Imagine 18 billion pieces of DNA from the pine tree genome scattered through an empty room like jigsaw pieces.
Scion scientists were given the task of joining them together, analysing each one to find the joins with their neighbouring pieces and ultimately create an image of the pine tree genome, similar to that done for the human genome in 2003.
For the past two years that has been the focus of Dr Emily Telfer and her science team in Rotorua.
They hoped mapping the pine tree genome in the same way the human genome was mapped would help breed trees that were future-proofed for everything the environment and climate were likely to throw at them.
“It is surprising to many but the pine tree genome consists of 25 billion base pairs of DNA nucleotides (consisting of A, C, G and T), considerably more than the three billion in the human genome.
“The genome itself is in 18 billion pieces and our job has been to try to get the puzzle together.”
The volume of extra data was partly created by the sheer age of the conifer genome.
“In the DNA sense, dating back 300 million years you tend to pick up a fair bit more genetic baggage along the way.
“This includes hybridisation events, something plants tend to do a lot more of than us simple humans.”
That genetic horsepower was no more evident than the ability to take a cutting off a plant, put it in water and see it regenerate with a root system, a luxury not afforded humans.
Similarly, the genetic deck of cards had been shuffled a lot more along the DNA chain for pine trees.
Where grasses might have only one or two copies of a gene, conifers might have six copies, for example, of the cellulose synthase gene responsible for converting carbon to cellulose, the main building block of wood.
The team made its first effort to sequence the genome in 2015.
After an eight-month assembly period through a genomics computer they had completed only 10% of it, making the completion date of 2023 seem a lifetime away as technology and genomic understanding raced onwards.
“It required us to be more ingenious about how we approached the technology to break apart and put the gene together.
“We decided to stick to gene islands, those genes among all the genetic DNA that are coded to perform certain tasks.”
There were 50,000-60,000 genes in a conifer that could be in a genome.
Telfer likened DNA to a library containing books on how to run an organism, whether it was how to replicate, how to acquire chlorophyll or how to breath.
“But there is also a huge amount of information in the library we do not know what it does.”
Researchers had to drop the image of DNA neatly lined up and linear and view it as a ball of tangled wool, complete with the pieces of DNA whose purpose was uncertain but which were thought to help create a 3D architecture of DNA.
“Some of that DNA has lost its purpose now and has been viewed as junk DNA initially.
“But with this 3D understanding of DNA we do know this DNA is responsible for helping other genes line up for duplication and activation.”
Researchers now knew simply grabbing the genes they thought mattered and discarding the junk DNA would not deliver a full genomic map.
“For example, if pine trees differ, maybe it is because it is not by the genes, but by the way this DNA has arranged those genes.”
The scientists convinced Scion accountants to invest in a computer to have another go at assembling the data. The three terabyte of RAM and 250Tb processor assembled the data over 10 days.
“This was a huge milestone. Something that was going to take almost eight years was down to 10 days.”
Telfer likened the resulting 18 billion pieces of DNA to words they were building into sentences and ultimately paragraphs then chapters describing the genome.
“We have gone from 18 billion words to 53 million sentences. We now have to go on to the paragraphs.”
But it was too early to highlight any particular findings about tree variations though the scientists appreciated there was some urgency about their work.
“We need to get to the point where we can quickly sequence trees cheaply then we can look individually at each tree’s traits, say seeking trees for dry areas or trees with certain disease resistance,” she said.
The 25-year span between identifying a tree with specific traits and harvesting it needed to be shortened and a genomic map would allow the best trees to be selected for the best areas without that long interval and with more certainty.
“We have climate change happening faster than their generational phase and we have disease incursions that so far have been more nuisance than fatal but that could change and we would need to identify trees capable of enduring such incursions.”
