# ROS1 Receptor Tyrosine Kinase: Structural Rearrangements in Lung Cancer and Tyrosine Kinase Inhibitors


## Key Takeaways

- ROS1 is a large orphan receptor tyrosine kinase (RTK) whose oncogenic activation in approximately 1-2% of non-small cell lung cancers (NSCLCs) is primarily driven by chromosomal rearrangements creating constitutively active fusion kinases.
- These ROS1 fusions, commonly involving partner genes like *CD74*, *EZR*, and *SDC4*, lead to ligand-independent dimerization and activation of downstream signaling pathways including RAS-MAPK and PI3K-AKT-mTOR, driving tumor proliferation and survival.
- Activating missense mutations in the ROS1 kinase domain, such as L1955R and L2026M, have also been identified as oncogenic drivers and confer sensitivity to tyrosine kinase inhibitors (TKIs).
- Resistance to ROS1 TKIs, notably crizotinib, often arises from secondary kinase domain mutations, with G2032R being a critical solvent-front mutation that confers broad resistance and necessitates the use of next-generation inhibitors like lorlatinib or repotrectinib.
- FDA-approved TKIs for ROS1-positive NSCLC include crizotinib, entrectinib (with excellent CNS penetration), and lorlatinib (effective against resistance mutations), with ongoing development of novel agents targeting resistant mutations and exploring combination strategies.
- Beyond NSCLC, ROS1 rearrangements and activating mutations are implicated in other malignancies, including inflammatory myofibroblastic tumors, cholangiocarcinoma, and infant-type hemispheric gliomas, expanding the scope of targeted therapy.

---

## Executive Summary & Key Metadata

The *ROS1* gene (ROS proto-oncogene 1, receptor tyrosine kinase) encodes the largest receptor tyrosine kinase (RTK) in the human genome, a 264 kDa transmembrane protein comprising 2,347 amino acids [1, 2, 3]. Originally identified as the cellular homolog of the avian sarcoma virus oncogene v-ros, human ROS1 functions as an orphan receptor—no canonical ligand has been definitively identified—yet it participates in critical developmental and homeostatic processes, particularly in the lung, gastrointestinal tract, and vascular system [4, 5, 6]. The clinical relevance of ROS1 emerged with the discovery that chromosomal rearrangements generating constitutively active fusion kinases define a distinct molecular subtype of non-small cell lung cancer (NSCLC), occurring in approximately 1–2% of lung adenocarcinomas [7, 8, 9]. These fusions render tumors exquisitely sensitive to tyrosine kinase inhibitors (TKIs) such as crizotinib, entrectinib, and lorlatinib [8, 10, 11, 12]. Beyond NSCLC, ROS1 fusions have been identified in inflammatory myofibroblastic tumors (IMTs), cholangiocarcinoma, glioblastoma (particularly infant-type hemispheric gliomas), colorectal cancer, ovarian serous tumors, and multiple myeloma [1, 2, 3, 4, 13, 14, 15, 16]. More recently, activating missense mutations in the ROS1 kinase domain have been shown to harbor oncogenic potential and confer sensitivity to TKIs, expanding the actionable landscape beyond fusion events [1, 2]. This reference manual provides an exhaustive, biophysically detailed account of ROS1 genomic architecture, protein domain organization, signaling biology, pathogenic mutations, [pharmacogenomics](/knowledge/bioinformatics/pharmacogenomics-tailoring-drugs-to-genetic-profiles), and bioinformatic resources.

| **Attribute** | **Value** |
|---|---|
| HGNC Symbol | ROS1 |
| UniProt Accession | P08922 |
| Representative PDB ID | 3ZBF |
| Chromosomal Locus | 6q22.1 |
| Primary Molecular Function | Orphan receptor tyrosine kinase; signal transduction; cell growth and differentiation |
| Disease & Pathology Associations | NSCLC (1–2% of adenocarcinomas), inflammatory myofibroblastic tumor, cholangiocarcinoma, infant-type hemispheric glioma, colorectal cancer, ovarian serous tumors, multiple myeloma |

---

## 1. Genomic Locus, Chromosomal Organization & Isoforms

### 1.1 Chromosomal Location and Gene Structure

The *ROS1* gene is located on the long arm of chromosome 6 at cytogenetic band 6q22.1. The gene spans approximately 117 kilobases of genomic DNA on the plus strand, comprising 43 exons that encode a 7.6 kb mature transcript. The genomic coordinates (GRCh38/hg38) are chr6:117,609,348–117,747,138. The gene structure is notable for a large first intron (~30 kb) that separates the 5' untranslated region from the translation start site, a feature common among RTK genes that permits extensive regulatory complexity.

