A tale of two multi‐focal therapies for glioblastoma: An antibody targeting ELTD1 and nitrone‐based OKN‐007

Abstract Glioblastoma (GBM) is the most common primary malignant brain tumour in adults. Despite a multimodal treatment response, survival for GBM patients remains between 12 and 15 months. Anti‐ELTD1 antibody therapy is effective in decreasing tumour volumes and increasing animal survival in an orthotopic GBM xenograft. OKN‐007 is a promising chemotherapeutic agent that is effective in various GBM animal models and is currently in two clinical trials. In this study, we sought to compare anti‐ELTD1 and OKN‐007 therapies, as single agents and combined, against bevacizumab, a commonly used therapeutic agent against GBM, in a human G55 xenograft mouse model. MRI was used to monitor tumour growth, and immunohistochemistry (IHC) was used to assess tumour markers for angiogenesis, cell migration and proliferation in the various treatment groups. OKN and anti‐ELTD1 treatments significantly increased animal survival, reduced tumour volumes and normalized the vasculature. Additionally, anti‐ELTD1 was also shown to significantly affect other pro‐angiogenic factors such as Notch1 and VEGFR2. Unlike bevacizumab, anti‐ELTD1 and OKN treatments did not induce a pro‐migratory phenotype within the tumours. Anti‐ELTD1 treatment was shown to be as effective as OKN therapy. Both OKN and anti‐ELTD1 therapies show promise as potential single‐agent multi‐focal therapies for GBM patients.


| INTRODUC TI ON
Glioblastomas (GBMs) represent approximately 57% of all gliomas and are the most common primary malignant central nervous system (CNS) tumour. 1 Currently, standard treatment includes surgical resection to remove the bulk tumour, radiotherapy, chemotherapy with temozolomide (TMZ) or bevacizumab, and supportive care. 2 However, overall survival is poor with a median survival time of 12-15 months. 2 The current problem lies with the chemotherapeutic agents.
is currently given during radiotherapy followed by another 6 cycles of TMZ. 3 TMZ is an alkylating agent that produces DNA lesions, leading to cell death. 4 Currently, TMZ is the only approved chemotherapeutic agent that has successfully prolonged the overall survival of patients. 5 However, resistance to TMZ is a key cause of treatment failure. High expression of O 6 -methylguanine-DNA methyl-transferase (MGMT) induces and contributes to TMZ resistance by restoring tumour cell DNA.
Bevacizumab, a humanized monoclonal antibody therapy against the vascular endothelial growth factor (VEGF), is a biologic that is used to combat GBMs. Bevacizumab selectively binds onto circulating VEGF to inhibit its binding onto a receptor (VEGFR) on the surface of endothelial cells. 6 Although pre-clinical studies showed promise, bevacizumab has not significantly increased overall patient survival in newly diagnosed and recurrent GBM patients. 6,7 Instead, tumours treated with bevacizumab show increased tumour metastasis and invasion alluding to a pro-migratory phenotype. [8][9][10] For example, loss of VEGF signalling has led to a more aggressive tumour phenotype in pre-clinical mouse models. 11 Clinically, this pro-migratory phenotype is seen by the development of invasive non-enhancing tumour progression on MRI. 12 Tumour angiogenesis is greatly upregulated in human highgrade gliomas in order to deliver nutrients and oxygen to the tumour core. 13 Pro-angiogenic factors such as VEGF and Notch have historically been examined as potential therapies for cancers that are characterized by unregulated angiogenesis. For example, various strategies for inhibiting the VEGF pathway have been investigated.
The most common therapy is bevacizumab, which inhibits the binding of VEGF onto its receptors, another is sunitinib which targets the VEGF receptor tyrosine kinase inhibitors (RTKIs). 9,14 Additionally, four clinical trials (NCT01122901, NCT01119599, NCT01269411 and NCT01189240) were conducted using RO4929097, a Notch signalling pathway inhibitor, against GBMs both as a single agent and in combination with TMZ or bevacizumab. 15 However, from these trials, only one phase 1 trial was completed, while the other three were terminated due to the termination of drug supply from the manufacturer. ELTD1 (epidermal growth factor, latrophilin and seven transmembrane receptor containing protein 1 on chromosome 1, ADGRL4) has previously been shown to be involved in brain angiogenesis and was shown to be regulated by the VEGF and DLL4/Notch signalling pathways. 16 ELTD1 has higher expression in human high-grade gliomas when compared to low-grade gliomas. 17 Moreover, targeting ELTD1 with an antibody was found to be effective in a G55 human GBM xenograft mouse model as demonstrated by decreasing tumour volumes, normalizing tumour vasculature and increasing survival. 18,19 Further optimization of the antibody therapy showed higher binding specificity against the tumour. 18,19 OKN-007 (OKN), which was recently found to target the transforming growth factor β1 (TGF β1) pathway, is also a small molecule that is effective in crossing the blood-brain barrier. 20 From previous pre-clinical studies in U87 and G55 GBM xenografts, and C6, GL261 and F98 high-grade glioma animal models, it was established that OKN is an effective therapy against GBM/high-grade gliomas by inhibiting cell proliferation and tumour necrosis, increasing apoptosis and increasing survival. [20][21][22][23][24] Recently, it was found that when OKN is combined with TMZ, it increases TMZ sensitivity, thus increasing a significant effect on TMZ-resistant GBM cells. 20 OKN is also known to target tumour-associated angiogenesis by targeting and decreasing both VEGFR2a and HIF-1α protein expression. 21

