Characterization of Staphylococcus aureus Small-Colony Variants Isolated from Lower Respiratory Tract Specimens

Introduction

Small colony variants (SCVs) are a subpopulation of nutrient-deficient bacteria that exhibit a unique phenotype as an adaptive response to environmental stress. They are characterized by small colony size, slow growth, reduced metabolic activity, and intracellular parasitism.¹ The first SCV was reported in Salmonella typhi in 1910 and, since then, SCVs have been identified in various bacterial species, including Staphylococcus aureus, Vibrio cholerae, and Escherichia coli.² Among these, S. aureus is a major pathogen responsible for hospital-acquired pneumonia (HAP) and hemorrhagic necrotizing pneumonia, and it can lead to severe conditions such as sepsis and septic shock.³

The formation of SCVs is primarily attributed to specific nutritional auxotrophies affecting the respiratory chain, including deficiencies in thymidine, menadione, or heme.⁴ These metabolic impairments lead to obstructed oxidative phosphorylation in the electron transport chain, resulting in significantly reduced bacterial replication and metabolic rates that ultimately manifest as the SCV phenotype. Unlike wild-type S. aureus, small colony variants of S. aureus demonstrate enhanced resistance to antibiotics such as aminoglycosides and trimethoprim-sulfamethoxazole. Both methicillin-resistant S. aureus (MRSA) and methicillin-susceptible S. aureus (MSSA) can develop SCVs. Notably, MRSA-SCVs are particularly challenging to treat, warranting heightened clinical vigilance among physicians and clinical microbiologists.⁵ Inappropriate antimicrobial therapy may facilitate SCV development.⁶ Prolonged use of sulfonamides such as trimethoprim-sulfamethoxazole is associated with thymidine-auxotrophic SCVs, as this antibiotic combination inhibits the synthesis of tetrahydrofolate, an essential cofactor for thymidylate synthase.⁷ Similarly, aminoglycoside usage correlates with the emergence of menadione- or heme-auxotrophic SCVs, although the precise underlying mechanisms remain incompletely understood. Furthermore, SCVs exhibit attenuated virulence and an enhanced biofilm-forming capacity. By downregulating host chemokines, including CCL5 and CXCL8, SCVs impair the rapid recruitment of neutrophils and other immune cells to infection sites, thereby facilitating persistent host colonization.⁸ Consequently, SCVs are frequently implicated in chronic and recurrent infections, such as periprosthetic joint infections, pulmonary cystic fibrosis-related infections, and refractory wound infections.^8–10

The rapid and accurate identification of SCVs is crucial for effective clinical treatment. However, owing to their slow growth, SCVs are frequently overlooked by clinical microbiologists in routine practice. Although certain chromogenic media can facilitate convenient screening of S. aureus SCVs, most clinical laboratories in China and other countries do not routinely test for these variants.^11,12 Furthermore, clinically isolated S. aureus SCVs often exhibit phenotypic instability and tend to revert to the wild-type phenotype.

Owing to the phenotypic instability of SCVs, researchers often employ stable site-directed mutant strains (hemB or menD mutants) to simulate SCV phenotypes.¹³ Consequently, studies on clinically derived SCVs remain limited. Existing evidence has established a significant association between S. aureus SCV formation and P. aeruginosa coinfection.¹⁴ However, epidemiological data on S. aureus SCVs in China are notably scarce. To address this knowledge gap, our study aimed to investigate the prevalence, antimicrobial susceptibility patterns, and genetic characteristics of S. aureus SCVs isolated from lower respiratory tract specimens of patients with P. aeruginosa pneumonia.

