The Territory

Half of all clinically used antibiotics target one structure: the bacterial ribosome. Aminoglycosides since 1944. Tetracyclines since 1948. Chloramphenicol, macrolides, lincosamides, oxazolidinones, pleuromutilins, streptogramins, thiopeptides. The 70S ribosome — 4,500 nucleotides of RNA plus 54 proteins, present in thousands of copies per cell — has been the most bombed target in antimicrobial history.

But here is what eighty years of bombing produced: nearly every drug converged on the same few sites. The peptidyl transferase center. The exit tunnel. The decoding center. A handful of pockets in a structure that spans 2.5 megadaltons. Resistance evolved in those pockets — not because bacteria are infinitely resourceful, but because that's where the pressure was. We created our own problem by attacking the same neighborhoods.

In 2025 and 2026, a single laboratory opened two new neighborhoods. Both are genuinely unoccupied. Both sit in regions where no antibiotic selection pressure has ever operated. This post maps the territory.

The Ribosome as Pharmacological Landscape

The Crowded Districts

The bacterial ribosome has ~7 pharmacologically exploited sites. But the exploitation is wildly uneven. One site — the peptidyl transferase center — hosts six distinct antibiotic classes that independently evolved or were designed to bind the same cavity.

District	Subunit	Classes Occupying	First Drug (Year)	Dominant Resistance
Peptidyl transferase center	50S	Chloramphenicol, lincosamides, oxazolidinones, pleuromutilins, streptogramins, phenicols	Chloramphenicol (1949)	Erm methylation, cfr modification, ABC-F ribosomal protection
Nascent peptide exit tunnel	50S	Macrolides, ketolides	Erythromycin (1952)	Erm methylation, mef/msr efflux, macrolide esterases
Decoding center (A-site)	30S	Aminoglycosides	Streptomycin (1944)	Aminoglycoside-modifying enzymes (AMEs), 16S rRNA methyltransferases (ArmA, RmtB)
Tetracycline pocket	30S	Tetracyclines (classical + glycylcyclines)	Chlortetracycline (1948)	Tet efflux (30+ genes), Tet(M)/Tet(O) ribosomal protection
GTPase-associated center	50S	Thiopeptides (thiostrepton, nosiheptide)	Thiostrepton (1955)	23S rRNA methylation (limited clinical use)
Spectinomycin site (helix 34)	30S	Spectinomycin	Spectinomycin (1961)	Point mutations, ANT(9) adenylyltransferase
E-site (H13/H21)	50S	Manikomycin (2026)	Manikomycin (2026)	None. No prior selection pressure.
Lariocidin site (16S/aa-tRNA)	30S	Lariocidin (2025)	Lariocidin (2025)	None. No prior selection pressure.

The pattern is immediately visible. The PTC alone — one cavity — has been independently targeted by at least six antibiotic classes over 75 years. This is not coincidence. It reflects both the PTC's functional importance (it catalyzes every peptide bond) and the medicinal chemistry bias toward known binding sites. When you know a pocket works, you design more drugs for it. When you design more drugs for it, you select for resistance against all of them simultaneously.

Erm methyltransferases are the signature consequence. A single enzyme — methylating A2058 of 23S rRNA — confers resistance to macrolides, lincosamides, and streptogramins simultaneously. One mutation, three classes neutralized. This is what convergent targeting produces: convergent resistance.

Why the E-Site Was Unmapped

The ribosomal E-site (exit site) is where deacylated tRNA leaves the ribosome after donating its amino acid. Functionally, it has always been the "departure lounge" — important for maintaining translation speed and reading frame, but never considered a bottleneck worth blocking. The logic seemed sound: why trap what's already finished?

Manikomycin answers this question. Published in Nature on June 3, 2026, it is a cyclic depsipeptide from Streptomyces rimosus — the same organism that produces oxytetracycline, studied since the 1950s. Seventy-five years of research on this species. The compound was there the whole time.

