Success and rejection

Reflecting in the spring of 2005 on his lab's recent successes, which he regarded as a culmination of decades of work, UC Irvine neuroscientist Gary Lynch said: "This will be a moment when all the tribes of neuroscience come to the same campfire."

He was wrong. There was no reaction. Nothing. Initially, he couldn't even get a short paper on a crucial visualization experiment published. Lynch envisioned the experiment as a grand confirmation of his notion that a change in the physical structure of brain cells at the connections between them was responsible for the encoding and persistence of memory.

It had taken 20 years to acquire the tools to execute, and when Eniko Kramar, a senior scientist in Lynch's lab, produced a series of spectacular microscopic photographs depicting where and how the change occurred, Lynch awaited the triumphal acclamation of the lab's success.

The tribes were not at the same campfire. Many apparently hadn't yet learned that fire had been discovered.

When a paper is submitted to a scientific journal, the journal editors send it for review to panels of scientists. Peer review is the backbone of contemporary scientific legitimacy and lauded by everyone involved. It is also an opportunity for mischief and misunderstanding.

Lynch's history of antagonizing his peers sometimes made peer review more a gantlet than a critique. Richard Thompson of USC, a renowned neuropsychologist, said he had more than once nominated Lynch to membership in the prestigious National Academy of Sciences, but was told by other members Lynch would not be elected so long as they lived.

"There's a reason for his paranoia. There are a lot of people out there who don't like him. Gary doesn't suffer fools gladly," Thompson said, then paused for a moment. He chuckled and said: "And there are a lot of fools in the world."

The reviews on Kramar's paper seemed not to even acknowledge its main point -- that the lab had for the first time demonstrated the physical reorganization of cells that occurred in the final stage of long-term potentiation, or LTP, which Lynch believed was the biochemical process underlying memory.

One reviewer, in recommending against publication, complained that the scientists had only looked at a specific set of synapses, which was inexplicable as criticism. They looked there because that's where they were doing the experiment, that was where the condition they were examining existed. It was as if a traffic engineer, having proposed adding carpool lanes to the San Diego Freeway, was asked why he hadn't examined four-way stop signs in Barstow.

Lynch was irate, and for a couple of days everybody avoided him. Then one morning, he was at his desk, smirking like a boy with the key to the cookie jar.

What happened? I asked.

"I can't tell you," he said, his grin growing.

But, of course, it was Lynch; he had to tell. He pointed at his computer monitor on which was displayed information on a company called Memory Pharmaceuticals, founded by Nobel laureate Eric Kandel, and a competitor of Lynch's biotech company, Cortex Pharmaceuticals.

"I'm shorting Eric's stock," Lynch said and cackled.

Kandel was the god king of contemporary neuroscience. He won the Nobel Prize in 2000 for investigations of synaptic activity that occurred during reflex learning in sea snails. He had also almost single-handedly made the study of protein synthesis a major focus of neuroscience.

Lynch thought the emphasis was wrongheaded, but there was little he could do. Kandel, for his part, had nothing but nice things to say about Lynch. He barely acknowledged they were competitors, despite the fact that the two had led opposing armies in a 1990s war over where at the synapse the crucial actions of LTP occurred.

Lynch turned out to be right and won that battle, but he lost the war. Kandel received the Nobel Prize; Lynch went to ground.

For reasons not entirely clear even to Lynch, he retreated to his Irvine lab. He focused on his research, continued to publish voluminously, but largely absented himself from the numerous academic conferences and symposiums at which neuroscience findings were presented and debated and, not insignificantly, reputations made and maintained. He declined to meet with visiting researchers and fought with administrators and colleagues.

"It got to be very hard for me to keep playing with the boys," he said.

Few things are more punishing to an ambitious man than to be right and unappreciated. The latest fight over the Kramar paper was but one more slight. In the end, Lynch reacted as he had before: He complained bitterly then went back to work.

One of the things he devoted time to was the development of a family of drugs, called ampakines, intended to enhance LTP.

Ampakines had followed a tortuous path, but finally in the spring of 2005, Cortex, the small biotech firm in Irvine that licensed them, had a viable drug candidate ready for federally mandated human trials. The drug, called CX717, had sailed through Phase I safety trials, and Cortex was now applying to the Food and Drug Administration to conduct separate Phase II trials for its effects on sleep deprivation, Alzheimer's disease and attention-deficit hyperactivity disorder (ADHD).