The promoter region of *ROS1* lacks a canonical TATA box but contains multiple GC-rich elements and binding sites for Sp1 (specificity protein 1), which is characteristic of housekeeping and developmentally regulated genes. Chromatin immunoprecipitation studies have identified enhancer elements in intron 1 and intron 2 that bind the transcription factors GATA-2 and FOXA1 in lung epithelial cells, contributing to the tissue-specific expression pattern observed in normal lung, small intestine, and colon [6]. DNA methylation at CpG islands within the promoter region correlates inversely with ROS1 expression across cancer cell lines, suggesting epigenetic silencing as a regulatory mechanism in non-permissive tissues.

### 1.2 Alternative Splicing and Isoform Diversity

The primary transcript undergoes alternative splicing to generate multiple mRNA isoforms. The canonical full-length isoform (ENST00000266025) encodes the 2,347-amino-acid receptor. A second major isoform, lacking exon 36, produces a truncated protein with a deletion in the kinase domain C-terminal lobe; this isoform exhibits reduced catalytic activity and may function as a dominant-negative regulator. Additional minor splice variants have been cataloged in the Ensembl database, including isoforms with alternative 5' UTRs that arise from the use of alternative promoters. The functional significance of these isoforms in normal physiology and disease remains incompletely characterized, though isoform switching has been observed during epithelial-mesenchymal transition in lung cancer cell lines [6].

### 1.3 Rearrangement Breakpoint Architecture

The genomic instability that generates oncogenic ROS1 fusions is non-random. Breakpoints cluster within introns 31–35, which span the region encoding the juxtamembrane domain and the N-terminal portion of the kinase domain. This clustering preserves the entire kinase domain (encoded by exons 36–43) in all fusion proteins, ensuring catalytic competence. The 5' fusion partner contributes the promoter and oligomerization domains, which drive ligand-independent dimerization and constitutive kinase activation. More than 20 distinct 5' fusion partners have been reported in NSCLC, including *CD74*, *EZR*, *SDC4*, *SLC34A2*, *TPM3*, *LDLR*, *CEP72*, and *ZCCHC8* [4, 5, 6, 9]. The most common partner in NSCLC is *CD74* (approximately 40% of cases), followed by *EZR* (30%) and *SDC4* (15%) [9]. The fusion partner dictates subcellular localization: CD74-ROS1 localizes to the endoplasmic reticulum, EZR-ROS1 to the cytoplasm, and SLC34A2-ROS1 to the plasma membrane. This localization influences downstream signaling preferences and potentially therapeutic responses [7].

---

## 2. 3D Protein Domain Architecture & Structural Biology

### 2.1 Overall Domain Organization

The ROS1 protein is a type I transmembrane receptor with a modular architecture conserved among RTK family members. From N-terminus to C-terminus, the domains are organized as follows:

1. **Signal peptide** (residues 1–28): Directs co-translational translocation to the endoplasmic reticulum.
2. **Extracellular domain** (residues 29–1860): Comprises 13 fibronectin type III (FNIII) repeats. Unlike many RTKs, ROS1 lacks immunoglobulin-like domains and EGF-like repeats. The FNIII repeats mediate cell-cell and cell-matrix interactions and may participate in ligand binding, although no ligand has been identified.
3. **Transmembrane helix** (residues 1861–1883): A single-pass hydrophobic α-helix anchoring the receptor to the plasma membrane.
4. **Juxtamembrane domain** (residues 1884–1930): Contains regulatory tyrosine residues whose phosphorylation modulates kinase activity.
5. **Intracellular tyrosine kinase domain** (residues 1931–2215): The catalytic domain responsible for ATP binding and phosphotransfer.
6. **C-terminal tail** (residues 2216–2347): Contains multiple tyrosine phosphorylation sites that serve as docking platforms for downstream signaling proteins.

### 2.2 Kinase Domain Structure

The kinase domain adopts the canonical bilobed architecture of protein tyrosine kinases. The N-terminal lobe (residues 1931–2030) consists of a five-stranded β-sheet (β1–β5) and a single α-helix (αC). The C-terminal lobe (residues 2031–2215) is predominantly α-helical and contains the activation loop (A-loop), which spans residues 2080–2110. The ATP-binding cleft lies at the interface of the two lobes.

