Cytokines, JAK-STAT Signaling and Radiation-Induced DNA Repair in Solid Tumors: Novel Opportunities for Radiation Therapy

Abstract

A number of solid tumors are treated with radiation therapy (RT) as a curative modality. At the same time, for certain types of cancers the applicable doses of RT are not high enough to result in a successful eradication of cancer cells. This is often caused by limited pharmacological tools and strategies to selectively sensitize tumors to RT while simultaneously sparing normal tissues from RT. We present an outline of a novel strategy for RT sensitization of solid tumors utilizing Jak inhibitors. Here, recently published pre-clinical data are reviewed which demonstrate the promising role of Jak inhibition in sensitization of tumors to RT. A wide number of currently approved Jak inhibitors for non-malignant conditions are summarized including Jak inhibitors currently in clinical development. Finally, intersection between Jak/Stat and the levels of serum cytokines are presented and discussed as they relate to susceptibility to RT.

1. Introduction

Radiation therapy (RT) is a critical and central component of treatment of numerous solid tumor types. For certain solid tumors, RT is the single curative modality. Radiation-induced DNA damage causes death of cancer cells by multiple mechanisms^1–5. Indeed, DNA damage imparted by RT exposure has been a robustly studied for many decades. There is a wealth of information on the mechanisms by which a cancer cell embarks on repair of DNA after radiation-induced damage including mitotic cell death, apoptotic cell death, and in some circumstances ablative necrosis^2–5. While there are multiple types of DNA damage induced by RT, double strand DNA breaks (DSDBs) constitute the most lethal events associated with radiation exposure^1,3,6–8. This makes targeting those pathways a highly appealing strategy in order to optimize radiation induced cell damage for treatment of solid tumors^4,8,9.

2. The Key Mechanisms of Repair of Double-Strand DNA Breaks in Solid Tumors

Two major categories of DSDB damage repair include homologous recombination (HR) repair and non-homologous end joining repair (NHEJ) (Fig. 1) ^1,5,6,10,11. In brief, NHEJ DNA repair mechanism, unlike HR DNA repair, does not require a long homologous sequence to guide the repair. NHEJ is a straightforward process of re-ligation of the broken DNA ends¹² and therefore does not require a template¹². The important repair protein responsible for NHEJ is DNA-dependent protein kinase (DNA-PK), a serine-threonine protein kinase consisting of three subunits (DNA-PKcs, Ku70 and Ku80)^1,13–15. The first protein to bind the DSB is Ku heterodimer and is capable of interacting with a nuclease (Artemis-DNA-PKcs), the polymerases (u and λ) and a ligase (XLF-XRCC4-DNA ligase IV)^12,16. In contrast, HR DNA repair is a relatively error free process that uses a DNA template of the same base sequence around the break site to perform accurate repair of the DNA break^17–19. The double strand DNA breaks are initially recognized by the MRN (MRE11, RAD50, NbS1) protein complex, which together with phosphorylated CtIP (C-terminal binding protein interacting protein) and BRCA1 (Breast Cancer Susceptibility 1), generate the necessary 3’ single-stranded DNA (ssDNA) overhangs for HR repair^1,11,20–23. BRCA1 promotes this resection by dephosphorylating the inhibitory 53BP1 protein^11,20–24. Following end-resection, the MRN complex recruits and activates ATM (ataxia telangiectasia mutated) kinase, which phosphorylates histone H2AX on regions around DNA DSDBs²⁵. Importantly, loading of RAD51 recombinase onto exposed ssDNA repair sites is carried out by BRCA1 and requires assistance of BRCA2^23,26. RAD51 is a requisite for HR DNA repair as it catalyzes DNA strand invasion and exchange of the DNA strands^23,27. RAD51 paralogues including RAD51B, RAD51C, RAD51D, XRCC2 and XRCC3 have only a supporting role in this process by assisting RAD51 in the initiation and execution of HR DNA repair²⁸. It is important to note that while BRCA1/2 assist RAD51 in the HR DNA repair, RAD51 is indispensable for HR DNA repair ^11,20–22 in its key role for strand invasion and exchange^{11,20–22,28}. BRCA1/2-deficient cancer cells have defective HR DNA repair (termed BRCA-ness), and consequently increased sensitivity to RT, platinum chemotherapy and PARP-inhibitors^29–31.

