Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2024 Feb 13;29(4):347–355. doi: 10.1111/gtc.13104

CdsA, a CDP‐diacylglycerol synthase involved in phospholipid and glycolipid MPIase biosynthesis, possesses multiple initiation codons

Runa Hikage 1, Yusei Sekiya 1, Katsuhiro Sawasato 1,2, Ken‐ichi Nishiyama 1,✉
PMCID: PMC11448367  PMID: 38351722

Abstract

CdsA is a CDP‐diacylglycerol synthase essential for phospholipid and glycolipid MPIase biosynthesis, and therefore for growth. The initiation codon of CdsA has been assigned as “TTG,” while methionine at the 37th codon was reported to be an initiation codon in the original report. Since a vector containing the open reading frame starting with “TTG” under a controllable promoter complemented the cdsA knockout, “TTG” could function as an initiation codon. However, no evidence supporting that this “TTG” is the sole initiation codon has been reported. We determined the initiation codon by examining the ability of mutants around the N‐terminal region to complement cdsA mutants. Even if the “TTG” was substituted with a stop codon, the clear complementation was observed. Moreover, the clones with multiple mutations of stop codons complemented the cdsA mutant up to the 37th codon, indicating that cdsA possesses multiple codons that can function as initiation codons. We constructed an experimental system in which the chromosomal expression of cdsA can be analyzed. By means of this system, we found that the cdsA mutant with substitution of “TTG” with a stop codon is fully functional. Thus, we concluded that CdsA contains multiple initiation codons.

Keywords: CDP‐diacylglycerol synthase, CdsA, initiation codon, MPIase, phospholipids

CdsA is a CDP‐diacylglycerol synthase involved in biosynthesis of phospholipids and glycolipid MPIase. The initiation codon of CdsA has been assigned as “TTG,” however, no experimental evidence has been reported. Mutagenesis analysis revealed that not only the “TTG” but also following several codons function as an initiation codon.

graphic file with name GTC-29-347-g001.jpg

1. INTRODUCTION

CdsA is a CDP‐DAG (diacylglycerol) synthase, which is an ubiquitous enzyme involved in phospholipid biosynthesis in all organisms (Raetz & Dowhan, 1990). CDP‐DAG is a precursor for the biosynthesis of phospholipids PS, PE, PG, and PI (Blunsom & Cockcroft, 2020). In Escherichia coli, all phospholipids are synthesized through CDP‐DAG (Icho et al., 1985; Sparrow & Raetz, 1985). In addition, CdsA is involved in the biosynthesis of a glycolipid MPIase, which catalyzes membrane protein insertion, and therefore is essential for cell growth (Sato et al., 2019; Sawasato, Sato, et al., 2019; Sekiya et al., 2021). Thus, CdsA and its homologs are essential for biogenesis of biomembranes.

The nucleotide sequence of the cdsA locus was determined first for E. coli (Icho et al., 1985). In this report, the initiation codon is assigned as the 37th codon in the current database. Searching of the DNA database on https://www.genome.jp/ revealed that the sequence around the initiation codon of E. coli cdsA (….TTG CTG AAG TAT CGC CTG ATA….) is highly conserved among Gram‐negative bacteria (Table S1), strongly suggesting that this “TTG” is the initiation codon for CdsA. On the other hand, in some bacteria, an upstream “TTG” (Enterobacter sp. FY‐07, Salmonella enterica subsp. Arizonae, etc.) (Table S1), downstream “GTG” (Shigella flexneri Shi06HN006, Raoultella terrigena, Klebsiella pneumoniae subsp. rhinoscleromatis SB3432, etc.), or “ATG” (Shigella boydii Sb227, K. pneumoniae PMK1, etc.) has been assigned as an initiation codon. The downstream “ATG” corresponds to the 37th “ATG” in E. coli cdsA, which was originally assigned as an initiation codon (Icho et al., 1985). Moreover, a subclone used to determine the cdsA sequence, which lacked the region containing the “TTG” codon, increased the CdsA activity (Icho et al., 1985), strongly suggesting that cdsA possesses an initiation codon(s) other than the “TTG.” Therefore, it is not clear whether or not the “TTG” codon is the sole initiation codon of CdsA.

In this study, we constructed mutants around the 5′ region of cdsA and assessed their ability to complement the cdsA knockout (Sawasato, Sato, et al., 2019) and the high pH‐sensitive mutant cdsA8 (Ganong & Raetz, 1982). We found that multiple codons, in addition to the “TTG” codon, can function as initiation codons. We also found that the truncated mutant of CdsA that lacked the first transmembrane region, the originally assigned ORF (open reading frame) starting from the 37th ATG (Icho et al., 1985), retained the CdsA function, suggesting that the function of CdsA is not entirely dependent on the integrity of its N‐terminus, and that the multiple initiation codons of CdsA may play an important role in maintaining its function.

