Antibiotic Resistance: Evolution in Real Time

Bacteria evolve so rapidly that resistance to a new antibiotic can emerge within months of clinical introduction. The WHO estimates antimicrobial resistance (AMR) contributed to 1.27 million deaths in 2019 — and without action, could kill 10 million per year by 2050. Understanding the biology and mathematics of resistance is the first step to combating it.

1. How Antibiotics Work

Antibiotics target structures or processes essential to bacteria but absent or different in human cells — explaining selective toxicity:

Cell wall synthesis inhibitors: β-lactams (penicillin, amoxicillin, carbapenems), glycopeptides (vancomycin). Inhibit transpeptidases that cross-link peptidoglycan → wall weakens → osmotic lysis. Human cells have no cell wall.
Protein synthesis inhibitors: Aminoglycosides, tetracyclines, macrolides. Target bacterial 70S ribosome (30S or 50S subunit). Human ribosomes are 80S → usually selective (mitochondrial 70S ribosomes → side effects).
DNA replication inhibitors: Fluoroquinolones (ciprofloxacin) inhibit bacterial gyrase (topoisomerase II). Sulfonamides block folate synthesis.
Cell membrane disruptors: Polymyxins (colistin) — last resort. Bind lipopolysaccharide in gram-negative outer membrane → disrupts permeability.

2. Resistance Mechanisms

Four main categories of resistance: 1. Enzymatic inactivation: β-lactamases: hydrolyse the β-lactam ring \to antibiotic inert Extended-spectrum β-lactamases (ESBL): inactivate 3rd-gen cephalosporins Carbapenemases (NDM-1, KPC, OXA-48): inactivate carbapenems (last resort) 2. Target modification: mecA gene: encodes PBP2a (altered penicillin-binding protein) \to MRSA (methicillin-resistant Staph aureus) PBP2a has 100\times lower affinity for β-lactams \to β-lactam drugs ineffective 23S rRNA mutation \to macrolide binding reduced (M. tuberculosis) 3. Efflux pumps: AcrAB-TolC (E. coli): expels many antibiotics before they act Can confer multi-drug resistance in one gene acquisition Reduced permeability: loss of OmpF porin \to reduced uptake of antibiotics 4. Target bypass / overproduction: vanA gene cluster (Enterococcus): reprograms cell wall precursor D-Ala-D-Ala \to D-Ala-D-Lac \to vancomycin 1000\times lower affinity "VRE" (vancomycin-resistant Enterococcus)

3. Mutation Rates and Selection

Bacterial mutation rates: ~10⁻⁹ to 10⁻¹⁰ mutations per base pair per replication (E. coli) Genome \approx 4\times10⁶ bp \to ~4\times10⁻³ to 4\times10⁻⁴ new mutations per genome per division "Hypermutators" (mutL, mutS, mutU knockouts): 100\times elevated mutation rate \to accelerated resistance evolution under selective pressure Population dynamics under antibiotic exposure: Sensitive cells: growth rate g_S, killed at rate k Resistant cells: growth rate g_R < g_S (fitness cost), not killed In absence of antibiotic: resistant mutants at frequency ~10⁻⁸ (selection against resistance cost keeps them rare) In presence of antibiotic: dN_S/dt = g_S \cdot N_S - k \cdot N_S \to if k > g_S: exponential decline dN_R/dt = g_R \cdot N_R \to exponential increase Time to dominant resistance: (1/g_R) \cdot ln(breakeven N_S/N_R) \approx days to weeks MEGA-plate experiment (Kishony lab, 2016): E. coli grew across antibiotic gradient (1000\times MIC at centre). Resistance evolved step-by-step: low-resistance mutants colonised medium zones, higher mutants then arose from that adapted population. Visible in 10 days. Demonstrates stepwise resistance evolution visually.

4. Horizontal Gene Transfer

Unlike vertical (parent-to-offspring) inheritance, horizontal gene transfer (HGT) allows resistance genes to spread between species — potentially crossing genus boundaries in hours:

Conjugation: Bacteria form physical contact via sex pili; plasmid DNA transfers through a channel. R-plasmids (resistance plasmids) can carry 5-10 different resistance genes simultaneously. One conjugation event spreads a multi-drug-resistant package.
Transformation: Bacteria take up naked DNA from the environment (e.g., from lysed bacteria). Streptococcus pneumoniae is naturally competent. Penicillin resistance in pneumococcus arose from genetic material of commensal streptococci.
Transduction: Bacteriophages accidentally package host DNA and inject it into the next host. Phage-mediated transfer of toxin genes → new virulent pathogen (MRSA virulence factors).
Transposons and integrons: "Jumping genes" that move between plasmids and chromosomes, capturing and rearranging resistance cassettes. Class 1 integrons are found in virtually all resistant Gram-negative bacteria.