Fig. 1.

Overview of repair mechanisms of DNA double-strand breaks (DSDBs). A. Non-homologous end joining repair (NHEJ) is a compact process of re-ligation of the broken DNA ends that requires minimal processing and does not require a template. It occurs throughout the cell cycle. NHEJ is initiated by the binding of the Ku 70–80 heterodimer, one of the subunits of DNA-PK to the double stranded DNA ends to protect it from nuclease digestion. DNA-bound Ku recruits the DNA-protein kinase catalytic subunit (DNA-PKcs), generating the DNA-PK complex which predominantly regulates NHEJ through autophosphorylation and facilitates recruitment of a ligation complex, which encompasses X-ray cross complementing Group 4 (XRCC4) and DNA ligase 4. DNA-PKcs further activate Artemis nuclease which cleave the DNA overhangs. The DNA polymerase λ or μ facilitates the DNA synthesis followed by ligation of the DNA gaps. B. Homologous recombination (HR) DNA repair is a predominantly accurate process that uses a sister chromatid as a template for DSDB repair and functions only in late S/G2 phase. DNA breaks are initially recognized by MRN complex which together with BRCA1and CtIP generates ssDNA overhangs. These overhangs are coated with replication protein A (RPA) which are then exchanged for RAD51 where BRCA1/2 assist the exchange. RAD51 loading promotes invasion onto the undamaged template and strand exchange followed by ligation of the DNA ends. RAD51 paralogues (RAD51B, RAD51C, RAD51D, XRCC2 and XRCC3) assist RAD51 in the HR DNA repair process. The DNA synthesis is performed by polymerase δ or ε.

Increased levels of RAD51 are commonly seen in solid tumors, and higher levels are associated with improved tumor repair capabilities. Numerous studies have suggested that higher levels of RAD51 can lead to worse clinical outcomes, which often are evident specifically in tumors that are commonly treated with RT. The correlation with RAD51 expression and clinical outcomes has been shown in esophageal cancer³², breast cancer³³, head and neck cancer³⁴, prostate cancer³⁵, soft tissue sarcoma³⁶, and pancreatic cancer³⁷, to name a few examples. In addition, cancer cells tend to have a higher ratio of cells in S and G2 phases of the cell cycle, which causes them to favor HR DNA repair over NHEJ repair^3,6,7,38.

Given the central role of RAD51 in the HR process, together with the fact that DSDBs represent a primary form of DNA damage secondary to RT exposure, RAD51 is potentially a critical protein mediating resistance of tumors to treatment with RT. Solid tumor patients with germ-line BRCA1/2 mutations, which incapacitate BRCA1/2 functions and HR DNA repair (termed BRCA-ness), are known to be more responsive to RT, platinum-based chemotherapeutics and PARP-inhibitors ^29–31,39. PARP1, through the addition of poly-ADP ribose (PAR) moieties to sites of single-strand (ss) DNA breaks, is critical for recognition and recruitment of DNA repair machinery for a variety of different DNA repair processes⁴⁰. If ssDNA breaks go unrepaired, DSDBs form during the cell cycle. In cells with intact repair pathways, HR DNA repair will function to repair the newly formed DSB of DNA. In the case of solid tumors with BRCA1/2 mutations, where BRCA-mediated HR repair is already deficient, the use of PARP inhibitors alone can promote the accumulation of DSDBs leading to selective death of tumor cells – a concept referred to as “synthetic lethality”^41–46. Given that RAD51 is overexpressed in a number of solid tumors treated with RT, targeting RAD51 represents an appealing strategy to preferentially sensitize cancer cells to treatment with RT and induce transient BRCA-ness to solid tumors^47,48.

