Photography by Josh Windsor
“French Cheese Under Threat” read the report from the French National Center for Scientific Research (CNRS) on January 16th, 2024. In the ensuing weeks the findings of the report have been picked up around the globe. As is often the case when reporting on the impact of discoveries in microbiology and population genetics, there has been a lot of confusion around how to understand and interpret the science. The biology at the heart of the report is complicated. The same thing can be said of our relationship with the microorganisms that we rely upon for cheese production. Rather than sound the alarm that these beloved French cheeses are on the road to extinction, what if we saw this as an opportunity to rethink what it is that we value in cheese? A chance to not only expand the diversity of rinds produced, but also to redefine our responsibility to our microbial coworkers.
But first, the science. To understand what’s at stake it is necessary to step back into the evolutionary past of a white, fluffy fungus called Penicillium camemberti (P. camemberti or PC for short). PC is the mold that forms a uniform cloudlike rind around soft ripened cheeses like Camembert de Normandie and Brie, to name just a few. Cheese is the natural habitat of P. camemberti, a domesticated species bred specifically to produce a host of “cheesy” aromas and produce that signature pillowy rind. Like other domesticated species (e.g., cows, dogs, corn, kale) its genetic makeup has been impacted through its relationship with humans. This means that the mold we have come to know and love today has an ancestor somewhere out in the wild. When we are comparing the genetic makeup of a domesticated species to its ancestor, the ancestor is called a “wild type”. It represents the genetic state of the species before humans affected any change. It turns out that there was an intermediary domestication step in PC’s evolutionary development—the wild type ancestor of PC is not its mother, but its grandmother.
You may have encountered P. camemberti’s mother, Penicillium biforme (P. biforme), in cheese at some point. Its aroma is musty and evocative of wet soil, some call it “cavey”—a term dear to my heart. This aroma predominantly comes from the production of geosmin, a volatile compound that shows up all over nature from beets to petrichor (the smell of rain). Its coloration ranges from a steely gray to the purplish blue of an impending storm. These lush and tempestuous descriptions should tip you off, we are getting closer to a wild ancestor, but not quite there yet. For producers looking for an orderly rind with milder aromas whose earthiness leans toward button mushrooms and not fresh humus, P. biforme is a pest: a defect to be avoided or eradicated. P. camemberti on the other hand is a commercial product purchased deliberately as an adjunct culture for rind formation. As far as I know, P. biforme is not available as a lab produced culture. In this way it is wild, however, it too is a domesticated mold.
Following this family tree back one step further, we arrive at Penicillium fuscoglaucum (P. fuscoglaucum). Its name comes from the Latin for dark gray. This cinereal ancestor is evocative of a cave—the primordial home of aged cheese—dim, dingy, fungal. Its appearance on food can cause alarm in the uninitiated. Jasper Hill for a time included a postcard explaining that its appearance on the spruce wrapping of Harbison was intentional and desirable. That card used the name Penicillium commune, which is a synonym for P. fuscoglaucum. Its appearance and aromas, more intensified than those of P. biforme, are not common in the food of the modern industrialized world. Which makes sense, it is the wild ancestor that has a life outside of a cheese rind. It has traits and features that allow it to exist in a bigger world. It can do more things than just live a simple life eating cheese. It is a little more removed from our food production systems.
We are only beginning to understand the mechanisms of how P. fuscoglaucum was domesticated and gave rise to P. biforme and in turn P. camemberti. This is important to know because it can provide a model for how to exist and work with our microbial partners. Domestication can take many paths, but there are two that are most likely to explain this process. The first is known as commensal domestication, which occurs when species live in close approximation of each other. By aging cheese in cool, dark caves and cellars, the natural environment of P. fuscoglaucum was changed. Since P. fuscoglaucum is a wild and diverse species, those with genetic variants (called “strains” in microbiology) that allowed it to best utilize this new food source, thrived. Species that are attracted to and flourish along with humans are given the beautiful name, synanthrope (being with humans). Commensal domestication is unintentional, it develops out of the serendipitous intermingling of species, of being with.
This form of domestication usually results in something called gene silencing. Genes are sequences of DNA that get transcribed into proteins and proteins are the building blocks of organisms. In simpler terms genes greatly impact the traits and features of a creature. There is a relationship between the collection of genes (called a genotype) and the features (called a phenotype) of an individual organism or species. There are two very important things to know about genes. The first is that they are hereditary. They can be passed on from one creature to another through reproduction. The other is that they can be either “off” or “on”, like a light switch. When a gene is on, it is directing the creation of proteins, which in turn are doing things in fungi like forming pigments (lawns of cosmic blue mycelium) or secreting metabolic enzymes (wafts of deep earthy aromas). When off, genes are doing nothing. If a gene is off and stays off, it is called gene silencing. This happens very frequently in commensal domestication. When food is abundant, the need to compete for it is lowered. Genes that regulate defensive competitive functions stop turning on and are silent. It takes energy to compete for food, so if a cave fills with delicious cheese, why fight for it? There is plenty to go around. When genes go silent, they can often be turned back on. So that if the cheese disappears, the competition is back, and the species has the tools to react. And since it is hereditary, a silenced gene can still be turned on in subsequent generations.
