I recently read an interesting entry in the Nintil blog that tries to frame our understanding of biology by asking several key questions. The questions were derived/inspired from Tinberg’s four questions, which are four general directions one can take, ( (evolutionary, proximate or individual) X (ontogenic, mechanistic) ), when studying biological traits. The questions were mainly written with animal behavior in mind (Tinberg was an ethologist) but are broadly applicable to any biological concept. This generality, however, comes at the cost, and I feel that they fail to capture the necessary detail when examining our understanding of molecular mechanisms in biology. But then, which would be the questions we need to ask when we want to test our understanding of molecular machines and how they relate to higher order phenotypes? When searching the literature for hints of these questions, I was reminded of a set of ten precepts that the legendary Arthur Kornberg published in his article Ten Commandments: Lessons from the Enzymology of DNA Replication. Kornberg’s ten commandments are a mix of recommendations of empirical frameworks and tools in biochemistry that have consistently yielded insights time and time again as well as warnings of the limits of intuition and rationalization when dealing with entities (like enzymes) that we can only interrogate and probe under limited contexts. While they are not by themselves attempts at probing our knowledge of molecular mechanisms, they are the distilled thoughts of the origin of modern biochemistry and perfectly capture the zeitgeist of the molecular thinking that is still prevalent today. In a way, we can reinterpret Kornberg’s words and try to capture what he was thinking for each commandment, effectively casting them into questions which could help test the limits of our understanding of molecular mechanisms. This is what I attempt to do below. As you go through each of these questions you might notice that evolution takes somewhat of a back seat. I ended up noticing this when I finished writing the post, and it struck me that enzymology, the discipline of origin for undestanding molecular mechanisms, stands firmly on top of evolution but provides its own set of guides and rationale to examine biology at the molecular scale. In a way, while evolution provides the main pillars for understanding in biology, enzymology stands as the arcs that hold the roof as in the end enzymes are what give life its spark. As Kornberg puts it “time and again, spontaneous reactions, such as the melting of DNA and the folding of proteins, are found to be driven and directed by enzymes”. On to the questions then.

1. What is the minimal system that maintains function?

Commandment I: Rely on enzymology to clarify biologic questions

In the first commandment, Kornberg explains that the preferred way to study an enzyme is to extract it from the cell and probe it preferably on a cell-free system, so it may be tested under many controlled conditions. In other words, to reconstitute is to understand. Be it in vitro, in vivo, or in silico, the extent in which we can isolate and probe a biological system demarcates very well how we understand it– if the molecular mechanism of interest cannot be reconstituted, then the system is missing a piece, or the function that we think the system performs is a subset of a much bigger or multitude of roles. This is perhaps the core thought behind the idea of biological molecular mechanisms: it is reductionsim incarnate that hopes to understand the whole by the sum of its parts. While this has notion has been challenged several times (e.g. can a biologist fix a radio?), and whole fields like systems biology have sprung to provide more holistic views, when it comes to actually testing predictions, you can’t escape the minimalist/reductionist approach.

2 How ubiquitous is the mechanism?

Commandment II: Trust the universality of biochemistry and the power of microbiology

Kornberg writes somewhat in jest that “what’s true for E. coli is true for elephants, and what’s not true for E. coli is not true.” Some crucial mechanisms (like replication machinery) are ubiquitous and can therefore be studied in a variety of systems. The conclusions in those systems are more robust and general than in other niche mechanisms that are only present in some organisms. The utility of those findings also depends on ubiquity. For example in cancer, some mechanisms are subtype specific, while others might be can be a general cancer strategy, with important therapeutic implications.

3. How many stories fit the data?

Commandment III: Do not believe something because you can explain it

Perhaps the most hazardous mental trap when studying biological systems is that it is many times very easy to come up with explanations of why something works. Indeed, we can and do create lengthy stories that make complete sense to explain the data at hand. Kornberg recalls how he was prey to this great garden of forking explanations when he missed the discovery of RNA polymerase when he and his collaborators interpreted results from an ATP/ADP incorporation assay to be the result of a phosphorylase. The explanation made sense, but it was incorrect since they were unaware of hidden components active in the assay. The number of models that make biologically sensible models contrasted with the number of those models that can be further falsified provides a good measure of how well we understand the system. This ratio is present in all of science, but it’s particularly skewed towards large number of explanations in biology.

4. How reliable are the measurements from which we infer mechanism?

Commandment IV: Do not waste clean thinking on dirty enzymes

One of the most crucial questions in biochemistry when interpreting results is “how pure is your mixture?”. Whether it be whole cell extracts to multiple fractionations of one specific molecule, the quality of the conclusions and even the insights that can be hoped to obtain will depend on the answer to that question. Analogously in other biological settings, different qualities of meausurements will provide different resolutions of insights, be it analyzing bulk measurements compared to single cell or extrapolating conclusions from a generic cell line vs a specific one with an engineered genetic background. This is the realm of model systems and their inductive limits.