Materials and Methods

Sample Collection and Identification of S. aureus SCVs

From October 2024 to September 2025, 1,832 lower respiratory tract specimens (including sputum and bronchoalveolar lavage fluid) were collected from patients with P. aeruginosa pneumonia at two major tertiary hospitals in Fuzhou, Fujian Province. All specimens suspected of harboring S. aureus SCVs were processed by isolating single colonies through quadrant streaking, followed by confirmatory identification using MALDI-TOF MS (Bruker Daltonik, Germany), according to the manufacturer’s protocol. Confirmed S. aureus isolates were cultured on blood agar plates for 24–48 hours. Colonies exhibiting characteristic SCV phenotypes, including approximately 10% reduction in size compared to normal morphology, grayish-white coloration, and diminished hemolytic activity, were identified as S. aureus SCVs.¹⁵ All verified S. aureus SCVs were subsequently lyophilized and preserved at −80°C for future research purposes.

DNA Extraction of S. aureus SCVs

DNA was extracted from S. aureus SCVs using previously described methods.¹⁶ Briefly, preserved SCVs were streaked onto blood agar plates and incubated overnight at 35°C. Subsequently, the bacterial colonies were suspended in 300 μL sterile distilled water, heated at 95°C for 10 min, and centrifuged at 12,000 × g for 5 min to remove cellular debris. The supernatant was stored at 4°C and used as the DNA template for amplification. DNA concentration and purity were measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA).

Molecular Typing of S. aureus SCVs

Multilocus sequence typing (MLST) analysis was performed as described previously.¹⁷ Seven S. aureus housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi, and yqiL) of S. aureus were amplified and sequenced. The sequencing data were compared with existing S. aureus sequences in the MLST database (https://pubmlst.org/) to determine the allelic profiles and sequence types (STs). Spa typing was performed based on polymorphisms in the X-region of the staphylococcal protein A (spa) gene.¹⁸ The X-region of each isolate was amplified by PCR, and the PCR products were subjected to Sanger sequencing. The obtained spa sequences were compared with the database (https://spa.ridom.de/) to determine the spa types of S. aureus SCVs.

Antimicrobial Susceptibility Testing of S. aureus SCVs

Antimicrobial susceptibility testing was performed using the Vitek^® 2 Compact system (bioMérieux, France). The results were interpreted according to Clinical and Laboratory Standards Institute (CLSI) guidelines.¹⁹ The study evaluated susceptibility to 13 antibiotics: penicillin, oxacillin, gentamicin, rifampin, ciprofloxacin, levofloxacin, moxifloxacin, clindamycin, erythromycin, linezolid, vancomycin, tetracycline, and tigecycline. Staphylococcus aureus ATCC 25923 was used as a quality control strain. Methicillin-resistant S. aureus (MRSA) strains were confirmed based on phenotypic resistance patterns and detection of the mecA gene.

PCR Detection of Antimicrobial Resistance and Virulence Genes

The detection of 15 virulence genes (sea, icaA, icaC, icaD, clfA, clfB, hla, hlg, pvl, fnbA, cna, sarA, sdrD, and sdrE, ebpS) and the resistance gene mecA in S. aureus SCVs was performed by PCR according to the methods described by a previous study.^20,21 The PCR amplification protocol consisted of: initial denaturation at 94°C for 5 min, followed by 30 cycles of denaturation at 95°C for 30 sec, annealing at 53°C for 30 sec, and extension at 72°C for 1 min, with a final extension at 72°C for 10 min. The PCR products were visualized using 2% agarose gel electrophoresis and subsequently sent to Sangon Biotech (Shanghai, China) for sequencing. The sequencing results were analyzed using SnapGene 7.1.2 software.

Detection of Biofilm Formation

Crystal violet staining measured biofilm formation.¹⁶ Briefly, fresh S. aureus SCVs were inoculated into LB broth, cultured at 37°C, and shaken (220 rpm) for 18 h. The bacterial suspension was adjusted to a McFarland standard of 0.5, and diluted at a ratio of 1:100 in fresh LB broth. Aliquots (200 μL) of the diluted culture were dispensed into eight wells of a 96-well microtiter plate and incubated statically at 37°C for 48 h. Sterile LB broth served as the negative control. After incubation, the plates were washed thrice with phosphate-buffered saline PBS (pH 7.0) and air-dried at room temperature. The biofilms were then fixed with methanol for 20 min, after which the methanol was discarded. The fixed biofilms were stained with 1% crystal violet solution for 15 min, followed by washing with PBS until the solution became colorless. After drying, the bound crystal violet was solubilized in 200μL of absolute ethanol and the solution was transferred to a new microtiter plate. Optical density (OD) was measured at 570 nm. Mean OD values were calculated for all tested strains and negative controls. The optical density cutoff (ODc) was defined as 3 standard deviations (SDs) above the mean optical density (OD) of the negative control. Classification: non-biofilm producers: OD ≤ ODc; weak-biofilm producers: ODc < OD ≤ 2ODc; moderate-biofilm producers: 2ODc < OD ≤ 4ODc; strong-biofilm producers: OD > 4ODc.