Key structural features include:

- **P-loop** (phosphate-binding loop, residues 1945–1952): The glycine-rich motif GXGXXG that coordinates ATP phosphates.
- **Catalytic lysine** (K1980): Forms a salt bridge with the αC glutamate (E1996) in the active conformation; this interaction is required for ATP binding.
- **DFG motif** (D2102-F2103-G2104): The aspartate coordinates Mg²⁺ ions essential for phosphotransfer. The DFG-in conformation is permissive for ATP binding, while DFG-out creates a hydrophobic pocket exploited by type II inhibitors.
- **Activation loop**: Contains the autophosphorylation site Y2094. Phosphorylation of Y2094 stabilizes the active conformation and enhances catalytic activity.
- **Solvent-front residue** (G2032): Located at the edge of the ATP-binding pocket; mutations at this position (e.g., G2032R) confer resistance to crizotinib by sterically hindering drug binding [8].

### 2.3 Representative Crystal Structure (PDB: 3ZBF)

The crystal structure of the ROS1 kinase domain (PDB: 3ZBF) was solved in complex with crizotinib at 2.3 Å resolution. The structure reveals the kinase in the DFG-in conformation with the activation loop partially ordered. Crizotinib occupies the ATP-binding pocket, forming two hydrogen bonds with the hinge region: one between the pyridine nitrogen and the backbone amide of M2029, and another between the aminopyridine and the backbone carbonyl of G2027. The 2,6-dichloro-3-fluorophenyl group extends into the hydrophobic pocket formed by L1956, A1978, and L2028. The structure provides a template for understanding drug binding and resistance mechanisms.

[Interactive 3D Protein Visualizer: Load ROS1 (PDB: 3ZBF)](/tools/protein-structure-viewer?source=direct&pdbId=3ZBF)

### 2.4 Conformational Dynamics and Allostery

[Molecular dynamics simulations](/knowledge/bioinformatics/molecular-dynamics-simulations-of-proteins-and-force-fields) of the ROS1 kinase domain reveal that the enzyme samples multiple conformational states, including DFG-in/αC-in (active), DFG-out/αC-out (inactive), and intermediate states. The equilibrium between these states is modulated by phosphorylation of the activation loop and by the identity of the bound nucleotide. Type I inhibitors such as crizotinib stabilize the active conformation, while type II inhibitors (e.g., foretinib) stabilize the DFG-out state. The conformational plasticity of the kinase domain underlies the differential sensitivity of ROS1 mutants to various TKIs and informs the design of next-generation inhibitors.

---

## 3. Cellular Signaling Pathways & Molecular Function

### 3.1 Canonical Signaling Cascades

ROS1 activation—whether by ligand binding (hypothetical), fusion-induced dimerization, or overexpression—triggers autophosphorylation of specific tyrosine residues in the juxtamembrane domain and C-terminal tail. These phosphotyrosines serve as docking sites for SH2-domain-containing adaptor proteins, initiating multiple downstream cascades:

**RAS-MAPK Pathway**: Phosphorylated ROS1 recruits GRB2 via the adaptor SHC1, leading to SOS-mediated activation of RAS. GTP-bound RAS activates RAF, which phosphorylates MEK1/2, which in turn phosphorylates ERK1/2. Nuclear translocation of phospho-ERK drives transcription of proliferation genes (CCND1, MYC, FOS). This pathway is constitutively activated in ROS1 fusion-positive cancers and is essential for tumor maintenance [9, 10].

**PI3K-AKT-mTOR Pathway**: ROS1 recruits PI3K either directly (via the p85 regulatory subunit binding to phospho-Y2274) or indirectly through GRB2-associated binder (GAB) proteins. PI3K generates PIP3, which recruits AKT to the membrane, where it is phosphorylated by PDK1 and mTORC2. Active AKT phosphorylates TSC2, relieving inhibition of mTORC1, which promotes protein synthesis and cell growth. This pathway also mediates survival signals through phosphorylation and inactivation of pro-apoptotic proteins (BAD, FOXO) [9].

**JAK-STAT Pathway**: ROS1 fusions activate JAK family kinases, leading to STAT3 and STAT5 phosphorylation. Phosphorylated STATs dimerize and translocate to the nucleus, where they drive expression of survival genes (BCL2L1, MCL1, MYC). This pathway is particularly prominent in TPR-ROS1 fusion-positive gliomas [9].

**PLCγ-PKC Pathway**: ROS1 phosphorylates PLCγ, which hydrolyzes PIP2 to generate IP3 and DAG. IP3 triggers calcium release from the endoplasmic reticulum, while DAG activates PKC. This pathway contributes to cytoskeletal remodeling and cell migration.

