he next revolution in medicine just might come from a new lab technique that makes neurons sensitive to light. The technique, called optogenetics, is one of the biggest breakthroughs in neuroscience in decades. It has the potential to cure blindness, treat Parkinson’s disease, and relieve chronic pain. Moreover, it’s become widely used to probe the workings of animals’ brains in the lab, leading to breakthroughs in scientists’ understanding of things like sleep, addiction, and sensation.
So it’s not surprising that the two Americans hailed as inventors of optogenetics are rock stars in the science world. Karl Deisseroth at Stanford University and Ed Boyden at the Massachusetts Institute of Technology have collected tens of millions in grants and won millions in prize money in recent years. They’ve stocked their labs with the best equipment and the brightest minds. They’ve been lauded in the media and celebrated at conferences around the world. They’re considered all but certain to win a Nobel Prize.
There’s only one problem with this story:
It just may be that Zhuo-Hua Pan invented optogenetics first.
Even many neuroscientists have never heard of Pan.
Pan, 60, is a vision scientist at Wayne State University in Detroit who began his research career in his home country of China. He moved to the United States in the 1980s to pursue his PhD and never left. He wears wire-rimmed glasses over a broad nose framed by smile-lines in his cheeks. His colleagues describe him as a pure scientist: modest, dedicated, careful.
Pan was driven by a desire to cure blindness. In the early 2000s, he imagined that putting a light-sensitive protein into the eye could restore vision in the blind — compensating for the death of rods and cones by making other cells light-sensitive.
That was the germ of the idea of optogenetics — taking a protein that converts light into electrical activity and putting it into neurons. That way, scientists could shine light and stimulate the neurons remotely, allowing them to manipulate brain circuits. Others had experimented with trying to make neurons light-sensitive before, but those strategies hadn’t caught on because they lacked the right light-sensitive protein.
That all changed with the first molecular description of channelrhodopsin, published in 2003.
Channelrhodopsin, a protein made by green algae, responds to light by pumping ions into cells, which helps the algae search out sunlight.
That “was one of the most exciting things in my life,” Pan said. “I thought, wow! This is the molecule we are looking for. This is the light sensor we are looking for.”
By February 2004, he was trying channelrhodopsin out in ganglion cells — the neurons in our eyes that connect directly to the brain — that he had cultured in a dish. They became electrically active in response to light. Over the moon with excitement, Pan applied for a grant from the National Institutes of Health. The NIH awarded him $300,000, with the comment that his research was “quite an unprecedented, highly innovative proposal, bordering on the unknown.”
Pan didn’t know it at the time but he was racing against research groups across the United States and around the world to put channelrhodopsin into neurons.
Deisseroth and Boyden were working at Stanford, where Deisseroth was finishing a postdoc and Boyden was finishing graduate school. At least two other groups were in the game as well, led by Stefan Herlitze and Lynn Landmesser, who were at Case Western Reserve University at the time, and Hiromu Yawo at Tohoku University in Japan.
And they were by no means the only scientists experimenting with ways to control neurons with light. By 2004, Gero Miesenbock and Richard Kramer had already published articles using other, more complicated molecules for that purpose. But channelrhodopsin was the tool that was about to revolutionize the field.
The Stanford group had been toying with the idea of controlling neurons with light for quite some time. They had also noticed the paper about the discovery of channelrhodopsin. Deisseroth got in touch with the paper’s author, Georg Nagel, in March 2004 and asked if Nagel would collaborate, sharing the channelrhodopsin DNA so Boyden could try it out in neurons. In August 2004, Boyden shined light on a brain neuron in a dish and recorded electrical activity from the channelrhodopsin.
Pan had done the same thing with retina neurons six months earlier. But then he got scooped.
‘We didn’t feel very lucky’
Boyden, who is now a professor at MIT, was surprised when told by STAT that Pan ran the experiment first.
“Wow. Interesting. I didn’t know that,” Boyden said.
“It’s funny to think about how science regards when something is proven,” he added, noting that scientists build on each others’ work, sometimes working together while at other times working in parallel, scrambling onto one another’s shoulders. “There’s both intentional and unintentional teamwork,” he said.