Clinical Data Collection

Clinical data of hospitalized and outpatient patients with P. aeruginosa pneumonia were collected from the Hospital Information System (HIS), including patient age, sex, department, treatment course, and underlying diseases. The data were organized using Microsoft Excel.

Statistical Analysis

All statistical analyses were performed using IBM SPSS Statistics 26.0. Categorical variables were analyzed using either the chi-square test or Fisher’s exact test, while continuous variables were evaluated using either the Student’s t-test or Mann–Whitney U-test. Statistical significance was set at a p-value of <0.05.

Results

Prevalence of S. aureus SCV Isolates

Thirty-eight S. aureus SCVs were isolated from 1,832 sputum and bronchoalveolar lavage fluid specimens collected from patients with P. aeruginosa pneumonia, yielding a detection rate of 2.1%. After 24 h of incubation on blood agar plates, S. aureus SCVs formed pinpoint-sized round white colonies. As shown in Figure 1, when cultured on blood agar plates at 37°C for 24 h, S. aureus SCVs demonstrated distinct morphological differences from normal S. aureus ATCC 29213.

Figure 1 Colony morphology of S. aureus strains ATCC 29213 and S. aureus SCV isolates. Notes: The clinical isolate of S. aureus SCVs was grown on the upper section of a blood agar plate, whereas the normal morphology S. aureus ATCC 29213 strain was cultured on the lower section.

Molecular Characteristics of S. aureus SCV Isolates

MLST typing of 38 S. aureus SCVs revealed 15 distinct sequence types (STs) (Figure 2). ST1 (15.8%, 6/38) and ST398 (15.8%, 6/38) were the predominant clones, followed by ST965 (13.2%, 5/38), ST72 (10.5%, 4/38), ST764 (10.5%, 4/38), and ST7 (7.9%, 3/38). The remaining nine ST types collectively accounted for 26.3% of isolates. Among the methicillin-resistant S. aureus SCVs (MRSA-SCVs), ST1 (25.0%, 5/20) and ST965 (25.0%, 5/20) were the most prevalent, followed by ST764 (20.0%, 4/20).

Figure 2 Distribution of sequence types (STs) of all isolates of S. aureus SCVs, MRSA-SCVs and MSSA-SCV isolates.

Spa typing revealed 23 distinct types, predominantly t128 (13.2%, 5/38). Notably, 25.0% (5/20) of MRSA-SCVs carried the t128 spa type (Table 1). The most prevalent S. aureus SCV in lower respiratory tract samples was ST1-t128 (13.2%, 5/38), followed by ST965-t062 (7.9%, 3/38), ST72-t3735 (7.9%, 3/38) and ST764-t1084 (7.9%, 3/38). Genotypic distribution analysis showed that ST1-t128 was the dominant MRSA-SCVs genotype, whereas ST72-t3735 was predominant among methicillin-susceptible S. aureus SCV (MSSA-SCVs) isolates. These findings suggest a potential association between specific molecular types of S. aureus SCV and methicillin resistance.