2. RESULTS

2.1. Multiple codons of CdsA can function as initiation codons

Plasmid pUSI2 carries the tac promoter followed by the ideal SD (Shine‐Dalgarno) sequence (Shine & Dalgarno, 1975). As a positive control, the cdsA ORF starting with “TTG,” as assigned in the database as an initiation codon, was cloned under the control of the tac promoter and the SD sequence in pUSI2 (pTac‐CdsA(TTG)) (Sawasato, Sato, et al., 2019). To evaluate the ability to complement the cdsA knockout, we used two cdsA knockout strains, KS23 (Sawasato, Sato, et al., 2019) and YS23 (Sawasato, Sekiya, & Nishiyama, 2019), both of which confer the cdsA::cat allele and a complementing plasmid, pAra‐CdsA. To critically examine the effects of mutants, we employed these cdsA knockouts, which also confer the ΔynbB mutation, since YnbB is a CdsA paralogue dedicated to MPIase biosynthesis (Sato et al., 2019). Both strains require arabinose for growth to induce CdsA from pAra‐CdsA, since cdsA is essential. The experimental system is illustrated in Figure 1a. When pUSI2‐based plasmids were introduced into KS23, the copy number was high, while in the case in YS23, it was very low or even one due to the pcnB mutation (Liu & Parkinson, 1989; Lopilato et al., 1986). When plasmid pTac‐CdsA(TTG) was introduced into KS23 and YS23, the plasmid complemented both cdsA knockout strains even in the absence of inducer IPTG through leaky expression (Table 1), indicating that “TTG” of codon 1 can function as an initiation codon. When plasmid pTac‐CdsA(ATG), in which codon 1 had been changed to “ATG,” was used, both KS23 and YS23 grew well, however, KS23 could not grow in the presence of IPTG (Table 1), indicating that efficient expression from codon 1 caused CdsA overproduction to a lethal level. To determine whether or not this “TTG” codon is the sole initiation codon, it was changed to leucine codon “TTA” (pTac‐CdsA‐1‐TTA) and termination codon “TAA” (pTac‐CdsA‐1‐TAA) (Table 1), both of which cannot function as an initiation codon, followed by the assessment of complementation (Table 1). Surprisingly, both mutants clearly complemented the cdsA knockouts in the presence of IPTG, indicating that a codon other than “TTG” can function as an initiation codon.

FIGURE 1.

FIGURE 1

Open in a new tab

The schematic representation of the strain used in this study, DNA sequence around the 5′ region of cdsA, and phospholipid profile and MPIase expression in YS23 harboring plasmid pCD1 and its derivatives. (a) The strain used to evaluate CdsA expression is illustrated. Upon the addition of arabinose, wild‐type CdsA is expressed from pAra‐CdsA. While a mutant is expressed from pTac‐CdsA xx upon the addition of IPTG, cdsA is expressed from the original promoter in pCD1 xx. (b) The putative initiation codons (“ATN” and “NTG”), which were mutated to a termination codon, “TAA,” are underlined. The positions for the insertion mutations are also indicated by arrows. Translation into amino acids was shown under the DNA sequence. The transmembrane (TM) regions are underlined. (c) YS23 cells harboring the specified plasmid, grown in LB medium supplemented with arabinose (0.2%) at 37°C, were washed with fresh LB medium three times, followed by inoculation to LB not supplemented with arabinose at the dilution of 1:100. Cultures were shaken at 37°C to the mid log‐phase or until cell growth ceased. Phospholipids prepared from specified strains were analyzed by TLC. The positions of PE, PG/CL, and PA are indicated. (d) YS23 cells harboring the specified plasmid were cultivated as described above. A whole cell extract was then prepared by treating the cell culture (500 mL) with 5% trichloroacetic acid. The precipitate was recovered by centrifugation (10,000 × g, 5 min, 4°C), followed by washing with acetone. The MPIase level was analyzed by SDS‐PAGE/immunoblotting. Extract equivalent to 0.5 mg and 5 mg proteins was analyzed to detect MPIase (upper panel) and SecB (lower panel), respectively. SecB was used as a loading control. The positions of MPIase and SecB are indicated. The uncropped images are presented in Figure S2.

TABLE 1.

Growth of KS23, YS23, and GN80 harboring the plasmid pTac series.

