STEAP Proteins in Cancer, Metabolism, and Therapy

STEAP proteins are a family of four cell-surface enzymes, found only in mammals, that help cells absorb iron and copper by converting these metals into forms that can cross cell membranes. The name stands for six-transmembrane epithelial antigen of the prostate, a nod to the prostate tissue where the first member was discovered. But the family’s reach extends far beyond one organ. STEAP proteins have become a focus of cancer research because certain members are dramatically overexpressed on tumor cells while barely showing up on healthy tissue, making them attractive targets for new immunotherapies already in clinical trials.

The Four Family Members

The human STEAP family has four members, STEAP1 through STEAP4, and while they share a common architecture, their individual roles differ in important ways. Three of them, STEAP2, STEAP3, and STEAP4, function as both iron and copper reductases, meaning they can strip electrons from intracellular donors and deliver those electrons to iron(III) or copper(II) ions on the outside of the cell, converting the metals to forms the cell can take in.¹ STEAP1 is the odd one out. It shares the same transmembrane scaffold as its siblings but is missing a crucial intracellular piece needed to drive the reaction on its own. That puzzle has kept researchers busy for years.

Each family member also shows up in different tissues under normal conditions. STEAP1 is expressed at low levels in a handful of tissues, STEAP2 appears in prostate and other epithelial cells, STEAP3 is heavily expressed in blood cell precursors where iron uptake is critical for making hemoglobin, and STEAP4 shows up prominently in fat cells and liver, where it connects to inflammatory and metabolic signaling. Despite these different home bases, all four are tied together by their shared ability to move metal ions across membranes.²

How They Work

The basic mechanism that STEAP proteins use to reduce metals was revealed in detail when researchers solved the three-dimensional structure of STEAP4 using cryo-electron microscopy. The protein sits in the cell membrane as a trimer, meaning three copies assemble together into a single functional unit. Each copy has two main parts: a cytoplasmic oxidoreductase domain that faces the inside of the cell and binds NADPH (a common electron-carrying molecule), and a six-helix transmembrane domain that threads through the membrane and houses two cofactors, FAD and a single heme group, stacked in a vertical chain.³

The electron relay works in steps. NADPH donates electrons to FAD at the intracellular face. The reduced FAD then migrates to a binding pocket in the transmembrane region of a neighboring subunit within the trimer, hands off one electron at a time to the heme, and the heme delivers the electron to a waiting iron(III) or copper(II) ion at the extracellular surface. The result is that the metal ion picks up an electron and becomes iron(II) or copper(I), the reduced forms that membrane transport proteins can shuttle inside the cell.⁴ This domain-swapped arrangement, where FAD from one subunit accepts electrons from the neighboring subunit’s NADPH domain, is an unusual design feature that distinguishes STEAP proteins from simpler reductases.

The STEAP1 Puzzle

STEAP1 has been the most intensely studied family member in cancer research, yet for years its biochemistry was baffling. It has the transmembrane scaffold, the heme-binding site, and the FAD pocket, but it completely lacks the intracellular NADPH-binding domain that provides electrons to the other three members. When researchers tested STEAP1 alone in cells, it showed no ability to reduce iron.⁵

The breakthrough came from a structural insight. Because STEAP proteins naturally form trimers, and because the electron relay is domain-swapped between adjacent subunits, STEAP1 does not necessarily need its own NADPH domain. If it pairs up with STEAP2, STEAP3, or STEAP4 in a mixed trimer, the partner’s NADPH domain could supply the electrons, and STEAP1’s transmembrane channel could still pass them through its FAD and heme to reduce metal ions on the outside. When researchers fused STEAP1 to the NADPH-binding domain of STEAP4, the resulting chimera was indeed able to reduce iron(III), confirming the idea that STEAP1 is a functional reductase when it forms heterotrimers with other family members.⁶ The conserved sequence motifs in STEAP1’s transmembrane domain had strongly suggested this latent capability all along.⁷

This matters beyond biochemistry. If STEAP1 depends on partners to function, then the composition of STEAP trimers in a given cell type could fine-tune metal uptake in ways we don’t yet fully understand. Tumors that overexpress STEAP1 alongside STEAP2, for instance, might have enhanced iron import, which could fuel the rapid cell growth that cancer demands.

