For decades, Scott Friedman, a hepatologist at the Icahn School of Medicine at Mount Sinai,  studied liver fibrosis in an effort to prevent or even reverse harmful scarring of the organ. Back in the 1980s, he identified the liver cells that, when activated, are most responsible for collagen production and scar tissue formation (1). Then as interest in those cells grew, he played an instrumental role in convincing the rest of the field to agree on a standard name: hepatic stellate cells. “It was 1992, so I mounted a vigorous fax campaign,” said Friedman.

Even though Friedman has been in the thick of liver fibrosis research for 40 years, a new Science Translational Medicine  study from University of Pittsburgh pharmacologist Wen Xie’s research group managed to surprise him. Xie’s team showed that inhibiting cytochrome P450 1B1 (CYP1B1) — an enzyme that metabolizes drugs, vitamins, and other substances — kept mice and humans from breaking down the plant sugar trehalose, which then prevented hepatic stellate cells from activating and creating scar tissue (2).

When Friedman saw the results, he was intrigued, but he wondered “why they started with CYP1B1. It sort of comes out of left field.”

According to Xie, it all started with one of his group’s earlier studies, which focused on the connection between liver fibrosis and the aryl hydrocarbon receptor (AhR), a major liver protein that recognizes environmental toxins and regulates metabolic enzymes like CYP1B1 (3). Xie’s team found that eliminating AhR activated hepatic stellate cells, and they concluded that the protein normally tells the cells to produce enzymes that discourage collagen production and scar tissue formation. 

It was a neat story, except for one unexpected finding. Generally speaking, less AhR means lower levels of the enzymes it regulates. But that wasn’t the case for CYP1B1. “Even though AhR expression decreased, CYP1B1 remained induced,” said Xie. He realized that there could be another unknown protein regulating CYP1B1 in hepatic stellate cells. “That got us excited,” said Xie.

In the present study, Xie’s team looked for evidence of an AhR-independent connection between fibrosis and high levels of CYP1B1 in hepatic stellate cells. First, they stained for CYP1B1 in sections of fibrotic human and mouse livers and found it in the same places as known indicators of activated hepatic stellate cells. 

Next, they eliminated the CYP1B1 gene in a group of mice and then used chemicals or surgery to induce fibrosis. Relative to wildtype mice, mice lacking CYP1B1  had less collagen in their livers and lower expression levels of genes associated with hepatic stellate cell activation, scarring, and inflammation. Treating wildtype mice with a CYP1B1 inhibitor had a similar protective effect.

Mother nature never made an enzyme — or rarely made an enzyme — doing a single reaction. 
- Wen Xie, University of Pittsburgh

However, Xie and his team could only produce these results in male mice. Eliminating the CYP1B1 gene in female mice did nothing to halt the progression of fibrosis, unless the mice lacked ovaries. Xie concluded that estrogens were responsible for the sex-specific outcomes.

The fact that estrogen levels could alter the success of CYP1B1-centered liver fibrosis interventions gave Friedman some pause, but he didn’t think that should dissuade researchers from pursuing treatments targeting the enzyme. “Most patients — women with liver fibrosis — are postmenopausal. Not all, but most,” said Friedman. “It might become a nonissue.”

According to Xie, the second half of his study pointed to a more promising therapeutic strategy anyway. Having shown that inhibiting CYP1B1 protects against liver fibrosis in mice with low estrogen levels, Xie and his team became curious about the treatment’s underlying mechanism. They exposed male mice with and without CYP1B1 genes to fibrosis-inducing chemicals, collected hepatic stellate cells from both groups, and then analyzed the molecular contents of the cells using mass spectrometry.

Initially, Xie was looking for differences in the substances that CYP1B1 is known to metabolize, such as retinol, which is another name for vitamin A. According to Friedman, hepatic stellate cells store a lot of the naturally fluorescent vitamin, and there appears to be a connection between the onset of fibrosis and the metabolism of retinol in the liver. “When stellate cells activate, they lose the vitamin A-containing droplets,” said Friedman.

