quantity and purity within the sample. Collect 50 μL of the eluent for future SDS-PAGE analysis.

14. Pool together all the fractions that have protein and measure the 280 nm absorbance of the pooled fraction again. Depending on the ACP, a reasonable titer is 15-25 mg of protein per liter of culture.

15. Dilute a small sample to 2-10 μg/mL in 50 mM sodium phosphate, pH 7.6 and determine the holo/apo ratio by LC-MS (see Note 31). See Section 3.10 for details.

16. Verify the over-expression of the target ACP by SDS-PAGE, including samples of the following: pre-induction and post-induction controls from Section 3.4, supernatant from Step 5, flow through from Step 9, wash from Step 12, and eluent from Step 13.

17. Dialyze the pooled protein overnight at 4°C toward 2-3 L of 50 mM sodium phosphate, pH 7.6. The next day, concentrate and store for future use (see Note 32).

3.6. Cyanylation of holo-ACP Ppant arm

1. Determine the desired amount of ACP to cyanylate. We typically aim to make 100 μL of 1.5 mM ACP-SCN (~1.5 x 10⁻⁷ mol) to ensure sufficient protein for analysis by IR and titration assays (see Note 33).

2. If necessary, thaw frozen holo-ACP samples on ice. If the holo-ACP is stored in multiple aliquots, combine and concentrate to a volume of ~1 mL or less. Reactions will occur in 1.5 mL microcentrifuge tubes.

3. Determine the concentration of the holo-ACP via a protein quantification assay and/or transmission UV-vis.

4. Some carrier proteins (such as DEBS holo-ACP2(2) and Yersiniabactin holo-ArCP) form disulfide bonds between solvent exposed Ppant arms, creating ACP-homodimers. This prevents reaction with Ellman’s reagent (5,5'-dithiobis-(2-nitrobenzoic acid) or DTNB). If your carrier protein does not form homodimers, skip Steps 5-7 and move to Step 8.

5. Reduce disulfide bonds between Ppant arms by adding 100 times molar excess dithiothreitol (DTT). Mix by gentle pipetting up and down and incubate for 30 minutes at room temperature.

6. Remove excess DTT using a PD-10 column (collect 0.25 mL fractions to maximize protein concentration of eluent).

7. Use a protein quantification assay or transmission UV-vis to identify fractions with protein. Pool these fractions and remeasure concentration. Move on to Step 8.

8. Add eight times molar excess of pre-dissolved DTNB (which is more soluble in a high ionic strength buffer) to the ACP sample. Mix gently by pipetting up and down. The solution should turn pale yellow due to the release of TNB²⁻. Incubate for at least 30 minutes at room temperature, or on ice for more sensitive ACPs.

9. Follow the reaction using UV-vis at A₃₂₅ for DTNB and A₄₁₂ for TNB²⁻. Refer to the Ellman’s Reagent Protocol by Thermo Fisher Scientific for details on how to calculate the free sulfhydryl concentration of the reaction over time: https://www.thermofisher.com/order/catalog/product/22582?SID=srch-hj-22582#/22582?SID=srch-hj-22582. Also note that the exact extinction coefficient for the side product has been shown to exhibit thermochromism and that this can affect the calculated yield. The reaction is considered complete once the moles of TNB²⁻ reach that of the ACP in solution (see Note 34).

10. Add 55x molar excess of dissolved NaCN to the mixed ACP/DTNB/TNB²⁻ solution. Mix by pipetting up and down. The solution will turn dark yellow/orange. Incubate the reaction at room temperature for 30 minutes. The activated ACP-TNB will react with CN⁻ ions to produce a cyanylated ACP, as illustrated in Figure 2 (see Note 35). Note that NaCN and aqueous cyanide ions are toxic chemicals. Be sure to refer to the material safety data sheet (MSDS) for information on how to properly store, handle, and dispose of NaCN solutions, which should always be made on the smallest scale possible and immediately disposed of after use.

11. While the reaction incubates, prepare a PD-10 desalting column, which uses Sephadex G-25 Medium as a size-exclusion chromatography matrix to separate high molecular weight materials (proteins) from low molecular weight materials (salts, small-molecule reagents, and solvents), for separation of excess cyanylation reagents. Details of this step should be obtained from the manufacturer’s protocol. Equilibrate column with 20 mL double distilled water (ddH₂O), followed by 20 mL of the same buffer used for the cyanylation reaction solution. Discard the flow through in a waste beaker.