Directed domestication is the path of intentional breeding programs. This is what most people think about when they hear the term, domestication. Here, humans intentionally select for specific traits within a species through breeding several generations of offspring. It is how we increased the udder size in Holstein cattle or ended up with short-snouted canine companions. If I find a spot of radiantly alabaster white mold growing on my cheese rind and I scrape a bit off and get it to grow successfully on the entirety of another cheese, I have started the process of directed domestication. I have selected for a trait (whiteness or lack of pigmentation) and have bred successive generations. With mold the whiteness is an absence of pigmentation. It is a result of the rich grays and blues no longer being produced. The genes for pigmentation are silent and I am now breeding for silent genes.
This is the point where directed domestication can become a problem. Gene inheritance is not an exacting process. Mistakes get made in making copies of genes for offspring to inherit and sometimes genes get lost. This is called gene deletion. Now if a gene is deleted and it controls something that is useful for the species, it usually won’t propagate for multiple generations and become a problem for entire populations. It may still be a part of the gene pool (the total collection of all genes present in a species), but there will be other more dominant variants. However, if the gene that is deleted was already silenced, there aren’t a lot of biological mechanisms to correct the problem. If I drive to work every day and lose my keys, I am made aware of the fact quickly and can often retrace my steps and find my keys quickly. If I lose my keys two days into hiking the Appalachian Trail, it may be months before I realize what has happened and finding those keys is a lost cause. The necessity of a gene impacts how likely it is to be deleted. When there are intensive outside pressures to select for specific traits, the likelihood of gene deletion increases.
And this is where we find ourselves today with our old friend Penicillium camemberti. After an initial cohabitation with Penicillium fuscoglaucum, we began silencing the genes for wild coloration and intense cavey aromas through a process of commensal domestication leading to the development of Pencillium biforme. We then sped up the process by selecting for even paler molds with milder aromas through directed domestication that resulted in the degradation of the genome (the entire collection of genes in a single organism). If these were the only traits that were affected by a degraded genome, then it would only be a problem of aesthetics. But it appears the issue is more complicated. P. camemberti is losing its ability to sporulate. In selecting for downy fields of white rinds we have created a mold that is losing the ability to reproduce. Or maybe already lost? Apart from the CNRS report, the underlying study has not been published to my knowledge. So, I am not entirely sure of the extent of the degradation. In any case, if it hasn’t happened already, it will. Selecting for unintentional traits in directed domestication happens all the time. Genes are highly interconnected and interdependent on one another. For some reason, when you breed foxes for more docile behavior, their ears become floppier. Don’t ask me why, I don’t know.
But before we wind our way out of this scenario, let’s talk about sex. Fungal reproduction is crazy. I don’t know any other way to put it. Fungi can pass on genetic material to successive generations in so many mind-boggling ways, but sporulation is the most common. Fungi can produce spores which are thick-walled vessels of genetic material that can grow to be a new organism. Spores become offspring. Penicillium can produce spores both sexually and asexually. The easier of the two to explain is asexual reproduction which creates a duplicate of the genetic material from the parent. If the parent fungi is blue, then the offspring is blue. The genome of the offspring is an exact replica (barring any mistakes or errors) of the parent. The child is a clone. Sexual reproduction occurs when the offspring contain a mix of genes from two separate donors. Which means that the offspring’s genome is different from both parents. Perhaps one donor is blue and one gray, the offspring may be either (or possibly a third; google
“Mendelian genetics”). Sexual reproduction allows for greater genetic diversity, and the diversity of a gene pool is an indicator of that species’ ability to survive catastrophe. From a Darwinian perspective, the more genetic diversity in a population the more fit the species is.
Since directed domestication of fungi is a targeted effort to yield a specific and narrow genome in a short amount of time, it relies entirely on cloning. This silences the genes that regulate sexual reproduction, which in turn leads to gene deletion. P. camemberti lost the ability to reproduce sexually early in its development, making it solely reliant on asexual reproduction (i.e. cloning) for propagation. Since cloning dramatically limits the genetic diversity of a species, the gene pool has reached a bottleneck and its Darwinian fitness is diminishing to the point that it cannot (or soon will not be able to) reproduce. SOUND ALARM BELLS!
Or maybe, don’t.
First, if soft, pillowy rinds that smell like raw mushrooms and fresh cream are your thing, they are probably not going to disappear. They will, however, become rarer and more expensive for a while. The technology to reproduce Penicillium camemberti in a lab exists, but it is more expensive, takes more time, costs more money, and has a lower yield. This will drive the availability of the commercial cultures down and the price up. There are also indications that the process to re-domesticate P. camemberti from P. biforme is possible. This happened with dogs who were domesticated in two separate events: one in eastern and one in western Eurasia, thousands of years ago. In the 2019 study, “Rapid Phenotypic and Metabolomic Domestication of Wild Penicillium Molds on Cheese” researchers demonstrated that Penicillium species can be domesticated very quickly when adapting to the surface environment of cheese. Like, a couple of generations over a few weeks, quickly. Domestication does not necessarily need to take thousands of years.