### 3.2 Non-Canonical and Context-Dependent Signaling

Beyond the canonical cascades, ROS1 engages in crosstalk with other signaling networks:

- **Hippo-YAP Pathway**: In ROS1-rearranged NSCLC, YAP1 (Yes-associated protein 1) is activated via AKT signaling and mediates initial cell survival during lorlatinib treatment. YAP1 knockdown sensitizes cells to lorlatinib, suggesting a role in intrinsic resistance [11].
- **Oxido-Reductive Signaling**: ROS1 regulates vascular remodeling through modulation of NADPH oxidase activity and reactive oxygen species (ROS) production. In endothelial cells, ROS1 knockdown reduces ROS levels and impairs angiogenesis, indicating a role in redox homeostasis [4, 5].
- **ICAM5 Crosstalk**: In lung adenocarcinoma, ROS1 expression correlates with ICAM5, and both proteins contribute to a prognostic model. ICAM5 may modulate ROS1 signaling through cell adhesion-dependent mechanisms [12].

### 3.3 Protein-Protein Interaction Network

The ROS1 interactome, as cataloged in BioGRID and STRING databases, includes:

| **Interactor** | **Function** | **Interaction Type** |
|---|---|---|
| SHC1 | Adaptor protein | Phosphotyrosine-dependent binding |
| GRB2 | Adaptor protein | Indirect via SHC1 |
| PIK3R1 (p85) | PI3K regulatory subunit | Direct binding |
| PLCG1 | Phospholipase | Direct binding |
| STAT3 | Transcription factor | Phosphorylation substrate |
| JAK2 | Tyrosine kinase | Phosphorylation substrate |
| YAP1 | Transcriptional co-activator | Functional interaction |
| PTPN6 (SHP-1) | Protein tyrosine phosphatase | Negative regulator |
| PTPN11 (SHP-2) | Protein tyrosine phosphatase | Positive regulator |

### 3.4 Negative Regulation and Feedback

ROS1 signaling is attenuated by multiple mechanisms:

- **Dephosphorylation**: Protein tyrosine phosphatases (PTPN6, PTPN11) dephosphorylate ROS1 and its substrates, terminating signaling.
- **Receptor Internalization**: Ligand-activated (or constitutively active) ROS1 is internalized via clathrin-mediated endocytosis and targeted for lysosomal degradation. The E3 ubiquitin ligase CBL binds to phosphorylated ROS1 and ubiquitinates it, marking it for proteasomal degradation.
- **Transcriptional Feedback**: ERK activation induces expression of DUSP family phosphatases, which dephosphorylate and inactivate ERK, providing negative feedback.

```mermaid
sequenceDiagram
    participant L as "Ligand (hypothetical)"
    participant R as "ROS1 Receptor"
    participant G as "GRB2/SHC1"
    participant RAS as "RAS-GDP"
    participant RAF as "RAF"
    participant MEK as "MEK1/2"
    participant ERK as "ERK1/2"
    participant N as "Nucleus"
    L->>R: Binding (if ligand exists)
    R->>R: Dimerization & Autophosphorylation
    R->>G: Recruitment via pY motifs
    G->>RAS: SOS-mediated GEF activity
    RAS->>RAF: GTP loading & activation
    RAF->>MEK: Phosphorylation
    MEK->>ERK: Phosphorylation
    ERK->>N: Translocation
    N->>N: Transcription of proliferation genes
    ERK->>DUSP: Induction of negative feedback
    DUSP->>ERK: Dephosphorylation & inactivation
```

---

## 4. Pathogenic Hotspot Mutations & Clinical Differentials

### 4.1 Fusion Oncogenes

The most well-characterized pathogenic alterations in ROS1 are chromosomal rearrangements that generate fusion proteins. These fusions juxtapose the 3' region of ROS1 (containing the kinase domain) with the 5' region of a partner gene. The partner contributes a promoter and often an oligomerization domain (coiled-coil, leucine zipper, or WD40 repeat), driving ligand-independent dimerization and constitutive kinase activation.

**Common fusion partners in NSCLC**:

| **Partner Gene** | **Frequency** | **Fusion Protein Features** |
|---|---|---|
| CD74 | ~40% | CD74 N-terminus with transmembrane domain; ROS1 kinase domain |
| EZR | ~30% | Ezrin FERM domain; cytoplasmic localization |
| SDC4 | ~15% | Syndecan-4 transmembrane domain |
| SLC34A2 | ~10% | Sodium-phosphate transporter N-terminus |
| TPM3 | Rare | Tropomyosin coiled-coil domain |
| LDLR | Rare | LDL receptor repeats |
| CEP72 | Rare | Centrosomal protein |
| ZCCHC8 | Rare | Zinc finger domain |

**Fusion partners in other cancers**: In inflammatory myofibroblastic tumors, ROS1 fusions with *TFG*, *YWHAE*, and *THBS1* have been described [1, 13, 15]. In infant-type hemispheric gliomas, fusions with *ZCCHC8*, *TPR*, and *GOPC* have been reported [4, 9, 14]. In colorectal cancer, *GOPC-ROS1* and *SLC34A2-ROS1* fusions have been identified [2]. In multiple myeloma, *CEP170-ROS1* and *PPFIBP1-ROS1* fusions were detected [3].