12. Apply the ACP-SCN reaction solution to the prepared PD-10 column. It is preferable to perform the following steps at 4 °C, although not necessary for most ACPs. Apply 0.5 mL of the reaction sample to the PD-10 column (see Notes 36 and 37). Allow the sample to enter into the packed bed and discard any flow-through.

13. Elute protein from the column. Prepare a rack of 15 microcentrifuge tubes labeled 1-15. Add 500 μL of equilibration buffer (50 mM Phosphate buffer) and collect in the first microcentrifuge tube. Repeat the addition of buffer to the column 14 times, collecting 500 μL eluate in separate microcentrifuge tubes each time. The protein will elute faster than the excess reagents and side products (in the yellow solution) because it is larger.

14. Determine the ACP-containing factions by measuring the concentration of protein via transmission UV-vis and discard fractions with substantial absorptions above 300 nm. These absorptions are the result of reaction side products.

15. Combine fractions that contain protein and concentrate using centrifugal concentration tubes with a 3 kDa molecular weight cut off (MWCO). Remember from Step 1 that a good target concentration for future assays is 1.5 mM.

16. Analyze the sample by liquid chromatography-mass spectrometry (LC-MS; see Section 3.10) and/or obtain a spectrum of the ACP-SCN via IR (see Section 3.7).

17. Flash freeze cyanylated proteins and store at −80°C (see Note 38).

18. Run buffer through the used PD-10 column until all cyanide is eluted (~25 mL), then rinse the column with ddH₂O for storage. Carefully dispose of the eluted cyanide waste as directed by the MSDS.

Figure 2.

Reaction scheme of the conversion of holo-ACP to the cyanylated ACP (ACP-SCN). The free thiol of the Ppant arm acts as a nucleophile to break up the disulfide bond of Ellman’s Reagent, activating the ACP for reaction with aqueous cyanide ions. Cyanide attacks the sulfur of the Ppant arm thiol, releasing TNB²⁻.

3.7. Collecting and Analyzing a Vibrational Spectrum of ACP-SCN [22]

1. Prepare the FT-IR instrument for data collection by cooling the photovoltaic mercuric cadmium telluride (MCT) detector with liquid nitrogen. Refill completely every 5-6 hours (see Note 39 for general instructions associated with other FTIR spectrometers).

2. Collect the IR background of the buffer as follows in Steps 3-13.

3. Rinse the CaF₂ windows of a Harrick 13 mm diameter demountable liquid cell (with 2 CaF₂ windows and a 56 μm Teflon spacer) with water and dry with a Kimwipe.

4. Place the Teflon spacer on one of the CaF₂ windows and place the two into the liquid cell.

5. Pipette 6-10 μL of the appropriate background buffer onto the CaF₂ window. Be very careful to avoid making bubbles as they will disrupt the IR signal.

6. Carefully drop the second window onto the sample inside the cell, again avoiding the formation of bubbles.

7. Quickly place the retaining ring and screw cap on top and seal the cell tight. Do not overtighten, as this will crack the brittle CaF₂ windows.

8. Place the entire cell into the Vertex 70 IR.

9. Close the IR lid and wait about 10-15 minutes for the chamber to automatically purge itself of CO₂.

10. Set up OPUS data acquisition. Open OPUS and click the “Measurement” tab to open the run parameters. Include scanning from 1000-4000 cm⁻¹ for 512 scans at 10 kHz scanner velocity and 2 cm⁻¹ resolution (see Note 40). Make sure that the signal is adjusted so that interferogram saturation is avoided when the blank (solvent+buffer) sample is present.

11. When all parameters are set, return to the basic tab and click “Background Single Channel” to collect the solvent reference spectrum.

12. Click “Sample Single Channel” to also collect a spectrum of the solvent blank (see Notes 41 and 42).

13. The (blank) spectrum file will appear on the left-hand column once it has finished processing. Choose the “ssc” sample block, Go to File → Save File As, and save your file as a .dpt text file in a desired location for further processing outside of OPUS.

14. Remove the cell from the IR chamber and rinse with deionized water.

15. Carefully add ~7 μL of ≥1 mM ACP-SCN in the same buffer solution onto the windows as done in Step 6 when adding the buffer, avoiding bubble formation and using the exact same windows and spacer.

