Business

Saturday May 16, 2009

Unravelling the process


Azhar... This is the first big step

SIME Darby Bhd’s announcement that it had completed the sequencing of the oil palm genome took many people by surprise because before that, there had been no indication of the conglomerate’s interest in genomics.

According to the company, research into oil palm genome began in 2003 after it recruited biotechnologist and restructured Sime Darby Technology Centre Sdn Bhd (STDC) for biotechnology research.

That eventually led to Tuesday’s news – that Sime Darby has completely sequenced, assembled and annotated the oil palm genome with 30 times coverage (repeated sequencing over 30 times) and 93.8% completeness.

That was Greek to most people. Similarly, it is not easy to grasp the significance of the project. To get a better idea, StarBizWeek spoke to Sime Darby Plantation Sdn Bhd managing director Datuk Azhar Abdul Hamid and STDC senior vice-president II K. Harikrishna. Excerpts:

StarBizWeek: Let’s talk about the genesis of the project.

Azhar: Let me put the whole thing into perspective. What are the issues of the oil palm industry? One, in Malaysia, we have strong regulation over the opening up of agricultural land, especially plantations, because as a nation, we have already said we want to protect our forests. Another problem is our high dependence on foreign labour.

The other thing is productivity. Over the years, we have never seen productivity move by leaps and bounds. We really have to do something about this. So it goes back to point one. We’re not going to have more land. The Government have said, “Look, we have four million plus hectares, and that’s all we’re going to have.”

Harikrishna... A lot of work needs to be done

When we look at our business, we took in all these factors. We came up with this idea (of using genomics) in 2003. But nothing serious came out of it initially. But as we got closer to the merger, we also had the advantage of having more people who were giving ideas on how we can do it.

Then my team decided that we had to aggressively pursue this. It became extremely focused after the merger.

We were caught then. We heard that Asiatic (Development Bhd) wanted to do it and that MPOB (Malaysian Palm Oil Board) was trying to do it. So I consulted my boss (Sime Darby CEO Datuk Seri Ahmad Zubir Murshid).

He said, “Let other people make noise. You keep quiet and do what you need to do. When you finish, then you talk about it.” That’s exactly what we did.

At that point, in 2003, there was already the human genome project. Wasn’t it a no-brainer to look at the oil palm genome?

Harikrishna: From a technology and cost standpoint, there were obstacles. The human genome project cost US$2bil.

At the time, it would have cost us probably about RM200,000 to RM500,000 to sequence the whole oil palm genome. And it would have taken years, using the same technology used in the human genome project.

However, in 2006 and 2007, new technologies were developed. These were using 10 times less reagents and cost to sequence a base was coming down from a few US dollars to a few cents. Then it becomes exciting because the oil palm genome is very large.

But the new technology is not as good as the old technology in certain respects. The old technology produces long reads, and the new one produces these small fragments.

So the next question is, how do we make sense of these small fragments. If you can’t piece them together, all you have is a very expensive jigsaw puzzle.

Then we have to build up our human resources. This is a knowledge-based field and the kind of people we can attract are different from those we can attract in the US or Europe.

Who we could attract are young guys who are very bright and enthusiastic, but with no exposure and skills.

What are the technical and intellectual challenges of sequencing?

Harikrishna: First, you have to make a BAC library. BAC is bacteria artificial chromosome. This is the means by which we can keep huge pieces of DNA in living form. That’s one challenge. Very few people in the world can make these BAC libraries.

The next challenge: Because we’re using second-generation sequencing technology, we get these so-called short sequence reads compared with those from the conventional method. And our mandate was to get the job done fast and done well.

Given these short sequence reads, we had to figure out how to stitch them together. It’s a very, very complicated task because the oil palm genome has a lot of repeat sequences.

Our collaborator has come up with a different strategy, which will get published in Nature (an international science journal). The strategy is a completely different approach. Nobody has tried this before. It was tested out and Pavel Pevzner (a US-based professor of computer science) was asked to look at it and to give an opinion.

He said it would work but it must have a very high coverage. That was why we had 30x coverage. The human genome sequencing was only done with 8x coverage.

We got it done superfast – in 15 months, which is the fastest I’ve heard of. The approach was not to look at the huge picture and to try to figure out where the pieces went because we didn’t even know what the picture was.

Instead, we built it up in small pieces, each in relation to the next. So when we got these small pieces all over the place, they began to overlap and eventually we could stitch them together.

Azhar: The quality of the material is also very important. Among the major players, Sime Darby is known to be a producer of good planting material.

So the sequencing is done. What’s next?

