PCR Verification of Escherichia coli Auxotrophic Expression Strains (Supporting Information)

Table S1 summarizes the current set of new C43(DE3) and BL21(DE3) auxotrophic expression strains of E. coli designed to facilitate the labeling of either membrane proteins or water-soluble proteins with selected amino acid types enriched with stable isotopes such as 2H, 13C and 15N. The use of a suitable auxotrophic expression strain with the corresponding input isotope labeled amino acid(s) in the growth medium ensures high levels of efficiency as well as selectivity in stable isotope labeling. This page provides the supporting information about these auxotrophic expression strains.

Back to the E. coli Auxotrophic Expression Strains Page.

Table S1. New E. coli amino acid auxotrophic host strains used for selective isotope labeling

PCR Primers for Verification of Each Target Gene

List of PCR primers used for verification of each knocked-out gene in Table S1 (typical results are shown below).

up

Colony PCR Data of Each Auxotrophic Strain in Table S1

Click to magnify each image (and either use browsers "back" button or double click to close the image window).
Left lane, standard marker; right lane, PCR-amplified target gene in the wild-type strain.

Auxotrophic stock strains having a few target genes deleted in an unexpected manner

The following stock strains should be used with caution! Colony PCR data (below) indicated that a few target genes (red) were probably deleted in an unexpected manner with the λ-Red recombination system (not tested by direct sequencing), whereas other target genes (black) were knocked out as originally designed. Click to magnify each image (and either use browsers "back" button or double click to close the image window).

up

PCR Primers Used for Construction of the RF Auxotrophic Strains

List of PCR primers used for construction of the RF/YM/EH/MS/ML auxotrophic strains (see Table S1) in steps 1, 2, Fig. S1.

up

Four General Transaminases (or aminotransferases) of Escherichia coli

For selective labeling with certain amino acids, multiple genetic lesions are required because of the reversible transfer reactions and the overlapping specificities of four general transaminases (or aminotransferases) of E. coli (encoded respectively by ilvE, avtA, aspC and tyrB genes) (Fig. S2). In principle, the potential use of strains with defects in all four general transaminase genes (ilvE, avtA, aspC and tyrB) of E. coli should help minimize possible dilution and scrambling of the input amino acid label whenever available (J. Biomol. NMR 8, 184-192 (1996)) - in practice, some of these strains appear to exhibit complicated patterns of amino acid combination requirements for growth or they grow very slowly (J. Biochem. (Review) 169, 387-394 (2021)), so careful planning of experimental conditions is required (Tables 1, S1):

• (i) neither the tyrB nor the aspC deletion by itself confers amino acid auxotrophy on the BL21(DE3) cells (Table 1);
• (ii) strains having knockouts in both aspC and tyrB genes (RF4, RF5, RF13, RF14, RF15, RF16, RF17, RF18 and RF21) require Asp (but not Glu) for growth in M63 minimal media, and some of them (RF15, RF16, RF17, RF18 and RF21) grow only slowly in M63 minimal media in the presence of the L-amino acids specified in Table 1 and Fig. S3;
• (iii) strain RF17, having knockouts in the aspC, tyrB and ilvE genes, requires the presence of Asp, Tyr, Phe, Ile and Leu for (slow) growth in M63 minimal media (Fig. S3);
• (iv) strains RF18 and RF21, having knockouts in all four of the general transaminase genes (aspC, tyrB, ilvE and avtA ), appear to require the presence of Asp, Tyr, Phe, Ile, Leu and Val for slow growth in M63 minimal media - of these, although strain RF21 has further knockouts in genes yfbQ (alaA) and yfdZ (alaC), it requires the presence of Asp, Tyr, Phe, Ile, Leu and Val for slow growth in M63 minimal media like RF18, and is not an Ala auxotroph either (Fig. S3) [cf., RF16, having knockouts in aspC, tyrB, trpA, trpB, glyA, serB and cysE genes and representing an ideal genotype to selectively label Ser, unexpectedly represents a new BL21(DE3) derived Ala auxotroph, which requires the presence of Asp, Tyr, Trp, Gly, Ser, Cys and Ala for very slow growth in M63 minimal medium, but does not grow in M63 minimal medium in the presence of Asp, Tyr, Trp, Gly, Ser and Cys like RF15; see Fig. 4] (J. Biochem. (Review) 169, 387-394 (2021));
• (v) considering also the regulation of the Asp degradation pathways by feedback inhibition by Thr and Lys, and some form of repression by Thr, Ile, Lys and Met, further deletions of asnA and asnB in RF18 (Fig. S3) would result in a BL21(DE3) derivative requiring the presence of Asp, Tyr, Phe, Ile, Leu and Asn for growth in M63 minimal media, which is expected to be applicable to selective labelling of Asp at least for short-term cultivations grown in medium supplied with sufficient amounts of Thr, Ile, Lys and Met (J. Biochem. (Review) 169, 387-394 (2021)).

Thus far, we have been able to obtain such "ideal" strains from E. coli BL21(DE3) (RF18 and RF21 in Tables 1, S1 and Fig. S3 (J. Biochem. (Review) 169, 387-394 (2021))) but not from C43(DE3) expression host cells, despite multiple attempts using the λ-Red recombination system (Fig. S1). For example, deletion of the ilvE, avtA and aspC genes from the C43(DE3) chromosome resulted in an auxotroph for Ile and Val (Tables 1, 3), and further knockout of the tyrB locus, which was not possible with the C43(DE3) cells for reasons unclear to us, would extend the auxotrophy to include Leu (J. Biochem. (Review) 169, 387-394 (2021)). For Leu auxotrophy of some C43(DE3) auxotroph strains in Table 1, it is therefore necessary to repress the tyrB gene expression by growing the cells in the presence of 0.4-1 mM Tyr in growth medium . Note that this strategy can only be applicable for a short-term cultivation but not suitable for a long-term cultivation for heterologous expression of foreign genes.

The complicated patterns of amino acid requirements for bacterial growth in these strains certainly reflect the significant overlap in cognate biosynthetic and biodegradation pathways (see Fig. 1), and our knowledge about the complicated regulation of the metabolic flow is still incomplete. It is therefore important to optimize the E. coli expression conditions for each target protein of interest before running selective amino acid isotope labeling experiments.

Practically speaking, as long as there is a high concentration of amino acids in the growth medium, the collection of the new auxotrophic C43(DE3) and BL21(DE3) expression host strains described here (Tables 1, S1) can solve many selective labeling problems, and can be used for cost effective, high-yield production of any recombinant water-soluble or membrane protein that can be expressed in E. coli.

up

• All ML strains were engineered by Dr. Gennis research group (Myat T. Linn, Robert B. Gennis) at University of Illinois at Urbana-Champaign, U.S.A.
• YM, MS, EH, and RF strains were engineered by our research group (Yoshiharu Miyajima-Nakano, Risako Fukazawa, Emi Hagiuda, Shin-ichi Matsushita, Toshio Iwasaki) at Nippon Medical School, Japan, in collaboration with Dr. Gennis research group.

• This strain bank project was supported in part by the International Collaborations in Chemistry Grant from JSPS (T.I.) and NSF (CHE-1026541 to S.A.D.), the JSPS Grant-in-aid 24659202 (T.I.), the Nagese Science and Technology Foundation Research Grant (T.I.), the DE-FG02-87ER13716 (R.B.G.) and DE-FG02-08ER15960 (S.A.D.) Grants from US DOE, NIH & NIGMS Roadmap Initiative (R01GM075937), and NIH grant GM062954 (S.A.D.).

up

Go to the ICC Project Details Page