The trials could be do-or-die events for Cortex.

That CX717 was to be tested in such a variety of diseases was in part a business strategy to give Cortex as many chances at success as possible. The strategy also reflected Lynch's belief that LTP was a fundamental brain process. Whatever the cause of many neurological diseases, poor communication between neurons was almost always one of the results.

The first CX717 results from a small sleep-deprivation trial came back in May. They were better than could have been hoped. The only drawback to the trial was its size -- just 16 men, who were deprived of a night's sleep, then given the drug and tested. Without the drug and without sleep, their test scores fell off the chart. With the drug and without sleep, they tested the same as they had when well-rested, and had none of the jitters commonly associated with stimulants.

It was a clear win for the company. Its stock sank.

"Try to lift the species out of the puddle of its own crap and what do you get?" Lynch joked. The other trials had yet to get underway, and there was too much going on to brood. The lab's incredible run of success in nailing down details of the biochemical processes underlying memory and forgetfulness, begun that March, continued through the spring and summer.

Lynch began experimenting with more potent versions of the ampakines. Julie Lauterborn, working in the lab of Christine Gall, Lynch's longtime significant other and collaborator, had earlier discovered that some ampakine variants increased production within the brain of compounds known as neurotrophins.

These molecules, in particular one known as brain-derived neurotrophic factor, or BDNF, were essential to the maintenance of brain function. So many claims had been made for BDNF over the years that Lynch tended to disregard them. He mocked it, referring to BDNF as the "big-deal growth factor."

"That's the story -- everything that's wrong with the brain, BDNF will fix it," Lynch said. "To me, it was ludicrous."

But when Kramar in his lab confirmed other reports that BDNF played a crucial role in the LTP process, Lynch began to examine the interaction between ampakines and BDNF.

He was astonished to learn that particular ampakines could be used almost as a switch to turn on BDNF production and thereby boost LTP.

There had been a long history of attempts to somehow get more BDNF delivered to the brain. A company in San Diego had gone so far as to drill holes in heads and pump BDNF directly in. To have found a simple, apparently painless and yet powerful means to turn on BDNF production would be like discovering a magic potion.

Lynch was convinced that many neurological diseases -- Alzheimer's, Huntington's, Parkinson's -- were in part caused by the normal wear and tear that accompanied aging.

Brain cells, unlike most of the cells in the body -- or most of the cells in most of living creatures in all the known world -- were more or less permanent. They did not die and get replaced by new cells. They lived, for a century if their host did, and accumulated all the damage anything that old might expect.

The combination of aging and specific diseases, some of them genetic in origin, led to mental difficulties -- memory loss among them, Lynch thought. Ampakines were supposed to help ameliorate many brain diseases by bulldozing through the problems the diseases created, compensating for aging. The drugs wouldn't cure the diseases, but would relieve the most debilitating symptoms.

Tools of the trade

The direction of science is largely determined by the tools available to pursue it. Problems arise when a tool dictates direction -- an illustration of the axiom that if you only have a hammer, every problem looks like a nail.

Grant-makers and journal editors demand that you use the popular tools to gain their approval. Tenure committees make decisions based on grants and publications. In very short order, research becomes normative. Interlopers are shunned, and risk-taking is constrained.

"If you've got a method that lets you look at something, then the answer must be what that method allows you to see," Lynch said.

In Lynch's view, an unsightly number of neuroscientists have been swinging hammers at a problem -- memory -- that didn't look to him like a nail. The field was enthralled by the tools of molecular biology and their ability to manipulate genes in experimental animals.

It was routine to reverse-engineer laboratory mice or rats -- knocking genes in or out of the animals -- so that they had or lacked certain qualities. In this way, the animals could mimic specific disease states. There were rats with Huntington's disease, Alzheimer's, Parkinson's.

One problem in using the animals for neuroscience was the complexity of the human brain, which in many of its actions was redundant. If one gene was knocked out, eliminating the protein that gene manufactured, other genes might make compensating proteins. It was often impossible to delineate precisely what caused what.

"They're nowhere near knowing what the machine is, so they can't know what the machine produces," Lynch said. "It's like a 747 crash-landed in the jungle. The monkeys are crawling all over it, having a hell of a time trying to figure out what it is."

