The two-phase study combined computational peptide screening with 72-hour stimulation of the donor’s cells. The donor was a 38-year-old man with a mid-deep dermal burn covering 35% of total body surface area. He was enrolled on day 14 after injury, and MRSA was documented through blood culture and antimicrobial-susceptibility testing. His PBMC preparation had 96.8% viability and a recorded yield of 2.1 × 10^7 cells per millilitre.

For the computational phase, investigators screened S. aureus and MRSA reference sequences and a nonredundant pangenome dataset containing 1,038 GenBank genomes. BLASTP criteria of at least 85% identity, a bit score of at least 150 and an E-value no greater than 1 × 10−5 were used to prioritize conserved targets. Linear B-cell epitope predictions, antigenicity, allergenicity, toxicity and three-dimensional surface-accessibility analyses then supported selection of four peptides: A1 and A2 from adhesin-related sequences, B from a biofilm-associated protein, and H from hyaluronidase. All four were predicted to be non-toxic and non-allergenic, and their selected regions were modelled as surface accessible.

The four peptides were synthesised at greater than 95% purity and applied individually and in eight combinations to PBMCs. Each condition used 1 × 10^6 cells per well and five technical replicate wells from the same donor preparation, with unstimulated cells serving as the negative control. Because every peptide component was supplied at 5 µg/mL, formulations containing more peptides also had a higher total peptide concentration, an important constraint when comparing combinations.

Flow cytometry was used to assess marker-defined immune populations after 72 hours, while ELISA measured IL-4, IFN-γ, total IgG, IL-2, IL-12 and TNF-α in culture supernatants. The reported immune-population measures included CD33+CD38+ myeloid cells, CD56+CD38+ cells, total CD3+ T cells, the CD3+CD16+ T-cell-associated subset, total CD20+ B cells and the CD19+CD20+CD38− B-cell phenotype. Total IgG was treated as a general humoral readout rather than evidence of peptide-specific antibodies.

A2-B-H generated several of the study’s strongest measured responses. The CD33+CD38+ myeloid phenotype reached 12.32%, a 73% increase from the 7.11% control value, and CD20+ B-cell frequency reached 42.36%, 42% above control. A2-B-H also produced the highest IL-4 concentration among the tested conditions at 1.38 pg/mL, compared with 1.01 pg/mL in the control, a 37% increase. Total CD3+ T-cell frequency remained comparable with control under A2-B-H, while A1-B-H produced the highest CD56+CD38+ phenotype at 13.20%, 21.4% above control.

Other combinations led specific endpoints rather than the overall ranking. A1-H produced the highest CD3+CD16+ subset, increasing it by 43%, and the highest TNF-α concentration at 4.11 pg/mL, 65% above the 2.49 pg/mL control. A2-B generated the highest IL-2 concentration at 1.01 pg/mL, twice the 0.50 pg/mL control. A2-H produced the highest total IgG concentration at 8.71 ng/mL, 31% above control, followed by A1 at 7.97 ng/mL, 20% above control. By contrast, IL-12 remained between 11.20 and 11.79 pg/mL across treatment groups, and IFN-γ ranged from 1.01 to 1.32 pg/mL; neither showed a statistically significant difference from control.

Adding all four peptides did not consistently improve the response profile. The A1-A2-B-H formulation scored 4.5 out of 20, below A2-B-H, and was associated with lower CD33+CD38+ and total IgG responses as well as an approximately 47% reduction in total CD3+ frequency relative to control. These observations show that the larger formulation did not dominate the three-peptide combination under the tested conditions, although the fixed 5 µg/mL dose for each component prevented a clean separation of peptide number from total dose.

The findings remain limited by the use of one biological donor and technical replication rather than independent participants. Statistical comparisons were conducted across 13 exploratory endpoints without an additional multiplicity adjustment across endpoints, and the integrated score used study-defined domain weights and significance multipliers for candidate ranking rather than as an efficacy measure. The results therefore identify A2-B-H as the leading candidate within this particular experiment for subsequent functional and translational validation, but they do not establish protection against MRSA, demonstrate vaccine performance, measure peptide-specific antibody function or show direct antimicrobial activity.

