Quorum sensing (QS) loss-of-function mutations that confer antibiotic resistance impose a fitness cost in polymicrobial environments because QS-deficient strains cannot exploit cooperative behaviours and are exploited as public-goods cheaters by QS-intact competitors. Combining QS-inhibitor therapy with a QS-dependent antibiotic creates an evolutionary landscape in which QS-resistant mutants are simultaneously antibiotic-resistant and ecologically disadvantaged in the polymicrobial infection environment; longitudinal competition assays will demonstrate that combined QS + antibiotic therapy drives population dynamics toward susceptibility restoration rather than resistance fixation, quantifiable as a negative selection coefficient for resistance alleles across ≥5 serial passage cycles.

In polymicrobial infection environments containing both QS-intact (wild-type) and QS-deficient (loss-of-function mutant) bacterial strains, QS-deficient mutants that exhibit antibiotic resistance (via QS-dependent resistance mechanisms, e.g., biofilm formation, efflux pump regulation, or virulence factor production) will display a measurable negative selection coefficient (s < −0.05 per passage) relative to QS-intact competitors across ≥5 serial passage cycles under combined QS-inhibitor + QS-dependent antibiotic treatment. Specifically: (1) QS-deficient mutants will be outcompeted by QS-intact strains in co-culture due to public-goods exploitation dynamics; (2) combined therapy (QS inhibitor + antibiotic) will produce a lower frequency of resistance fixation compared to antibiotic monotherapy; and (3) the resistance allele frequency will decline monotonically across passages under combined therapy, quantifiable by allele-specific qPCR or whole-genome sequencing with ≥10× coverage.

Minimum Viable Test (MVT) Design — 3-Phase Protocol

Phase 1: Strain Construction and Characterization (Weeks 1–6)

• Generate isogenic QS-deficient mutants (ΔlasR, ΔrhlR, or Δagr) in P. aeruginosa PAO1 or S. aureus USA300 via lambda Red recombineering or CRISPR-Cas9.
• Confirm loss of QS function by measuring AHL/AIP production (LC-MS/MS), virulence factor output (elastase, pyocyanin, RNAIII), and biofilm formation (crystal violet assay, OD590).
• Confirm antibiotic resistance phenotype: MIC determination (broth microdilution, CLSI guidelines) for tobramycin, ciprofloxacin, and colistin. Target: ≥4-fold MIC increase in QS-deficient mutant vs. wild-type.
• Introduce distinguishable neutral fluorescent markers (GFP in QS-deficient, mCherry in QS-intact) for competition tracking without fitness artifacts (confirm marker neutrality in monoculture: |s_marker| < 0.01).

Phase 2: Serial Passage Competition Assays (Weeks 7–18)

• 4 treatment arms × 6 biological replicates × 10 passages = 240 culture vessels
  • Arm A: Antibiotic monotherapy (tobramycin at 0.5× MIC of wild-type)
  • Arm B: QS inhibitor monotherapy (furanone C-30 at 50 μM, sub-bactericidal)
  • Arm C: Combined therapy (tobramycin 0.5× MIC + furanone C-30 50 μM)
  • Arm D: No treatment control
• Initial inoculum: 90% QS-intact (mCherry) + 10% QS-deficient (GFP), total 10^7 CFU/mL in artificial sputum medium (ASM).
• Passage every 24 hours (1:100 dilution into fresh medium + treatment); sample at passages 0, 1, 3, 5, 7, 10.
• Quantify strain frequencies by: (a) flow cytometry (GFP/mCherry ratio, ≥10,000 events/sample), (b) allele-specific qPCR (ΔlasR vs. lasR+), (c) selective plating on antibiotic-containing agar.

Phase 3: Mechanistic Validation (Weeks 19–26)

• RNA-seq on sorted QS-deficient and QS-intact cells from co-culture at passages 1, 5, and 10 (n = 3 biological replicates per timepoint per arm).
• Supernatant proteomics (LC-MS/MS) to quantify public-goods secretion (elastase, LasB, pyocyanin) across passages.
• Mathematical modeling: Fit experimental data to a two-strain Lotka-Volterra competition model with public-goods term; estimate selection coefficients and validate against empirical allele frequencies.