Plasmid Complementation a
KS23 YS23 GN80
−IPTG +IPTG −IPTG −IPTG −IPTG +IPTG
pUSI2 − − − − − −
pTac‐CdsA(TTG) +++ +++ ++ +++ +++ ++
pTac‐CdsA(ATG) ++ − +++ ++ +++ −
pTac‐CdsA‐1‐TTA ++ +++ − +++ +++ +++
pTac‐CdsA‐1‐TAA − +++ − +++ +++ +++
pTac‐CdsA‐1,2‐TAA + ++ − + +++ +++
pTac‐CdsA‐1,2,6‐TAA − ++ − + +++ +++
pTac‐CdsA‐1,2,6,7‐TAA − − − − +++ +++
pTac‐CdsA‐1,2,6,7,11‐TAA − − − − +++ +++
pTac‐CdsA‐1,2,6,7,11,13‐TAA − − − − +++ +++
pTac‐CdsA‐1,2,6,7,11,13,17‐TAA − − − − +++ +++
pTac‐CdsA‐1,2,6,7,11,13,17,20‐TAA − − − − +++ +++
pTac‐CdsA‐1,2,6,7,11,13,17,20,22‐TAA − − − − +++ +++
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23‐TAA − − − − − −
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26‐TAA − − − − − −
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26,30‐TAA − − − − − −
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26,30,33‐TAA − − − − − −
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26,30,33,34‐TAA − − − − − −
pTac‐CdsA Δ1‐5 − +++ − − +++ +++
pTac‐CdsA Δ1‐5ATG + − − +++ +++ −
pTac‐CdsA Δ1‐6 − + − − +++ +++
pTac‐CdsA Δ1‐7 − − − − +++ +++
pTac‐CdsA Δ1‐36 − − − − +++ −
pTac‐CdsA‐A‐ins2CTG − ++ − ++ +++ +++
pTac‐CdsA‐C‐ins6CTG − − − − +++ +++
pTac‐CdsA‐G‐ins7ATA − − − − +++ +++
pTac‐CdsA‐A‐ins8TCT − − − − +++ +++
Open in a new tab
a

Cell growth was assessed as described in Materials and Methods. GN80 cells were grown in LB at pH 6.0, and then streaked onto LB plates of which the pH had been adjusted at 8.5. Images of growth on LB agar plates are shown in Figure S1.

To determine the putative initiation codons, we constructed two types of mutants. One involves substitution at the possible positions of initiation codons of “NTG” and “ATN” (Hecht et al., 2017; O'donnell & Janssen, 2001; Reddy et al., 1985) with the termination codon “TAA” up to the 37th “ATG,” and the other involves deletion from the “TTG” (Figure 1b). When plasmid pTac‐CdsA‐1,2‐TAA was tested, clear complementation of both KS23 and YS23 was observed in the presence of IPTG. Plasmid pTac‐CdsA‐1,2,6‐TAA also complemented both cdsA knockouts in the presence of IPTG, albeit at lower efficiency. On the other hand, plasmid pTac‐CdsA‐1,2,6,7‐TAA and those with further introduction of “TAA” lost the ability to complement the knockouts, indicating that codons 1 (TTG), 2 (CTG), 6 (CTG), and 7 (ATA) can function as initiation codons (Table 1). Consistent with these results, the deletion mutants pTac‐CdsA Δ1‐5 and pTac‐CdsA Δ1‐6 complemented the growth of KS23, but pTac‐CdsA Δ1‐7 did not. In the case of YS23, in which the plasmid copy number becomes very low due to the pcnB mutation (Liu & Parkinson, 1989; Lopilato et al., 1986), these plasmids could not support the growth of it. On the other hand, pTac‐CdsA Δ1‐5ATG, in which codon 6 (CTG) had been substituted with ATG, supported the growth of both strains (Table 1), indicating that the starting efficiency of codons 6 and 7 is so low that a high level of transcription is necessary for complementation. Consistent with these results, insertion of a base before codon 2 (CTG) (Figure 1b) did not affect the complementation ability (pTac‐CdsA‐A ins2CTG), but insertion after codon 2 hampered the ability (pTac‐CdsA‐C ins6CTG, pTac‐CdsA‐G‐ins7ATA and pTac‐CdsA‐A ins8TCT; Figure 1b) (Table 1).

The above‐mentioned plasmids were also tested for the ability to complement the growth of GN80 (cdsA8), a high pH‐sensitive mutant (Ganong & Raetz, 1982). Essentially all the mutants that complemented the cdsA knockout could support the growth of GN80 at pH 8.5. In addition, the TAA mutants up to codon 22 complemented the growth of GN80 at pH 8.5, but those after codon 23 did not, indicating that those up to codon 23 (TTG) can function as an initiation codon (Table 1). In the case of the deletion mutant, pTac‐CdsA Δ1‐7 could complement the growth of GN80 at pH 8.5, consistent with the above results. Plasmid pTac‐CdsA Δ1‐36, which encodes the originally assigned ORF (Icho et al., 1985), complemented the growth of GN80 at pH 8.5, while the “TAA” mutants after codon 23 (TTG) did not. Note that induction of CdsA Δ1‐36, which starts with the “ATG” codon, was lethal (Table 1) presumably due to overproduction, as seen for pTac‐CdsA(ATG). Since pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26,30,33,34‐TAA can encode CdsA Δ1‐36, we further examined whether or not CdsA Δ1‐36 is functional when combined with CdsA8. When plasmid pTet‐CdsA8 was introduced in YS23, the cells exhibited the high pH‐sensitive growth in the absence of arabinose. On co‐introduction of pTac‐CdsA Δ1‐36, the growth at pH 8.5 recovered even in the absence of IPTG, indicating that CdsA Δ1–36 is functional. On the other hand, the “TAA” mutants could not support the growth at pH 8.5 (Table 2). Therefore, it was found that CdsA Δ1–36 was partially functional, however, the SD sequence at an appropriate location was necessary for sufficient expression.