16. Place the liquid cell into the chamber and purge the instrument’s internal atmosphere.

17. Return to the basic tab and click “Sample Single Channel”.

18. Look again to the left column for your file. Select the absorbance block and right click to save the absorbance file (“abs” block) in .dpt format. If you plan to attempt manual averaging, you should also save the “ssc” block in the same way. The “rsc” block contains the reference spectrum collected above in Step 11.

19. To correct the baseline of the absorbance spectrum in the CN stretching region and highlight the CN peak for further analysis (see Note 43), follow Steps 20-25.

20. Open Origin and go to File → Import → Single ASCII and select the absorbance file saved in Step 13 to open the .dpt file of the scan data.

21. Delete all data points far away from the nitrile stretching region. Do not delete 2100-2210 cm⁻¹.

22. Duplicate the data twice by right clicking the tab at the bottom and clicking “Duplicate”. The first data set is for record keeping. The second data set is for making a “spectral hole” window. The third data set is for subtracting the reference background.

23. The window in which any ACP-SCN will appear is from 2145 to 2180 cm⁻¹. Highlight these data points in the window. Right click and “Clear” to make a hole in the data where the CN signal should be. Graph the data with the hole by selecting the forward slash at the bottom left-hand corner. At the top of the window select Analysis → Fitting → Polynomial Fit → Open Dialogue. Fit the remaining data (without the CN peak region) to a high-order polynomial (at least 7th degree).

24. Go to the third duplicated window and highlight both columns [A(X) and B(Y)]. Select Analysis → Data Manipulation → Subtract Reference Data → Open Dialogue. A Data Manipulation subtract_ref window will pop up. The Input 1 set will already be filled with the data you highlighted from the third window. Fill the empty Reference Data by clicking on the red button. On the bottom tabs, click FitPolynomialCurve1 (which is the fitted polynomial to the data with the hole). Select the first two columns. Return to the Data Manipulation Window and hit OK. Return to the third duplicated window. The B(Y) column should now contain Subtracted values. Plot the two columns.

25. Save the resulting absorbance spectrum in the CN region and export the graph, if desired, using the file dropdown menu.

26. If interested in further improving signal-to-noise ratio beyond what the instrument can provide in single scans, manually average multiple spectra of 512 scans as follows in Steps 27-31.

27. Collect five (or more) solvent spectra and the same number of sample spectra, as outlined in Steps 2-18. These can be from new refills of the same cell or from different illumination spots or orientations of the same sample.

28. Import the “ssc” data blocks, exported to .dpt files, from Steps 13 and 18 into Origin using the “Import Multiple ASCII” dialogue. The x data (frequencies in cm⁻¹) should all be the same; place all of the y data into one worksheet.

29. Create two new columns. In the first new column, use “Set Column Values” to create a sum of the blank single channel data blocks, and in the second new column, do the same for the sample (ACP-SCN) single channel data blocks.

30. Add one more new column, and use “Set Column Values” to create a new absorbance spectrum by manually calculating the absorbance spectrum, which is equal to

    $$-{\log }_{10}\left(\frac{\text{Sum of}\text{Samples Column}}{\text{Sum of}\text{Blanks Column}}\right).$$

31. Plot this last column, which is the signal-averaged absorbance spectrum from repeated samples and repeated blanks. Follow the baseline procedure in Steps 26-31 to further reveal and isolate the CN stretching peak.

3.8. Colorimetric ACP-KS Mechanistic Crosslinking Assay Utilizing Ellman’s Reagent (DTNB) (see Note 44)[23]

3.8.1. Preparation of ACP-TNB⁻:

1. Add 8 molar equivalents of 25-mM Ellman’s reagent (5,5'-dithiobis-(2-nitrobenzoic acid) or DTNB) in 50 mM phosphate buffer, pH 7.6, to holo-ACP (0.8–1.1 mM in 50 mM sodium phosphate buffer, pH 7.6) for conversion (see Note 45). Incubate at room temperature for 60-90 minutes. The reaction mixture will turn bright yellow, as shown in Figure 3.

2. While the reaction mixture incubates, wash a PD-10 column with 5-10 CVs of 50 mM phosphate, pH 7.6 buffer to remove any ethanol or antimicrobial solution remaining from storage of the column.

3. Label 15 1.5 mL microcentrifuge tubes from 1-15.

4. After the DTNB reaction mixture has incubated for 90 minutes, apply 1-2 mL of the mixture to the column (see Note 46).

