Professor of Proteomics

Professor Mark Baker has built a dynamic career in molecular cell proteomics and gives some insights into the Human Proteome Project, the thrills that proteins provide in life and why we need to stop and smell the roses.

Lab+Life Scientist: How did you become interested in studying science?

Professor Mark Baker: I was the eldest in a family of seven Maroubra kids and my father died of a heart attack when I was 12. From then on I was interested in trying to find out why people died early and what was the mechanism behind disease.

That has stayed with me ever since.

LLS: What inspired you to focus on proteins?

MB: I went to Macquarie University because it had molecular biology - I think it was the first place to teach it in Australia. I ended up doing an honours degree and a PhD there on proteins and free radical biochemistry with the ‘guru’, Professor Jan Gebicki.

Jan taught third-year biochemistry and I became very wrapped up in his course. We had to randomly pick an enzyme ‘out of the hat’ to purify and I picked Cu/Zn superoxide dismutase (SOD), which had only been discovered the year before by a young guy called Joe McCord.

After that I was well and truly into proteins and free radicals.

Subsequent postdocs gave me the opportunity to start looking for protein oxidation products in vivo and that was what I really got interested in - the pathology of disease and mapping free radical damage in pathological tissues: the signatures of disease.

I was pretty lucky to draw that SOD out of a hat.

LLS: Being ill led you to change direction in your research?

MB: Yes - I contracted Guillain-Barre syndrome after the flu, which left me paralysed for six months. While I was sick I read about Jessie Bradman, Donald Bradman’s wife, passing away from cancer. This inspired me to switch from colon and breast cancer research to ovarian cancer, primarily because of its lethality.

At the time I was working at the University of Wollongong and they couldn’t support the work because they didn’t have a medical school. So I left Wollongong and established an ovarian cancer research centre with Michael Quinn, a surgeon at the Royal Women’s Hospital (RWH) in Melbourne. I was the chief scientist and Michael was the clinical director - it was an amazing collaboration with a truly great bloke.

We had no money - the NHMRC wasn’t really interested in a low-impact disease - so we looked at alternative approaches to funding our research.

We went to philanthropists, we ran opera in the Queen Vic Markets with Opera Australia as a big fundraising event. It was a success and the centre grew into what became the Women’s Cancer Research Centre at the RWH.

All along I had watched proteomics with great interest because of my roots from Macquarie, but it really got going in Melbourne.

LLS: Is that when you started looking at the cell membrane proteome?

MB: It was then when we started thinking about the cell membrane proteome as a serious target.

Because the protein receptors that we were working on were membrane bound, the idea struck me that we should focus on doing cell membrane proteomics.

Identifying cell surface changes was a niche compared to looking at intracellular changes in cancer cells, but when we looked at the proteomics of membrane-bound proteins it was really difficult. Membrane-bound protein receptors are hydrophobic and in low abundance, and weren’t easily studied by 2D gel electrophoresis, which was the core technology in use at the time.

That’s when we switched to developing new technologies for analysing cell membrane proteins and the cancer surface proteome.

LLS: What made you switch to working in industry?

MB: While we were setting up ovarian cancer proteomics at RWH, I was offered a position in industry in San Francisco with a biomarker discovery company called Lumicyte. My UNSW-trained brothers had always pushed me to consider working seriously in industry.

So, I moved to San Francisco in 2000 and had three of the best years of my life.

At that time proteomics was going through a mindset technological shift and I got to work in a company that had hand-picked 50 of the best multidisciplinary minds to address problems around the use of MALDI mass spectrometry for biomarker discovery, including most cancers.

Lumicyte had 20 mass spectrometers in one room - it was the most any lab anywhere in the world had at that time. It was very cool to develop a chip-based platform with world-class ‘Silicon Valley’ big science informatics from scratch.

Whilst this next-gen technology was being developed at Lumicyte, the world was starting to adopt proteomics. John Yates in San Diego published his ideas about shotgun proteomics, which is where you take the whole proteome without separating it, digest the proteins up into peptides, separate them and then analyse them by mass spec.