STEAP3 and Red Blood Cell Maturation

STEAP3, originally called TSAP6, has the clearest link to normal physiology. It is the primary ferrireductase in developing red blood cells, where enormous amounts of iron must be imported to build hemoglobin. Mice engineered to lack STEAP3 develop microcytic anemia, a condition in which red blood cells are abnormally small because they cannot load enough iron.⁸ Beyond iron, STEAP3 has a surprising second job: it is required for normal exosome secretion. Exosomes are tiny vesicles that cells release to communicate with neighbors or dispose of unneeded surface proteins. In STEAP3-null cells, exosome production drops sharply, including a specific secretory pathway triggered by DNA damage through the tumor suppressor p53.⁹ How exactly a metal reductase also governs vesicle trafficking remains an open question, but it illustrates how STEAP proteins participate in cell biology beyond their metal-handling role.

STEAP4 at the Crossroads of Inflammation and Metabolism

STEAP4, also known as STAMP2, occupies different territory. It is highly expressed in adipose tissue and liver and appears to sit at a junction between inflammatory signaling and metabolic regulation. When researchers silenced STEAP4 in fat cells, insulin signaling weakened: insulin-stimulated glucose transport dropped, the glucose transporter Glut4 moved less efficiently to the cell surface, and the phosphorylation cascade downstream of the insulin receptor was blunted.¹⁰

The working hypothesis is that STEAP4 helps protect fat cells from the metabolic damage caused by chronic low-grade inflammation, the kind associated with obesity. Inflammatory signals ramp up STEAP4 expression, and STEAP4 in turn counteracts the insulin resistance those same signals would otherwise produce. Without STEAP4, the feedback loop breaks, and metabolic homeostasis deteriorates. This places STEAP4 in the broader story of how obesity-driven inflammation leads to type 2 diabetes, though most of what we know still comes from cell and animal models rather than human clinical data.

STEAP Proteins in Cancer

The cancer angle is what has drawn the most attention and funding to STEAP research. STEAP1 is overexpressed across a wide range of tumor types, including prostate, bladder, colon, pancreatic, ovarian, testicular, breast, and cervical cancers, as well as Ewing sarcoma, while appearing at only minimal levels in most normal tissues.¹¹ That expression gap is exactly what drug developers look for: a surface protein abundant on cancer cells and scarce on healthy ones, reducing the chance that a targeted therapy will cause collateral damage.

The relationship between STEAP expression and tumor behavior is more nuanced than “more protein equals worse cancer.” In lung cancer, for example, STEAP1 is upregulated and its expression correlates with greater migration and invasion of cancer cells, while STEAP2 is actually downregulated and lower levels are associated with worse prognosis.¹² In breast cancer, the picture can flip: one analysis found that higher levels of STEAP1, STEAP2, and STEAP4 were associated with better survival, while low expression correlated with higher mortality.¹³ In oral squamous cell carcinoma, overexpressing STEAP1 actually inhibited tumor cell proliferation, migration, and invasion while reducing intracellular reactive oxygen species.¹⁴

The same duality appears at the molecular level with epithelial-mesenchymal transition (EMT), the process by which tumor cells become more mobile and invasive. In breast cancer cell lines, knocking down STEAP1 enhanced invasion and boosted markers of EMT, while increasing STEAP1 reversed those effects.¹⁵ So in some contexts STEAP1 acts more like a tumor suppressor than a promoter. The simplest way to make sense of this is that STEAP proteins’ roles depend heavily on the cellular environment. A protein that helps regulate iron and copper can either restrain or accelerate cancer depending on what other pathways are active, what metals are available, and what tissue the tumor originates from.

STEAP2 has its own cancer storyline. In prostate cancer cells, it promotes proliferation through activation of the ERK signaling pathway and acts as a survival factor: when researchers knocked it down, more cancer cells underwent programmed cell death.¹⁶ The evidence here is less ambiguous than for STEAP1; in prostate cancer at least, STEAP2 seems to consistently favor tumor growth.

Drugs and Immunotherapies Targeting STEAP1

The most advanced therapeutic efforts target STEAP1, primarily in metastatic castration-resistant prostate cancer (mCRPC), a stage of the disease where standard hormone-blocking treatments have stopped working and options are limited. Several distinct strategies are in various stages of development and clinical testing.