5. Collect 400-500 μL fractions, adding reaction mixture to the column as necessary to prevent the column from running dry (see Note 47).

6. Measure the OD₂₈₀ of each fraction. Two products should be isolated: the target product, ACP-TNB⁻ (which absorbs at 280 nm), and the byproduct, free DTNB/TNB²⁻ (which absorbs at 315 and 412 nm) (see Note 48).

7. Pool fractions containing pure ACP-TNB⁻ and measure the absorbance. Save 10 μL for analysis by liquid chromatography-mass spectrometry (LC-MS). Concentrate the rest of the purified protein using a centrifugal concentration column with a 3 kDa MWCO.

8. Measure the concentration of ACP-TNB⁻ in the sample using a protein quantification assay (following the manufacturer’s specifications) or by UV-vis spectroscopy.

Figure 3.

Reaction scheme of activating holo-ACP to obtain/visualize ACP-KS interactions. The ACP Ppant arm is converted into a mix disulfide using DTNB. ACP-TNB⁻ is then activated to form a covalent cross-link with the KS active site thiolate upon mechanistically relevant binding of the ACP and KS. The cross-linking reaction can be observed by SDS PAGE or by the release of aqueous TNB²⁻, which absorbs at 412 nm and turns the solution yellow.

3.8.2. Quantification of crosslinking activity:

1. Ensure that the DTNB reaction successfully attached the TNB⁻ probe to the holo-ACP by LC-MS using the sample saved in Step 7 above. See Section 3.10 for details.

2. Combine purified ACP-TNB⁻ with the desired ketosynthase (either a ketosynthase homodimer, KS, or a ketosynthase-chain length factor heterodimer, KS-CLF) to make a 20 μL solution in 50 mM phosphate, pH 7.6 buffer with target concentrations of 25 μM KS and 50 μM ACP-TNB⁻.

3. Incubate the reaction mixture at room temperature for exactly 20 minutes. A yellow color-change is an indication of reactivity. Upon crosslinking of ACP-TNB⁻ with the KS, TNB²⁻ is released, which absorbs at 412 nm. After 20 minutes, measure the absorbance of the reaction mixture at 412 nm. Use the TNB²⁻ extinction coefficient (14,150 M⁻¹ cm⁻¹ at 25 °C at 412 nm) to calculate the number of moles of TNB²⁻ released for each reaction.

4. Crosslinking can additionally be observed via SDS-PAGE under non-reducing conditions as a band with a molecular weight of the crosslinked ACP-KS complex. Under reducing conditions (5% BME), the crosslinking disulfide bond is reduced and the complex separates, appearing as two separate bands (one for the ACP and one for the KS, or one each for the KS and CLF, if applicable).

3.9. Tracking Ppant Sequestration Activity Using Raman Spectroscopy

3.9.1. Loading of alkyne probe-labeled acyl chain [24]:

1. In a glass vial, combine the following, with a total volume on a 1 mL scale and final concentrations as given (see Note 49):

   1. 275 μM ACP in 50 mM sodium phosphate buffer, pH 7.6
   2. 2.5 mM DTT
   3. 23 mM MgCl₂
   4. 18 mM ATP, pH adjusted to 7.6 (see Note 50)
   5. 0.8 μM Acyl-acyl carrier protein synthetase (AasS) in Tris buffer, pH 7.6 (see Note 51)
   6. 4.6 mM carboxylic acid probe (4-pentynoic acid (C5), 7-octynoic acid (C8), or 12-tridecynoic acid (C13)), stock in isopropanol

2. Shake the reaction mixture at 100 RPM for 16 hours at 37 °C.

3. Centrifuge at 17,000 x g for 5 minutes to precipitate a pellet of insoluble species.

4. Decant the supernatant onto a PD-10 desalting column to separate the probe-loaded ACP from salts and excess substrate.

5. Identify fractions containing protein using transmission UV-vis. Unloaded carboxylic acid can be identified by a peak at 260 nm and should elute after the probe-loaded protein. Combine any fractions containing protein and lacking unloaded carboxylic acid.

6. Concentrate pooled fractions to 1-3 mM using a centrifugal filter with a 3 kDa MWCO.

3.9.2. Data collection:

The methods described here are for a home-built CW-Raman spectrometer[25]. However, similar methods can be applied to other similar dispersive Raman instruments, substituting instrument-specific instructions as required by the manufacturer and exact hardware and optical setup, which can be highly variable (see Note 52).