Why was this site overlooked? Three reasons. First, the structural biology bias: most ribosome-antibiotic crystallography focused on the PTC and decoding center because that's where existing drugs bound. You find what you look for. Second, the E-site's role in translation was considered "passive" — tRNA exiting, not entering. But translation is not an assembly line with independent stations; it is a coupled cycle where exit events constrain entry events. Third, natural product screening historically used growth inhibition assays that don't reveal which ribosomal site is targeted. You need the cryo-EM after finding the compound.

Fourteen Months Earlier: Lariocidin Opens Another Front

In March 2025, the same laboratory — Gerry Wright's group at McMaster University — published lariocidin in Nature. A lasso peptide isolated from a Paenibacillus bacterium that grew for one year in a soil sample from Hamilton, Ontario. One year. Most screening protocols use weeks. Wright's team waited longer, and the slow-growing producer revealed itself.

Lariocidin binds the 30S subunit at a site distinct from every existing aminoglycoside or tetracycline pocket. It interacts with 16S rRNA near the aminoacyl-tRNA, but its mechanism is dual: it both inhibits translocation (the movement of mRNA through the ribosome) and induces miscoding (incorrect amino acid incorporation). Two effects from one binding event.

Two novel ribosomal binding sites. Fourteen months apart. Same lab. Both natural products from soil. Both targeting sites where no pharmaceutical selection pressure has ever operated.

The Selection Pressure Principle

This is the conceptual core of the discovery. Resistance is not a fixed property of bacteria — it is an evolutionary response to selective pressure. Where pressure concentrates, resistance concentrates. Where pressure has never been applied, no resistance has been selected for.

The PTC has been under pharmaceutical attack since 1949. Seventy-seven years of selective pressure. The result: Erm methyltransferases, cfr eight-carbon methylation (confers PhLOPSA resistance — phenicols, lincosamides, oxazolidinones, pleuromutilins, streptogramin A — from a single gene), ABC-F ribosomal protection proteins that physically pry drugs off the ribosome, enzymatic drug inactivation (chloramphenicol acetyltransferases). An entire resistance ecology, evolved specifically because we kept attacking the same site.

The E-site has been under pharmaceutical attack for zero years. Zero selection pressure. Zero evolved resistance mechanisms. The only resistance manikomycin encounters in laboratory settings are loss-of-function mutations in uptake transporters (sbmA/YejABEF) — blunt instruments that reduce drug entry at fitness cost, not the sophisticated target-site modifications that decades of selection produce.

This is not a guarantee of future invulnerability. Bacteria will adapt. But the starting position is categorically different from introducing yet another PTC binder into a resistance ecology that already contains dozens of countermeasures.

The Hybrid Dimension

A parallel development exploits the ribosome's geography differently. In a 2025 paper in ACS Central Science, researchers demonstrated hybrid antibiotics — molecular conjugates that link two ribosome inhibitors via a chemical linker designed to let both drugs simultaneously bind their respective sites.

The most advanced example: an azithromycin-tedizolid conjugate active against MDR Gram-positives. Azithromycin occupies the exit tunnel; tedizolid occupies the PTC. Connected by a linker calibrated to the 5.2 Å distance between the binding pockets. A single molecule that occupies two addresses.

The pharmacological logic: resistance that dislodges one drug from one site does not dislodge the other from the adjacent site. Erm methylation (which blocks macrolides) leaves the oxazolidinone half functional. cfr modification (which blocks oxazolidinones) leaves the macrolide half functional. Only simultaneous modification of both sites — far less likely as a single evolutionary step — neutralizes the hybrid.

Now map this concept forward. When the E-site and lariocidin sites are pharmacologically developed, the ribosome offers not five but seven or eight sites for combination targeting. The geometric space for hybrid design expands dramatically. Conjugates spanning novel and classical sites would face resistance ecologies that evolved for only one of their two binding events.

The Wright Pattern

Gerry Wright's laboratory at McMaster University has produced four major antibiotic discoveries in approximately one year. Lariocidin. Manikomycin. Butyrolactol A (an antifungal targeting phospholipid flippase). A fourth compound not yet fully disclosed. Two of the four target the ribosome at novel sites.

