Exploring how evolutionary byproducts, or genomic spandrels, challenge the adaptationist view of genome evolution and transform our understanding of DNA.
When we look at the human genome, it's tempting to see every element as a precisely engineered adaptation, fine-tuned by millions of years of natural selection. This "adaptationist program" has long dominated evolutionary biology, but what if we're missing a crucial part of the story? In 1979, evolutionary biologists Stephen Jay Gould and Richard Lewontin proposed a radical alternative in their famous paper, "The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme" 3 . They borrowed an architectural metaphor: the spandrels—the triangular spaces between arches that support domes in churches—are not the primary design elements but inevitable byproducts of the architectural blueprint. Yet, these spaces often become beautifully decorated, giving the appearance of purposeful design.
These molecular "spandrels" might later be co-opted for useful functions, but their origin lies in structural necessity rather than adaptive design.
As we delve deeper into genomic research, recognizing these spandrels becomes essential for truly understanding how evolution has sculpted our DNA and why some genomic elements exist without apparent function. This perspective challenges us to look beyond mere utility to appreciate the complex, sometimes messy, architectural constraints that shape living organisms.
In Renaissance architecture, spandrels are the inevitable geometric spaces created when arches meet a rectangular frame. While architects often decorate these spaces with intricate mosaics or paintings, the spandrels themselves were not the reason for the architectural design—they were necessary byproducts 3 . Gould and Lewontin argued that biology similarly contains many features that arise as necessary consequences of structural constraints rather than direct products of natural selection.
When this concept is applied to genomics, we move from cathedral ceilings to the molecular architecture of DNA. The emerging field studying genomic spandrels suggests that many DNA sequences, structural elements, and even functional pathways may exist primarily as evolutionary byproducts 2 5 . These elements emerge not because they were directly selected for their current utility, but because they were dragged along by other adaptive changes or forced into existence by structural constraints.
Just as architectural spandrels are byproducts of structural design, genomic spandrels emerge from evolutionary constraints rather than direct selection.
Genomic architecture creates inevitable byproducts
Features emerge without direct selective advantage
Spandrels may later acquire functional roles
Gould and later co-author Elizabeth Vrba expanded this concept by introducing the term "exaptation" to describe characteristics that enhance fitness in their present role but were not originally built for that purpose by natural selection 3 . Exaptations can be divided into two subcategories: pre-adaptations (features adapted for one function later used for another) and spandrels (features that originated as non-adaptive byproducts) 3 .
"Evolutionary biology needs such an explicit term for features arising as byproducts, rather than adaptations, whatever their subsequent exaptive utility" 3 .
This distinction is crucial for understanding genomic evolution. As Gould noted, this perspective helps explain why genomes contain so many elements that appear functional but may have originated as structural necessities.
The concept of genomic spandrels connects to a broader revolutionary framework in evolutionary biology: the importance of neutral evolution 6 . Proposed by Motoo Kimura and others, the neutral theory suggests that the majority of evolutionary changes at the molecular level are caused not by natural selection but by random genetic drift of mutant alleles that are effectively neutral 6 .
This provides a crucial null hypothesis for genomic analysis: instead of asking "what is this sequence for?" we should first ask "how has this sequence evolved?" 6 This shift in perspective has profound implications for how we interpret genomic data, especially in the face of discoveries like "junk DNA"—the recognition that much of our genome doesn't code for proteins and may lack any specific function 6 .
| Genomic Element | Traditional Adaptationist View | Spandrel/Neutral Evolution View |
|---|---|---|
| Non-coding DNA | "Junk" DNA must have cryptic functions | Most is true junk: evolutionary debris |
| Introns | Necessary for alternative splicing | Mostly junk; persist due to weak selection |
| Gene order | Optimized arrangement | Largely neutral; reflects historical contingency |
| 5' slow translation | Adaptive "ramp" to prevent ribosome collisions | Byproduct of evolutionary turnover 5 |
The power of random genetic drift is particularly strong in small populations, such as those typical of many multicellular eukaryotes, including humans 6 7 . In these populations, selection is too weak to eliminate many non-functional elements, allowing them to persist as evolutionary "spandrels" or true junk DNA.
In 2015, a team of researchers tackled a fundamental question in evolutionary biology: can non-functional protein interactions emerge spontaneously as evolutionary spandrels? 2 The study addressed how proteins evolve new binding interactions while maintaining their structural stability—a classic challenge in molecular evolution.
The researchers hypothesized that the structural coupling between folding and binding—the fact that proteins must be folded to bind their targets—creates an evolutionary coupling between the traits of folding stability and binding strength 2 . This coupling could cause strong binding interactions to emerge as evolutionary spandrels even when they provide no fitness benefit.
Combined biophysical modeling with directed evolution to test the spandrel hypothesis
The team developed a sophisticated biophysical and evolutionary model that simulated how protein sequences evolve under various selective regimes 2 . Their approach included several key components:
They modeled proteins as having two-state folding kinetics, with free energies of folding (Ef) and binding (Eb) determined by additive contributions from individual amino acids.
They created a mathematical fitness function based on the probabilities of a protein being in one of three states: folded and bound, folded and unbound, or unfolded and unbound.
