There are around a trillion species of microorganisms on Earth, the great majority of which are bacteria.
Bacteria are composed of a single cell. They lack bones and are not like large animals that leave unambiguous traces in the geological record for paleontologists to investigate millions of years later.
Because of this, scientists have had an extremely difficult time creating a timeline of their early evolution. However, machine learning has allowed us to fill in a lot of the blanks. Additionally, our new study, which was published today in Science, shows that some bacteria acquired the capacity to use oxygen some 2.4 billion years before the Earth became saturated with it.
A significant event in Earth’s history
The moon created about 4.5 billion years ago. violently. Earth’s surface turned to molten rock after it collided with an object the size of Mars. Life was most likely wiped out if it existed prior to this apocalypse.
Then single-celled microorganisms, the present ancestors of all living things, emerged. These bacteria were the only life on Earth throughout the first 80% of life’s history.
Theodosius Dobzhansky, an evolutionary scientist, famously stated in 1973 that nothing in biology makes sense unless it is understood in the context of evolution. But what was the course of life’s evolution during the Earth’s early history?
We can learn about the relationships between various groups by comparing DNA sequences from the amazing diversity of life that exists today. For example, compared to apple trees, humans are more closely linked to mushrooms. Similar comparisons can also reveal the relationships between various bacterial groupings.
However, comparing DNA sequences can only go us so far. When evolutionary events occurred on Earth is not determined by DNA comparisons. There once was an organism that produced two offspring. One produced humans (as well as many other species), and the other produced mushrooms. However, when was that organism alive? How long ago was it?
The fact that there was another significant event in Earth’s history 2.4 billion years ago is one thing geology teaches us. At that time, Earth’s atmosphere underwent a significant transformation. Photosynthesis was a trick developed by a group of bacteria known as the cyanobacteria that would permanently change the course of existence.
Their cells were fuelled by energy harvested from the sun. However, oxygen gas, an inconvenient waste product, was also produced.
Oxygen gradually built up in the atmosphere over millions of years. Almost no oxygen was present on Earth before to this “Great Oxidation Event,” thus life was not prepared for it. When oxygen is released into the atmosphere, it likely caused a catastrophic extinction because, to uninitiated microbes, it is a toxic gas. The microorganisms that survived either adapted to use oxygen or moved into areas of the globe where it is inaccessible.
The bacterial tree of life
We are particularly interested in the Great Oxidation Event since it can be precisely pinpointed in addition to its significance in the history of life. We know that it occurred approximately 2.4 billion years ago, and that the majority of bacteria that developed oxygen adaptations had to survive beyond this time. Using this data, we added dates to the bacterial tree of life.
We began by developing an artificial intelligence (AI) model to predict whether a bacteria thrives with or without oxygen based on its genes. Many microorganisms that we see now, including cyanobacteria and those found in the ocean, require oxygen. However, many do not, including the bacteria that dwell in our gut.
This was a relatively simple machine learning exercise. The chemical power of oxygen significantly alters a bacteria’s genome because a cell’s metabolism gets organized around oxygen utilization, and so there are several signals in the data.
We then used our machine-learning models to forecast which bacteria used oxygen in the past. This was made possible by current tools that allow us to determine not just how the species we observe today are linked, but also which genes each ancestor carried in its DNA.
A surprise twist
Our method provided a thorough timeline of bacterial evolution by effectively employing the Great Oxidation Event as a “fossil” calibration point on a global scale.
By integrating findings from phylogenetics, geology, paleontology, and machine learning, we were able to greatly improve the chronology of bacterial evolution.
Another unexpected finding from our research was that there were bacterial lineages that could use oxygen about 900 million years prior to the Great Oxidation Event. This implies that even in situations where atmospheric oxygen was limited, these bacteria developed the capacity to use oxygen.
Our research revealed that cyanobacteria actually evolved the capacity to consume oxygen prior to the development of photosynthesis.
In addition to reshaping our knowledge of the evolutionary history of bacteria, this paradigm shows how the capacities of life changed in response to the shifting circumstances on Earth.