The methodology is instructive. Wright's approach is not AI-driven molecular design (the dominant paradigm of 2024-2026). It is patient natural product screening with three specific modifications to the classical Waksman platform:

1. Extended cultivation time. Lariocidin's producer took a full year to grow to detectable levels. Standard screens discard plates at weeks. Most soil bacteria are slow growers that never dominate fast cultures.
2. Re-examining "known" organisms. Manikomycin came from Streptomyces rimosus — a species studied for 75 years. The compound was missed not because the species was unknown but because the biosynthetic gene cluster (6.7 kb NRPS, manA/manB) was silent under standard conditions. Cryptic cluster activation through non-standard media or stress conditions revealed it.
3. Mechanism-first characterization. Both compounds were immediately subjected to cryo-EM structural biology, revealing their binding sites before any optimization chemistry began. Target identification is not an afterthought — it is the first question.

This is worth emphasizing in a moment when AI antibiotic discovery receives overwhelming attention and funding. APEX, GNEprop, SyntheMol, Phare Bio — all computationally brilliant. All targeting known mechanisms (membrane disruption, LptA, known PTC interactions). Wright's approach is orthogonal: let nature's combinatorial chemistry do the library generation, but apply modern structural biology to identify where the molecules bind before asking whether they're good drugs.

Clinical Reality

Neither manikomycin nor lariocidin is a drug. Both are early-stage natural products with significant optimization required.

Parameter	Manikomycin	Lariocidin
MIC vs E. coli	32 μg/mL (high)	Active (specific MICs not fully disclosed)
In vivo efficacy	Limited murine model effect (bioavailability issue); C. elegans positive	100% mouse survival vs CRAB
Selectivity window	15-fold (bacterial vs eukaryotic IC₅₀)	Not disclosed; lasso topology suggests proteolytic stability
Derivatives	60+ synthesized; optimizing residency time	Self-resistance enzyme (lrcE) characterized (ACS Infect Dis 2026)
Stage	Early preclinical	Early preclinical
Key limitation	High MICs, poor bioavailability, modest selectivity	Scale-up of fermentation, formulation for systemic delivery

Manikomycin's 32 μg/mL MIC against E. coli is too high for a clinical candidate. For context, ceftriaxone works at 0.06 μg/mL. The 15-fold selectivity window (bacterial vs eukaryotic) is adequate but not comfortable — most clinically used ribosome inhibitors achieve >100-fold. The murine model showed limited effect, suggesting rapid clearance or poor distribution. These are real problems that optimization may or may not solve.

But the significance of manikomycin is not the molecule. It is the target. The E-site is now a validated pharmacological address. Other groups will screen against it. Medicinal chemistry campaigns will optimize for it. The 2.4 Å cryo-EM structure is public — anyone can now do structure-guided drug design against the E-site. The gate is open.

What the Map Tells Us

Eighty years. Billions of doses. Trillions of bacteria exposed. And through all of that, we bombed the same four or five neighborhoods — building resistance ecologies so sophisticated that a single gene (cfr) can neutralize five antibiotic classes at once.

The ribosome was never the problem. The ribosome is an enormous structure with many functional sites. Our imagination was the problem. We kept returning to the places that worked before, optimizing chemistry against known pockets rather than asking whether the pockets we hadn't tried might offer clean terrain.

In 14 months, Wright's laboratory demonstrated that the terrain exists. Two sites. Both natural products — compounds that evolved in soil ecosystems where bacteria have fought each other for billions of years, finding vulnerabilities we never looked for. The E-site and the lariocidin site are not theoretical targets. They are proven functional sites where small molecules bind and kill bacteria. The pharmacology is real; only the clinical optimization remains.

The ribosome, it turns out, still has territory left. We just had to stop looking where everyone else was standing.

Sources: Kaur et al., "A natural depsipeptide antibiotic that targets the E site of the bacterial ribosome," Nature (June 2026). Lariocidin, Nature (March 2025). Hybrid antibiotics targeting the bacterial ribosome, ACS Central Science (2025). Mechanistic insights into clinically relevant ribosome-targeting antibiotics, Biomolecules (2024). Structural signatures of antibiotic binding sites on the ribosome, Nucleic Acids Research (2010). Prior coverage: Post #5: The New Arsenal, Post #25: The Blueprint.