The Stanford press office said Deisseroth was unavailable. In response to questions provided by STAT, spokesman Bruce Goldman wrote that Pan’s study was “a far cry from the use of optogenetics … to open up a new world of precision neuroscience. That’s the potential revealed in Dr. Deisseroth’s widely cited 2005 publication.”
Pan said he might have mentioned the timing of his experiment to Boyden once several years ago, but, Pan said, “I didn’t want to take too much time to talk about this because people feel uncomfortable.”
That sentiment is in keeping with Pan’s wider approach — diligent, reserved, outside the limelight. Wayne State is a small university not known for its scientific research. Pan had gone to a state school for his PhD, then done mostly obscure research for decades. These things may have contributed to what happened next, when he tried to get his invention out into the world: It wasn’t seen as the big advance it was.
Pan spent the summer of 2004 figuring out how to get the channelrhodopsin protein into a living eye. He settled on the idea of using a virus, which could infect cells in the eye and sneak the channelrhodopsin DNA inside. His colleague, Alexander Dizhoor, a professor at Salus University, engineered the channelrhodopsin DNA to add the gene for a protein that fluoresced green under blue light, so they could track where the channelrhodopsin ended up.
In July 2004, Pan dosed his first rat with the virus. About five weeks later, he looked at the retinas to see if it had worked. What he saw was a sea of green — thousands of ganglion cells had the green protein coupled to channelrhodopsin in their membranes. And when he stuck an electrode in one of those cells and turned on a lamp, the cell responded with a flurry of electrical activity. The channelrhodopsin was working. It was just a first step, but it was a revolutionary step — indicating that Pan’s method may just be able to restore sight to the blind.
“Everything turned out beautifully,” Pan said.
So Pan and Dizhoor wrote a paper about their work and submitted it to Nature on November 25, 2004, according to the submission letter Pan shared with STAT. The editors at Nature suggested they send it on to a more specialized journal called Nature Neuroscience, which rejected it. Early the next year, Pan sent the paper to the Journal of Neuroscience, where it was reviewed but then again rejected.
Disheartened, Pan set to work revising his paper, and in May 2005 traveled to Fort Lauderdale, Fla. for the Association for Research in Vision and Opthamology conference, where he described his work using channelrhodopsin in neurons. That single lecture, lasting just 15 minutes, would come to be his clearest stake along the timeline of invention.
It was what came next that would make that stake matter. A few months later, in August of 2005, Nature Neuroscience published a paper about using channelrhodopsin to make neurons sensitive to light. The paper was by Edward Boyden and Karl Deisseroth.
Pan heard the news from a colleague who emailed him the paper. “I felt terrible. I felt terrible,” Pan said, pausing. “We didn’t feel very lucky.”
Met with a shrug
Deisseroth and Boyden’s paper was slightly different than Pan’s. They simply demonstrated that they could use channelrhodopsin to control neurons’ activity in a dish; Pan had waited to publish until he could make it work in a live animal. And Deisseroth and Boyden had shown incredibly precise time control, by turning the light on for just a millisecond. But their technical feat was essentially the same: They had used channelrhodopsin to successfully make neurons in a dish respond to illumination.
The Stanford paper took a little while to take off, but take off it did. The work jump-started both Deisseroth’s and Boyden’s careers, landing them big money grants and talented students for their labs — Deisseroth at Stanford and Boyden at MIT. The New York Times started writing about Deisseroth’s breakthroughs with optogenetics in 2007, and the citations of the research paper took off exponentially.
By the time Pan finally managed to publish his paper, in Neuron in April 2006, it was mostly met with a shrug. Richard Kramer, a neuroscientist at UC Berkeley who was also studying vision, remembers, “It wasn’t that creative, it was just ‘Oh look, you can put channelrhodopsin in neurons from the brain, you can also put it in neurons from the retina.’ Was it impressive? No.”
Those handful of months seem to have made all the difference.