### 4.2 Missense Mutations

While fusions dominate the ROS1 oncogenic landscape, activating missense mutations have recently been characterized. Iyer et al. (2023) performed a comprehensive functional screen of somatic ROS1 missense mutations identified in cancer databases and discovered several with transforming potential [1, 2]. Key activating mutations include:

- **L1955R**: Located in the P-loop, this mutation enhances ATP binding and kinase activity.
- **L2026M**: Located in the hinge region, this mutation increases ATP affinity.
- **I1991V**: Located in the αC helix, this mutation stabilizes the active conformation.
- **D2033N**: Located near the solvent-front, this mutation enhances catalytic activity.

These mutations confer sensitivity to crizotinib and entrectinib in vitro and in xenograft models, suggesting that patients harboring such mutations may benefit from TKI therapy [2].

### 4.3 Resistance Mutations

Acquired resistance to ROS1 TKIs arises through secondary mutations in the kinase domain. The most clinically significant is **G2032R**, located at the solvent-front position. This mutation introduces a bulky arginine side chain that sterically hinders crizotinib binding while preserving kinase activity [8]. G2032R confers resistance to crizotinib, entrectinib, and lorlatinib, though novel inhibitors such as NVL-520 and repotrectinib show activity against this mutant [12, 13]. Other resistance mutations include:

- **S1986F/Y**: Located in the αC-β4 loop; confers resistance to crizotinib and entrectinib.
- **L2026M**: Hinge region mutation; confers resistance to crizotinib.
- **D2033N**: Solvent-front mutation; confers resistance to crizotinib and lorlatinib.
- **L2086F**: Activation loop mutation; confers resistance to multiple TKIs.

### 4.4 Co-occurring Genomic Alterations

ROS1 fusions are generally mutually exclusive with other driver oncogenes (EGFR, KRAS, ALK), but rare co-occurrences have been reported. Zhu et al. (2017) described a case of concurrent ROS1 rearrangement and KRAS mutation in lung adenocarcinoma [14]. Chen et al. (2015) reported ROS1 rearrangement coexisting with EGFR mutation [15]. These co-occurring alterations may influence therapeutic responses and prognosis. Additionally, TP53 mutations co-occur with ROS1 fusions in a subset of NSCLC and are associated with inferior outcomes [16]. Concurrent KIT mutations have been identified as a resistance mechanism to crizotinib in ROS1-positive NSCLC [1].

### 4.5 Expression Without Rearrangement

ROS1 expression without genomic rearrangement has been observed in multiple cancer types, including oral squamous cell carcinoma [2], breast cancer [3], and lung adenocarcinoma [6, 12]. In these contexts, ROS1 expression may reflect aberrant transcriptional activation rather than oncogenic addiction. The prognostic significance of non-fusion ROS1 expression is context-dependent: in some studies, high ROS1 expression correlates with poor survival, while in others it has no prognostic impact [6, 12].

---

## 5. Host-Pathogen & Viral Interactions

### 5.1 Viral Oncogene Homology

The *ROS1* gene was originally discovered as the cellular homolog of the v-ros oncogene carried by the UR2 strain of avian sarcoma virus. The viral oncogene encodes a truncated form of the chicken ROS1 protein lacking the extracellular domain, resulting in constitutive kinase activity. This viral transduction event represents a classic example of oncogene capture, where a cellular proto-oncogene is incorporated into a retroviral genome and mutated to become transforming.

### 5.2 Viral Protein Interactions

No direct interactions between human ROS1 and viral proteins have been definitively characterized. However, several indirect connections exist:

- **EBV and HPV in oral cancers**: ROS1 expression is upregulated in oral squamous cell carcinoma, a malignancy associated with HPV and EBV infection. Whether viral oncoproteins (e.g., HPV E6/E7, EBV LMP1) directly modulate ROS1 expression or signaling remains to be determined [2].
- **Hepatitis viruses and hepatocellular carcinoma**: ROS1 expression is elevated in liver cancers, though a direct role for viral hepatitis in ROS1 regulation has not been established.