# Experimental Validation Package — Implementation Sketch
# QS Loss-of-Function Fitness Cost in Polymicrobial Environments

# ============================================================
# MODULE 1: STRAIN MANAGEMENT
# ============================================================
class BacterialStrain:
    def __init__(self, name, qs_status, fluorescent_marker, antibiotic_resistance):
        self.name = name  # e.g., "PAO1_WT", "PAO1_deltaLasR"
        self.qs_status = qs_status  # "intact" or "deficient"
        self.marker = fluorescent_marker  # "GFP" or "mCherry"
        self.resistance = antibiotic_resistance  # dict: {"tobramycin": MIC_value}
        self.selection_coefficient = None  # calculated post-experiment

def construct_mutant_library():
    """
    Returns validated isogenic strain pairs for competition assays.
    Validates: deletion confirmation, MIC fold-change >= 4x, marker neutrality |s| < 0.01
    """
    strains = {
        "PAO1_WT_mCherry": BacterialStrain("PAO1_WT", "intact", "mCherry", {"tobramycin": 1.0}),
        "PAO1_deltaLasR_GFP": BacterialStrain("PAO1_deltaLasR", "deficient", "GFP", {"tobramycin": 4.0}),
        "USA300_WT_mCherry": BacterialStrain("USA300_WT", "intact", "mCherry", {"vancomycin": 1.0}),
        "USA300_deltaAgr_GFP": BacterialStrain("USA300_deltaAgr", "deficient", "GFP", {"vancomycin": 4.0}),
    }
    return strains

# ============================================================
# MODULE 2: SERIAL PASSAGE EXPERIMENT
# ============================================================
class SerialPassageExperiment:
    def __init__(self, n_passages=10, n_replicates=6, initial_qs_deficient_freq=0.10):
        self.n_passages = n_passages
        self.n_replicates = n_replicates
        self.initial_freq = initial_qs_deficient_freq
        self.treatment_arms = {
            "A": {"antibiotic": 0.5, "qs_inhibitor": 0.0},   # monotherapy
            "B": {"antibiotic": 0.0, "qs_inhibitor": 50.0},  # QSI only (uM)
            "C": {"antibiotic": 0.5, "qs_inhibitor": 50.0},  # combined
            "D": {"antibiotic": 0.0, "qs_inhibitor": 0.0},   # control
        }
        self.results = {}  # arm -> replicate -> passage -> frequency

    def run_passage_cycle(self, arm, replicate, passage_num, current_freq):
        """
        Simulates one passage cycle; in practice, executes:
        1. 24h incubation at 37C, 200 rpm in ASM
        2. Flow cytometry sampling (10,000 events)
        3. qPCR allele frequency measurement
        4. 1:100 dilution into fresh medium + treatment
        Returns: new_frequency (float), raw_data (dict)
        """
        # Placeholder for experimental execution
        raw_data = {
            "flow_cytometry_freq": None,  # measured GFP/(GFP+mCherry)
            "qpcr_freq": None,            # measured by allele-specific qPCR
            "cfu_total": None,            # total CFU/mL
            "cfu_resistant": None,        # CFU on antibiotic plates
        }
        return raw_data

    def calculate_selection_coefficient(self, freq_trajectory):
        """
        s = slope of ln(f/(1-f)) vs. passage number
        Uses linear regression; returns (s, 95_CI_lower, 95_CI_upper, p_value)
        """
        import numpy as np
        from scipy import stats
        log_odds = np.log(np.array(freq_trajectory) / (1 - np.array(freq_trajectory) + 1e-10))
        passages = np.arange(len(freq_trajectory))
        slope, intercept, r_value, p_value, std_err = stats.linregress(passages, log_odds)
        ci_lower = slope - 1.96 * std_err
        ci_upper = slope + 1.96 * std_err
        return {"s": slope, "ci_lower": ci_lower, "ci_upper": ci_upper, "p_value": p_value}

# ============================================================
# MODULE 3: MATHEMATICAL COMPETITION MODEL
# ============================================================
def two_strain_public_goods_model(y, t, params):
    """
    ODE system for QS-intact (N1) vs QS-deficient (N2) competition
    with public goods exploitation dynamics.
    
    dN1/dt = r1*N1*(1-(N1+N2)/K) + alpha*G(N1,N2)*N1
    dN2/dt = r2*N2