Table 1 Molecular Characteristics of S. aureus SCV Isolates

Antibiotic Susceptibility of S. aureus SCV Isolates

The VITEK^® 2 Compact results revealed that 86.8% of the isolated S. aureus SCVs were resistant to at least one antibiotic. However, all S. aureus SCVs remained susceptible to linezolid, vancomycin, and tigecycline. The isolated S. aureus SCVs showed high resistance rates to the β-lactam antibiotics, penicillin (81.6%) and oxacillin (52.6%). Lower resistance rates were observed for other antibiotics, including gentamicin (26.3%), tetracycline (34.2%), rifampin (2.6%), ciprofloxacin (36.8%), levofloxacin (36.8%), and moxifloxacin (36.8%). Methicillin resistance was observed in 52.6% (20/38) of S. aureus SCVs, with the remaining (18/38) being MSSA-SCVs. Compared to MSSA-SCV strains, MRSA-SCV strains demonstrated significantly higher resistance rates to clindamycin, erythromycin,ciprofloxacin, levofloxacin, moxifloxacin and oxacillin (P<0.05) (Table 2). Additionally, it should be noted that all ST764-t1084 MRSA-SCVs were resistant to penicillin, oxacillin, gentamicin, ciprofloxacin, levofloxacin, moxifloxacin, erythromycin, clindamycin, and tetracycline. ST764 S. aureus SCVs showed high resistance to multiple antibiotics, whereas ST72 and ST9 remained susceptible to all tested antibiotics (Figure 3).

Table 2 Susceptibility of S. aureus SCVs to Antimicrobials

Figure 3 Heatmap of antimicrobial resistance rates and virulence gene carrier rates in different ST S. aureus SCV isolates. Abbreviations: PEN, penicillin; OXA, oxacillin; GEN, gentamicin; RIF, rifampin; CIP, ciprofloxacin; LEV, levofloxacin; MFX, moxifloxacin; CLI, clindamycin; ERY, erythromycin; LNZ, linezolid; VAN, vancomycin; TCY, tetracycline; TGC, tigecycline.

Detection of Virulence and Resistance Genes

The distribution of virulence genes among the isolated S. aureus SCVs is summarized in Table 3 and Figure 3, demonstrating that most S. aureus SCVs co-harbored multiple virulence determinants. The analysis revealed universal positivity (100%) for icaD, hla, hlg, and fnbA, whereas other virulence genes, including icaA, icaC, clfB, cna and pvl, showed positivity rates exceeding 70%. In contrast, the enterotoxin gene sea and immune evasion-associated gene sdrE exhibited lower positivity rates of 36.8% and 52.6%, respectively. Notably, MRSA-SCVs displayed significantly higher positivity rates for cna, sea, and pvl than their MSSA-SCV counterparts (P<0.05), with no other statistically significant differences observed. Furthermore, strain typing analysis revealed distinct genotype-specific patterns: pvl was predominantly associated with the ST1, ST5, and ST764 lineages, whereas sarA and sdrD genes were primarily detected in the ST6, ST25, and ST672 strains.

Table 3 Carrying Rates of Virulence Genes in S. aureus SCV Isolates

Biofilm Formation Capacity of S. aureus SCV Isolates

All S. aureus SCVs obtained in this study exhibited biofilm production, as observed by crystal violet staining. Among these, 25 isolates (65.8%) exhibited strong biofilm-forming capacity, 10 isolates (26.3%) showed moderate biofilm formation, and the remaining isolates (7.9%) displayed weak biofilm formation (Table 4). No significant difference in biofilm-forming capacity was observed between MRSA-SCVs and MSSA-SCVs.

Table 4 Biofilm Formation of S. aureus SCV Isolates

Clinical Characteristics

Clinical data from P. aeruginosa pneumonia were extracted using the HIS. S. aureus SCV carriers showed significant differences in epidemiological and clinical features compared with non-carriers (Table 5). Patients with P. aeruginosa pneumonia coinfected with S. aureus SCVs experienced significantly longer hospital stays and antimicrobial therapy duration than those without SCVs (P = 0.010 and 0.002, respectively). S. aureus SCV carriers also required fluoroquinolone or vancomycin treatment more frequently (P = 0.022). However, no significant differences were observed between the two groups in terms of sex (P = 0.156), age (P = 0.071), PCT level (P = 0.673), or prevalence of underlying conditions, such as diabetes (P = 0.082) and hypertension (P = 0.100).