Subject of Research: Single-donor ex vivo evaluation of MRSA virulence-factor peptide combinations in burn-associated bacteraemia

Article Title: Exploratory immunoinformatic and single-donor ex vivo evaluation of methicillin-resistant Staphylococcus aureus virulence-factor peptide combinations in burn-associated bacteraemia: Implications for MRSA vaccine development

Article References: Arviansyah, Sumarno, Mustamsir, E., Ratridewi, I., Barlianto, W., & Santosaningsih, D. (2026). Exploratory immunoinformatic and single-donor ex vivo evaluation of methicillin-resistant Staphylococcus aureus virulence-factor peptide combinations in burn-associated bacteraemia: Implications for MRSA vaccine development. Pharmacia, 73, Article e196054. https://doi.org/10.3897/pharmacia.73.e196054

Image Credits: AI Generated

DOI: 10.3897/pharmacia.73.e196054

Keywords: MRSA, burn-associated bacteraemia, peptide combinations, ex vivo PBMC study, immunoinformatics, immune-response profiling, virulence factors, single-donor research, A2-B-H, Exploratory, immunoinformatic, single-donor

The research team, led by corresponding author Ying Xu of the Department of Clinical Laboratory, enrolled febrile patients admitted to the hospital between 2021 and 2023. From this population, 457 eligible patients were classified into three etiological groups—bacterial, viral, or fungal infection—based on final clinical diagnoses corroborated by microbiological results. Rather than hunting for the pathogen itself, the investigators turned the diagnostic lens inward, asking whether the patient’s own immune response could reveal the identity of the invader. The rationale rests on a well-established biological principle: different classes of pathogens trigger distinct patterns of cytokine signaling, the chemical messengers through which immune cells coordinate their defense.

The biomarker panel the team evaluated combined six cytokines—interleukin-2 receptor (IL2R), interleukin-6 (IL-6), interleukin-8 (IL-8), interleukin-10 (IL-10), tumor necrosis factor-alpha (TNF-α), and interleukin-1 beta (IL-1β)—with three routine inflammatory markers: procalcitonin (PCT), C-reactive protein (CRP), and the neutrophil-to-lymphocyte ratio (NLR). Each component carries distinct immunological meaning. IL-6 is an early-response cytokine that drives the acute-phase reaction and is often markedly elevated in bacterial infections. IL-10, an anti-inflammatory cytokine, rises when the immune system attempts to restrain excessive inflammation, and the balance between the two—captured by the IL-6/IL-10 ratio—reflects the net inflammatory tone of the host. IL-8 is a chemokine that recruits neutrophils to sites of infection, while PCT, a prohormone of calcitonin, is released by extra-thyroid tissues under bacterial endotoxin stimulation and has long been regarded as one of the more specific markers of bacterial insult. The NLR, a simple calculation from a standard complete blood count, indexes the relative dominance of innate, neutrophil-driven immunity over adaptive, lymphocyte-driven immunity.

To translate these measurements into clinically usable decision rules, the researchers performed receiver operating characteristic (ROC) curve analysis, a statistical technique that plots diagnostic sensitivity against the false-positive rate across all possible cut-off values. The area under the ROC curve (AUC) quantifies overall discriminative ability, with 0.5 indicating performance no better than chance and 1.0 representing perfect classification. From these analyses, the team derived optimal threshold values for each marker and then combined them into multivariable models tailored to specific diagnostic questions. The result is not a single test but a sequential pathway: first determine whether the infection is bacterial or viral, then, if bacterial, determine whether the pathogen is Gram-negative or Gram-positive, and finally assess whether the bacterial infection has progressed to sepsis.

The first step of this hierarchy addresses the most common and consequential dilemma in febrile patients—separating bacterial from viral disease. A model combining an IL-6/IL-10 ratio greater than 1.13, IL-8 above 28.70 picograms per milliliter, and an NLR above 2.96 achieved an AUC of 0.80 (95% confidence interval: 0.74–0.87) for this distinction. In practical terms, an AUC of 0.80 indicates good discriminative ability, meaning the model correctly ranks a randomly chosen bacterial case above a randomly chosen viral case about four times out of five. The logic behind the chosen variables is coherent: a high IL-6/IL-10 ratio signals unopposed pro-inflammatory activity typical of bacterial stimulation, elevated IL-8 reflects intense neutrophil recruitment, and a high NLR confirms the neutrophil-dominant blood picture that accompanies acute bacterial infection. Viral infections, by contrast, tend to provoke lymphocyte-dominated responses with comparatively restrained IL-6 and IL-8 production.