TABLE 2.

Growth of YS23 harboring both a plasmid with mutations in cdsA and pTet‐CdsA8.

Complementation a
YS23
Plasmid pTet‐CdsA8 −IPTG +IPTG
pUSI2 + − −
pTac‐CdsA Δ1‐36 + ++ +
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26,30,33‐TAA + − −
pTac‐CdsA‐1,2,6,7,11,13,17,20,22,23,26,30,33,34‐TAA + − −
pCD1 − +++ nd
pCD1‐CdsA‐1‐TAA − +++ nd
pCD1‐CdsA‐1,2‐TAA − − nd
pCD1‐CdsA‐1,2,6‐TAA − − nd
pCD1 + +++ nd
pCD1‐CdsA‐1‐TAA + +++ nd
pCD1‐CdsA‐1,2‐TAA + − nd
pCD1‐CdsA‐1,2,6‐TAA + − nd
Open in a new tab
a

When YS23 cells harbored pTet‐CdsA8, they were grown in LB at pH 6.0, and then streaked onto LB plates of which the pH had been adjusted at 8.5. “nd,” not determined. Images of growth on LB agar plates are shown in Figure S1.

Taking these results together, it is concluded that codons 1 (TTG) and 2 (CTG) have the full ability to complement the cdsA knockout, codons 6 (CTG) and 7 (ATA) weak ability, and codons up to 23 (TTG) and 37 (ATG) partial activity.

2.2. Determination of the CdsA initiation codons at the chromosomal level

The experiments described above involved an artificial system using a strong promoter and an ideal SD sequence, with which the obtained results might not reflect the in vivo conditions. Therefore, we developed an experimental system to determine the initiation codon of cdsA at the chromosomal level. Plasmid pCD1 carries the cdsA gene flanked by 2.5 kbp upstream and 2.2 kbp downstream regions (Sawasato, Sato, et al., 2019). When this plasmid was introduced into YS23 (ΔcdsA ΔynbB ΔpcnB), the cells grew as well as the parent strain EK413 even in the absence of arabinose (Table 2), indicating that analysis of the regulation of cdsA expression at the chromosomal level could be possible since the copy number of pCD1 in YS23 is very low or even one due to the pcnB mutation. When pCD1‐CdsA‐1TAA, in which codon 1 (“TTG”) was changed to “TAA,” was introduced into YS23, the cells grew as well as YS23/pCD1. On the other hand, plasmids pCD1‐CdsA‐1,2TAA and pCD1‐CdsA‐1,2,6TAA, in which codons 1 and 2, and 1, 2, and 6 were substituted with “TAA,” respectively, were unable to complement the cdsA knockout in YS23 (Table 2). These results indicate that “CTG” at codon 2 alone functions as an initiation codon at the chromosomal level. Other codons, which are possible as initiation codons, require high expression from a strong promoter and ideal SD sequence. Essentially the same results were obtained as to the complementation of the high pH‐sensitive growth due to CdsA8 (Table 2), supporting the conclusion that “CTG” at codon 2 alone functions as an initiation codon at the chromosomal level.

Finally, YS23 cells harboring the pCD1‐derived plasmids were examined as to PA (phosphatidic acid) accumulation and MPIase expression (Figure 1c,d). YS23 cells harboring pCD1 and pCD1‐CdsA‐1TAA did not accumulate PA (Figure 1c) and expressed MPIase (Figure 1d) like wild‐type cells (EK413). On the other hand, YS23 cells harboring pCD1‐CdsA‐1,2TAA and pCD1‐CdsA‐1,2,6TAA accumulated a significant amount of PA (Figure 1c) and expressed almost no MPIase (Figure 1d). These results coincided with the growth complementation.

3. DISCUSSION

In this study, we determined the initiation codon for CdsA. Mutagenesis analyses revealed that codon 2 (CTG) functions as an initiation codon as effectively as codon 1 (TTG). In addition to these two codons, codons 6 (CTG) and 7 (ATA) were found to also function as initiation codons albeit at low efficiency. Moreover, the codons up to 23 (TTG) and 37 (ATG) have the potential to express the partially functional CdsA, since they could complement the cdsA8 mutation. Thus, it was found that cdsA possesses multiple codons that can function as initiation codons. Since cdsA is essential for cell growth, backup codons for CdsA initiation would be present in cdsA so that CdsA should be expressed even if the N‐terminal region has been mutated or deleted. Other than codon 37 (ATG), all the putative initiation codons are not canonical ATG ones, suggesting that the expression level of CdsA is kept appropriately low. As seen for pTac‐CdsA(ATG), overproduction of CdsA is rather toxic for cells. Such regulation of CdsA expression seems to be conserved among Gram‐negative bacteria, since the sequence at the 5′ region of cdsA is highly conserved (Table S1). Nonetheless, CdsA is upregulated at low temperature through an increase in the level of mRNA (Sawasato, Sekiya, & Nishiyama, 2019; Sawasato, Suzuki, & Nishiyama, 2019).