Table 5 Epidemiological and Clinical Characteristics of Study Patients

Discussion

In this study, we isolated 38 S. aureus SCVs from 1,832 lower respiratory tract specimens obtained from patients with P. aeruginosa pneumonia, yielding a detection rate of 2.1%. Similarly, Min¹⁰ screened three S. aureus SCVs from 278 S. aureus isolates obtained from wound specimens, with a detection rate of 1.1%. Zheng²² identified three stable S. aureus SCVs from 41 rifampin-resistant S. aureus isolates, with a detection rate of 7.3%. The higher detection rate among rifampin-resistant S. aureus may be attributed to mutations in the rpoB gene, which reduce the transcription of tricarboxylic acid (TCA) cycle and oxidative phosphorylation-related genes, forcing the bacteria to rely on fermentation pathways and thereby adopting SCV characteristics. Cervantes-García²³ detected four S. aureus SCVs from 47 pus samples of diabetic foot ulcer patients infected with S. aureus, with a detection rate of 8.5%. Further investigation revealed that all four patients carrying S. aureus SCVs had received multiple antibiotic treatments, including trimethoprim-sulfamethoxazole, for one month or longer. We propose that the variation in detection rates of S. aureus SCVs across different studies may be linked to its survival environment. When S. aureus is exposed to adverse conditions, such as prolonged antibiotic exposure, the small-colony variant phenotype may provide a survival advantage for it. Additionally, we investigated the carriage rate of S. aureus SCVs in patients with P. aeruginosa pneumonia in Southeast China, which distinguishes our study from previous research in terms of population and geographic location. The carriage rate of S. aureus SCVs may vary across different populations and regions.

Among the S. aureus SCVs isolated, ST1-t128 (13.2%, 5/38) were the predominant clones. The ST1 lineage, originally prevalent in North America (the United States and Canada) before spreading to Europe and other regions, had not been previously reported as circulating in China.^24,25 Fujian Province’s status as a southeastern coastal region with extensive exchanges with North America may explain ST1 emergence in this area. Notably, ST1 represents one of the major community-associated MRSA (CA-MRSA) clones.²⁵ Unlike multidrug-resistant hospital-associated MRSA (HA-MRSA) strains such as ST5 and ST239, ST1 typically exhibits a narrower resistance profile. Our study confirmed this pattern, with ST1 S. aureus SCVs showing resistance only to penicillin and oxacillin, while remaining susceptible to the other tested antibiotics.

Globally, the ST5 lineage is one of the most widely disseminated HA-MRSA lineages. ST764 is a single-locus variant of the ST5 HA-MRSA lineage, exhibiting characteristics of community-associated MRSA, and was first identified in Japan in 2006.²⁶ The ST764 clone has become increasingly prevalent in China, Japan, and other Asian regions, which may be associated with its strong biofilm formation and cell adhesion capabilities.²⁷ Our study found that ST764-t1084 is resistant to all tested antibiotics, including penicillin, oxacillin, gentamicin, ciprofloxacin, levofloxacin, moxifloxacin, erythromycin, clindamycin, and tetracycline. These results highlight the necessity for increased surveillance of ST764-t1084 S. aureus SCVs in the clinical setting.