3. Jak-Stat Signaling Pathway

Signal Transducer and Activator of Transcription (STAT) proteins are critical mediators of the cellular actions of a number of peptide hormones, growth factors and cytokines (Fig. 2)⁴⁹. The canonical pathway for activating of STAT involves tyrosine phosphorylation of STAT by a JAK tyrosine kinase. Activation of STAT is a two-step process where, first, the STAT monomer docks transiently to a phospho-tyrosyl moiety of the tyrosine kinase-receptor complex, which results in the phosphorylation of a specific tyrosine residue in the C-terminus of STAT protein by a tyrosine kinase⁵⁰. In the next step, the SH2-domain of a phosphorylated STAT monomer binds the phosphorylated tyrosine residue of the partner STAT to form a transcriptionally active parallel dimer, which translocates to the nucleus to regulate transcription⁵⁰. In other words, STAT dimerization is also mediated by the SH2-domain of STAT molecule. This leads to nuclear translocation of dimerized STAT which binds to DNA to directly regulate transcription (Fig. 3)⁵¹. STAT5 has been shown to regulate growth and differentiation of cells in several solid tumors, including hormone-regulated cancers such as prostate cancer and breast cancer⁵². In addition, STAT5 has been shown to be expressed in approximately 50% of pancreatic adenocarcinomas⁵³. In prostate tumors, STAT5 is associated with more biologically aggressive cancer⁵⁴ and development of epithelial-to-mesenchymal transition^32,55. Numerous independent studies have provided proof-of-concept that Stat5 drives prostate cancer growth^52,54–77. Stat5 sustains prostate cancer cell viability in vitro and induces prostate tumor growth in mice. Consequently, blockade of Stat5 signaling induces death of prostate cancer cells and blocks growth of prostate tumors^{55,57,58,60–62,64,67,68,78}. Overexpression of active Stat5, in turn, induces growth of PC cells in culture and tumor growth in mice⁶⁵. Recent work demonstrates a novel concept that the second-generation anti-androgen Enzalutamide induces sustained Jak2-Stat5 phosphorylation in prostate cancer, which promotes prostate cancer growth during treatment of prostate cancer with Enzalutamide ⁷⁸. Enzalutamide-induced Jak2 phosphorylation was shown to occur through a process involving Jak2-regulating protein SHP2 which leads to Stat5 phosphorylation and a formation of a hyperactivated feed-forward loop⁷⁸. An appealing aspect of the STAT5 pathway is that in 30–40% of advanced prostate cancers, cytological analysis of the chromosome 17 locus encompassing STAT5A and STAT5B genes has revealed STAT5A/B gene amplification, resulting in increased STAT5 protein levels⁶⁵, especially when compared to normal prostate epithelium⁵⁴. Notably, high STAT5 expression in clinical prostate cancers, at the time of the initial treatment, predicted development of castrate-resistant disease in three independent cohorts totaling 1,035 patients^74,75,77. A recent study shows that combined positive status for both STAT5 protein expression and gene amplification in prostate cancer at the time of radical prostatectomy is a powerful predictor of prostate cancer recurrence in a multivariate analysis even when comparing to the variables of the CAPRA-S nomogram⁷⁷. The finding of STAT5 as a strong predictor of clinical progression of prostate cancer to lethal castrate-resistant state^74,75,77 demonstrates the significance of STAT5 in prostate cancer growth and progression of prostate cancer in patients, corroborating the results obtained using preclinical prostate cancer model systems^52,55–73.

Fig. 2.

Schematic structure of STAT proteins and their key functional domains. Signal Transducer and Activator of Transcription (STAT) proteins are intracellular transcription factors that regulate various aspects of cellular immunity, proliferation, differentiation and apoptosis. Seven members of STAT family have been identified in mammals: STAT1, STAT2, STAT3, STAT4, STAT5 (STAT5A and STAT5B) and STAT6. STAT proteins range between 750–900 amino acids and are encoded by different genes. All STAT proteins share conserved structural motifs consisting of an N-terminal domain followed by a coiled-coil, DNA-binding, linker, Src homology 2 (SH2), and a C-terminal transactivation domain. The N-terminal and SH2 domain mediate homologous or heterodimer formation, while the coiled-coil domain functions as a nuclear localization signal (NLS). DNA binding and transcriptional activity are determined by transactivation and DNA binding domains. Y-S within the C-terminal end are the phosphorylation sites for regulation of STAT activation.

Fig. 3.

Canonical JAK-STAT signaling pathway for STAT activation. STAT activation is initiated by the binding of an extracellular ligand to transmembrane receptors which induces the receptors to dimerize bringing receptor associated JAKs to close proximity leading to their activation. The activated JAKs phosphorylate tyrosine residues in the cytoplasmic tails of the receptors that serve as docking sites for cytoplasmic STAT transcription factors recruited to the receptor-JAK protein complex. JAK tyrosine kinases phosphorylate STAT proteins at the C-terminus at specific tyrosine residues leading to nuclear translocation of STAT dimers to regulate the transcription of their target genes.