Originally, the initiation codon was suggested to be codon 37 (ATG) (Icho et al., 1985), but it was not the true initiation codon, since CdsA Δ1‐36 could not complement the cdsA knockouts. However, the ORF starting with codon 37 expresses a partially functional CdsA, as it complemented the CdsA8 mutation. The mutation point of CdsA8 is mapped at the C‐terminal domain, Y207H (Sawasato, Sato, et al., 2019). Since CdsA forms a homodimer (Liu et al., 2014), the heterodimer of CdsA8 and CdsA Δ1‐36 would give a functional dimer. Considering that CdsA8 retains the ability to biosynthesize MPIase (Sawasato, Sato, et al., 2019), it is plausible that the N‐terminal region of CdsA is involved in MPIase biosynthesis. The CdsA derivatives that lack the N‐terminal region should express the functional CdsA, if they had been inserted into membranes properly with the N‐terminal region responsible for MPIase biosynthesis, and they possessed the conserved C‐terminal domain (Sato et al., 2019; Sawasato, Sato, et al., 2019; Sekiya et al., 2021). While the fully functional mutant of 1‐TAA lacks only one amino acid, the mutants 1,2‐TAA and 1,2,6‐TAA lack the positive charged amino acids in the N‐terminus (Figure 1b). Since positive charges at the N‐terminal region are involved in topology determination (Luirink et al., 2005; Virkki et al., 2014), these mutants might have trouble in the process of membrane insertion, explaining the lower efficiency in complementation.

YnbB is a CdsA paralogue (Sato et al., 2019; Sawasato, Sato, et al., 2019). While the N‐terminal half is not homologous at all between CdsA and YnbB, the C‐terminal half is highly homologous (Sato et al., 2019, Sawasato, Sato, et al., 2019). In spite of no homology in the N‐termini, codons 2 (CTG) and 6 (CTG) are the same in CdsA and YnbB, suggesting that a similar backup system for expression is shared by the two proteins. This type of the backup system might be utilized for other genes whose expression is essential but should be kept at an appropriately low level. In the case of fabZ, ftsK, and yihA, some non‐canonical initiation codons occur after the initiation codon “TTG.”

4. EXPERIMENTAL PROCEDURES

4.1. Bacterial strains and plasmids

E. coli strains EK413 (F− Δ[argF‐lac]U168 rpsL150 relA1 thi deoC7 ptsF25 flbB5301) (Nishiyama et al., 1996) and its derivatives with cdsA knockout KS23 (EK413 ΔcdsA::cat ΔynbB) (Sawasato, Sato, et al., 2019) or YS23 (EK413 ΔcdsA::cat ΔynbB ΔpcnB::Tn10) (Sawasato, Sekiya, & Nishiyama, 2019) were used. KS23 and YS23 harbor plasmid pAra‐CdsA, from which CdsA is expressed upon arabinose addition (Sawasato, Sato, et al., 2019). E. coli GN80, which confers a cdsA8 mutation (Ganong & Raetz, 1982), was also used. E. coli ME9783 (Nozaki & Niki, 2019), a strain for in vivo E. coli cloning (iVEC), was obtained from NBRP (NIG, Japan): E. coli. Plasmid pCD1 carries the cdsA gene flanked by 2.5 kbp upstream and 2.2 kbp downstream regions (Sawasato, Sato, et al., 2019). While plasmid pTac‐CdsA(TTG) (referred to as “pTac‐CdsA” (Sawasato, Sato, et al., 2019)) carries the cdsA gene starting from codon 1 (TTG) under the control of tac promoter/lac operator, followed by the ideal SD sequence in plasmid pUSI2 (Shibui et al., 1988), and plasmid pTac‐CdsA(ATG) (Sawasato, Sato, et al., 2019) carries the cdsA gene, in which codon 1 was changed to ATG in pTac‐CdsA(TTG) (Sawasato, Sato, et al., 2019). Plasmid pTet‐CdsA8 (Sawasato, Sato, et al., 2019) carries the cdsA8 gene, which is constitutively expressed from the tet promoter in plasmid pACYC‐Km (Chang & Cohen, 1978; Sawasato, Sato, et al., 2019).