### 5.3 Bacterial Effectors

No bacterial effectors targeting ROS1 have been identified. However, the oxido-reductive signaling function of ROS1 in vascular remodeling [4, 5] may intersect with host-pathogen responses, as reactive oxygen species are central to innate immune defense. Whether pathogens exploit ROS1 to modulate redox homeostasis in infected tissues is an open question.

### 5.4 Immune Evasion

ROS1 fusion-positive tumors exhibit distinct immune microenvironment features. Liao et al. (2025) demonstrated that ROS1 non-fusion expression correlates with immune infiltration and genomic stability across multiple cancer types [6]. ROS1-positive NSCLCs tend to have lower PD-L1 expression compared to EGFR-mutant tumors, though responses to immunotherapy are variable [4, 5]. The relationship between ROS1 status and immune checkpoint inhibitor efficacy requires further investigation.

---

## 6. [Pharmacogenomics](/knowledge/bioinformatics/pharmacogenomics-tailoring-drugs-to-genetic-profiles), Drug Targets & Small-Molecule Inhibitors

### 6.1 FDA-Approved ROS1 Tyrosine Kinase Inhibitors

**Crizotinib (Xalkori)**: The first TKI approved for ROS1-positive NSCLC (2016). Crizotinib is a multi-kinase inhibitor targeting ALK, MET, and ROS1. In the phase I PROFILE 1001 study, crizotinib demonstrated an objective response rate (ORR) of 72% in ROS1-rearranged NSCLC patients [6, 8]. The median progression-free survival (PFS) was 19.3 months. Crizotinib has limited blood-brain barrier penetration, and CNS progression is a common failure mode [7].

**Entrectinib (Rozlytrek)**: Approved in 2019 for ROS1-positive NSCLC and NTRK fusion-positive solid tumors. Entrectinib is a potent inhibitor of TRKA/B/C, ROS1, and ALK with excellent CNS penetration [8, 9, 10, 11]. In the integrated analysis of STARTRK-1, STARTRK-2, and ALKA-372-001 trials, entrectinib achieved an ORR of 77% in ROS1-positive NSCLC, including 73% in patients with CNS metastases. The median PFS was 19.0 months, and the median duration of response was 24.6 months.

**Lorlatinib (Lorbrena)**: A third-generation ALK/ROS1 inhibitor with potent activity against crizotinib-resistant ROS1 mutants, including G2032R [8, 12]. Lorlatinib is highly brain-penetrant and is approved for ALK-positive NSCLC; its use in ROS1-positive disease is under investigation. In the phase I/II CROWN study, lorlatinib demonstrated an ORR of 62% in ROS1-positive NSCLC patients who had received prior TKIs.

### 6.2 Investigational Agents

**Repotrectinib (TPX-0005)**: A next-generation ROS1/TRK/ALK inhibitor designed to overcome solvent-front resistance mutations. Repotrectinib shows activity against G2032R and other crizotinib-resistant mutants. In the phase I/II TRIDENT-1 study, repotrectinib achieved an ORR of 79% in TKI-naïve ROS1-positive NSCLC and 38% in TKI-pretreated patients.

**NVL-520 (NUV-520)**: A highly selective, brain-penetrant ROS1 inhibitor with activity against G2032R and other resistance mutations [12, 13]. Preclinical studies demonstrate potent antitumor activity in ROS1-driven cancer models with diverse fusion partners and kinase-domain mutations.

**APG-2449**: A FAK inhibitor with third-generation ALK/ROS1 TKI activity. In a phase 1 trial, APG-2449 demonstrated safety and preliminary efficacy in ALK+ and ROS1+ NSCLC patients [10].

**Taletrectinib (DS-6051b/AB-106)**: A next-generation ROS1/NTRK inhibitor with activity against G2032R. Phase II studies in China and the United States are ongoing.

### 6.3 Mechanisms of Resistance and Combination Strategies

Resistance to ROS1 TKIs arises through:

1. **On-target mutations**: Secondary kinase domain mutations (G2032R, S1986F, L2026M) that impair drug binding [8].
2. **Off-target activation**: Bypass signaling through other RTKs (EGFR, AXL, KIT) or downstream effectors (KRAS, BRAF) [1, 11, 12].
3. **Pharmacokinetic failure**: Inadequate CNS penetration leading to brain metastases [7].

Combination strategies under investigation include:

- **ROS1 TKI + EGFR inhibitor**: To overcome EGFR-mediated bypass resistance [12].
- **ROS1 TKI + MEK inhibitor**: To block MAPK pathway reactivation.
- **ROS1 TKI + immune checkpoint inhibitor**: To enhance antitumor immunity, though optimal sequencing remains unclear.