In regards of radiation response of solid tumors, RAD51 is central in the repair of RT induced DNA damage via HR and is upregulated in most solid tumors. This has recently lead to a highly compelling finding that STAT5 inhibition sensitizes prostate cancer cells to radiation-induced cell death and increased exposure to RT, without a similar effect on neighboring normal tissues⁷⁶. The molecular mechanisms underlying STAT5-regulation of radiation-induced prostate cancer cell death involves up-regulation of RAD51 levels by JAK2-STAT5 signaling pathway⁷⁶. Specifically, the regulatory regions of the RAD51 gene have been shown to contain STAT5 response elements, and STAT5 directly induces the levels of RAD51 mRNA and protein in prostate cancer⁷⁹. Such a finding supports a persuasive concept to be clinically translated into early phase clinical trial as a novel method of radiation sensitization of prostate cancer^42,80.

4. STAT3 and DNA repair in solid tumors

Activation of STAT3 has been demonstrated in several different human cancer cell lines along with actual tumor specimen. Tumor examples in which activation of STAT3 has been demonstrated include pancreatic adenocarcinoma, prostate cancer, head and neck cancer, lymphoma, and breast cancer⁸¹. Aberrant STAT3 activation in tumor cells has been shown to be associated with increased cell survival, increased cell growth, along with potentially invasion and the development of metastatic disease⁸¹.

The role of STAT3 in DNA repair is supported by the following findings: STAT3 deficient mouse fibroblast cells showed reduced activity of the ATM-Chk1 and ATR-Chk2 pathways, both important in sensing DNA damage⁸². Cells lacking STAT3 also failed to induce mRNA expression of MDC1, a regulator of the ATM-Chk1 pathway and facilitator of the DNA damage response⁸². In conjunction with reduced active ATM, STAT3-deficient cells exhibited reduced phosphorylation of H2AX⁸². Reduced activity of H2AX slows down the rate of accumulation of DNA repair factors around the site of strand breakage, thereby hampering the DNA repair. In DU145 human prostate cancer cells, enhanced expression of BRCA1 resulted in the constitutive activation of JAK-STAT3 signalling pathway, providing critical survival signal for tumor formation⁸³.

DSDBs are the most critical type of DNA damage induced by radio/chemotherapy and activate ATM-ATR-DNA PKcs signalling to promote DNA repair⁸⁴. Lung, bone, and prostate cancer cells demonstrate that DSDBs upregulate PD-L1 expression in an ATM/ATR/Chk1-dependent manner and lead to the activation of STAT3 signalling together with STAT1/IRF1 for DSB-dependent PD-L1 upregulation⁸⁴. The major role of PD-1/PD-L1 is to regulate autoimmune response in the peripheral tissue to maintain a balanced immune response in the body.

Recent studies also revealed that Cetuximab, monoclonal antibody designed to inhibit EGFR-ligand interaction led to the activation of STAT3 in several tumor cell lines as well as primary glioblastoma cells⁸⁵. Cetuximab mediated STAT3 activation has been shown to promote elevated levels of the regulatory subunit Eme1 of the heterodimeric endonuclease Mus81/Eme1 complex, that is involved in DNA repair pathways that remove UV light–induced DNA lesions and cross-links between DNA strands⁸⁵. Cetuximab treatment increased the phosphorylation of the checkpoint kinases Chk1 and Chk2 and its downstream effectors p53 and Histone H2AX⁸⁵. In summary, these findings support the concept that cancer cells which overexpress STAT3 may be resistant to chemotherapeutics through their increased efficiency of DNA repair and greater resistance to genotoxic stress.

5. Cytokines and JAK-STAT pathway in solid tumors

The JAK and subsequent STAT activation pathway is known to be triggered by a variety of cytokines, particularly IL-10 and IL-6. These cytokine levels are highly relevant to consider in the context of several solid tumors, as both have been associated with higher rates of metastatic disease, together with potentially resistance to treatment with RT. Engagement of IL-10 receptor leads to JAK1 activation, which induces phosphorylation of STAT1, STAT3, and STAT5⁸⁶. Such an association may imply that IL-10 is associated with poor response to RT and associate with the development of distant metastatic disease in patients with high risk prostate cancer⁸⁷. In addition, IL-6 has been shown to activate JAK family members JAK1, JAK2, and TYK2, which leads to the activation STAT3⁸⁸. Similarly serum levels of IL-6 have been associated with poor response to RT, which could be potentially mechanistically explained by the IL-6 activation of STAT3 and JAK1/2 signaling⁸⁹. In addition, it has recently been shown that IL-6 leads to prostate cancer resistance to RT through upregulation of DNA repair⁹⁰.

