The second of four authors on Mojica’s groundbreaking report was graduate student César Díez-Villaseñor. He watched the accolades showered on Mojica with a mixture of pride and envy. In the early days of CRISPR, “it seemed likely that spacers didn’t have any function at all,”15 he recalled, because each spacer had a unique sequence, minimizing the likelihood of any sort of common function. But Díez-Villaseñor was puzzled: “If the sequence of the spacers is not really relevant, why bother?” Perhaps their uniqueness meant they were somehow toxic to similar sequences. “The proposition was immediately dismissed, although it didn’t leave my mind.” He says wistfully, “it was hinting the right direction.” Another possibility was that the spacers were being generated by some peculiarly sloppy form of DNA replication. But then adjacent spacers ought to be more similar to each other than distant ones, which was not the case. Díez-Villaseñor charted the CRISPR spacers from E. coli to show his boss, refuting the idea of mutation incorporation. He remembers it was the day before Mojica’s eureka moment.
“I said with a bit of frustration that spacers had to be taken from previous sequences but, obviously, they had to be present inside the cell at that time.” As he said that aloud, he realized it made perfect sense. “Of course—it must be an immune system!” There were known examples of RNA interference being used as an immune system. “Suddenly, everything that had looked so strange made perfect sense.” Díez-Villaseñor asked Mojica if he was running sequence searches with the spacers. Mojica responded briskly: “Doing that is my job! Don’t do anything.”
To find a microbial immune system that could recognize invading phage was a very big discovery. The next day, doubts started to be dispelled. “Francis came to the lab elated and directly talked to me. He told me he had found the first homology of an E. coli spacer in phage P1.” Later he told another professor that CRISPRs were like “memorabilia from past hosted genetic elements.” Díez-Villaseñor says: “That was probably the most satisfying professional moment in my entire life.”
Mere weeks after Mojica began his publication odyssey,16, 17 Gilles Vergnaud in Paris submitted his own CRISPR story, and experienced similar frustrations. With concerns growing about Saddam Hussein’s use of biological weapons, Vergnaud, working for the French Ministry of Defense, was tasked with improving microbial detection methods. At the end of 2002, Vergnaud obtained access to DNAs from dozens of strains of Yersinia pestis isolated during a plague outbreak in Vietnam in the mid-1960s. Using the best genetic methods available at the time, he found the strains were identical except for one region, which they named minisatellite number 6 (MS06). When a graduate student, Gregory Salvignol, sequenced MS06 in more detail, it turned out to be a CRISPR. Moreover, the French team found that it acquired new spacers from viral DNA. They too proposed that these structures were part of a bacterial immune system. The first draft of the manuscript, written in July 2003, included the notion of a “defense mechanism”—a record of “past genetic aggressions.”
But like Mojica, Vergnaud endured his own depressing runaround. He submitted his paper—the first to include “CRISPR” in the title—just behind Mojica, but in November 2003, PNAS dismissed it without review. It was the same story at the Journal of Bacteriology (twice), Nucleic Acids Research, and Genome Research. In July 2004, he submitted to Microbiology, which eventually published his paper in 2005.18
The early history of CRISPR would look very different if any one of those journals had said yes. As it was, Vergnaud applied for grants from the French National Research Agency over three consecutive years without success. A third report on CRISPR came out from Alexander Bolotin, a Russian microbiologist working for the French ministry of agriculture. Bolotin noted a correlation between the number of spacers and phage sensitivity, deducing that “spacer elements are the traces of past invasions by extrachromosomal elements.”19
We now know that some 90 percent of Archaea contain CRISPR elements, but only 40 percent of bacteria. Mojica explains that bacteria have a larger repertoire of defense systems at their disposal. CRISPR serves as a genetic barrier of sorts to microbial evolution because it discourages horizontal gene transfer. “Do you prefer to have a barrier to genetic transfer or defend against viruses?” says Mojica.
Spain has only celebrated two Nobel Prize winners in science—Santiago Ramón y Cajal in 1906 and Severo Ochoa in 1959. That puts a lot of pressure on Mojica, even as he laughs off such idle speculation. “It’s good to know that some people think I deserve it, I really appreciate that, but thinking about the possibility of getting the Nobel Prize is—how you say—crazy! There’s no way one could expect to get the Nobel Prize.”
Before heading back to Alicante, Mojica and I stop for a beer before lunch. He is about to embark on a three-week lecture tour of Australia, the sort of career success that most academics covet. I’m curious about how the microbiologist who has been nicknamed el padrino—the godfather—is handling his newfound celebrity. Mojica pauses, reaching for another olive. The restaurant is deserted, but in almost a whisper, he says, “I hate it… I hate it.” This was not the reaction I was expecting. “I just want a quiet life,” he says, shaking his head. “I want to do my research and go home to my wife.”
Anyone taking bets on where the next pivotal step in the CRISPR story would occur could have found extremely long odds on a Danish yogurt company. But for scientists at Danisco, ensuring bacteria used in starter cultures can ward off the constant threat of phage infection is a commercial priority. The next CRISPR breakthrough came in two parts of the world where cheese making is revered—France and Wisconsin.
Philippe Horvath was born about fifty miles south of Strasbourg, close to the German border. As we walk to a restaurant in Vilnius, Lithuania, he stops to catch the score of a World Cup game displayed on a giant outdoor screen. Croatia are winning en route to their surprise appearance in the final of the 2018 tournament. “Did you know my name means ‘from Croatia’ in Hungarian?” Horvath asks.20 (I mean, why would I know that?!)
During his PhD at the University of Strasbourg, Horvath studied the genome of Lactobacillus plantarum, which is traditionally used in food fermentation including sourdough, kimchi, pickles, and sauerkraut. I was skeptical this could sustain an entire PhD thesis, but Horvath shoots me a look. “This is not a trivial bacterium like E. coli,” he says sternly. Sauerkraut is serious business in Alsace, where it forms the base of the famous choucroute garnie.
Horvath saw an ideal job advertisement for a molecular biologist in industry and sent off the only application letter he has ever written. He was hired in December 2000 by Rhodia Food (formerly Rhone-Poulenc, a famous French chemical company).IV Horvath’s expertise in bacterial genetics helped improve the quality of starter cultures—the seed bacteria used to ferment milk into yogurt and cheese, which Rhodia sold to food giants like General Mills, Danone, and Nestlé. Phages that prey on the bacteria used in fermentation are found naturally in milk. “When you have a tank containing 10,000 liters of milk and you add a starter culture that is sensitive to a phage that is present, it’s a disaster!” says Horvath. “The milk remains milk.”
A typical starter culture consists of a high density of three to eight strains of bacteria—about 1 trillion bacteria per gram. Common examples include Lactobacillus acidophilus, Lactococcus lactis, and Streptococcus thermophilus. Horvath explains a freeze-dried starter culture pouch or brick is added to some 2,000 liters of milk. The goal is to minimize the number of rounds of cell division the bacteria need to produce enough lactic acid to lower the pH of the milk. Acidification must occur quickly to protect the milk from spoilage bacteria such as Salmonella and Listeria. “The higher the number of bacteria, the fewer generations you’ll need and the less risk you’ll take in terms of phages,” says Horvath. Humans have fermented milk in this fashion for millennia without knowing the molecular minutiae.