But why? Why would we spend so much time and energy to recreate the same genetic bottleneck we face today? For many European producers this is entwined with the Protected Designation of Origin (PDO) system. That Camembert de Normandie mentioned at the beginning of this essay is a PDO cheese. Its production is governed by a series of regulations that control what can and can’t be done to make the cheese. At its heart the PDO system is designed to protect traditional practices and products, which prior to the Industrial Revolution and subsequent globalization were tied to localities and had geographic boundaries – hence the Origin in PDO. Terroir is the philosophical guiding principle of PDO regulations. The taste and aroma of a cheese is uniquely embedded in the land and traditions from which the cheese originated. In the past thirty years there has been increasing research that has drawn correlations between the microbiome (the collection of bacteria and fungi in an ecosystem) of a region and the flavor of a cheese. This developed into the concept of “microbial terroir”. Not only are the traditions, practices, and land responsible for the cheese, but so are the organisms that populate the microhabitats in these places, from fields to barns to creameries to caves. We see this reflected directly in the PDO designation for Camembert de Normandie, “The rind is thin with surface mold constituting a white felt (Penicillium candidum) which can, depending on the degree of ripening, reveal red pigmentation (Brevibacterium linens).” Without launching into another lengthy scientific digression, consider P. candidum a synonym for P. camemberti. So now the genetic identity of a microbe is baked into the identity of a cheese—a microbe that can no longer reproduce. So, the French also have a legal problem on their hands.
As an American producer who does not participate in a PDO system, I am not obligated to work with a specific microbe to produce a rind. P. biforme is a willing and available partner in my endeavors. Peering back into evolutionary history reveals a host of fungi with a plethora of genetic diversity and biological adaptations making them fit to thrive for years in caves full of cheese. So, what’s the problem? I believe this boils down to aesthetics. Many people are averse to wildly colorful rinds full of deep earthy aromas. In the years of producing whiter rinds and milder aromas, we have also created consumers who are comfortable with only a limited subset of what cheese can be. Instead of investing in efforts to preserve or recreate a genetically degraded and bottlenecked fungi, we should find a way to change minds about the types of food we want to eat. Now is the time to work hard to promote acceptance in the diversity of sensory experience that is possible in cheese. This work is both internal to the cheese industry and reaching out to cheese lovers everywhere.
The evolutionary history of P. camemberti presents two models for how we interact with the microbial world. Directed domestication has dominated our microbial relationships since the advent of the Industrial Revolution. This model relies on a philosophy of control where fungi is treated as subservient to our needs. Its genome has been stripped down to produce a precise collection of traits. On the other hand, we have also seen a model of co-existence that has yielded delicious cheese through commensal domestication. If diversity (of rinds, of genomes, of sensory experience, of practitioners) is a goal—and I strongly believe it is—then we need to adopt metaphors that look much more like the commensal and less like the directed. Microbes are our coworkers, not our servants. We will need to find a way to relinquish some elements of control from the cheese making and aging process. This may seem like a leap to many. But I believe it necessary, particularly as we think about how to promote acceptance of a greater diversity of rinds. Who is going to love a rind that we do not love ourselves? Between the two models of microbial coworking we have been examining, only one resembles love. If nothing else, adopting a broader model of partnership in our microbial relationships will keep us from recreating the situation we are in today.
Lastly, and certainly not least, how do we work to build a world of people who are more tolerant of rind pigmentation and wild aromas? One major step is to look closely at how we talk about cheese and its rind. Many of the colors and aromas that come from P. biforme are disparagingly referred to as defects, whether it is a deduct in a cheese competition or a vocabulary lesson in cheese education. If we keep training mongers to pull bespeckled rinds from the case and mark down brash cheeses in competitions, what is the message we are sending to consumers? We should work hard to instill poetry into how we portray what is traditionally described as browning, blue spotting, musty, moldy, etc. Reframing these sensory experiences as rustic, wild, even mystical is one way to start changing the paradigm.
As we learn more about our complex relationship with the microbial world and how intertwined it is with our love of cheese, we should be prepared to correct the mistakes we have made in the past and create a better future. Simply put, free your rind and the rest will follow.
Josh Windsor
Josh Windsor is an affineur and the Senior Caves Manager at Murray’s Cheese in New York City, where he actively nurtures and ages the Cave Aged line of specialty cheeses including the 2019 American Cheese Society (ACS) “Best in Show” winner Stockinghall Clothbound Cheddar and the 2022 World Cheese Awards “Best American Cheese” winner Greensward. Always curious about how cheese ages, Josh is an active member of the community biology lab, Genspace, where he explores the microbial world of cheese rinds. Josh shares his love of all things dairy by teaching cheese appreciation, science, and history at Murray’s Cheese and sensory evaluation at Cornell University’s cheese short courses.