We were starting to discover that these technologies could accurately and simultaneously look at hundreds to thousands of proteins rather than the old story of one or two at a time. And we (and many other colleagues) were beginning to think we could use these technologies to map the whole human proteome.

LLS: Did that work lead to the patents you hold?

MB: One of the patents came from initial work I did at RWH and that I shared with Greg Rice. Basically that idea got grandfathered with other more advanced biomarker discovery ideas I had at Lumicyte.

When I was at Lumicyte I had an idea that has led now to three patents. Very generously, the company released the technology to me and when I came back to Australia I patented it back at Macquarie.

The patents involve making more visible many of (… not all) the low-abundance proteins found in biofluids using solid-phase polyclonal chicken antibodies raised against chromatographically separated plasma to deplete the most abundant proteins - this allows us to reproducibly see proteins previously not seen in samples like human plasma, as our recent submission to PeptideAtlas demonstrates.

LLS: What brought you back to Australia?

MB: The biotech boom in 2003 was almost over in the States and things were getting pretty tough. Lumicyte looked like it was going to be sold - it did end up getting sold to Qiagen in Germany - and I saw an ad for the CEO position at the Australian Proteome Analysis Facility (APAF) back at Macquarie University.

I was also missing my two kids Matt and Tegan a lot, as well as the surf at Maroubra and watching the mighty Rabbitohs play - so I decided to bring the technology, business acumen, biomarker discovery skills and other loves back home to Australia and focus on making a significant difference to the emerging science of proteomics back home.

At that stage a number of key proteomics people had left APAF and Macquarie to have their shot at private industry through Proteome Systems Ltd.

APAF was in pretty bad shape, losing significantly every year, and was luckily refunded for national functional proteomics services in an expanded commercial format.

LLS: And you managed to turn APAF around?

MB: Yes. Rather than being an intellectual and curiosity driven research centre set up by Keith Williams, it became a fully fledged national service provision company reliant on modern high-throughput mass spec-based proteomics. And that’s when APAF really started to flourish.

It was a cooperative between the University of Sydney, UNSW, Macquarie University and TGR Biosciences in Adelaide. Those four organisations were what was called APAF Ltd up until NCRIS Bioplatforms Australia was launched, which I was closely involved in.

APAF Ltd was very much a forward-looking company. It had an independent, nationally focused board governing it - rather than any university - with an independent chairman, Geoff Grigg, who was an ex-CSIRO dual divisional head.

Australian researcher and industry costs were offset heavily, giving people access to world-leading services at a reasonable price.

APAF Ltd reinvested any profit back into buying brand new infrastructure - as mass spectrometers were turning over about every 4 to 5 years. This attracted new business and new researchers who wanted access to state-of-the-art technologies.

We also had world-class technicians, most of whom had PhDs.

LLS: Why did you leave APAF?

MB: In 2009, Macquarie decided to ‘roll’ APAF Ltd and its staff back into the university, cut previous affiliations and focus on other benefits.

They saw the recently won NCRIS investment as a grant. Knowing my preference for the commercial model I decided it was best to focus on other outcomes.

LLS: Do you think Australia is falling behind by not supporting researchers with up-to-date infrastructure?

MB: It’s really quite staggering the investment that is going into the proteomics field in Asia and some parts of Europe now. Undoubtedly, this means Australia will continue to fall behind.

It’s not good enough to have a 2- or 5-year plan, we need a 20-year coordinated plan in place regarding infrastructure investments, along with a commitment up front that R&D infrastructure in both human health and basic research is absolutely critical.

We have the opportunity and the will to work collaboratively as teams across the Australian sector - we are so small the only way to have a real impact globally is to work together.

We need big projects, big teams, big infrastructure and far more centres of excellence.

It’s frustrating - the maximum you can put in for a centre of excellence is $28 million over 7 years, but how can that compete with the Chinese who are putting in $100 million a year for the next 10 years?

If we want to find out how to treat a particular cancer or a particular disease, we need to understand that disease in greater detail and resolution than we currently do. We haven’t completed the human proteome so medicine still has a long way to go before we fully understand the organism that we call the human being.