| Generation of anti-ELDT1 antibodies
Human ELTD1 (Glu20-Leu406) and mouse ELTD1 (Glu20-Leu455) genes were used to create the extracellular domains of human and mouse ELTD1 expression vectors as previously reported. 19 Briefly, the expression vectors were transfected into HEK293F cells (Invitrogen, Carlsbad; CA, USA) and the C kappa fusion proteins were purified from the supernatant using KappaSelect resin (GE Healthcare).
Chickens (white leghorn) were inoculated with a human ELTD1 C kappa fusion protein, where the RNA was obtained from bone marrow, spleen and bursa of the chickens as previously described. 18 Briefly, phage-display library was created from the clones and the positive clones that showed cross-reactivity with human and mouse ELTD1 were selected and enriched determined by Sanger sequencing.
Human or mouse ELTD1 C κ fusion proteins were pipetted onto 96-well microtiter plates (Corning Inc., Corning, NY, USA) in a coating buffer as previously described. 19 0.1 M NaHCO 3 buffer (pH 8.6) was used to coat the ELTD1 C κ fusion proteins, and then subsequently blocked with 3% (w/v) BSA in phosphate-buffered saline (PBS).
Microtitre plate wells were subsequently incubated and washed as previously reported. 19

| Glioma model and treatment
Animal studies were conducted with the approval (protocol 17-48) of the Oklahoma Medical Research Foundation Institutional Animal Care Use Committee policies, which follow NIH guidelines. Two-month-old male Athymic Nude-Foxn1nu mice (Harlan Inc., Indianapolis, IN) were implanted with human G55 cells as previously described. 26

| Morphological imaging
All mice were subjected to MR imaging while under anaesthesia and restrained in a cradle which was inserted into a 30-cm horizontal bore Brucker Biospin magnet (7T, Bruker BioSpin GmbH; Karlsruhe, Germany). The animals were first imaged 10 days post-tumour implantation and then every 3-4 days depending on treatment administration with a BA6 gradient set and mouse coil as previously described. 26

| Perfusion imaging
The perfusion imaging method, arterial spin labelling, was used as previously described. 28 For perfusion quantification, five region of interests (ROIs) were outlined in the tumour as well as in the contralateral side of the brain as a control. Blood perfusion rates (BPRs) values were determined by subtracting late and early tumour BPRs.
This difference was then normalized to BPRs in the contralateral brain region of corresponding animals.