1. Focus a 532 nm DPSS CW laser attenuated to 80 mW incident power vertically through a 1 mM diameter glass capillary filled with 1-5 μL of sample.

2. Collect scattered light at 90° to the incoming excitation using an f/1.2 camera lens, collimate the collected light, and reject the Rayleigh scattering using an appropriate filter.

3. Focus the remaining collected light into the slit of a 500 mm single monochromator with a 600 grooves/mm diffraction grating blazed at 500 nm and oriented at an angle such that the OH stretching band of water appears at the long-wavelength edge of the spectrum.

4. Collect the spectrum on a back-illuminated, liquid-nitrogen cooled CCD camera. Bin the camera output to produce a single spectrum (scattered light counts vs wavelength).

5. Collect spectra in exposures of 1 minute (exposure time should be kept below detector thresholds for saturation) for up to 2 hours of total accumulation time.

6. Convert the wavelength to Raman shift for each column of the camera output.

7. Analyze spectra as described in Section 3.7, Steps 19-25, importing the raw data from the scans directly into Origin. The same assumptions apply to using this analysis for Raman data as for IR (see Note 43).

3.10. Liquid Chromatography-Mass Spectroscopy (LC-MS) Analysis of ACPs

1. ACPs can be analyzed using a single quadrupole LC-MS designed for routine mass-selective detection. While ACPs are typically ~8 to 10 kDa, they can be detected within the <1500 m/z range by deconvoluting sets of peaks observed in the mass spectrum representing multiple changed states. First, calculate the expected molecular weights of the ACPs of interest (in the apo form) using the ExPASyProtParam Tool, found here: https://web.expasy.org/protparam/. Note that the addition of the Ppant arm leads to a Δ + 340 Da, so add 340 Da to the weight calculated for the apo form to find the predicted molecular weight of the holo form. Conversion of holo-ACP to ACP-SCN (Δ + 26 Da) or ACP-TNB- (Δ + 196 Da) can also be readily observed by LC-MS analysis.

2. Inject 20 μL of a 0.1 mg/mL ACP solution (in 50 mM sodium phosphate buffer, pH 7.6) to an LC-MS equipped to perform electrospray ionization-mass spectroscopy (ESI-MS) with a Waters XBridge Protein BEH C4 Column (or similar 300Å, 3.5 μm, 2.1 mm × 50 mm column) heated to 45°C (see Note 53). Use the following solvent gradient (where solvent A = water + 0.1% formic acid and solvent B = acetonitrile + 0.1% formic acid): 0–1 min 95% A, 3.1 min 5% A, 4.52 min 5% A, 4.92–9 min 95% A.

3. ACPs typically elute at ~4-4.5 minutes, which can be observed by peak in the LC trace that absorbs at 280 nm. The corresponding mass spectrum will likely look like a single set (for ACPs in a single form) or multiple sets (for ACPs in multiple forms) of 8-10 peaks corresponding to the +7 to +14 charge states.

4. Deconvolute the mass peaks produced by the run using ESIprot, a free online software found at https://www.bioprocess.org/esiprot/esiprot_form.php[26]. The resulting value can be compared to molecular weight predictions for the proteins of interest to determine if the ACP has been converted to the holo form (see Notes 54 and 55).

4. Notes

1. Two particularly helpful resources for understanding and applying Gibson Assembly® are Gibson, et al., Nature Methods (2009) and https://www.addgene.org/protocols/gibsonassembly/. It is advisable to understand the method before applying the primer design algorithm above.

2. We clone our genes of interest into pET-28a (+) at the NdeI/EcoRI cut sites to obtain the N-terminal His6-tagged ACP. The His6-tag can then be cleaved using thrombin.

3. Note that the program will run prematurely if carriage returns (new paragraphs) are included within the copied-and-pasted sequence. This will result in “Command not found” errors and can be solved by removing carriage returns before copying the sequence.

4. The primers produced by this algorithm are 49 base pairs in length. Typically, this length is within the less expensive range of oligonucleotides offered by manufacturers such as Eurofins Genomics.

5. While the melting temperatures of the primers should ideally be within ~5 °C of each other, we have observed success with wider gaps (up to 10 °C). Nonetheless, there are several ways to alter the melting temperature of a primer. Generally, adding more base pairs increases the temperature, while decreasing the length of the primer decreases the temperature. An A-T pair forms only two hydrogen bonds and thus has a lower melting temperature than a G-C pair, which forms three hydrogen bonds. Through trial-and-error, these guiding principles can be combined to produce primers that accurately reflect the gene sequence while also having an acceptable melting temperature.