We know most of the human genome now, but we only know about 70-75% of the proteome.

If you were a Ferrari car manufacturer and blocked out 25% of the parts list of the car and then tried to put one together only having/knowing 75% of the parts, it would be a pretty ugly Ferrari indeed - it’s the same with a human being.

LLS: Can you talk about your involvement with HUPO and the Human Proteome Project?

MB: After I left APAF Ltd I refocused my efforts around the Human Proteome Project (HPP) and became much more involved with the Human Proteome Organisation (HUPO).

I was in the Bay Area when HUPO started and have been involved with the organisation from day one.

HUPO was set up as the worldwide governing body overseeing three regions - the Americas, Europe/Africa and Asia/Oceania.

I was also involved in setting up the Asia/Oceania region and established the Australasian Proteomics Society (APS) with Richard Simpson in 2003. I am an invited speaker at this year’s Lorne APS Conference and plan to discuss how my team’s cancer cell membrane protein data shows that specific interactions differentially drive changes in colorectal cancer signalling.

In 2010, with Ian Smith, Ed Nice and Marc Wilkins, we ran the 9th HUPO World Congress here in Sydney, which is when HUPO launched the ambitious HPP.

The first thing HUPO agreed to do was to look at where we had strong evidence either at a mass spec or antigen-antibody level for protein-coding genes - because we hadn’t even mapped onto the proteome what proteins we knew anything about.

So HUPO decided to take a multipronged approach and assemble all the information in a big database to look at the human proteome from as many angles and perspectives as we could. As well as evidence for protein-coding genes, we looked at where the proteins were located, what they do, what they interact with, what modifications they have, what switches them on and off.

We also agreed to give every protein equal value and not just chase someone’s favourite protein/s which, by the way, for me is the urokinase receptor uPAR - a protein I’ve worked on for 25 years now. Instead each protein became just one of the 20,300 proteins - or far more if you consider all possible derivatives.

Over the last three years HUPO has taken this approach and identified the proteins for which there is really good evidence and the ones where there isn’t much evidence.

As part of these efforts, two recent Nature papers with draft maps of the proteome were published this year and a new version of the Human Protein Atlas was also released in November 2014 - all of which have been coordinated with HUPO’s standards and database efforts and were discussed at length at the recent HUPO Madrid Congress.

LLS: How do you get your head around the mind-bogglingly huge number and variation of proteins in the human proteome?

MB: This is hard - I think one can comprehend that proteins work in teams or social groups, a bit like us humans. If you understand the individual components and dynamics of the team - who works best with who and how they work together with what resources - then you understand how to optimise your team.

The current draft map of the human proteome is simply one representation of our combined current knowledge about human proteins using just mass spec and/or antibodies. It’s like saying here’s the baseline, now we can really start to understand protein variants (PTMs) and how the protein teams get put together. Collectively, some have termed this the interactome - how members of the proteome work together.

LLS: And it involves a lot of data management?

MB: Yes, with proteomics we are talking about simultaneously identifying and quantitating thousands of proteins and modifications - that’s why it’s called the big science-big data revolution.

We now have a vast array of available tools - mass spec-based quantitative assays, biochips and multiplexed ELISA and affinity platforms (some combining each) that can simultaneously measure hundreds of proteins instead of just a single individual protein.

LLS: Has the HPP provided any new insights into what we don’t know about humans yet?

MB: Yes - there are some big gaps in our knowledge.

We’ve found families of proteins that we don’t know very much about. For example, the olfactory receptors that Linda Buck and Richard Axel found and won the Nobel Prize in Physiology or Medicine in 2004 is the most ‘missing’ of all the protein families.

We think each of these receptors occur in very low abundance in only a few cells at a time, so it’s really hard to detect them by mass spec - one nerve cell expresses one olfactory receptor and is responsible for detecting and sending a signal to the brain about one chemical. Next door to that cell is a different neuron expressing a different olfactory receptor that picks up a different smell.

Research groups worldwide are now finding these olfactory receptors in other organs of the body, not just the nose, where they appear to be involved more broadly in chemosensation.