Everything seemed to suggest Xie had stumbled onto a story about vitamin A metabolism. He wasn’t particularly surprised when his team saw higher levels of retinol and lower levels of retinoic acid, its metabolized form, in the hepatic stellate cells that lacked CYP1B1 enzymes. “That’s more or less expected,” said Xie. “But probably not as exciting as I hoped.”

Xie decided that wasn’t the end of the story. “Mother nature never made an enzyme — or rarely made an enzyme — doing a single reaction,” he said. “Are there metabolites other than retinol we can identify?” When they compared all the metabolites in the hepatic stellate cells with and without CYP1B1 proteins, they identified a difference in levels of trehalose, a sugar that mammals cannot synthesize but can metabolize. 

It was 1992, so I mounted a vigorous fax campaign. 
- Scott Friedman, Icahn School of Medicine at Mount Sinai

But when Xie’s team incubated trehalose with CYP1B1, nothing happened. So, they turned their attention to trehalase (TREH), the enzyme that metabolizes the sugar. Using real-time PCR and Western blotting, the scientists showed that mice without CYP1B1 enzymes did have less TREH, just not in the expected location. “It’s not decreased in the liver,” said Xie. “But in the intestine.”

Xie realized that he and his team were no longer telling a simple story about isolated liver cells. They had to figure out how the gut and the liver were communicating with each other. There wasn’t much information available about TREH’s regulation, so Xie’s team used chromatin immunoprecipitation sequencing to figure out what protein could bind to the promoter region of the TREH  gene. That’s when everything really started to come together, because the protein they identified was none other than retinoic acid receptor alpha (RARa), which promotes TREH  expression upon detecting vitamin A’s metabolized form.

Without CYP1B1 to metabolize vitamin A in the hepatic stellate cells, there wasn’t enough retinoic acid making its way to the gut to maintain normal levels of RARa activity there. Consequently, there were fewer TREH enzymes in the intestines, and more trehalose moved into the bloodstream and traveled to the liver unmetabolized. Using a fluorescent dye that responds to acidity, Xie’s team showed that high levels of trehalose prevent hepatic stellate cells from digesting their fat stores and activating.

This was a story that could satisfy Xie’s curiosity, and it pointed to a simple intervention for liver fibrosis. Xie’s team put trehalose in a group of mice’s drinking water. After exposure to fibrosis-inducing chemicals, mice that consumed trehalose had less collagen and fewer markers of hepatic stellate cell activation in their livers than mice that drank water alone. The same was true for both male and female mice, suggesting that estrogens would not interfere with a trehalose-based treatment for liver fibrosis.

Friedman was intrigued by the idea of using trehalose to treat liver fibrosis, but he noted that the mice in Xie’s study seemed to consume a lot of the sugar. “It’s great if trehalose helps, but not if you have to give three gallons a day,” said Friedman. He thought it might make more sense to focus on other targets among the enzymes and metabolites implicated in Xie’s study.

While Xie acknowledged that patients with liver fibrosis would need to take a dose of trehalose that would exceed their normal dietary intake to see any therapeutic benefit, he could not envision any such regimen causing harm to a patient. “Trehalose is very safe,” he said.

But Xie agreed that his study pointed to other potential therapeutic strategies as well. At the same time, he noted that the enzymes and metabolites involved in hepatic stellate cell activation could aid in diagnosis, too. “Early treatment is always a challenge,” Xie said. “As a biomarker, the value is equally important.”

References

1. Friedman, S. et al. Hepatic lipocytes: the principal collagen-producing cells of normal rat liver. PNAS   82, 8681-8685 (1985).
2. Tung, H. et al. Inhibition of heme-thiolate monooxygenase CYP1B1 prevents hepatic stellate cell activation and liver fibrosis by accumulating trehalose. Sci Transl Med   16, eadk8446 (2024).
3. Yan, J. et al. Aryl hydrocarbon receptor signaling prevents activation of hepatic stellate cells and liver fibrogenesis in mice. Gastroenterology   157, P793-806.E14 (2019).