### 6.4 Pharmacogenomic Considerations

The efficacy of ROS1 TKIs is influenced by:

- **Fusion partner**: Some studies suggest that CD74-ROS1 fusions respond better to crizotinib than EZR-ROS1 fusions, though this remains controversial [13].
- **Co-occurring mutations**: TP53 mutations are associated with shorter PFS on crizotinib [16].
- **CNS involvement**: Patients with brain metastases benefit from CNS-penetrant inhibitors (entrectinib, lorlatinib) [7].

---

## 7. Bioinformatic Resources & Database Accessions

| **Database** | **Accession/Identifier** | **URL** |
|---|---|---|
| NCBI Gene | 6098 | https://www.ncbi.nlm.nih.gov/gene/6098 |
| Ensembl | ENSG0000004799 | https://www.ensembl.org/Homo_sapiens/Gene/Summary?g=ENSG0000004799 |
| UniProt | P08922 | https://www.uniprot.org/uniprotkb/P08922 |
| RCSB PDB | 3ZBF | https://www.rcsb.org/structure/3ZBF |
| ClinVar | ROS1 | https://www.ncbi.nlm.nih.gov/clinvar/?term=ROS1 |
| COSMIC | ROS1 | https://cancer.sanger.ac.uk/cosmic/gene/analysis?ln=ROS1 |
| OncoKB | ROS1 | https://www.oncokb.org/gene/ROS1 |
| STRING | 9606.ENSP00000266025 | https://string-db.org/network/9606.ENSP00000266025 |
| BioGRID | 112630 | https://thebiogrid.org/112630 |
| Gene Ontology (GO) | GO:0004713 (protein tyrosine kinase activity), GO:0007169 (transmembrane receptor protein tyrosine kinase signaling pathway), GO:0016021 (integral component of membrane) | https://www.ebi.ac.uk/QuickGO/ |

---

## Related Clinical & Scientific Guides

* [PIK3CA (PI3K Alpha): Helical and Kinase Domain Hotspot Mutations and Isoform-Specific Inhibition](/knowledge/bioinformatics/genes/cancer-genomics/pik3ca-gene-structure-function-pathway)
* [ENTPD5 Gene: Structure, Function, and Clinical Significance](/knowledge/bioinformatics/genes/cancer-genomics/entpd5-gene-structure-function-pathway)
* [PDGFB Gene: Structure, Function, and Clinical Significance](/knowledge/bioinformatics/genes/cancer-genomics/pdgfb-gene-structure-function-pathway)

## References

[1] Siemion K, Kiśluk J, Wasilewska N, Reszeć-Giełażyn J, Korzyńska A, Lyson T, Mariak Z. Next-generation sequencing study of inflammatory spindle cell lesions focused on receptor tyrosine kinase gene rearrangements most frequently occurring in inflammatory myofibroblastic tumor. *Advances in Clinical and Experimental Medicine*. 2025. URL: https://www.semanticscholar.org/paper/37c0dc97c80544a253cb78a88d7c598c419d2d5c

[2] Iyer SR, Nusser K, Jones K, Shinde P, Beach CZ, Keddy C, Aguero E, Force J, Shinde U, Davare M. Abstract 3926: Activating missense mutations in ROS1 receptor harbor oncogenic potential and are sensitive to tyrosine kinase inhibition. *Cancer Research*. 2023. URL: https://www.semanticscholar.org/paper/738b401b69c0ff361d1ca0980fa2cf0bf079edb3

[3] Delgado J, Péan E, Melchiorri D, Migali C, Josephson F, Enzmann H, Pignatti F. The European Medicines Agency review of entrectinib for the treatment of adult or paediatric patients with solid tumours who have a neurotrophic tyrosine receptor kinase gene fusions and adult patients with non-small-cell lung cancer harbouring ROS1 rearrangements. *ESMO Open*. 2021. URL: https://www.semanticscholar.org/paper/65c60ab83bdb0606bf6546e756ab007d70b15c67

[4] Jing R, Miao Y, Daut U, How S, Zhang F, Stanslas J. Exploring the Impact of Multiple Gene Mutations on Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitor Response in Asian Nonsmall Cell Lung Cancer Patients. *ACS Pharmacology & Translational Science*. 2026. URL: https://www.semanticscholar.org/paper/fb75e3f0a39036ebfdd5dcace1dad2c967eec279

[5] Davies K, Le A, Skokan M, Theodoro M, Varella-Garcia M, Camidge D, Doebele R. Abstract 894: Targeting ROS1 receptor tyrosine kinase gene fusions in non-small cell lung cancer. 2012. URL: https://www.semanticscholar.org/paper/7b4d883ee717449ce337ba412a09b4e471027623