6. PCR can be performed with a variety of polymerases. Information and protocols for the use of one such polymerase, the Q5® High-Fidelity DNA Polymerase, can be found here: https://www.neb.com/protocols/2013/12/13/pcr-using-q5-high-fidelity-dna-polymerase-m0491.

7. Actinomycete genomes (and other type II PKS BGCs) are often rich in GC content (>60%). The increased bond strength between this pair of nucleic acids (as opposed to AT pairs) can lead to difficulties in PCR. If our sequence is particularly high in GC, we add up to 5% DMSO (v/v) or a “GC enhancer” (volume according to the manufacturer’s guidelines) to the PCR reaction mixture. Several companies produce GC enhancers. We use the Q5 High GC Enhancer by New England BioLabs.

8. Information, including protocols, for the Zymo DNA Clean & Concentrator™ Kit can be found here: https://www.zymoresearch.com/collections/dna-clean-concentrator-kits-dcc.

9. Additional useful information about restriction digestion can be found on the New England BioLabs website: https://www.neb.com/protocols/2012/12/07/optimizing-restriction-endonuclease-reactions.

10. To ensure linearization of the vector, we recommend transforming chemically competent cells with the doubly digested (linearized) vector and plating on LB agar plates supplemented with kanamycin. Do the same with intact pET-28a (+). The sample transformed with the linearized vector should not exhibit any growth on the antibiotic-treated plate, while the sample transformed with the intact plasmid should grow many colonies. DNA gel electrophoresis can also be used to verify the successful digestion of the vector, as circularized DNA runs differently than linear DNA.

11. An easy way to assure the correct ratio between the DNA segments is to dilute the more concentrated component to match the concentration of the lesser concentrated component and simply add 1 μL of plasmid, 3 μL of insert, and 1 μL of nuclease-free water.

12. New England BioLabs, the producer of Gibson Assembly® materials, suggests that a total of 0.02 to 0.5 pmol of DNA be used when assembling 1 or 2 inserts into a vector.

13. Any host strain for routine cloning applications using plasmid vectors can work for this step. We typically use XL1 Blue or DH5ɑ cells.

14. Target concentrations (in μg/mL) for several widely used antibiotics: Ampicillin - 100; Kanamycin - 50; Carbenicillin - 100; Chloramphenicol - 25; Spectinomycin - 50; Tetracycline - 10. Note: Tetracycline is a type II polyketide!

15. The replication of a large plasmid (i.e., a plasmid with a gene insert larger than a single ACP) is a resource-intensive process for cells, so expect the colonies of strains containing larger plasmids to take longer to appear. However, plasmids appearing after 72 hours of incubation are typically “satellite colonies” and very likely do not contain the desired plasmid.

16. Growth for greater than 16 hours is not recommended. Cells will begin to lyse and plasmid yields will be reduced.

17. Be sure to use a tube or flask with a volume of at least 4 times the volume of the culture and that the cap is loose enough to allow appropriate aeration (although not so loose as to allow contamination).

18. Information, including protocols, for the QIAprep Spin Miniprep Kit can be found here: https://www.qiagen.com/us/products/discovery-and-translational-research/dna-rna-purification/dna-purification/plasmid-dna/qiaprep-spin-miniprep-kit/#orderinginformation

19. We use the tube sequencing services provided by Eurofins Genomics: https://www.eurofinsgenomics.com/en/products/dna-sequencing/all-sequencing-options/

20. This step can be carried out before sequencing results are received, just be sure to keep track of which strains have been transformed with confirmed plasmids and which have not. Sequencing results are often returned the day immediately following shipment, so it is unlikely that strains with a faulty plasmid will get too far down the pipeline before the results come back. Restriction digestion can also be utilized (using NdeI and EcoRI) to cut the putative plasmid at the NdeI and EcoRI cut sites. These fragments can then be run on an agarose gel to confirm the size of the insert.

21. Production of the target protein can easily be scaled up by doubling (tripling, etc.) the number of production cultures grown on Day 2. In this case, inoculate at least one seed culture for every planned production culture.