Why didn’t Pan’s paper get published first? He may never know the answer. After Boyden’s paper came out, Pan wrote to the editor at Nature Neuroscience asking how they could have rejected his paper but published Boyden’s.
In her response, the editor replied that while the papers were similar, Boyden et al. presented theirs as a new technology rather than as a scientific finding. Pan’s paper, it seemed, was too narrow, only focusing on using channelrhodopsin to restore vision, while Boyden’s paper took the broad view of thinking of channelrhodopsin as a tool for neuroscience in general.
The reviews that other researchers submitted to the Journal of Neuroscience shed some more light on what people thought of Pan’s paper. One reviewer liked it and had some minor suggestions for improvement. The other, in a single long paragraph, said the research was “ambitious” and “very preliminary” and concluded that “there is too little here to entice most neuroscientists.”
In hindsight, Pan’s coauthor Dizhoor can’t help but laugh while reading that. Reviewers would ultimately greenlight an expanded version of Pan’s paper, in 2006, with minimal revisions.
But that hasn’t elevated Pan to the optogenetics pantheon. In terms of publication, he was quite late to the party, with three different groups publishing papers about channelrhodopsin before he did. He didn’t share in two big prizes that recently went to Deisseroth and Boyden, the Brain Prize in 2013 (1 million euros split between six inventors of optogenetics) and the Breakthrough Prize in 2015 ($3 million each to Boyden and Deisseroth).
Since 2005, Deisseroth has been awarded over $18 million in NIH grants for his work on optogenetics, and Boyden has received more than $10 million. Both have other major projects that bring in additional funding to their labs each year. Boyden is a prolific speaker who’s given multiple TED talks; Deisseroth was the subject of an in-depth profile in the New Yorker in 2015.
Pan, on the other hand, has cumulatively received just over $3 million over the past 10 years and holds one NIH grant — the bare minimum to keep a research program going. Most of the accolades for his work have come from Wayne State University. According to his website, he’s been invited to give a couple of talks — most recently at a technology show in Russia.
Rules of the invention game
The whole saga raises the question of what it means to invent something in science. It’s a question that has plagued scientists in recent years — including the ongoing CRISPR patent fight — as research becomes ever more global and the spoils of biotechnology and medical discoveries become ever more valuable.
The answer, it turns out, shifts depending on context.
Fellow academics often consider the first scientists to publish a paper on a technique the discoverers or inventors of that technique.
But that metric can be problematic, as Pan’s experience shows. In a recent essay in the journal eLife, Ronald Vale and Anthony Hyman, two biologists, laid out the problem. They point out that “the delay between the submission of a paper and its publication can range from a few weeks to more than two years,” adding that journals “slow down and create inequities in how knowledge is transferred from the scientist to the worldwide scientific community.”
And reviewers can be biased toward familiar names or prestigious institutions. Blinded review, in which the author’s name is redacted, has been suggested as a way to minimize that effect, but many scientists are skeptical that it would work, since research is often discussed ahead of time at conferences.
Vale and Hyman advocate, instead, for scientists to post drafts of their work on “preprint servers” such as bioRxiv before they submit it to journals. If such a server had been widely used by neuroscientists in 2004, Pan could have posted his rejected findings there, staking his claim.
But whether that would mean he would be on the short list for the Nobel Prize is unclear. Kramer thinks that even if Pan had published on bioRxiv, he’d be shut out because he wasn’t the first to publish a peer-reviewed paper on the technique. That’s what will matter if and when the inventors of optogenetics win the Nobel.
The legal system doesn’t play by quite the same rules. According to an American Bar Association representative specializing in patent law, to prove precedence for a patent in the early 2000s, most of the time you needed to show both “when someone had actually conceived of the invention — that’s sort of in your mind the lightbulb going off, ‘Aha! I have it!’ — and when the invention was reduced to practice — that means you’ve actually done it and you’ve proven that your idea can work.”
By those standards, a discovery happens at the time of its demonstration in the lab, even before it’s been posted on a preprint server.
Then there’s the court of public opinion. Scientists are increasingly public personalities, running Twitter accounts and appearing on late-night talk shows.
