A  growing body of literature ties the gut microbiome to symptoms of depression in a seemingly circular relationship where each affects the other. However, many of the studies on this relationship merely link certain bacterial populations or diets to major depressive disorder—leaving open critical questions about the underlying mechanisms of how the gut microbes might influence depression.

Research published last month (May 3) in Cell Metabolism takes an important step toward filling such gaps, demonstrating in multiple animal species that there is likely a causative relationship between depression severity and serum levels of the nonessential amino acid proline, which the study finds depend on both diet and the activity of proline-metabolizing bacteria in the gut.

"To the best of my knowledge, this is the first time that a team actually demonstrates a causal relationship between proline intake and depressive behavior," King’s College London metabolism researcher Sandrine Claus, who didn’t work on the study and is also chief scientific officer of the microbiome therapeutics company YSOPIA Bioscience, tells The Scientist over email. "I am unaware of a proline-mediated gut-brain axis. This is therefore a completely novel mechanism of action."

Depression diet: the effects of proline

Previous research had found that proline, among other dietary compounds, seems to play a role in major depressive disorder, but "we found increased levels not only [in] major depression but also in subjects with moderate depression," study coauthor José Manuel Fernández-Real, a researcher at the Girona Biomedical Research Institute and Dr. Josep Trueta Hospital, both located in Spain, explains. Indeed, the severity of the symptoms correlated with the subjects’ circulating proline.

Fernández-Real and his colleagues uncovered this when they compared people’s responses on an 80-item food intake questionnaire with scores on the Patient Health Questionnaire-9 (PHQ-9), a common clinical survey for diagnosing and measuring the severity of a person’s depression. Out of all the dietary nutrients in the questionnaire, Fernández-Real says, the one "most associated with depressive traits was precisely proline." Blood tests in the same participants solidified the correlation between proline and depressive traits.

However, some discrepancies emerged within the data that demanded a closer look. "Not all subjects with increased proline in the diet had increased proline in the plasma," hinting that some yet-undiscovered factor was involved, Fernández-Real explains. In search of that explanation, he and the other researchers determined the microbiome compositions of the human participants.

The paper notes that most previous studies attempting to do the same failed to achieve bacterial species-level resolution and have reached inconclusive and conflicting findings. But Fernández-Real and colleagues employed a multi-omics approach that allowed them to link microbial function to the specific biological pathways associated with depression, granting their study a level of resolution that Fernández-Real says was lacking from what he describes as underpowered previous studies.

In the study participants, plasma proline levels were associated with the presence and activity of specific gut bacteria—people with high proline consumption and higher plasma proline levels had different microbiome compositions than those who consumed the same amount of proline but had less circulating in their blood. Furthermore, the team found that the microbial communities of the former were associated with more severe depression.

How the gut microbiome influences depression

To determine whether there’s a direct link between proline and depression, the researchers revisited and modified mouse and Drosophila melanogaster models that they’d previously used to study how the microbiome influenced cognitive abilities.

The researchers fed 10 mice a standard diet and another 10 a proline-supplemented diet, then subjected them to stressors typically used to trigger depression-like behaviors. After six weeks, the experimental group had significantly higher proline levels circulating in their plasma and exhibited more signs of depressive behaviors, such as a disinterest in sugar water and decreased mobility during a tail suspension test.

To see how the microbiome factored in, the researchers took fecal samples from 20 human volunteers (nine of whom had high proline levels and all of whom demonstrated a direct correlation between their PHQ-9 score and circulating plasma proline) and put them into antibiotic-treated mice, effectively transferring the human microbiomes into the animals. When the mice were subjected to another test meant to induce depressive behaviors, the researchers found that the mice’s behavior correlated with the PHQ-9 scores—and therefore circulating proline levels—of their donors as well as the mix of microbes now residing in their guts.

The data demonstrated that "a particular microbiota metabolizes proline and is critical to develop more or less depressive symptoms," says Fernández-Real.

The researchers also conducted RNA sequencing of the animals’ prefrontal cortex, a region of the brain associated with cognition. That revealed that genes related to depressive behaviors had been upregulated following fecal transplantation—and that expression of the proline transporter gene Slc6a20 in the brain correlated with the mice’s behavior and their microbe donors’ PHQ-9 scores.

"The microbiota from subjects with the highest depression scores induced emotional traits in the mice," says Fernández-Real. "Interestingly, the prefrontal cortex of transplanted mice showed increased expression of genes . . . that we also found in the intestine of subjects with increased proline intake." 

From there, the researchers moved on to Drosophila experiments, subjecting both wild type control flies and those with downregulated CG43066—the Drosophila version of sl6a20—to stressors to see if the transporters affect whether the animals exhibit depressive behaviors. They then ran the same tests on Drosophila colonized with the bacteria found to increase or decrease proline metabolism in the prior experiments. Downregulating the proline transporter gene or colonizing the Drosophila with specific bacteria, especially certain Lactobacillus species, seemed to protect the flies from depressive behavior, the study found.

Animal depression, human questions

The researchers weren’t able to conduct similar experiments in people, which they concede limits the conclusions that can be drawn from their work. Going forward, Fernández-Real says it will be important to test, for example, "whether diets with different proline contents influence depressive traits and depressive symptomology."

Chrysi Sergaki, a microbiome researcher at the Medicines & Healthcare products Regulatory Agency in the UK who did not work on the study, tells The Scientist over email that "using these [animal] models is a start. They can help us understand the impact of the microbiome on brain function, but that doesn’t necessarily mean that it will work the same way in humans." Still, she says that because similar experiments can’t be performed on humans, the animal models used in the new study can grant researchers "a deeper understanding of how the microbiome can influence the functions of the organism they live in," adding that "that knowledge can be valuable in the way we think about the microbiome when we move to humans."

Claus expresses similar sentiments. "Modeling depressive behaviors in animals is . . . very challenging," she writes. "I actually thought that the drosophila model was interesting despite the fact that we cannot directly translate behavioral observations from drosophila to humans. These are useful to study mechanisms of action though."

Still, Claus adds that a lack of data on circulating proline levels in the mouse model, combined with repeated reanalysis of the same cohort of people, make it difficult to draw definitive conclusions about the mechanism of microbial proline metabolism and its link to depression.

"The authors keep reanalyzing the same cohort, insisting that they always find a consistent microbial signature with PHQ-9 and proline," Claus writes. "But this is not surprising since proline is correlated to PHQ-9 score in this cohort, and PHQ-9 score is correlated with a microbial signature."

Sergaki applauds the study authors for describing the limitations of their work, adding that microbiome studies are notoriously difficult to reproduce and therefore validate. "I think all microbiome scientists look at these studies with a critical eye," she tells The Scientist. "The authors mention certain limitations of their study which are quite important. The biggest question is always this: correlation or causation? Due to the complexity of the system, this is very difficult to answer."