22. We keep 1 mL aliquots of a number of different antibiotic stock solutions at 1000x the target concentration so we can simply add 1 aliquot of antibiotic to 1 L of LB to reach the correct concentration. For example, we keep 1 mL aliquots of kanamycin at 50 mg/mL, which results in a 50 μg/mL final concentration upon addition to 1 L of LB.

23. Typically, cultures supplemented with a single antibiotic will reach an OD of 0.4-0.6 in about 2-3 hours. Note that E. coli has a doubling time of ~20 min.

24. You can blank with water, as LB does not absorb at 600 nm.

25. We use a Beckman-Coulter Avanti floor centrifuge with a JA-10 rotor at 5,000 RPM to harvest the cells.

26. There are a number of other lysing methods, such as freeze-thaw cycling, French press, and commercially available lysis reagents. We recommend sonication due to its speed and ease of use.

27. We use a Beckman-Coulter Avanti floor centrifuge with a JA-18 rotor at 13,000 RPM to remove cell debris.

28. If low on materials, this reaction can be run with as little as 2 mL of reaction mixture. Simply scale ingredients accordingly.

29. DTT is a strong reducing agent and must be made fresh every time. It is normal for DTT to turn the Ni resin brown.

30. It usually takes about 30-50 mL of Wash Buffer to bring the absorbance at 260 nm close to zero.

31. The Sfp reaction is quite robust, so the goal is to reach 100% holo-ACP. If the target ACP is not efficiently converted from apo- to holo- form by R4-4 Sfp, the efficacy of AcpS-type phosphopantetheinyl transferases and/or use of the native 4'-phosphopantetheinyl transferase can be explored [19].

32. We generally store our ACPs in 500 μL aliquots at ~1-2 mM.

33. The Harrick 13 mm IR liquid cell that we use requires 7 μL of (at least) 1 mM ACP per run. Be aware that conversion of holo-ACP to ACP-SCN does not always go to completion, according to LC-MS results. The protocol outlined here cyanylates 1-2 mg of ACP.

34. A₄₁₂ measures TNB²⁻, the side product of the reaction between DTNB and the free thiol of the Ppant arm. In general, for every mole of TNB²⁻ in solution, a mole of ACP has been converted to the target mixed disulfide. If the target ACP has solvent-exposed thiols from surface cysteines, it is possible for these to react with the DTNB, although we tend to observe preferential modification of the Ppant arm due to its steric accessibility. If off-target modification becomes a problem, any solvent-exposed cysteine residues can be mutated to serine residues.

35. CN⁻ ions also react with remaining DTNB, which is largely what leads to the orange-colored solution. Keep in mind that CN⁻ is a strong base, so sufficient buffer capacity is required to keep the pH from rising and interfering with the successful formation of the covalently linked SCN functional group.

36. Excess cyanide and other species will not be effectively separated from the protein if more than 0.5 mL of the reaction solution is applied to the column at a time.

37. Repeated for emphasis: NaCN and aqueous cyanide ions are toxic chemicals. Be sure to refer to the MSDS for information on how to properly store, handle, and dispose of NaCN solutions.

38. Some less-stable ACPs require flash freezing with a cryoprotectant. If precipitation, typically white and wispy, is observed upon freeze-thawing, we add 5-30% glycerol (v/v) before freezing.

39. Detecting a single nitrile probe signal at millimolar concentration (or even below) can be accomplished with a number of different commercial FTIR spectrometers, but what is absolutely necessary is a very stable and high signal-to-noise detector. We use a photovoltaic HgCdTe (MCT) detector produced by Kolmar, Inc. because it provides an excellent balance between sensitivity and robustness in that the signal does not easily saturate but very small differences in transmitted light can be detected. Other groups use InSb detectors, which are very sensitive in this particular spectral region, are NOT generally useful at lower frequencies, and saturate very easily (thus the source intensity must be dialed way back to work with InSb in a way that is less necessary with MCT). The most important hardware to make this kind of FTIR experiment work successfully is a really excellent, liquid N₂-cooled detector and a stable N₂ purge source.

40. In H₂O samples, the region from approximately 1900-2800 cm⁻¹ will contain usable data for probe detection, but on either side of this region there will be nonsensical data due to complete absorption of the IR source by H₂O. While it can be useful to collect data outside of this narrower window to view other signals from the protein, best signal-to-noise is obtained (and faster scanning can be accomplished) by limiting each scan to 1900-2700 cm⁻¹. The nitrile stretching band is broad enough that scanning at high resolution (i.e., lower than 1 cm⁻¹) is not productive, but 2 cm⁻¹ provides an upper bound on being able to collect a high-fidelity lineshape that is not affected by instrument resolution.