5. How reliable are the measurements which we use to test predictions?

Commandment V: Do not waste clean enzymes on dirty substrates

Related to the commandment and question above, after we are set with a mechanism and want to test it, it is sensible to ask if we can rely on the measurements to test our predictions While many times we could use the same experimental framework where we inferred the mechanism to make our predictions, as we test a model in increasing generality we will switch experimental approaches as well (use different measurement types [proteomics vs transcriptomics], use an in vivo model, etc.) Each of these approaches will have limitations that will narrow the conclusions that we can make when testing predictions. Kornberg relays this perfectly when testing new enzymes, if your substrate is dirty, the conclusions will be too. No matter how pristine your inferences are, when testing your molecular mechanism it’s garbage in garbage out.

6. What other minimal systems in nature also exhibit the function?

Commandment VI: Depend on viruses to open windows

Phages were an important tool in Kornberg’s work to understand how DNA and RNA were synthesized, primed, and replicated. After all, viruses are minimal replication machines: a small strand of nucleic acids surrounded by the minimal proteins to protect it and bootstrap their replication. If a function is biologically important, it will likely be extracted, repurposed, or duplicated elsewhere. These naturally occurring minimal systems provide smoking gun evidence for suspected mechanisms.

7. Under what biological conditions has a minimal system been tested?

Commandment VII: Correct for extract diluation with molecular crowding

One of the banes of (cell free) in vitro work in biochemistry is (the lack of) molecular crowding. In the cell, enzymes function under a very crowded environmnet where a myriad of other reactions are happening and local concentrations of substrates can be sky high. Kornberg recalls how he was stumped for 10 years attempting to replicate an intact chromosome…until he added PEG to the mixture, making it more ‘cell-like’. Similarly, the conclusions extracted from probing function in any minimal system will depend on how close the conditions are to the actual biological context. How similar is that cell line to the actual tumor? What naturally occurring molecular partners are missing in that single-molecule experiment?

8. Can we model the dynamics of the molecular mechanism?

Commandment VIII: Respect the personality of DNA

If there is one pattern of understanding that repeats itself over and over in molecular biology, it’s that everything is far more dynamic than we realized. As Kornberg puts it “DNA was regarded as a rigid rod devoid of personality and plasticity…Then we came to realize that the shape of DNA is dynamic in ways essential for its multiple functions…Especially noteworthy is breathing, the transient thermodynamic-driven opening (melting) of the duplex that facilitates the binding of specific proteins such as the helicase responsible for priming and the onset of replication.”. Proteins fold into multiple states to perform function. RNA twists and turns to activate/deactivate genes, and single-cell measurements have shown the intricate transcriptional paths that cells undergo from one cell state to the next, revealing a continuum orather than discrete cell states. In the most desired end, full understanding of a biological function will be engraved in the set of differential equations of its dynamics.

9. How is our mechanism altered or ablated genetically?

Commandment IX: Use reverse genetics and genomics

The ability to go from enzyme function to protein sequence to gene – to then probing the enzyme’s function by changing the underlying gene was something magical in Kornberg’s eyes. Indeed, being able to predict how genetic alterations in the system’s components affects function is an ultimate test of understanding it. Molecular tools have advanced considerably in the past decade and we now have CRISPR and friends to edit genes as we please to understand downstream function. Even in more challenging contexts, large-scale efforts to collate genetic and phenotypic data, ranging from thousands of genomic and metagenomic sequencing projects (metagenomes, microbiomes, you name it!), and clinical biobanks provide a solid foundation to test hypothesis on almost any organism that we know.

10. What new questions does our mechanism allow to us to answer?

Commandment X: Emply enzymes as unique reagents

In the last commandment, Kornberg lays out why modern biochemistry became so powerful: with each new enzyme discovered, new functions become possible, and avenues for discovering new enzymes open up. In a virtuous feedback loop, enzymes aid discovering other enzymes that yield new reactions to be repurposed for engineering. Understanding our mechanism in isolation is only part of the journey. Rather, grasping how it fits in the grand scheme of things, what other questions it allows us to answer, and better yet, how can it be repurposed, engineered, and tweaked is the true mark of the maturity of understanding. Interestingly, the questions extracted from the commandments are in a way ordered in the level of understanding: starting from the minimal knowledge required to even probe the system and ending with the trifecta that is the pinnacle of understanding molecular mechanisms: dynamics, genetics, and engineering.

Back to posts