| RE SULTS
In this study, we aimed to compare OKN-007 therapy and monovalent monoclonal antibody treatments against ELTD1 against the clinically used agent, bevacizumab. Tumour growth was monitored via MRI every 3-4 days, and the animals were either left untreated or treated via tail vein upon tumour detection with either mmAb anti-ELTD1 Ab or bevacizumab (2 mg/kg), or given OKN via the drinking water. Once the tumours reached 150 mm 3 (via MRI) mice were sacrificed, and their brains were harvested for histology. The untreated animal and ELTD1 treatment group data were obtained from cumulative studies. As shown in Figure  To assess microvasculature alterations within the tumour region, we used MRI perfusion to measure the relative change in BPRs. As shown in Figure 2A, normal tissue has an organized vasculature and therefore has a set BPR. However, due to increased angiogenesis, the vasculature becomes chaotic and decreases the perfusion rates within the tissue. 29 The untreated animals had a drastic decrease in BPR, shown in the quantitative graph in Figure 2B, which depicts an increase of angiogenesis within the tumour region. Anti-ELTD1 and OKN treatments both normalized tumour vascular perfusion rates compared to untreated and bevacizumab (anti-ELTD1: UT ****p < 0.0001, bevacizumab **p = 0.0061, OKN: UT ****p < 0.0001, bevacizumab **p = 0.0027). There was significant normalization of the perfusion rates between the combined therapy against both untreated and bevacizumab-treated animals; however, there was no additive effect with the combined therapy (UT ****p < 0.0001, bevacizumab **p = 0.0012). Therefore, we can conclude that anti-ELTD1 and OKN treatments are more effective in normalizing the microvasculature alterations associated with GBMs.
To further examine the effect of each treatment on tumourassociated vasculature, we analysed microvascular density (MVD) of each tumour sample staining for CD34. CD34 is a well known endothelial cell marker for the quantification of angiogenesis. 30,31 Representative CD34 IHC images are shown in Figure 2C UT ***p = 0.0008, bevacizumab ****p < 0.0001, OKN **p = 0.0038).
There was no significant difference in Notch1 positivity staining when comparing OKN to untreated and bevacizumab-treated animals (see representative images in Figures C-G). The next angiogenic marker in question was VEGFR2. Similar to Notch1, VEGFR2 positivity levels were upregulated, with an average of 61%, in the untreated animals as shown in Figure 3B. As expected, bevacizumab significantly decreased VEGFR2 positivity levels within the tumour region compared to untreated (UT ****p < 0.0001). Additionally, anti-ELTD1, OKN and combined therapies also significantly decreased VEGFR2 positivity compared to untreated controls (anti-ELTD1 and combined: UT ****p < 0.0001; OKN: UT ***p = 0.0002). There was no significant difference between the four treatment groups as shown through the representative images ( Figure 3H-L).
We wanted to determine if the effects we were seeing were due to the direct targeting of the human tumour cells (G55 GBM cells). Therefore, we stained the tumour tissue against the human mitochondrial antibody. Figure 4A,B is a representative untreated IHC slice in 1× and 5×, respectively, showing that there is no human mitochondrial antibody staining in the healthy contralateral tissue.
The black box and Figure 4B shows the distinct tumour boundaries between the tumour/infiltrating cells and the healthy mouse tissues. Bevacizumab studies have demonstrated that therapies that target angiogenesis in GBMs cause an increase of invasiveness by promoting a migratory phenotype to ensure sufficient delivery of oxygen. 8,14 This, therefore, led us to investigate if our other treatments favoured a pro-migratory phenotype. The transient receptor potential melastatin family member 8 (TRPM8), bone morphogenetic protein 2 (BMP2) and L1CAM are upregulated in GBMs compared to normal brain tissue and are associated with glioma cell proliferation, migration and invasion. [32][33][34][35] To assess tumour cell migration, we first stained the tumour tissue against TRPM8.
As shown in Figure 5A We then sought to determine if the various treatments had an effect on cellular proliferation within the tumour region. Ki-67 is an established cell proliferation marker that has also been strongly correlated with GBM tumour growth and metastasis. 36 As shown in of apoptosis within a given tissue. As shown in Figure 6B, the un-