41. Single scans may be used to visualize the purge and the disappearance of water and CO₂ from the spectrum. CO₂ will show up on the spectrum as a strong doublet centered at about 2345 cm⁻¹ overshadowing the area of interest corresponding to the thiocyanate probe, and water vapor appears as a series of sharp peaks spread across two broad spectral regions. No light will escape the sample from 3300-2900 or 1500-1650 cm⁻¹, so those regions should be ignored when H₂O is the solvent. Removal of these signals via a continuous dry N₂ purge provides much cleaner signals, especially from small peaks like the nitrile stretch of ACP-SCN at low-mM or lower concentrations.

42. If you plan to manually average multiple spectra in the following ACP scans, collect the blank sample scan 5 times and give each a different name (i.e., Buffer 1, Buffer 2, etc.). This is particularly advisable when working with a small concentration of ACP with poor signal-to-noise ratio. Flip the orientation of the cell at least once and remember to allow the chamber to re-purge after moving or replacing the cell.

43. This baseline subtraction scheme assumes that there are no signals outside the “hole” region, which is not technically the case; the CN stretching band appears on top of the very broad, solute-sensitive combination band of liquid H₂O that is expected to shift between the solvent reference spectrum and the protein-containing sample spectrum. The baseline-subtraction scheme outlined here (which fits the regions immediately outside the peak region to an arbitrary but very flexible polynomial function, then subtracts the fit from the actual data) can introduce artifacts in the CN stretching lineshape. These artifacts can appear as “wiggles” or negative absorption features at the edges of the post-processed lineshape. The presence of such artifacts suggests that the “hole” might be incorrect or that otherwise the assumptions made in this rudimentary baseline correction procedure are breaking down.

44. This colorimetric assay was developed and rigorously tested using the ketosynthase FabF from the E. coli fatty acid synthase. Whether it can or cannot be applied to other KSs and KS-CLFs is still under investigation, although preliminary results indicate that it likely can be.

45. “Molar equivalent” refers to the ratio of the number of molecules of one species relative to the number of molecules of another. In this case, we want there to be 8 DTNB molecules for every ACP molecule.

46. The column should be run almost dry of buffer before the sample is added; adding the sample with more than 50-100 μL of buffer on top of the column will dilute the sample.

47. Collecting larger fractions will make it more difficult to isolate a pure sample of ACP-TNB⁻.

48. In our experiments, ACP-TNB⁻ typically elutes in or around fractions 4, 5, and 6, while free DTNB/TNB²⁻ elutes after fraction 7.

49. Prepare in a glass vial, as plastic tubes may contain stray carboxylic acids that compete with the desired probe for loading onto the ACP.

50. The addition of more ATP may help if the reaction does not run to completion.

51. Acyl-acyl carrier protein synthetase (AasS) is derived from Vibrio harvei and is a promiscuous ligase for the addition of fatty acids (such as C5, C8, and C13) to the terminal thiol of the ACP’s Ppant arm.

52. Successful detection of single-probe signals from aliphatic alkynes in low-mM (or lower) concentration samples via CW Raman depends mainly on two factors: a relatively intense scattering light source (visible is best as UV could lead to sample damage for long exposure) and a back-illuminated camera (usually CCD or EM-CCD). FT-Raman instrumentation will typically not be able to collect these signals with high fidelity: a dispersive instrument with an excellent detector is required. Time-resolved or pulsed higher-order Raman techniques should be able to acquire these probe signals. Dispersion gratings should be selected to provide a balance between spectral resolution (which should be on the order of 2-3 cm⁻¹ in the alkyne CC stretching region, at maximum) and over-spreading the few scattered photons across the available camera pixels.

53. We use an Agilent G6125BW LC-MS for our analysis of proteins.

54. Several post-translational modifications to the ACP are often observed via LC-MS. We most commonly observe the loss of methionine (-Met) and the gluconoylation of the His tag (+gluconoylation). These and other modifications can be found in the given Excel spreadsheet of post-translational modifications.

55. A large error in the ESIprot result indicates the presence of multiple species in the sample, such as both apo- and holo-ACP, and the sets of peaks corresponding to each species have been mixed when selecting peaks for deconvolution.