So we’ve realised we know very little about the olfactory receptors.

I suspect that this part of the human proteome is going to open up all sorts of amazing knowledge about human behaviour, emotions and how memory is implanted simultaneously with experience. For example, the memory of a family member or friend can be driven by a smell you associated with them - like walking past a bakery and smelling a particular type of bread can remind a person of their grandmother and make them cry when they smell that smell. Human history is full of stories just like that.

There’s over 500 olfactory receptor genes in humans, which is a small number compared to mice or elephants, who are more reliant on smell than we are - we are more reliant on sight.

The olfactory receptor also is a part of our genome that we appear to be losing the fastest in a long-term evolutionary sense. I think what we (and HUPO) need to do is smell the roses a bit more, so to speak, maybe to save our olfactory receptors.

LLS: You also have an interest in the truffle proteome?

MB: Funny - that’s a bit of a hobby. I’m also interested with other mates like Ed Nice, Ian Smith and James Whisstock at Monash (and I’m sure quite a few other scientists around the planet) in epicure and the science of great cuisine and wine.

It’s something a group of us did for the fun of it and it’s been a runaway success.

A French group and colleagues published the genome for the black Périgord truffle - so I thought about analysing the truffle proteome.

And of course you smell truffles with your olfactory receptors, and because we were interested in that proteome we wanted to work out what was the biochemistry of the truffle that led to the production of the famous truffle perfume.

Also, we were ahead of Europe because truffles grow in the other six months of the year in the Southern Hemisphere - so we had a six-month lead on them. And Australia is pretty fast at proteomics analysis.

So with Shoba Ranganathan’s marvellous bioinformatics analyses we published the Périgord truffle proteome in the Journal of Proteome Research, and recently, whilst visiting the south of France, we found that the French populist discovery magazine Sciences et Avenir had picked up the story!

That’s what you can do when you’ve got the technology - and we have it here in Australia - you can help an industry like the Australian truffle growers understand their product a lot better. Hopefully, this will help them find markers of authenticity in case there are cheap species brought onto the market - which does happen overseas - and we can develop an understanding of the biochemistry of an organism when previously it was a black box.

LLS: What do you see proteomics being applied to in the future?

MB: I think we are going to end up with something like a wellness health biochip, and I know that there are a number of people around the world working on just this concept now.

Maybe we’ll go to the doctor, have a blood test simultaneously measuring 30 or 40 different proteins on the chip and this will give your doctor an index of how well you are for someone your age, sex and genotype.

One thing we are realising in doing plasma proteomics is that as men and women age there are differences in their plasma proteomes. So, instead of there being a reference range for all humans, there should be ‘personalised’ reference ranges established for each decade of life, for each sex and maybe for different races on the planet.

We are going to come to a time not so far away where molecular pathology could simultaneously look at one hundred proteins at the same time on a tissue slice of a tumour, or any other disease, whereas previously using immunohistochemistry we would just identify one ‘marker’.

This will generate a molecular signature of disease and will allow treatments to be based on this particular molecular signature rather than what we currently do, which is treat cancer with reference to what organ it emanates from.

In the future we will have treatments approved for particular molecular signatures of cancers rather than because it happens to be in the brain or the gut or the breast or the skin.

LLS: Since you will take up the role as president of HUPO in 2015, do you have plans for the organisation’s future?

MB: Yes, I was elected president of the global organisation in 2014 and will take up this role in January 2015.

One thing moving forward for HUPO is to develop a set of success examples that very clearly explain to the public and funding agencies what proteomics has already delivered and what it is promising to deliver.

HUPO has been very much run on an academic basis and we need to make proteomics understandable by generating examples the average person can understand - like people are now starting to understand their genomes.

It’s about familiarity. We’ve sold the genome - actually we’ve probably oversold it because the public wrongly believes genes are the bits that actually do things in their body when in fact they are the coding instructions for life and it’s the proteins that do the work - do the living.

The thrill of driving a Ferrari is actually in driving it (the proteins), not looking at it on a computer as a blueprint (the genome). Like Ferraris, proteins are the thrill of life.