NDM-1 (New Delhi Metallo-β-lactamase): A carbapenemase gene first identified in Klebsiella pneumoniae from a patient who had travelled from India in 2008. Within 2 years, NDM-1 appeared on multiple plasmid types in bacteria from dozens of countries on 6 continents. This illustrates how HGT enables global spread of resistance genes independently of the bacterium — a "resistance gene pandemic" that follows human mobility and healthcare travel.

5. ESKAPE Pathogens and Clinical Reality

The ESKAPE pathogens — named by the IDSA as those most likely to "escape" antibiotic treatment — represent the most urgent AMR threats:

E — Enterococcus faecium: VRE (vancomycin-resistant). Common in hospital ICUs, especially immunocompromised patients. Intrinsic resistance to many antibiotics.
S — Staphylococcus aureus: MRSA, resistant to all β-lactams. Community-MRSA now epidemic. HA-MRSA deaths exceed 10,000/year in the US.
K — Klebsiella pneumoniae: Carbapenem-resistant (CR-Kpn). Mortality 40-50% for bacteraemia. NDM, KPC, OXA-48 carbapenemases.
A — Acinetobacter baumannii: Extensive drug-resistant (XDR) strains. Survives on hospital surfaces for weeks. Colistin often last resort.
P — Pseudomonas aeruginosa: Multidrug-resistant (MDR) with intrinsic efflux pumps. Dangerous in cystic fibrosis and burn patients.
E — Enterobacteriaceae: ESBL-producing E. coli. Most common AMR pathogen. UTIs with no oral treatment options.

6. Antibiotic Discovery Pipeline

The antibiotic discovery pipeline is critically depleted. Most antibiotics approved since 2000 are derivatives of existing classes — resistance to which exists or develops quickly:

Antibiotic discovery timeline (last novel classes): 1945: β-lactams (penicillin \to widespread clinical use) 1948: Aminoglycosides (streptomycin, neomycin) 1952: Macrolides (erythromycin) 1960: Glycopeptides (vancomycin) 1962: Quinolones (nalidixic acid \to fluoroquinolones 1980s) 1987: Lipopeptides (daptomycin — approved 2003) 2003: Oxazolidinones (linezolid — approved 2000) 2010+: No truly new classes in clinical use for >30 years Market failure: Antibiotic course: $20-100, patient takes for 5-10 days Oncology drug: $50,000-200,000/course \to No market incentive for antibiotic R&D \to Most large pharma exited the field (Pfizer, AstraZeneca, Eli Lilly, Novartis) \to Only small biotech remain + academic research

7. Strategies to Combat AMR

Antibiotic stewardship: Use the right antibiotic at the right dose for the right duration. Rapid diagnostics (PCR, MALDI-TOF) to identify pathogen and resistance profile in hours (not days). Each course that's unnecessary is avoided — selection pressure reduced.
Phage therapy: Bacteriophages are viruses that kill bacteria with high specificity. Personalised phage therapy against MDR infections has shown dramatic individual successes (Strathdee UCSD cases). Challenges: phage resistance can also evolve, immune clearance, regulatory pathway unclear.
Antivirulence strategies: Rather than killing bacteria — block their virulence factors (quorum sensing inhibitors, toxin inhibitors). Less selective pressure for resistance since bacteria can survive but can't harm the host.
Combination therapy: Two drugs with different targets — probability of simultaneous resistance mutations is product of individual probabilities (~10⁻⁸ × 10⁻⁸ = 10⁻¹⁶). Basis of tuberculosis 4-drug RIPE regimen.
Push/pull incentives: Government contracts (US BARDA, UK AMR fund) that pay for development regardless of sales volume. "Market entry rewards" for successful new antibiotics. WHO "priority pathogens" list to direct funding.
Vaccines: Pneumococcal, Hib, meningococcal vaccines reduce bacterial infections → fewer antibiotics prescribed. S. aureus vaccine remains an elusive goal (multiple clinical trial failures).