4.2. Materials

DNA oligonucleotides used in this study are listed in Table S2. IPTG (isopropyl‐β‐D‐thiogalactopyranoside) was purchased from Calbiochem. L‐(+)‐Arabinose, anisaldehyde and E. coli polar phospholipids were obtained from Sigma. Anti‐MPIase (Nishiyama et al., 2012) and anti‐SecB (Shimizu et al., 1997) antisera were raised in rabbits.

4.3. Plasmid construction

To construct pTac‐CdsA Δ1‐36, the PCR product amplified with the specified primers was digested with SalI and BglII, followed by ligation with pUSI2 digested with the same enzymes. For plasmids pTac‐CdsA‐1‐TTA, pTac‐CdsA‐1‐TAA, and pTac‐CdsA‐1,2‐TAA, the PCR products amplified with the respective forward (Fw) primers and primer pTac‐CdsA‐1 (Rv) were digested with BamHI and SalI, followed by ligation with pUSI2 digested with the same enzymes.

Plasmids pTac‐CdsA and its derivatives, or pCD1 were used as template for PCR, to introduce mutations around the 5′ region of cdsA. Forward (Fw) and reverse (Rv) primers were designed such that mutation points were franked by upstream (15–30 bp) and downstream regions (15–30 bp), giving the overlapping sequence of ~20 bp. The PCR products were then introduced into ME9783, allowing the recombination through the overlapping sequence.

The introduction of mutations was confirmed by DNA sequencing.

4.4. Assessment of complementation

The cdsA knockout strains KS23 and YS23 harboring the specified plasmids were streaked onto LB agar plates supplemented with appropriate antibiotics or inducers (0.2% arabinose and 1 mM IPTG), followed by incubation at 37°C for ~14 h. For the strains with CdsA8, cells, grown in LB supplemented with 50 mM 2‐(N‐morpholino)ethanesulfonic acid‐KOH (pH 6.0), were streaked onto LB agar plates of which the pH had been adjusted at pH 8.5, as described (Sato et al., 2019; Sawasato, Sato, et al., 2019). The plates were incubated at 37°C for ~14 h. When the single colonies were as big as those of EK413, the growth was assessed as “+++.” When they were smaller, but clearly formed, it was assessed as “++.” When they were much smaller like pin‐hole colonies, it was assessed as “+.” When no isolated single colonies were observed, it was assessed as “−.”

4.5. Other methods

Phospholipids were extracted as described (Bligh & Dyer, 1959), and then analyzed by TLC using a solvent system (chloroform/methanol/water: 9/5/1) for development, followed by visualization using anisaldehyde‐H2SO4, as described (Nishiyama et al., 2010). SDS‐PAGE/immunoblotting were performed as described (Sasaki et al., 2019).

AUTHOR CONTRIBUTIONS

All authors have participated in the design of the study. Runa Hikage and Ken‐ichi Nishiyama wrote the manuscript. Runa Hikage, Yusei Sekiya, and Katsuhiro Sawasato performed the experiments. All the authors reviewed and edited the manuscript.

FUNDING INFORMATION

This work was supported by KAKENHI grants to KN (Nos. 22H02567, 22 K19262, 22H05392, and 23H04536).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

FIGURE S1. Growth of KS23, YS23, and GN80 harboring the specified plasmid. LB agar plates, on which the respective cells were streaked, were incubated at 37°C for ~14 h.

GTC-29-347-s001.pdf (3.9MB, pdf)

FIGURE S2. The uncropped images used in Figure 1c,d. The areas surrounded with dotted boxes are used in Figure 1c,d. The positions of MPIase, SecB, and molecular weight markers are indicated.

GTC-29-347-s003.pdf (1MB, pdf)

TABLE S1. Sequence alignment of the 5′ flanking region of cdsA in Gram‐negative bacteria.

GTC-29-347-s004.docx (45KB, docx)

TABLE S2. DNA oligonucleotides used to construct plasmids encoding the cdsA mutants.

GTC-29-347-s002.xlsx (12.8KB, xlsx)

ACKNOWLEDGMENTS

We thank the NBPR (E. coli) for the iVEC strain, and E. coli Genetic Resources at Yale (CGSC) for E. coli GN80.

Hikage, R. , Sekiya, Y. , Sawasato, K. , & Nishiyama, K.‐i. (2024). CdsA, a CDP‐diacylglycerol synthase involved in phospholipid and glycolipid MPIase biosynthesis, possesses multiple initiation codons. Genes to Cells, 29(4), 347–355. 10.1111/gtc.13104

Runa Hikage and Yusei Sekiya contributed equally to this study.