The antibody-drug conjugate DSTP3086S was one of the first to reach patients. In a phase I trial of 62 patients who received therapeutic doses, roughly one in five showed a 50% or greater decline in prostate-specific antigen (PSA), and about 6% of those with measurable tumors achieved a partial radiographic response. Among patients with high circulating tumor cell counts at the start of treatment, about 59% converted to favorable counts.¹⁷ These are modest numbers for a first-in-human dose-escalation study, but they were enough to confirm the basic idea: drugs could reach STEAP1-expressing tumors and produce measurable antitumor effects.

A more recent approach is xaluritamig (AMG 509), a bispecific T-cell engager. This molecule has two arms that grab STEAP1 on the tumor cell and one arm that grabs CD3 on a T cell, physically bridging the two so the immune cell can kill the cancer cell. Its design exploits avidity, meaning it binds more tightly to cells with lots of STEAP1 than to normal cells with little, providing a margin of selectivity. In preclinical models, xaluritamig caused tumor regression, and a first-in-human study has reported objective responses in mCRPC patients.¹⁸¹⁹

CAR-T cell therapy against STEAP1 is also moving forward. Researchers have engineered T cells with chimeric antigen receptors that recognize STEAP1 on prostate cancer cells. In mouse models, these CAR-T cells demonstrated antitumor activity even when tumor cells expressed relatively low levels of the target, and a human STEAP1 knock-in mouse model suggested the approach was safe.²⁰ One limitation that has already emerged, however, is antigen escape. Tumors can downregulate or lose STEAP1 expression under therapeutic pressure, which blunts the treatment’s effectiveness and has been linked to reduced antigen processing and presentation by the tumor cells.²¹ This is a familiar problem in targeted immunotherapy and is one reason some groups are combining STEAP1 CAR-T cells with tumor-localized cytokine delivery, such as IL-12, to boost the broader immune response.

Imaging STEAP1 to Guide Treatment

If you are going to target a protein with a drug, you need to know which patients actually have tumors that express it. Radiolabeled antibodies against STEAP1 are being developed for PET/CT imaging that can map STEAP1-positive disease throughout the body. In a clinical study of patients with mCRPC, an antibody labeled with zirconium-89 (89Zr-DFO-MSTP2109A) showed excellent tumor uptake, with median peak standardized uptake values of about 21 in bone lesions and 17 in soft-tissue lesions.²² Those are high numbers, indicating strong binding and clear tumor visualization on scans.

Newer work is refining the chemistry. A head-to-head comparison of two different chelators, DFO and HOPO, for attaching zirconium-89 to an anti-STEAP1 antibody found that the HOPO version achieved comparable or better tumor uptake while significantly reducing off-target accumulation in bone, which is a common nuisance with zirconium-89 tracers because free zirconium naturally homes to bone.²³ Less nonspecific bone signal is important in prostate cancer specifically, because bone metastases are extremely common and you don’t want the tracer’s natural bone affinity to obscure or mimic real disease. These imaging agents could eventually serve as companion diagnostics, identifying patients most likely to benefit from STEAP1-directed therapy before treatment begins.

Evolutionary Origins

STEAP proteins are unique to mammals, but the enzymatic domain they use is much older. Phylogenetic analysis suggests that the ferric reductase domain in STEAP proteins is related to YedZ, a protein found in bacteria, and that the STEAP lineage may have entered the animal kingdom through ancient horizontal gene transfer rather than straightforward vertical inheritance.²⁴ Plants and fungi use an entirely different family of ferric reductases (called eFREs) to import iron. Animals never acquired an eFRE ortholog. Instead, STEAP proteins fill that niche, carrying a telltale structural signature that distinguishes them from the plant enzymes: a conserved arginine replaces one of the canonical histidines that coordinate the heme in other ferric reductase superfamily members.²⁵

This evolutionary backstory helps explain why STEAP proteins look so different from ferric reductases studied in yeast or plant biology, and why researchers working on iron metabolism in model organisms like yeast sometimes overlook them. Mammalian iron reduction at the endosomal membrane is fundamentally a STEAP job, not an eFRE job, and understanding how STEAP proteins diverged from their bacterial ancestor continues to shed light on which structural features are essential for function and which have been repurposed for the additional roles these proteins play in immunity, metabolism, and cancer.