Australian experts urge cautious, sector-specific AI policy for early

Salmonella enterica is one of the most important bacteria that cause disease in humans. Experts estimate that it causes between 200 million and one billion infections each year. These infections range from gastroenteritis, which often gets better on its own, to life-threatening typhoid fever. In 2024, the World Health Organization (WHO) placed fluoroquinolone-resistant Salmonella Typhi among the highest-priority pathogens that spread in communities. Resistance to these antibiotics is increasing. Against this background, a new open-access review in Discover Bacteria by Ronan C. Jensen and Gregory D. Fairn calls for renewed attention to a protein injected by the bacteria into host cells. This protein, called SopB or SigD, helps the bacteria at almost every stage of infection, from entering a cell to creating a place inside it where they can multiply.

Scientists first identified SopB in the 1990s. They found that the secreted protein was needed for fluid to build up and neutrophils, a type of immune cell, to gather in a model using a tied-off section of a cow's small intestine. Independently, another research team found SopB while searching for genes that help bacteria enter the cells lining the intestine. Researchers later discovered that a SopB protein made in the laboratory has inositol polyphosphate 4-phosphatase activity. This means it can break down both soluble inositol phosphates and a membrane lipid called PtdIns(3,4,5)P3 in laboratory tests. This was the first demonstration that an effector delivered by a bacterial type III secretion system could act as a phosphoinositide phosphatase. The discovery established a family of related proteins that now includes IpgD from Shigella and VPA0450 from Vibrio parahaemolyticus. SopB is a protein made of 561 amino acids and weighs about 62 kilodaltons. Its C-terminal region contains the enzyme's active site and a conserved Cys-X5-Arg motif. Its N-terminal region contains a part that binds to the host cell's Cdc42 protein, which helps control cell activities.

The most important recent discovery, and the main focus of the review, is that SopB does more than act as a phosphatase. Walpole and colleagues showed that SopB also has its own phosphotransferase activity. This activity does not depend on the host cell's PI 3-kinases or on ATP. During the reaction, a temporary chemical link forms between a phosphate group and the amino acid cysteine at position 460 of SopB. In a typical phosphatase reaction, water removes this phosphate. But SopB can also transfer the phosphate to the 3-position of another phosphoinositide molecule. This produces PtdIns(3,4)P2 directly and can briefly produce PtdIns(3,4,5)P3 from PtdIns(4,5)P2. Researchers recreated the reaction using laboratory-made SopB and liposomes, which are tiny membrane-like structures made from lipids. The test used a cell-free system without ATP, divalent metal ions or host enzymes. The researchers also confirmed the activity by making SopB in budding yeast. This organism lacks all class I and class II PI 3-kinases, yet it still produced PtdIns(3,4)P2 when SopB was switched on. The same activity was found in IpgD, suggesting that this ability existed in an early ancestor of this family of bacterial proteins.

This discovery changes how scientists can understand more than 20 years of research. SopB has long been known to strongly activate Akt, a host cell protein that helps cells survive. This response depends on SopB's enzyme activity and can continue for many hours after the bacteria enter a cell. The earlier explanation was that SopB activated the host cell's phosphoinositide 3-kinases, which then increased the lipids that activate Akt. Several findings supported a role for these host enzymes. Drugs that blocked PI3K activity partly reduced Akt activation, producing extra PTEN protein stopped it, and a broad screening study found more than 50 host kinases whose removal reduced Akt phosphorylation during bacterial invasion. However, no single host kinase could fully explain the response. The review argues that the new discovery fills this gap: SopB itself makes the lipid that activates Akt. The earlier findings showing that PI3K activity partly affects Akt activation can be explained by additional contributions from host enzymes working alongside SopB's own activity.

Studies using high-performance liquid chromatography and mass spectrometry over time show that SopB's two enzyme activities are not separate options. Instead, they happen in sequence as part of one reaction. When SopB acts on liposomes containing PtdIns(4,5)P2, it first creates a short-lived burst of lipids with a phosphate at the 3-position. These lipids help cells survive and support bacterial invasion. SopB then gradually removes phosphate groups, eventually leaving phosphatidylinositol as the main product. During real infection, about half of the cell's PtdIns(4,5)P2 is broken down during bacterial entry. This creates a carefully timed sequence: first, a brief rise in the 3-phosphorylated lipids, followed by a lasting reduction in negatively charged phosphoinositides as the vacuole containing Salmonella matures and loses its surface charge. Other amino acids in SopB, including Lys525 and Lys528, affect its phosphotransferase activity in specific ways. These residues give researchers tools to study the two enzyme activities separately, which was not possible using only the conserved Cys-X5-Arg motif.

SopB also has important effects on the survival of infected cells. Knodler and colleagues showed that SopB is needed to prevent the cutting and activation of caspase-3, a protein involved in programmed cell death, in infected cells lining the intestine. They also found that SopB protects these cells against apoptosis caused by camptothecin, a drug that can trigger cell death. Both effects depend on ongoing Akt signalling. This protection also occurs in B lymphocytes, a type of immune cell. There, SopB activates Akt, which causes the Hippo pathway protein YAP to gain a phosphate group. This stops YAP from entering the cell nucleus and helps prevent the assembly of the NLRC4 inflammasome, a cell structure involved in immune responses. As a result, the cell releases less IL-1β, an inflammatory signalling molecule. In intestinal lining cells, SopB provides another form of protection by binding to and trapping the host E3 ligase TRAF6. This prevents TRAF6 from moving to the mitochondria and reduces cell death caused by reactive oxygen species (ROS). This effect works at least partly independently of Akt. The review explains why these actions benefit the bacteria: if an infected cell dies too soon, immune cells outside the cell can attack the bacteria, and their ability to multiply inside the cell ends. Keeping Akt active helps the infected cell survive longer, giving the bacteria time to complete several rounds of replication.