PVL, encoded by the lukF-PV and lukS-PV genes, forms pores in target cell membranes leading to membrane damage and subsequent leukocyte necrosis.²⁸ The ST1 strain of S. aureus frequently carries PVL and represents a highly virulent form of CA-MRSA. PVL-mediated destruction of white blood cells and cytokine storms by ST1 S. aureus are key factors in necrotizing pneumonia.²⁹ In our study, 71.1% of S. aureus SCVs carried PVL genes, with 100% carriage among ST1 isolates. Previous research has shown that PVL-positive MRSA strains can evade host immunity and induce keratinocyte apoptosis, thereby promoting local inflammatory spread.³⁰ However, some studies found no correlation between PVL production levels and clinical infection severity, indicating the need for further investigation into PVL’s role in disease progression.³¹ The intercellular adhesion gene cluster (icaADBC) collectively encodes enzymes for the polysaccharide intercellular adhesin (PIA) synthesis essential for biofilm formation, while icaR encodes a repressor protein regulating icaADBC expression.³² In this study, the positivity rates of icaA, icaC, and icaD genes in S. aureus SCVs all reached 90.0% or higher. However, in normal-phenotype S. aureus infections, the positivity rates of icaA and icaD genes were only 77.6%.³³ This suggests that, compared to normal-phenotype S. aureus, the high expression of the intercellular adhesion gene cluster icaADBC may contribute to the strong biofilm-forming capacity of S. aureus SCVs. The cna gene encodes a collagen-binding protein that mediates bacterial adhesion. Among our SCVs, the cna positivity rate was 71.1%, with a significantly higher prevalence in MRSA-SCVs than in MSSA-SCVs (P = 0.011). The hemolysins encoded by hla and hlg genes can penetrate eukaryotic membranes and lyse target cells. In our SCVs, both hla and hlg genes showed 100.0% positivity rates, that is, markedly higher than the 83.5% and 69.5% rates observed in normal-phenotype S. aureus.²⁰ Additionally, the immune evasion-related gene sdrD demonstrated a positivity rate of 73.7%, exceeding the 63.5% rate reported for normal-phenotype strains in previous studies.²⁰ The elevated expression of these virulence factors in S. aureus SCVs likely plays a crucial role in their pathogenic processes, including adhesion, invasion, and host immune evasion.

Finally, S. aureus SCVs displayed enhanced biofilm-forming capabilities. Our data showed that 92.1% of S. aureus SCVs exhibited moderate to strong biofilm production, with 65.8% of isolates demonstrating particularly robust biofilm formation. This pronounced biofilm-forming capacity may contribute to chronic and persistent infections associated with SCVs. Through analysis of clinical data from the Hospital Information System, we observed that P. aeruginosa pneumonia patients co-infected with S. aureus SCVs experienced significantly prolonged hospitalization and required extended courses of antimicrobial therapy. These findings suggest that S. aureus SCVs may play a critical role in sustaining persistent pulmonary infections in P. aeruginosa pneumonia patients.

In China and other countries, routine screening through specialized methods such as chromogenic media for S. aureus SCVs may be beneficial but substantially increase patient treatment costs. Extended culture duration may represent a cost-effective alternative for detecting S. aureus SCVs. In future studies, we will employ whole-genome sequencing to elucidate the formation mechanisms of S. aureus SCVs in patients with P. aeruginosa pneumonia, which will contribute to improving the diagnosis and treatment of these patients, particularly those with refractory infections.

Conclusion

In Fujian, China, S. aureus SCVs were detected in 2.1% of lower respiratory tract specimens from P. aeruginosa pneumonia cases, representing missed coinfection in clinical practice. Among these S. aureus SCVs, the ST1-t128 type was predominant and most strains carried the PVL gene. The ST1 S. aureus SCVs showed susceptibility to all tested antibiotics except penicillin and oxacillin. All ST764-t1084 MRSA-SCVs exhibited pan-resistance to tetracycline, β-Lactam antibiotics (penicillin and oxacillin), fluoroquinolones (ciprofloxacin, levofloxacin, and moxifloxacin), and macrolide-lincosamides (erythromycin/clindamycin), highlighting the critical need for clinical vigilance. Notably, the strong biofilm-forming ability of S. aureus SCVs may be associated with prolonged treatment duration and extended hospitalization.

There were some limitations in this study. (1) Although we detected previously overlooked S. aureus SCVs in patients with P. aeruginosa pneumonia at two tertiary hospitals in Fujian Province, our findings are intended primarily as a reference for clinicians and microbiologists in other regions managing refractory P. aeruginosa pneumonia. (2) Future studies utilizing whole-genome sequencing and animal models to elucidate the mechanisms underlying the formation and persistent colonization of S. aureus SCVs in the lower respiratory tract of these patients will be crucial for alleviating disease burden.