Lynch realized he suddenly possessed new tools of his own: Whatever the fate of the unpublished Kramar paper (which was eventually published), the method it described to visualize the late stage of LTP was an important new laboratory tool; and the ampakines themselves were a tool that could probe the inner workings of LTP.

The outline of Lynch's LTP hypothesis was this: When you experienced a sensation in the outside world -- seeing, smelling or touching something -- the sensation was translated by the sensory organs into an electrical signal that was routed to the brain, where it caused the brain cells, or neurons, that received the stimulus to release chemicals to neighboring neurons. A cascade of chemical events inside those neighboring neurons resulted in their interior reorganization. That reorganization strengthened the connection between cells at the points where they meet, called the synapses. Networks of those neurons with strengthened connections constituted the underpinning of memory.

In a normal LTP experiment, a slice of a rat's brain was subjected to a precisely timed and measured electrical stimulus, mimicking the electrical signal produced by a real-world sensory stimulus. The experimenter measured the strength of the electric signal as it traveled through the slice. If a larger-than-usual signal exited the slice, that meant LTP had occurred and connections between the neurons in the slice had been strengthened. All you really needed for these experiments were a microscope, a chemical catalog and a pair of electrodes. That didn't mean they were easy, just that the tools to do them were straightforward.

Using her visualization method and the ampakines, Kramar examined the BDNF-LTP interaction. Over the course of months, what emerged was a picture that was at once immensely complicated and impossibly elegant.

"Endogenous BDNF does everything they said. It's all true. It's all true," Lynch said. "But I was too lazy to read the papers carefully."

BDNF, Lynch now thought, was crucial to the physical restructuring inside a neuron during LTP. In essence, it was on the "on" switch. Lynch thought another molecule, adenosine, was the "off" switch. A fine balance of the two was needed for the brain to work. Through the fall, Lynch and company elaborated on this line of thinking.

Chris Rex, a grad student who had discovered LTP deficits in middle-aged rats, set up an experiment to see what would happen if he used ampakines to instigate BDNF production in the same rats.

It worked. The ampakine turned on the BDNF, and the BDNF promoted LTP. The age-related deficit disappeared. As Lynch put it later: "Middle-aged aging cured."

Lynch, at Gall's persistent urging, began to reintegrate himself into the wider world of science. He accepted invitations to speak at a few conferences.

The progress was not without drama, some of it self-inflicted. Lynch continued his war with the university administration. Finally, feeling he wasn't being treated with sufficient respect, he shut down his lab at 101 Theory Drive, dispersing his researchers to Gall's lab. Some of the scientists were by then -- or would soon be -- gone for good. Ted Yanagihara went to med school in New York. Laura Colgin left on a postdoc to Europe. The lab's computer expert left for private industry.

Lynch's health teetered between bad and calamitous. He showed up at a conference in Vancouver fevered and shot full of antibiotics. He gave his lecture, listened to others and made small talk, all the while looking to be on the verge of collapse. "Just to have one damned thing that works," he said.

His various maladies never seemed to go away. The problems he'd had with balance didn't completely resolve. His neurologist held to the initial diagnosis -- a viral infection. Lynch suspected something more serious but never pursued another diagnosis. "I put that into the 'Forget About It' file," he said. "Don't have time to worry about it."

Through it all, he continued to publish. In a universe where most papers are written in a combination of dense chemical symbolism and genre jargon, Lynch's work stood out for its sometimes whimsical, often literary tone. Some examples: "Consolidation: A View From the Synapse," "Long-Term Potentiation in the Eocene," "Spandrels of the Night?" and "Ampakines and the Three-Fold Path to Cognitive Enhancement."

Not that this endeared Lynch to everyone in the field. In addition to the Kramar paper, he began to have others rejected at a rate he had never experienced. Said Lynch: "Uneducated reviewers to the left, pygmies to right, but on came the army of science."

The army's accumulating evidence was producing a rich portrait of the LTP process that seemed to Lynch to have far-reaching implications. LTP seemed to be a fundamental brain process, perhaps the fundamental brain process.

The end of the artifact

A

disquieting aspect of LTP research had long been that Lynch, his colleagues and thousands of other scientists had devoted decades of research to it, attempting to describe its details, yet all the while there was no assurance it had anything to do with memory. They hoped it did. Some believed it did. None of them knew.