“The quality rising to the top is a little more influenced by non-scientific things than it used to be,” said Richard Masland, a professor at Harvard Medical School, who also holds patents on gene therapy for blindness.
Being at Wayne State University might have meant that Pan didn’t have the resources to get a high-profile paper published. There’s the actual costs of doing high quality of research, but in addition, senior researchers at top universities usually mentor junior professors, reading their work and helping them take it to the next level.
Pan agrees that fact may have put him at a disadvantage compared with scientists at prestigious institutions like MIT or Stanford. “Of course, I cannot prove that with evidence,” he said. And Pan’s modesty and non-native language abilities may have kept him from promoting himself as well as Boyden and Deisseroth did.
“He’s just not as public a speaker and presenter as other people in the field. And this is an important part of the whole game of being able to get out there and sell yourself,” Kramer, the UC Berkeley vision researcher, said.
That publicity can be self-reinforcing. Landmesser, the Case Western professor who worked on channelrhodopsin in the beginning, said, “I think there’s always a tendency [that] whoever gets there first gets more publicity, let’s put it that way.”
A university PR video can spawn a national news article, which spurs someone to think of your name in nominations for a nice cash prize, which leads to some TV appearances. The word “inventor” gets used at some point and before you know it you’re Google’s automatic answer to the question “Who invented optogenetics?”
Ultimately, both Pan and the team of Boyden and Deisseroth won patents for their discoveries.
Pan’s May 2005 lecture threatened to derail the Boyden-Deisseroth patent for a while — the US patent office rejected it multiple times because Pan’s abstract was published more than a year before they got around to filing.
Eventually, Deisseroth and Boyden signed a document stating that they had invented this method of using channelrhodopsin privately in the lab before Pan’s conference abstract was published. The relevant patent was issued in March 2016, almost 10 years after they filed.
Now, Deisseroth is a cofounder and scientific advisor at Circuit Therapeutics, a company developing a wide range of therapies based on optogenetics, presumably using Deisseroth’s patented inventions. (Circuit Therapeutics declined to comment on specifics of their intellectual property licenses.)
Pan won a patent as well, to use channelrhodopsin to restore vision in the eye. His patent was licensed by RetroSense, which won an award from the Angel Capital Association in 2015. Retrosense — whose CEO in passing told STAT about Pan’s role in the invention of optogenetics — began clinical trials this year to put the algae proteins in blind people using gene therapy. It’s the first application of optogenetics in humans and the first time a non-human gene is being used in a gene therapy trial.
Right now, there are blind people in Texas walking around with algae DNA and proteins in their eyes. And that was what Pan was in it for all along. “One thing I still feel glad about is that even right now our clinical study is still ahead of anyone,” Pan said.
But given that there are no gene therapies approved for clinical use in the United States, the road to successfully using optogenetics in humans will likely be a long one. Yang Dan, a professor of neuroscience at UC Berkeley who uses optogenetics to study sleep, isn’t betting on optogenetics cures being in the clinic any time soon. “I believe that these safety checks will take a long, long time,” she said.
As for the invention itself, some scientists say Pan may not have had the big, award-worthy vision that Deisseroth and Boyden had. Stefan Herlitze, one of the others who was scooped for the first publication about channelrhodopsin in neurons, said, “Of course I have to say, Deisseroth and Boyden, they really developed the field further.”
Boyden echoed this. “Karl and I were very interested in the general question of how to control cell types in the brain,” he said. “In recent years, we worked to push these molecules to their logical limits.”
So maybe it doesn’t matter who invented optogenetics, just who has stretched science’s boundaries the furthest.
Asked whether he deserves the recognition that Boyden and Deisseroth have enjoyed, Pan declined to answer. He later told STAT that Deisseroth “also did a very excellent job, no doubt. But he’s also very lucky because if our paper was ahead of him, the story would be different. We would have gotten more credit.”
That is about as much as Pan is willing to say about the way his cards fell. Today he’s still in Detroit. He’s been working on new versions of channelrhodopsin that could be used to cure blindness. “My lab is a very small lab,” Pan said, “We’re mainly interested in trying to restore vision.”