Using exact numerical algorithms, they characterized adaptive evolutionary paths on this fitness landscape, examining how coupled protein traits affect evolutionary outcomes.
The researchers then tested their model's predictions through directed evolution experiments in which they selected for either folding stability alone, binding function alone, or various combinations of both traits 2 .
The experiments yielded striking evidence for evolutionary spandrels. When selection acted only on folding stability (with no intrinsic fitness advantage for binding), proteins nevertheless frequently evolved strong but non-functional binding interactions 2 .
| Selective Pressure | Emergence of Non-functional Binding |
|---|---|
| Folding stability only | Frequent |
| Binding function only | Rare |
| Both traits | Occasional |
| Characteristic | Functional Binding | Spandrel Binding |
|---|---|---|
| Evolutionary predictability | High | Low |
| Dependence on random mutations | Weak | Strong |
| Functional role | Direct fitness advantage | Stabilizes protein |
The research demonstrated that what appears to be optimized binding function could instead be a molecular spandrel—a byproduct of selection for stability rather than direct selection for function 2 . This may explain the abundance of apparently non-functional interactions observed in high-throughput protein assays.
Furthermore, the study revealed dramatically different evolutionary predictability between functional and spandrel traits. Evolution of functional binding followed a predictable pattern: proteins first gained extra folding stability, then partially lost it as the new binding function improved. In contrast, evolution of non-functional spandrel binding was highly unpredictable, dependent on random mutation events 2 .
Studying genomic spandrels requires specialized methods and reagents that can distinguish adaptive functions from non-adaptive byproducts.
| Reagent/Method | Function | Example Use in Spandrel Research |
|---|---|---|
| Ribosome Profiling | Measures translation speed codon-by-codon | Identifying slow translation regions that may be spandrels 5 |
| tRNA-adaptation index (tAI) | Proxy for codon translation speed | Comparing 5' vs internal codon preferences 5 |
| Population genetic simulations | Models evolutionary processes | Testing neutral vs adaptive hypotheses 7 |
| Directed evolution | Artificial selection in laboratory | Evolving proteins under controlled selective regimes 2 |
| Ribosome Residence Time (RRT) | Direct measurement of codon translation speed | Quantifying translation speed variations along genes 5 |
| Phylogenetic comparative methods | Comparing sequences across species | Distinguishing conserved vs rapidly evolving elements 5 |
Laboratory techniques to measure molecular interactions and functions
Bioinformatics approaches to analyze genomic data and model evolution
Methods to distinguish signal from noise in evolutionary patterns
These tools have been instrumental in shifting how we interpret genomic features. For instance, ribosome profiling and RRT measurements helped researchers re-evaluate the "translational ramp" hypothesis—the idea that slow translation at the beginning of genes is an adaptation to prevent ribosome collisions 5 . Instead, this pattern appears to be a spandrel resulting from the evolutionary turnover of 5' gene ends 5 .
The concept of genomic spandrels forces a major reconsideration of functional genomics. The once-common assumption that most genomic elements exist for adaptive reasons has been challenged by the recognition that biological systems are constrained in ways that generate inevitable byproducts . This is particularly relevant in interpreting projects like ENCODE, which famously claimed that 80% of the human genome is "functional" 6 .
ENCODE's claim of "functional" genome
Evolutionary conservation estimate
Evolutionary biologists quickly countered that evolutionary conservation suggests only about 6-9% of our genome is under selective constraint 6 . The discrepancy may be explained by genomic spandrels—features that have biochemical activity but emerged as non-adaptive byproducts rather than selected functions.
Understanding genomic spandrels has profound implications for medicine and disease research. Many diseases of aging may result from antagonistic pleiotropy, where genes beneficial early in life become harmful later—a form of evolutionary spandrel . As we age, the same biological mechanisms that enabled growth and development may become detrimental, leading to pathology.
In conservation biology, researchers caution against what they call "modern spandrels"—the overemphasis on preserving measurable adaptive variation while ignoring the vast unknown majority that cannot be measured 4 . This represents a modern twist on the adaptationist program that Gould and Lewontin critiqued nearly 40 years ago.
The concept of genomic spandrels represents a fundamental shift in how we view evolution's creative power. Rather than seeing natural selection as an omnipotent designer crafting every genomic element for a purpose, we're learning to appreciate the complex interplay of adaptation, constraint, and chance in shaping genomes. Some genomic features are indeed finely tuned adaptations, but many others are architectural byproducts—molecular spandrels that might be decorated with biochemical activity but originated as inevitable consequences of genomic architecture.
This perspective doesn't diminish the power of evolution; instead, it enriches our understanding of evolutionary processes. As we continue to explore the human genome and those of other species, recognizing these spandrels will be crucial for distinguishing true function from evolutionary decoration. The challenge ahead lies in developing better tools to identify which genomic elements are adaptive masterpieces and which are structural necessities—the spandrels of our genomic cathedrals.
"Causes of historical origin must always be separated from current utilities; their conflation has seriously hampered the evolutionary analysis of form in the history of life" 3 .
In the genomic era, this wisdom has never been more relevant.