Communicated by: Hiroji Aiba

REFERENCES

  1. Bligh, E. G. , & Dyer, W. J. (1959). A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology, 37(8), 911–917. 10.1139/o59-099 [DOI] [PubMed] [Google Scholar]
  2. Blunsom, N. J. , & Cockcroft, S. (2020). CDP‐diacylglycerol synthases (CDS): Gateway to phosphatidylinositol and cardiolipin synthesis. Frontiers in Cell and Development Biology, 8, 63. 10.3389/fcell.2020.00063 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chang, A. C. , & Cohen, S. N. (1978). Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. Journal of Bacteriology, 134(3), 1141–1156. 10.1128/jb.134.3.1141-1156.1978 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Ganong, B. R. , & Raetz, C. R. (1982). Massive accumulation of phosphatidic acid in conditionally lethal CDP‐diglyceride synthetase mutants and cytidine auxotrophs of Escherichia coli . The Journal of Biological Chemistry, 257(1), 389–394. 10.1016/S0021-9258(19)68376-6 [DOI] [PubMed] [Google Scholar]
  5. Hecht, A. , Glasgow, J. , Jaschke, P. R. , Bawazer, L. A. , Munson, M. S. , Cochran, J. R. , Endy, D. , & Salit, M. (2017). Measurements of translation initiation from all 64 codons in E. coli . Nucleic Acids Research, 45(7), 3615–3626. 10.1093/nar/gkx070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Icho, T. , Sparrow, C. P. , & Raetz, C. R. (1985). Molecular cloning and sequencing of the gene for CDP‐diglyceride synthetase of Escherichia coli . The Journal of Biological Chemistry, 260(22), 12078–12083. 10.1016/S0021-9258(17)38988-3 [DOI] [PubMed] [Google Scholar]
  7. Liu, J. D. , & Parkinson, J. S. (1989). Genetics and sequence analysis of the pcnB locus, an Escherichia coli gene involved in plasmid copy number control. Journal of Bacteriology, 171(3), 1254–1261. 10.1128/jb.171.3.1254-1261.1989 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Liu, X. , Yin, Y. , Wu, J. , & Liu, Z. (2014). Structure and mechanism of an intramembrane liponucleotide synthetase central for phospholipid biosynthesis. Nature Communications, 5, 4244. 10.1038/ncomms5244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Lopilato, J. , Bortner, S. , & Beckwith, J. (1986). Mutations in a new chromosomal gene of Escherichia coli K‐12, pcnB, reduce plasmid copy number of pBR322 and its derivatives. Molecular & General Genetics, 205(2), 285–290. 10.1007/BF00430440 [DOI] [PubMed] [Google Scholar]
  10. Luirink, J. , von Heijne, G. , Houben, E. , & de Gier, J. W. (2005). Biogenesis of inner membrane proteins in Escherichia coli . Annual Review of Microbiology, 59(6), 329–355. 10.1016/j.bbabio.2011.12.006 [DOI] [PubMed] [Google Scholar]
  11. Nishiyama, K. , Maeda, M. , Abe, M. , Kanamori, T. , Shimamoto, K. , Kusumoto, S. , Ueda, T. , & Tokuda, H. (2010). A novel complete reconstitution system for membrane integration of the simplest membrane protein. Biochemical and Biophysical Research Communications, 394(3), 733–736. 10.1016/j.bbrc.2010.03.061 [DOI] [PubMed] [Google Scholar]
  12. Nishiyama, K. , Maeda, M. , Yanagisawa, K. , Nagase, R. , Komura, H. , Iwashita, T. , Yamagaki, T. , Kusumoto, S. , Tokuda, H. , & Shimamoto, K. (2012). MPIase is a glycolipozyme essential for membrane protein integration. Nature Communications, 3, 1260. 10.1038/ncomms2267 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Nishiyama, K. , Suzuki, T. , & Tokuda, H. (1996). Inversion of the membrane topology of SecG coupled with SecA‐dependent preprotein translocation. Cell, 85(1), 71–81. 10.1016/S0092-8674(00)81083-1 [DOI] [PubMed] [Google Scholar]
  14. Nozaki, S. , & Niki, H. (2019). Exonuclease III (XthA) enforces in vivo DNA cloning of Escherichia coli to create cohesive ends. Journal of Bacteriology, 201(5), e00660. 10.1128/JB.00660-18 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. O'Donnell, S. M. , & Janssen, G. R. (2001). The initiation codon affects ribosome binding and translational efficiency in Escherichia coli of cI mRNA with or without the 5′ untranslated leader. Journal of Bacteriology, 183(4), 1277–1283. 10.1128/JB.183.4.1277-1283.2001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Raetz, C. R. , & Dowhan, W. (1990). Biosynthesis and function of phospholipids in Escherichia coli . The Journal of Biological Chemistry, 265(3), 1235–1238. 10.1016/S0021-9258(19)40001-X [DOI] [PubMed] [Google Scholar]
