Enzyme Selection
The core intuition behind enzyme evolution is simple: if an enzyme variant improves survival or reproduction, the backing code can spread. The hard part is that selection never “chooses” code directly— it filters phenotypes (observable effects), and only indirectly winnows genotypes (the sequences that produced those effects).
Selection can only “see” what an enzyme does
A nucleotide sequence is not selectable by itself. It becomes selectable only when it is expressed into a protein and that protein’s activity changes the organism’s outcomes in a heritable way.
Many origin narratives quietly assume a smooth gradient from “almost functional” to “functional.” For enzymes, that gradient can be broken by thresholds, toxicity, regulation requirements, and multi-component dependencies.
Why this matters for your broader arc
- Once translation is stable, enzyme families can plausibly diversify and optimize.
- Before translation is stable, enzyme-like functionality can be fragile, rare, and hard to preserve.
- So the “enzyme question” loops back into the Phase-2 bottleneck: translation, folding, and error control.
Three levels of difficulty
It helps to separate three questions that are often blended:
- Optimization: improving an already-working enzyme.
- Innovation: obtaining a genuinely new catalytic activity.
- Integration: making the new activity useful inside a living system (timing, dosage, pathway fit).
Demonstrations of cumulative selection often succeed at “optimization.” The origin story you’re interrogating is “innovation + integration” under tight prebiotic constraints.
Five necessary conditions for selecting enzyme sequences
These five conditions make precise the gap between “a sequence exists” and “a sequence can be selected.” They don’t prove that enzyme evolution is impossible—but they clarify why it is often harder than popular summaries imply.
Sequences can’t be selected unless they are expressed visibility
If DNA/RNA isn’t translated (or otherwise realized as functional chemistry), it yields no phenotype, and selection has nothing to act on.
Below thresholds, “partial” may not mean “partly useful” discontinuity
Many catalytic functions require a minimum fold stability, pocket geometry, or binding arrangement. 85% of the way there may be no better than 72% if the structure never crosses the functional threshold.
Intermediates can be deleterious toxicity
Expressed but non-functional (or misfolded) proteins can interfere with existing processes, aggregate, consume resources, and impose a fitness cost—meaning they are actively selected against.
Regulation is part of function timing & dosage
Even a beneficial activity may be useless—or harmful—if produced at the wrong time, in the wrong place, or at the wrong concentration. Regulation is often a prerequisite for selection.
Enzymes rarely act alone pathway context
A single enzyme typically sits inside a larger network (transport, cofactors, upstream/downstream steps). Without complementary components, a “new” enzyme may provide little or no selective advantage.
What selection can do well
Once you have a functioning enzyme and a stable cellular environment, selection can tune:
- reaction rate (kcat)
- binding specificity (Km / affinity)
- stability under new temperatures or pH
- expression levels and resource economy
What selection struggles to “see”
Selection is blind to latent potential. If an intermediate yields no phenotype, it is effectively invisible.
- silent code that isn’t expressed
- folds that never stabilize
- proto-functions that require missing cofactors
- activities that are only beneficial in combination
How this connects to Phase 2
Translation is the gateway that makes enzyme selection possible at scale. But translation itself requires a tightly coordinated system (charging, initiation, elongation, termination, folding, and quality control).
In other words: enzyme evolution is often easiest to explain after the machinery exists—yet the machinery is made of enzymes (and enzyme-like activities).
Common confusions worth disarming
“If it evolved, there must be a smooth path.”
Not necessarily. A system can exist today even if the path to it was narrow, contingent, and dependent on rare intermediate states. The question becomes: were the necessary intermediates sufficiently frequent, stable, and selectable?
“Enzymes are lock-and-key, so any change breaks them.”
The lock-and-key picture is a useful first pass, but many enzymes exhibit flexibility (“induced fit”). That flexibility can help evolutionary tuning—but it also complicates the origin story because specificity and control must eventually be enforced.
“So are you saying enzymes cannot evolve?”
No. The point here is narrower: selecting new enzyme sequences requires a chain of expressed, non-lethal, meaningfully beneficial intermediates in a cellular context. That’s a higher bar than most pop narratives acknowledge—and it becomes especially acute when you’re talking about origins rather than optimization.
Selection is powerful — but it’s not magic
Cumulative selection can climb impressive slopes when each step is both real and visible to selection. For enzymes, “visibility” depends on expression, thresholds, toxicity management, regulation, and integration into pathways. Those constraints don’t end the discussion—but they tell you exactly where the discussion must become concrete.