Kandel, the godfather himself, said he was far from convinced LTP had real-world significance. Lynch, in the early years of LTP research, seldom used the word memory to describe its relevance. He said it was a presumed "substrate of behavioral plasticity," a phrase nearly perfect in its obfuscation.

The great final task for Lynch would be linking LTP unequivocally to learning and memory. There was a chance he had spent 30 years chasing a mere curiosity, "an interesting little piece of biology," as he put it.

This worried Lynch; it scared the hell out of him.

Lynch determined he might be able to use the new visualization method to do an experiment that could show without question that LTP was memory.

The visualization technique involved staining structural molecules inside neurons so that their reorganization could be seen. The reorganization occurred only on the extensions of the neurons known as dendritic spines. If particular spines had no reorganization, the dye would wash right through without sticking. It would only stick to the spines where LTP had occurred.

Lynch's lab had introduced and used the technique in normal LTP bench experiments. Lynch wanted to try it in actual animals. Rats would be trained in a new task, something that would be encoded into memory. Then the dye would be injected into the rat's brain.

If it worked, the staining pattern would illuminate a neural trace of memory. This was not a modest undertaking. Scientists for more than a century had been trying to find such traces, often referred to as engrams.

For a time in the mid-20th century, the search had been a furious chase. It had since been all but abandoned as a sort of pipe dream.

His goal, Lynch said, was to see "spines encode memory in real rats after real learning . . . maps of memory encoding sites in the brain. . . . The crowd goes wild!!!"

The crowd, such as it was, would need patience. It took nearly half a year to get the experiment up and running.

By fall, the lab had demonstrated that the technique could work in a rat learning to navigate a new environment. The next test would be to compare the brains of rats that had been allowed to roam free with rats kept in cages. Presumably, the roaming rats would have learned and remembered something of their environment and converted this learning into memories that would have caused more spines to restructure.

After the roaming period, the rats would be sacrificed and the brains of the roaming and the caged rats examined. If successful, it would confirm that the method worked in living animals, and the team could proceed to do specific learning and memory experiments. Lynch thought that might take a few days. Days turned to weeks, weeks to months.

Given that Lynch had been at this for three decades, a few months hardly mattered, unless you were living them.

Then, on a Saturday evening in early spring, Kramar was at the microscope when she saw it -- an actual trace of an actual memory. She started screaming.

"Gary was in the bathroom. I was so excited I almost ran in there to get him," Kramar said.

Now that he knew the method worked, Lynch gathered the group the next Monday to prepare for the final push. He had decided the rats would learn pairs of odors such as lemon-orange and strawberry-peppermint.

"The olfactory advantage is that we understand where odor memory is encoded. We know a priori that if the animal learns, it has to be here. In the visual cortex, you don't know where to start," he said.

Vadim Fedulov, a graduate student, was assigned to run the rats in a maze in the odor-learning experiment. He had been in the lab for just a couple months. Lynch had agreed to take him on when it looked as if he might be tossed out of grad school altogether. He was young, very bright, somewhat unpredictable and not punctual at all. He was, in other words, a typical Lynch recruit.

One morning not long after, before the markets opened, Cortex announced the results of a clinical trial in which the ampakine CX717 had been given to adults diagnosed with ADHD. The results were an unqualified success. The drug reduced ADHD symptoms across the board, almost equal to existing medications -- mainly stimulants -- without any of their deleterious side effects.

ADHD affects an estimated 4% of children in the United States. More than 30 million prescriptions are written for the disorder annually.

Cortex stock doubled in value over the next week. Chief Executive Roger Stoll announced the company was in negotiations with at least eight big pharmaceutical companies that wanted to license CX717. Such a deal, Stoll said, would be worth immediately as much as $30 million to the company and eventually several hundred million dollars.

Lynch was ecstatic. It was the first big public demonstration of the power of the ampakines. This was a day he had waited 15 years for.

"It's immensely gratifying. It really is," he said. "It validates the principle that you can treat neurological diseases by increasing cortical communication."

Lynch, feeling magnanimous, patched up his relationship with the university and moved back to 101 Theory. Kramar received a job offer from an Irvine biotech company. Although she hated the timing, in tears, she took the job.

Danielle Simmons took over her role in the engram search. Fedulov built his T-maze -- which was outfitted with sliding doors and flashing lights and apertures through which he inserted cotton swabs soaked in various scents -- and began running rats.