Azhar: In the next phase, we identify what we want to do. Let’s say we want shorter palms, we bring together the team or whoever we want to collaborate with, and decide our objectives. Now we are able to know what we need to touch to influence the future growth of the palms, that will result in shorter palms. The next few phases will be very important.

We need to commercialise the data from the sequencing and that’s where collaboration comes into play. We want to collaborate with others. We want to make sure that the Malaysian industry benefits.

Having said that, would Sime Darby collaborate with others based on their data?

Azhar: We have to see what they bring to the table. For us, it’s just the beginning. Now I want an aggressive business plan to take this forward in the shortest time possible. This is all about dollars and cents. We’re not doing this as a hobby.

So do you already know the markers?

Harikrishna: To a certain extent. A lot more work needs to be done.

Azhar: For example, if you take the DNA of a high-yielding palm and compare it to what we have, you can see the difference. You take the palm with the characteristic you want, that you’ve already seen out there on the field, and you compare it to what we have, you’ll know what to touch in future planting material.

Harikrishna: You go into the field, where the palms are said to be high-yielding. The first thing you do is to check if they are really high-yielding.

Lucky for us, we have good records and good breeders. We go through the data to validate the yields. Then, the question is, why are the palms high-yielding?

Some others of similar parentage are low-yielding. So what’s special about the high yielders? So we need to do a comparison of the high-yielding and low-yielding palms with the same genetic background, that is, the same crosses. In order to know what is excellent, you need to compare it with something mediocre or poor. You need a reference point. And you need to have a population to interrogate and you need to ask the right questions.

How does the genome sequence fit into this exercise?

Harikrishna: The genome sequence is to help us identify many of the genes involved. So we’ve done transcriptomes.

If you say the genome is a library, the transcriptome is a book in the library. So you figure out which books have been taken out of the library. If you’re in Boston, which has over 20 universities. Using probabilities, you can work out that the books taken out are some form of academic books rather than novels.

If you are in, say, Kuala Lumpur, the likelihood is that the books taken out are novels maybe. In that manner, you can focus down on what is there.

We extract messenger RNA from the high-yielding and low-yielding palms and we compare this to the genome information to determine where and what they are.

Otherwise, they are just bits and pieces of sequence. We also look at the annotations to figure out what are the genes’ functions. We are basically building up a picture.

Some of the genes have no known functions and only exist in oil palm. We don’t know what they do. That’s where all the work comes in. And there are hundreds of these genes, not just two or three.

Azhar: This is a big first step. Now we are opening up to the possibilities and that will not come tomorrow or next week. It will take time and effort. Okay, we celebrated two days ago because it was a milestone. But the journey doesn’t end there. There’s a long way to go. People need to appreciate that.

We still need to pursue the real goal of making use of this asset. And it’s important for people to understand that there’s no point in creating too many of these databases.

Harikrishna: We should work together and move forward. You have a 10m platform and you want to go to 100m. Another person is building his own 10m platform. It’s a waste of time. We might as well work together and raise the existing 10m to 100m.

With other crops whose genome has been sequenced, have we seen the kind of leaps and bounds you’re talking about for oil palm?

Harikrishna: At the presentation on Tuesday, Dr Mani Subramaniam, (a professor of chemical and biochemical engineering) who comes from Iowa, in the US Corn Belt, said the corn industry there were now going for 300 bushels per hectare. The current yield is about 140-150 bushels.

Doubling the yield of a hybrid crop that has been bred intensively since 1950, where you can get four cycles per year – palm oil takes 10 years per cycle – is an enormous achievement. And it’s all done through biotechnology and knowledge of the genome.

How do you produce this high-yielding palms using the genomic data?

Harikrishna: The first thing you do is to validate the markers. We have to be sure that the markers correlate to the phenotype (an observable characteristic of an organism) that we want.

Then we have to figure out how many of the palms we can produce. What is the size of the parentage of the material, the mother palm? The likelihood is that you don’t have that many.

So what are you going to do? The management needs to decide then whether to clone them or propagate them through other means. From there, after amplifying the mother palm population, we do the crosses and produce the seeds. After that, we push them out into the field.

This is to tap into the spectrum of diversity. Today we have palms with many different features in terms of oil yields, oil quality, disease resistance and so on. They all look the same to us. Even the breeders have to spend many years collecting data before they can say one palm is better than another. When you have a series of markers, you can go through a population quickly and identify which ones are different from the others.

Azhar: There’s also luck. With the merger, we have more material to choose from. It’s a big advantage.

  • E-mail this story
  • Print this story