| DISCUSS ION
This study sought to compare anti-ELTD1 and OKN therapies in an in vivo G55 xenograft mouse study. MRI is a powerful tool used by clinicians to monitor tumour progression and clinical effectiveness of prescribed treatment regimens. 38 In this study, we used morphological and perfusion MRI to compare anti-ELTD1 and OKN therapies.
Both anti-ETLD1 and OKN treatments were significantly better at increasing overall survival when compared to untreated, and there was no significant difference between the two groups. Furthermore, there was also no significant difference between the two groups with overall tumour volumes.
Typically, the tumour-associated microvasculature alterations associated with tumour angiogenesis is measured by perfusion MRI, which measures a relative change in BPR, and is used to assess the efficacy of anti-angiogenic therapies. 39 In GBMs, the tumour-associated vasculature has irregular and leaky vessels that cause the BPR to be lower than the surrounding normal brain. 40 Once treated with anti-angiogenic therapies, the BPRs within the tumour tend to increase in patients. An explanation to this observation is that there is a decrease of the leaky vessels created by the tumour and instead the remaining vessels mimic normal vessels which results in increased blood perfusion. 40 In regard to tumour vascularization, our results showed a drastic decrease in perfusion rates in untreated animals, which is characteristic sign of increased angiogenesis. Although bevacizumab was capable of decreasing this drastic change in perfusion rate, it seems that bevacizumab therapy may not be as effective in normalizing microvascularity within the tumour regions compared to anti-ELTD1 and OKN therapies.
Anti-ELTD1 therapy was more successful in decreasing microvessel density levels. Although there was no additive effect with the combined therapy in any of these aspects, MVD levels were significantly decreased compared to untreated, bevacizumab and OKN therapies. Again, our results demonstrate that bevacizumab therapy has no significant effect on the microvascular density levels within the tumour region. In normal vasculature, ELTD1 expression is regulated by VEGF and Notch/DLL4 pathways. 16 We previously examined the relationship between VEGF and a polyclonal anti-ELTD1 antibody or OKN-007 treatment.
In those studies, we demonstrated that ELTD1 and VEGFR2 were co-localized within blood vessels and glioma cells in G55 glioma tumours. The polyclonal anti-ELTD1 treatment significantly reduced the expression of ELTD1 and VEGFR2, but had no significant effect on VEGF. 27 In orthotopic rat F98 and human U87 xenograft glioma models, OKN-007 was shown to decrease microvessel density levels as shown through CD31; however, OKN-007 did not alter VEGF levels. 41 In this study, we further demonstrated that in a GBM model, targeting ELTD1 results in decreased expression of both VEGFR2 and Notch1. This further demonstrates that anti-ELTD1 treatment works to decrease several pro-angiogenic factors (although not VEGF) within the tumour region. OKN-007 treatment was also effective in decreasing VEGFR2 expression levels, however, had no effect on Notch1. This further demonstrates that anti-ELTD1 treatment is directly associated with the Notch signalling pathway.
Previous studies have shown that anti-angiogenic therapies, such as bevacizumab, cause GBMs to become progressively invasive and invade normal brain tissue leading to the formation of satellite tumours. 6 In this study, we sought to determine whether various treatments affected the invasiveness of the tumours by examining the proliferative rate and the expression of various migratory markers.
Bevacizumab treatment had no effect on various migration markers compared to untreated and had a high Ki-67 positivity staining. This suggests that bevacizumab treatment is not effective in decreasing tumour invasiveness.
However, anti-ELTD1 and OKN treatments both significantly decreased Ki-67 positivity suggesting that the tumour is has decreased invasive properties. Additionally, we see that anti-ELTD1 and OKN treatments successfully targeted and decreased migration in our G55 pre-clinical model. Anti-ELTD1 therapy decreased TRPM8 and BMP2, while OKN therapy was successful in decreasing L1CAM positivity levels. This suggests that while anti-ELTD1 and OKN therapies both target migratory pathways, they may have effects on different migration cellular pathways. It was previously shown from microarray analysis that OKN affects tumorigenesis by targeting the TGF-β1 pathway by downregulating key-associated genes has. 20 OKN was shown to downregulate key genes associated with   Notch1. 18,19 Anti-ELTD1 therapy has targeted key genes such as c-MET, Ki-67, BMP2 and VEGFR2 which are correlated with Notch1.
Therefore, we have reason to believe that Notch1 may be the master regulator of ELTD1. 18,19 Further studies will need to be done to confirm the role of Notch 1 in ELTD1 antibody therapy.

| CON CLUS ION
In preliminary assessments, we saw that anti-ELTD1 and OKN treatments were effective as therapies in a human G55 pre-clinical model.
In this paper, we validated those claims by demonstrating that both treatments increased survival, decreased tumour volumes, normalized the tumour-associated vasculature, decreased migratory markers and specifically targeted human tumour cells. Anti-ELTD1 and OKN-007 seem to be similar in most instances; however, combined therapy seems to be more effective than either regarding NF-κB or better than OKN-007 in inducing apoptosis (cleaved caspase 3). James Battiste: Validation (supporting); writing -review and editing (equal). Rheal A. Towner: Conceptualization (lead); funding acquisition (lead); investigation (lead); project administration (lead); resources (lead); supervision (lead); validation (lead); writing -review and editing (lead).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.