Lynch was scheduled to speak at a pharmaceutical conference in San Francisco. He prepared to be welcomed as a conquering hero.

Two days before the conference, the FDA called Cortex and said it had found unspecified problems in preclinical results -- that is, lab tests on animals -- with CX717. It ordered an immediate halt to all human trials.

Monday morning, as Lynch was scheduled to speak, the clinical hold was announced. Lynch had a bronchial infection and was loaded with antibiotics and steroids; his plane was two hours late because of bad weather. Cortex stock fell 60%. The subject of Lynch's talk was the failure of translation from preclinical lab work to clinical trials in memory drugs.

Irony wasn't quite strong enough to describe the circumstances.

Lynch had no objection to the FDA's action, even though he thought the hold would be resolved painlessly. "They're doing what they have to do," he said. "We're putting stuff in people's brains, and they should be careful."

Lynch returned to Irvine, the lab and the odor-learning experiment.

Fedulov had trained three rats and was ready to inject the dye and have the brains prepared for examination. Two of the rats, for reasons unknown, died from the injections. The sole remaining rat was sacrificed, its brain sliced and set on slides. Fedulov had the slides at 101 Theory. Lynch wanted to look at them at Gall Lab, where Lauterborn, an expert microscopist, could read and photograph the images. He called Fedulov and asked that he bring the slides.

Tracey Shors, a neuropsychologist from Rutgers who early in her career had written a much-discussed paper casting a skeptical eye toward the role of LTP in memory, happened to be on the UCI campus. She and Lynch were old friends and met to discuss their various researches. Lynch told her about the afternoon's prospects; she was interested, but skeptical. Bring me an engram, she said. Bring me an engram.

Lynch went to lunch, just about dying from anxiety. When he returned, Fedulov was nowhere to be found. There was Lynch, the big experimental result waiting -- no rat brains, no scientists.

"You'd think I was trying to launch the space shuttle," he said.

Lynch called Fedulov on his cell. He'd come and gone and left the slides in a refrigerator, neglecting to tell anyone. To further darken the atmosphere, a short paper Lynch had written describing the preliminary results of this work had come back from a journal editor, declined with scathing reviews.

Fedulov finally showed up, retrieved the slides and mounted the first in the microscope. One of the great difficulties in finding the markings of memory, assuming you even knew what they looked like, was knowing where to look. The brain, at the nano-scale, was a very big place. Even a rat has hundreds of billions of synapses. USC neuropsychologist Thompson has estimated that the human brain, with as many as 10 quadrillion synapses, is capable of more distinct neural patterns -- memories -- than there are atoms in the universe. The reason Lynch had chosen an odor-learning exercise for this experiment was because the olfactory cortex was the simplest system to navigate.

Lauterborn searched through the slides, found the olfactory cortex and moved aside to let Lynch take a look. He immediately began oohing and ahhing. "Oh. Oh. Yes, yes, they could." He continued to scan across the slides. "Yes, yes, yes, YES. I'm almost convinced. Almost."

Lauterborn took digital photographs of the slides through the computer and brought up the images for more-detailed examination. The slides were a guided tour through the brain from the olfactory bulb to the cortex to the hippocampus, to the region thought to control emotion, the amygdala.

Lynch scanned across the brain regions. "That's gorgeous. . . . That's the way the world's supposed to be. Look at that. . . . All this time, that is the picture I wanted to see. Right there."

The following week they replicated the experiment with four more rats.

Lynch was again at the microscope. The image on the monitor showed a vast gray field of brain matter. The gray was lighted up here and there sparsely, but intensely.

"You see 'em. You see 'em. Look at them," he said. He traced on a piece of paper the path from the nose to the point on the image in the olfactory cortex being viewed. "It's a direct connection between the olfactory cortex and the hippocampus [the memory center of the brain]. Four synapses from the nose to the hippocampus."

Back at the microscope, he followed the path through to the hippocampus and murmured, "Vadim, Vadim, I'm going to make you famous. . . . That's it. That is the first demonstration that LTP is engaged in memory."

Is memory? I asked.

"Is memory," he said. "It couldn't have been much better. All those years, all those arguments -- it's all gone."

He went back to the microscope. After a minute or so of further scanning and examination, he shouted, "You see that, boys and girls? Science works."

