In vivo mutations of thymidylate synthase (encoded by thyA) are responsible for thymidine dependency in clinical small-colony variants of Staphylococcus aureus

Abstract

Trimethoprim-sulfamethoxazole (SXT)-resistant Staphylococcus aureus thymidine-dependent small-colony variants (TD-SCVs) are frequently isolated from the airways of cystic fibrosis (CF) patients, often in combination with isogenic normal strains if patients were treated with SXT for extended periods. As SXT inhibits the synthesis of tetrahydrofolic acid, which acts as a cofactor for thymidylate synthase (thyA), the survival of TD-SCVs depends exclusively on the availability of external thymidine. Since the underlying mechanism for thymidine dependency is unknown, we investigated if alterations in the thyA nucleotide sequences were responsible for this phenomenon. Sequence analysis of several clinical TD-SCVs and their isogenic normal strains with reference to previously published S. aureus thyA nucleotide sequences was performed. Three clinical TD-SCVs were complemented by transforming TD-SCVs with the vector pCX19 expressing ThyA from S. aureus 8325-4. Transcriptional analysis of metabolic and virulence genes and regulators (agr, hla, spa, citB, thyA, and nupC) was performed by quantitative reverse transcription-PCR. The previously published sequences of thyA and two normal clinical strains were highly conserved, while thyA of four normal strains and four SCVs had nonsynonymous point mutations. In 8/10 SCVs, deletions occurred, resulting in stop codons which were located in 4/10 SCVs close to or within the active site of the protein (dUMP binding). Complementation of TD-SCVs with thyA almost fully reversed the phenotype, growth characteristics, and transcription patterns. In conclusion, we demonstrated that mutations of the thyA gene were responsible for the phenotype of TD-SCVs. Complementation of TD-SCVs with thyA revealed that a functional ThyA protein is necessary and sufficient to change the SCV phenotype and behavior back to normal.

FIG. 1.

Clinical TD-SCV is phenotypically complemented by a functional thyA gene. (A) Normal S. aureus grown on Columbia blood agar. (B) TD-SCV grown on Columbia blood agar exhibiting the SCV phenotype with small, less pigmented colonies. (C) TD-SCV transformed with pCX19thyA expressing functional ThyA, leading to the normal phenotype with large, pigmented colonies. (D) TD-SCV transformed with pCX19empty with the SCV phenotype, indicating the importance of functional ThyA for reversion to the normal phenotype. The images are representative of three sets of TD-SCV, normal S. aureus, and SCV/pCX19thyA strains.

FIG. 2.

Analysis of growth of normal S. aureus, TD-SCV, SCV/pCX19thyA, SCV/pCXthyA induced with xylose, and SCV/pCX19empty. Bacteria were grown in BHI broth with or without 0.5% xylose on a rotary shaker at 35°C. Every hour, the OD₅₆₈ of the cultures were determined. The data are representative of two sets of TD-SCV, normal S. aureus, and SCV/pCX19thyA strains.

FIG. 3.

Real-time RT-PCR quantification of expression in normal, TD-SCV, and SCV/pCX19thyA with or without 0.5% xylose of RNAIII as the effector gene for agr (A), hla (C), and citB (D) in stationary phase (SP) and of spa in late log phase (LL) (B). The levels of mRNA expression of the different genes were normalized using the expression of an internal control, yqiL. The transcript quantities are expressed as increases (n-fold) relative to the values for the internal control. The data are means and standard errors of the means obtained in three independent experiments. Asterisk, P < 0.05 compared to the normal strain (t test).

FIG. 4.

Real-time RT-PCR quantification of expression in normal, TD-SCV, and SCV/pCX19thyA with or without 0.5% xylose of thyA (A) and nupC (B) in early log phase (EL), late log phase (LL), and stationary phase (SP). The levels of mRNA expression of the different genes were normalized using the expression of an internal control, yqiL, and the transcript quantities are expressed as increases (n-fold) relative to the values for the internal control. The data are means and standard errors of the means obtained in three independent experiments Asterisk, P < 0.05 compared to the normal strain (t test).