Surprisingly, SopB can extend this period of protection through a second, separate wave of protein delivery by type III secretion system 1. Finn and colleagues showed that Salmonella already inside the host cell's cytosol can deliver SopB and SipA into cells containing rapidly multiplying bacteria, long after the first wave of protein delivery during invasion has ended. This late delivery activates Akt again and helps the infected cells survive longer. Unlike the usual process during invasion, this later delivery does not depend on the pore-forming structure called a translocon. This shows that the secretion system works in two distinct phases over time. SopB's unusually long lifetime makes this possible. Most SPI-1 effectors are broken down quickly, but SopB can still be detected up to 12 hours after invasion. It therefore links activities that happen at different times in the two secretion systems and helps prepare the Salmonella-containing vacuole for the later network of SPI-2 effectors.

SopB can also affect host cells without using its enzyme activity. Zhao and colleagues showed that, in U2OS cells, SopB uses its N-terminal region to bind to and activate Cdc42, a host protein that controls cell shape and other activities. This leads to the MEK1/2-dependent formation of a cage made of vimentin around the maturing vacuole. The cage helps keep the vacuole intact and supports bacterial multiplication. The review authors point out two important limitations. Most of the evidence comes from one study using an osteosarcoma cell line that is not widely used in this field. The findings therefore need to be confirmed in intestinal lining cells and macrophages. Also, scientists have debated whether SopB directly activates Cdc42 or simply binds to it. Structural studies show that SopB's N-terminal region resembles a host protein called a guanine nucleotide dissociation inhibitor. It binds to parts of Cdc42 called switch regions and slows the exchange of nucleotides, small molecules that help control the protein's activity. SopB also works with other bacterial effectors, and some effectors counteract its actions. For example, it cooperates with SopD, a protein that switches off the Rab10 GTPase, to help separate membranes as the closing ruffle forms around invading bacteria. At the same time, SopB changes membrane lipids in ways that can weaken the newly formed vacuole. SopF counteracts this effect by helping preserve the vacuole and preventing ATG16L1 from gathering there. This helps the bacteria avoid xenophagy, a process in which cells break down invading microbes.

SopB is also controlled by the host cell's ubiquitin system. Ubiquitin is a small protein that cells attach to other proteins to change their location or activity. Mass spectrometry identified six lysine residues in SopB's N-terminal region that receive multiple single ubiquitin tags after SopB enters the host cell. Researchers identified TRAF6 and the E2 enzyme UbcH5c as the enzymes responsible for adding these tags. Importantly, this modification appears to control SopB rather than destroy it: changing the lysine residues does not alter how long SopB survives inside the cell. Two studies published around the same time proposed a model called "ticket-to-ride." In this model, ubiquitin tags move SopB from the cell's outer membrane to the vacuole and switch its enzyme activity according to which compartment it occupies. A version of SopB that cannot receive ubiquitin remains stuck at the cell's outer membrane and causes stronger, longer-lasting Akt activation. Jensen and Fairn suggest a more detailed explanation, which they call a ubiquitin-mediated electrostatic switch. They propose that adding ubiquitin reduces the positive charges that SopB uses to attach to negatively charged membrane lipids. This allows SopB to move to compartments rich in Cdc42. The authors also suggest experiments to test this idea, including changing SopB's amino acids to reduce its charge and checking whether those changes produce the same effects as ubiquitination.

The review ends by identifying unanswered questions and possible medical uses. A cryo-electron microscopy (cryo-EM) structure of SopB is now available in the Protein Data Bank, although the paper describing it has not yet been published. Researchers still need to measure how much each of SopB's phosphatase and phosphotransferase activities contributes inside living cells. They also need to clarify SopB's roles in macrophages, dendritic cells and primary intestinal lining cells. The enzyme's active site could be a useful target for drugs that weaken the bacteria's ability to cause disease without killing them directly. Such anti-virulence drugs could avoid the selective pressure that drives antibiotic resistance. SopB is also conserved across the roughly 16 core effectors shared by typhoidal and non-typhoidal Salmonella types, making it a possible target for next-generation vaccines. The authors also discuss possible links to cancer. Epidemiological studies have connected long-term typhoid infection with gallbladder cancer. Repeated exposure to non-typhoidal Salmonella has also been identified as an environmental risk factor for colon cancer. Bacterial effectors, including SopB, may contribute by keeping Akt and MAPK signalling active for long periods. The authors argue that SopB's ability to make lipids that encourage cell growth directly provides a simpler explanation than the idea that it only changes the activity of host enzymes. They suggest that further study of SopB could also reveal important details about parts of cell biology that are otherwise difficult to investigate.

Subject of Research: The different ways the Salmonella protein SopB works to change phosphoinositide signalling in host cells

Article Title: Enzymatic versatility of the salmonella effector SopB enables manipulation of host phosphoinositide signalling

Article References: Jensen, R. C., & Fairn, G. D. (2026). Enzymatic versatility of the salmonella effector SopB enables manipulation of host phosphoinositide signalling. Discover Bacteria, 3(1), Article 19. https://doi.org/10.1007/s44351-026-00056-8

Image Credits: AI Generated

DOI: 10.1007/s44351-026-00056-8

Keywords: Salmonella enterica, SopB, type III secretion system, phosphoinositides, phosphotransferase, Akt signalling, Salmonella-containing vacuole, vimentin, ubiquitination, Cdc42, bacterial effector, host-pathogen interaction