A theory of a lot

One day not long after, Arvid Carlsson, a Nobel laureate pharmacologist from Sweden, visited. Lynch briefed him on the engram experiments. Carlsson was enthralled.

"To me, it seems so absolutely surprising and convincing," he said. "It makes so much sense. It seems to be a fundamental discovery."

The approbation of a learned and widely respected old hand like Carlsson was gratifying to Lynch, but to a surprising extent he had moved beyond needing it. The science had yielded. All he wanted now was more.

As the results continued to come in, he got it. An unrelenting problem in memory research for decades had been determining the location of long-term memory storage. "One of the darkest areas of research," Alcino J. Silva of UCLA called it. "We know nothing about it, literally nothing."

Researchers almost unanimously agreed that the hippocampus played a fundamental role in acquiring new memories, but they seemed to be moved later to the cortex for storage. If that were true, who or what moved it?

Because Lynch Lab could now see actual memory traces, the scientists were able to plot them onto brain maps. The maps showed how complicated a phenomenon even a simple memory was and the degree to which a single memory trace was distributed across brain regions. When the rats learned new smells, the trace of the learning landed in the olfactory bulb, then in the cortex, the hippocampus, the amygdala.

In a single stroke, Lynch's LTP research seemed to have yielded a realistic hypothesis for long-term storage. The memory wasn't moved from the hippocampus but was encoded simultaneously in both the hippocampus and the cortex. Presumably, a signal could later be sent from the hippocampus to keep or discard that particular memory.

Last October, the FDA lifted its hold on ampakine clinical trials, but it imposed a dose limit for patients, and the trials didn't restart until the limits were lifted this July. In the meantime, other researchers around the globe began experimenting with ampakines in a variety of indications, ranging from breathing disorders to mental retardation.

Rex and Lulu Chen, a new graduate student, devised an alternative means for the visualization experiments. The new method was easier to execute and produced stunning, unambiguous, easily replicable results. Publication of their work early this year provided the validation Lynch had anticipated almost two years earlier.

Offers -- pleas, even -- to collaborate rained in from around the globe. There was talk with the National Institutes of Health about setting up an engram project, a sort of national memory- mapping program to extend Lynch's work.

Lynch took gleeful satisfaction in the way in which the lab's methods -- described disdainfully as producing results that were "simply not credible" and contradictory to "all the assembled cell biological knowledge" a year before -- suddenly seemed poised to become standard practice in dozens of labs.

While declaring almost daily that he was about to quit the whole enterprise, Lynch couldn't help himself. The experiments seemed to provide interesting insights into Huntington's, retardation and, out of left field, menopause. Kramar, bored with the pace of work in private industry, returned to the lab. Hypotheses were being born by the dozen, and Lynch began planning the next expeditions onto the reedy shores of the unknown.

Lynch had in his career tried hard, he thought, even if no one shared the opinion, to remain modest, shying from big ideas and theories. He had thought it vain to suppose he could formulate an overarching explanation of memory and cognition. Now he was greedy. He wanted to shake the entire apparatus and demand all its secrets fall to the ground.

A physicist, when grandiose, will talk about forming a TOE -- a Theory of Everything. Lynch wasn't ready quite for that. This was still biology, after all. He would settle for a TOAL -- Theory of a Lot.

"I need to solidify the breakthrough -- get my arms around how big the thing is, how much of brain biology can be folded into it," he said.

The brain, he thought, represented a devil's bargain. You get memory storage almost beyond measure, but because that memory required more or less permanent neurons, you could not routinely replace them with new cells. If neurons broke, and they do, you were stuck with the results. As he so bluntly put it: You get stupid. And, because the brain controls so much of what the body does, when neurons fail you lose much more than memory.

"The evolutionary idea of stability as the cost of memory fascinates me," Lynch said. "And maybe, at the end, there lies the answer for how to get my broken-down brain going again."

Most of the work by this time was being done in Gall's lab, a few hundred yards away from 101 Theory, where, some days, Lynch worked alone.

It seemed fitting, somehow. There he sat at the end of the great, long chase, often sick as a dog, the entry locked, the clamorous tribes of the neurosciences a low hum in the distance; no phone, no e-mail, not even a name on the door to betray his presence. The only way you would know he was there at all was the blue Corvette out front. And, of course, the science, which, no matter the circumstance, difficulty or hour, had poured out for 30 years like water from the well. And poured still.

terry.mcdermott@latimes.com