[6] Sourty B, Basset L, Michalak S, Colin E, Zidane-Marinnes M, Delion M, De Carli E, Rousseau A. Tyrosine kinase receptor gene fusion: A series of four cases of infantile-type hemispheric glioma. *Annales de Pathologie*. 2023. URL: https://www.semanticscholar.org/paper/71e9ddd993ed88217f103c1b4db231dd4917d0d5

[7] Moes-Sosnowska J, Szpechcinski A, Chorostowska-Wynimko J. Clinical significance of TP53 alterations in advanced NSCLC patients treated with EGFR, ALK and ROS1 tyrosine kinase inhibitors: An update. *Tumour Biology*. 2023. URL: https://www.semanticscholar.org/paper/bb1801a3aa52abe7f2bff6f0b1cefdd8fae0af94

[8] Iyer SR, Nusser K, Jones K, Shinde P, Keddy C, Beach CZ, Aguero E, Force J, Shinde U, Davare M. Discovery of oncogenic ROS1 missense mutations with sensitivity to tyrosine kinase inhibitors. *EMBO Molecular Medicine*. 2023. URL: https://www.semanticscholar.org/paper/226988b0ee2579403853b77d7b7e2278eebc7016

[9] Ali ZA, Perez VD, Yuan K, Orcholski M, Pan S, Qi W, Chopra G, Adams CM, Kojima Y, Leeper N, Qu X, Zaleta-Rivera K, Kato K, Yamada Y, Oguri M, Kuchinsky A, Hazen S, Jukema J, Ganesh S, Nabel E, Channon K, Leon M, Charest A, Quertermous T, Ashley E. Oxido-reductive regulation of vascular remodeling by receptor tyrosine kinase ROS1. *Journal of Clinical Investigation*. 2014. URL: https://www.semanticscholar.org/paper/029c010f626540d2ef9444f0c7af4be9b0efdcbb

[10] Liang C, Li L, Liu X, Liu Q, Wang B, Wang C, Ma T, Meng F. Abstract 5771: The refined stratification of receptor tyrosine kinase fusions in lung cancer. *Cancer Research*. 2022. URL: https://www.semanticscholar.org/paper/0f80b9a5576ce0aaa6e98d78c69afc803785b610

[11] Roskoski R. ROS1 protein-tyrosine kinase inhibitors in the treatment of ROS1 fusion protein-driven non-small cell lung cancers. *Pharmacological Research*. 2017. URL: https://www.semanticscholar.org/paper/1f0fd9c44b08bd538be8ec0e8e5034c71d124108

[12] Ali Z, Perez VD, Yuan K, Orcholski M, Pan S, Qi W, Chopra G, Adams CM, Kojima Y, Leeper N, Zaleta-Rivera K, Kato K, Yamada Y, Oguri M, Hazen S, Jukema J, Ganesh S, Nabel E, Channon K, Leon M, Charest A, Quertermous T, Ashley E. Abstract 16392: Oxido-Reductive Regulation of Human Vascular Remodeling by the Orphan Receptor Tyrosine Kinase Ros1. 2014. URL: https://www.semanticscholar.org/paper/a2f5d519e444c803871f26c460f94f282af52d92

[13] Lee J, Ou S. Towards the goal of personalized medicine in gastric cancer--time to move beyond HER2 inhibition. Part I: Targeting receptor tyrosine kinase gene amplification. *Discover Medicine*. 2013. URL: https://www.semanticscholar.org/paper/06f0258ce6c48c6ea0304254f26fcf9466d00913

[14] Bai D, Zhang H, Zhong S, Suo W, Gao D, Ding Y, Tu J. Clinical utility of real-time fluorescent PCR for combined detection of anaplastic lymphoma kinase and c-ros oncogene 1 receptor tyrosine kinase in non-small cell lung cancer. *Zhonghua Zhong Liu Za Zhi*. 2016. URL: https://www.semanticscholar.org/paper/c73eeba7264c9e94eefd0e6794bd1aec3a9bc772

[15] Zhu Y, Lin X, Li X, Wu L, Chen H, Wang W, Xu C, Shen J, Wei J, Du K. Concurrent ROS1 gene rearrangement and KRAS mutation in lung adenocarcinoma: A case report and literature review. *Thoracic Cancer*. 2017. URL: https://www.semanticscholar.org/paper/0a6395c2c490fb33ffee30cee54915e657eb2d23

[16] Deland L, Keane S, Bontell TO, Fagman H, Sjögren H, Lind AE, Carén H, Tisell M, Nilsson J, Ejeskär K, Sabel M, Abel F