6. Oncological Relevance

The ability to preferentially sensitize certain tumors to RT without simultaneous sensitization of the surrounding normal tissues is highly attractive approach for the treatment of several malignancies. This is because RT is associated with a potential for toxicities to the neighboring organs at risk. In addition, RT can not be given at sufficient doses to induce effective killing of cancer cells in a number of malignancies and therefore requires the use of surgery or concurrent chemotherapy. Furthermore, many tumors progress despite treatment with RT. Oftentimes these progression events happen directly within the RT-treated volume of the tumor tissue. Combining RT with novel mechanisms of radiation sensitization would enable higher efficacy of RT while maintaining acceptable toxicity levels. This would both improve the therapeutic ratio and limit the potential for side-effects of radiation for multiple malignancies. These include certain types of prostate cancer, such as high risk, node positive prostate cancer which both have a poor prognosis with the current standard of care treatment strategies using RT⁹¹. It is important to note that effective RT sensitization strategies for this malignancy are currently absent.

Inoperable pancreatic cancer presents another highly appealing malignancy for improved attempted RT sensitization as current treatment strategies with concurrent chemotherapy and RT are ineffective and novel RT sensitization strategies are lacking⁹². Some tumors that currently have poor response rates to RT, such as glioblastomas, have known associations with STAT transcription factors, including STAT3 which has been shown to be elevated in glioblastoma tissue⁹³. These tumors also have high in field failure rates when treated with RT and also require the use of surgery⁹⁴. In addition, patients with colon cancer positive for JAK-1 and STAT have been shown to have shorter survival⁹⁵. This introduces a potentially attractive option for inhibition of JAK proteins and subsequent treatment with RT for colon cancer. The malignancies addressed in this section are just a fraction of the potential cancers that could benefit from increased sensitization to RT via exploitation of the JAK-STAT pathway.

There exists a number of inhibitors of JAK tyrosine kinases, several of which are FDA approved for treatment of non-malignant conditions or under clinical investigation. Table 1 provides a list of compounds that are either currently approved, or in current clinical development.

Table 1.

Summary of existing JAK inhibitors

Compound Name	Target	Clinical Indications	Citations
Approved:
Baricitinib (trade name Olumiant)	JAK1/JAK2	Rheumatoid Arthritis	28
Ruxolitinib (trade names Jakafi/Jakavi)	JAK1/JAK2	Psoriasis, Myelofibrosis, Rheumatoid Arthritis, Polycythemia Vera	96
Tofacitinib (trade names Xeljanz/Jakvinus	JAK3	Psoriasis and Rheumatoid Arthritis	97
Peficitinib (ASP015 K, JNJ-54781532; trade name Smyraf)	JAK3	Rheumatoid Arthritis	98
Fedratinib (SAR302503; trade name Inrebic)	JAK2	Myelofibrosis	99
Upadacitinib (trade name Rinvoq)	JAK1	Rheumatoid Arthritis	100
In Trials:
Filgotinib (G-146034, GLPG-0634)	JAK1 (Phase III)	Rheumatoid Arthritis, Chron’s disease	101
Gandotinib (LY-2784544)	JAK2 (Phase II)	Myeloproliferative neoplasm	102
Lestaurtinib (CEP-701)	JAK2 (Phase II-III)	Acute Myeloid leukemia	103
Pacritinib (SB1518)	JAK2 (Phase III)	Relapse lymphoma, advanced myeloid malignancies, myelofibrosis	104
Momelotinib(GS-0387, CYT-387)	JAK1 and JAK2 (Phase I/II)	Myelofibrosis	105

7. Conclusions

Relatively little research has been conducted to examine the combination of JAK or STAT inhibition with RT to sensitize solid tumors to treatment with RT. This is despite compelling pre-clinical and clinical evidence that the synergy between these treatment modalities could potentially be very strong. Clinical trials are needed to evaluate the combination of RT and JAK-STAT inhibition in a variety of solid tumors.