  17. Reddy, P. , Peterkofsky, A. , & McKenney, K. (1985). Translational efficiency of the Escherichia coli adenylate cyclase gene: Mutating the UUG initiation codon to GUG or AUG results in increased gene expression. Proceedings of the National Academy of Sciences of the United States of America, 82(17), 5656–5660. 10.1073/pnas.82.17.5656 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Sasaki, M. , Nishikawa, H. , Suzuki, S. , Moser, M. , Huber, M. , Sawasato, K. , Matsubayashi, H. T. , Kumazaki, K. , Tsukazaki, T. , Kuruma, Y. , Nureki, O. , Ueda, T. , & Nishiyama, K. (2019). The bacterial protein YidC accelerates MPIase‐dependent integration of membrane proteins. The Journal of Biological Chemistry, 294(49), 18898–18908. 10.1074/jbc.RA119.011248 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Sato, R. , Sawasato, K. , & Nishiyama, K. (2019). YnbB is a CdsA paralogue dedicated to biosynthesis of glycolipid MPIase involved in membrane protein integration. Biochemical and Biophysical Research Communications, 510(4), 636–642. 10.1016/j.bbrc.2019.01.145 [DOI] [PubMed] [Google Scholar]
  20. Sawasato, K. , Sato, R. , Nishikawa, H. , Iimura, N. , Kamemoto, Y. , Fujikawa, K. , Yamaguchi, T. , Kuruma, Y. , Tamura, Y. , Endo, T. , Ueda, T. , Shimamoto, K. , & Nishiyama, K. (2019). CdsA is involved in biosynthesis of glycolipid MPIase essential for membrane protein integration in vivo . Scientific Reports, 9(1), 1372. 10.1038/s41598-018-37809-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Sawasato, K. , Sekiya, Y. , & Nishiyama, K. (2019). Two‐step induction of cdsA promoters leads to upregulation of the glycolipid MPIase at cold temperature. FEBS Letters, 593(14), 1711–1723. 10.1002/1873-3468.13460 [DOI] [PubMed] [Google Scholar]
  22. Sawasato, K. , Suzuki, S. , & Nishiyama, K. (2019). Increased expression of the bacterial glycolipid MPIase is required for efficient protein translocation across membranes in cold conditions. The Journal of Biological Chemistry, 294(21), 8403–8411. 10.1074/jbc.RA119.008457 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Sekiya, Y. , Sawasato, K. , & Nishiyama, K. (2021). Expression of Cds4/5 of Arabidopsis chloroplasts in E. coli reveals the membrane topology of the C‐terminal region of CDP‐diacylglycerol synthases. Genes to Cells, 26(9), 727–738. 10.1111/gtc.12880 [DOI] [PubMed] [Google Scholar]
  24. Shibui, T. , Uchida, M. , & Teranishi, Y. (1988). A new hybrid promoter and its expression vector in Escherichia coli . Agricultural and Biological Chemistry, 52(4), 983–988. 10.1271/bbb1961.52.983 [DOI] [Google Scholar]
  25. Shimizu, H. , Nishiyama, K. , & Tokuda, H. (1997). Expression of gpsA encoding biosynthetic sn‐glycerol 3‐phosphate dehydrogenase suppresses both the LB− phenotype of a secB null mutant and the cold‐sensitive phenotype of a secG null mutant. Molecular Microbiology, 26(5), 1013–1021. 10.1046/j.1365-2958.1997.6392003.x [DOI] [PubMed] [Google Scholar]
  26. Shine, J. , & Dalgarno, L. (1975). Determinant of cistron specificity in bacterial ribosomes. Nature, 254(5495), 34–38. 10.1038/254034a0 [DOI] [PubMed] [Google Scholar]
  27. Sparrow, C. P. , & Raetz, C. R. (1985). Purification and properties of the membrane‐bound CDP‐diglyceride synthetase from Escherichia coli . The Journal of Biological Chemistry, 260(22), 12084–12091. 10.1016/S0021-9258(17)38989-5 [DOI] [PubMed] [Google Scholar]
  28. Virkki, M. T. , Peters, C. , Nilsson, D. , Sorensen, T. , Cristobal, S. , Wallner, B. , & Elofsson, A. (2014). The positive inside rule is stronger when followed by a transmembrane helix. Journal of Molecular Biology, 426(16), 2982–2991. 10.1016/j.jmb.2014.06.002 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

FIGURE S1. Growth of KS23, YS23, and GN80 harboring the specified plasmid. LB agar plates, on which the respective cells were streaked, were incubated at 37°C for ~14 h.

GTC-29-347-s001.pdf (3.9MB, pdf)

FIGURE S2. The uncropped images used in Figure 1c,d. The areas surrounded with dotted boxes are used in Figure 1c,d. The positions of MPIase, SecB, and molecular weight markers are indicated.

GTC-29-347-s003.pdf (1MB, pdf)

TABLE S1. Sequence alignment of the 5′ flanking region of cdsA in Gram‐negative bacteria.

GTC-29-347-s004.docx (45KB, docx)

TABLE S2. DNA oligonucleotides used to construct plasmids encoding the cdsA mutants.

GTC-29-347-s002.xlsx (12.8KB, xlsx)

Articles from Genes to Cells are provided here courtesy of Wiley

RESOURCES