The second step of the pathway tackles a subtler question: once a bacterial infection is confirmed, is the culprit Gram-negative or Gram-positive? This distinction matters because the two bacterial classes differ in cell-wall structure, antibiotic susceptibility profiles, and clinical behavior. The study found that no single universal model sufficed; instead, site-specific models were required, reflecting the fact that the immune signature of a bacterial infection is shaped by both the pathogen’s Gram status and the anatomical location of the infection. Notably, in bloodstream infections, a model incorporating IL-6 above 44.91 picograms per milliliter and PCT above 2.30 nanograms per milliliter yielded an AUC of 0.82 (95% confidence interval: 0.70–0.93). The pairing is immunologically sensible—Gram-negative organisms release endotoxin, a potent inducer of both IL-6 and PCT, so elevated values of these two markers together point toward a Gram-negative etiology.

The third and final step addresses severity stratification: among patients with confirmed bacterial infection, which ones have developed sepsis, the life-threatening dysregulated response that demands immediate aggressive intervention? Here the combination of IL-8 above 72.45 picograms per milliliter and PCT above 1.60 nanograms per milliliter achieved an AUC of 0.80 (95% confidence interval: 0.72–0.89). Very high IL-8 concentrations likely reflect the massive neutrophil activation and endothelial inflammation that characterize septic physiology, while elevated PCT tracks the systemic bacterial burden. Together, the three steps of the pathway achieved AUC values of 0.80 to 0.82 across the key diagnostic decisions—a consistent level of performance that, while not replacing microbiological confirmation, is sufficient to guide early therapeutic choices in the critical first hours of illness.

The practical appeal of this strategy lies in its accessibility. Every marker in the panel—cytokines, PCT, CRP, and the NLR—can be measured with routine laboratory instrumentation, and many are already part of standard workups for febrile patients. The study’s authors emphasize that the approach constitutes a cost-effective, rapid decision-support tool designed to facilitate early clinical decisions and antibiotic stewardship, with particular relevance for medical institutions that lack access to expensive molecular diagnostics such as multiplex PCR panels or next-generation sequencing. In resource-limited settings, where blood cultures may take days or be unavailable altogether and where antibiotic misuse is often most severe, a tiered pathway built from existing assays could meaningfully shorten the time to targeted therapy. The work was supported by the National Natural Science Foundation of China and by Science and Technology Projects funding from Guangzhou.

As with any biomarker study, the findings come with caveats that will shape future validation. The analysis was retrospective, drawing on existing medical records from a single center, and the classification of patients depended on final clinical diagnoses, which themselves carry inherent uncertainty. Cut-off values derived from one population may require recalibration in others, and the confidence intervals around some models—particularly the Gram-negative versus Gram-positive bloodstream infection model—indicate a degree of statistical uncertainty that larger cohorts could narrow. Nonetheless, the study demonstrates a compelling concept: that the immune system’s own chemical conversation, read through a carefully chosen panel of cytokines and routine markers, can be decoded into a structured diagnostic pathway. By sequencing the questions—etiology, pathogen subtype, severity—the strategy converts a scatter of laboratory values into an actionable clinical narrative, offering a template for how host-response diagnostics might complement, and in some cases temporarily stand in for, the slower tools of classical microbiology.

Subject of Research: Host immune response biomarkers for rapid differentiation of bacterial, viral, and fungal infections

Article Title: A rapid diagnostic strategy of pathogen infection using host markers of immune response

Article References: Cai, C., Zhong, R., Zheng, S., Li, J., Lin, D., Zhang, R., Duan, C., & Xu, Y. (2026). A rapid diagnostic strategy of pathogen infection using host markers of immune response. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14518-6

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14518-6

Keywords: biomarkers, cytokines, interleukin-6, interleukin-8, procalcitonin, neutrophil-to-lymphocyte ratio, bacterial infection, viral infection, sepsis, antibiotic stewardship, diagnostic markers, ROC analysis