[RS] – Exploitation of the Hypoxia-Inducible Factor (HIF) pathway as an anti-cancer strategy for treatment of kidney cancer

Sarah Welsh was looking at the potential of a drug she identified called CL67 which may be able to slow kidney cancer.

Kidney image (2)

Dr Sarah Welsh, CRUK Cambridge Institute, University of Cambridge

Exploitation of the Hypoxia-Inducible Factor (HIF) pathway as an anti-cancer strategy for treatment of kidney cancer

December 30th 2014 – December 31st 2017

The hypoxia-Inducible Factor (HIF) pathway is the main pathway used by cells to detect, and respond to, low levels of oxygen (as well as some other nutrients) in their environment. Cancer cells have exploited this pathway by ensuring that it is constitutively switched on. Over 95% of kidney cancer cells have high levels of activity of the HIF pathway suggesting that inhibiting this pathway with drugs might be an effective strategy to control kidney cancer.

The HIF pathway is activated when two different HIF subunits (HIF-alpha, and HIF-beta) come together to form the HIF complex. There are 2 different alpha subunits (HIF-1α and HIF-2α) of which either can form the HIF complex. When low levels of oxygen (or nutrients) are detected, an alpha and a beta subunit bind to each other and the HIF factor becomes active and switches on over 100 different genes. These genes are involved in either helping the cells to survive by switching off pathways by which cells die, or switching on pathways by which cells survive, or to bring the cells more oxygen or nutrients by increasing blood flow to the area by increasing blood vessels, amongst other mechanisms. This project investigates whether inhibiting the HIF pathway using a drug called CL67 might be an effective anti-cancer strategy in kidney cancer.

CL67 was developed in Professor Stephen Neidle’s laboratory at University College London by myself and his team. CL67 binds to 3D structures, called G-quadruplexes, which form in the DNA structure of the HIF-1α and HIF-2α genes. These structures stop HIF-1α and HIF-2α protein being made, and since kidney cancer cells rely on the HIF pathway to survive this inhibits their ability to grow. Although we already knew from our previous work that CL67 could inhibit the HIF pathway in kidney cancer cells, this project investigates in more detail how CL67 works, and whether it works in animals; these are important experiments to o predict how safe and effective this drug might be in patients with kidney cancer.

Aim 1: To determine the specificity of CL67 for the HIF-1α and HIF-2α G-quadruplexes in kidney cancer

Work was needed to determine how specifically CL67 binds to HIF-1α and HIF-2α compared to other genes in which G-quadruplex structures are also formed. This is important because if CL67 affects other genes, it may increase its toxicity in patients. DNA from a number of different genes known to form G-quadruplexes was incubated with and without CL67 in test tubes and the temperature at which the DNA “melted” was determined (Table 1). The difference between the temperatures at which the DNA melted (ΔTm), with and without CL67, was calculated. Significantly higher ΔTm were seen when CL67 was added to HIF-1α and -2α compared to other genes known to form G-quadruplexes. The majority of the ΔTm values for other genes were very low (<2oC). This shows that CL67 binds to, and stabilises, G-quadruplexes in HIF-1α and -2α more effectively than other G-quadruplexes suggesting that CL67 is likely to act mainly by stabilising HIF-1α and -2α reducing the likelihood for toxic side effects.

Table 1 – Specificity of CL67 for HIF-1α and -2α G-quadruplexes compared to G-quadruplexes in other genes

Gene known to form G-quadruplexes ΔTm (oC)
HIF-1α 20
HIF-2α 17
c-KIT-1 1.5
c-KIT-2 6.6
T-loop (telomeric DNA) 1.8
c-MYC 1.6
MYB29 1.2
BCL-2 1.5
RB1 1.5
hTERT 1.7
PDGFRβ 1.6
K-Ras 2.0

The second part of Aim 1 proposed to investigate whether CL67’s effects on kidney cancer cells are due to effects on the HIF-1α and HIF-2α G-quadruplexes.  I attempted to generate pairs of kidney cancer cell lines which have either normal HIF-1α and HIF-2α DNA sequences such that they form normal G-quadruplexes, or abnormal HIF-1α and HIF-2α DNA sequences such that they do not form G-quadruplexes.  However, despite considerable time and effort, unfortunately this has proved very difficult as the cell lines with abnormal HIF-1α and HIF-2α DNA sequences genes do not generate high enough levels of abnormal protein to enable testing as planned.

Instead, I have compared the effect of CL67 on the levels of 168 different genes involved in metabolism (energy generation) or angiogenesis (the process of making blood vessels).  Some of these genes are known to be under the direct control of the HIF pathway, and some are known NOT to be under the direct control of the HIF pathway.  I hypothesised that if CL67 acts mainly by inhibiting the HIF pathway, treatment of kidney cancer cells with CL67 should lead to greater suppression of HIF-pathway genes, compared to non-HIF pathway genes.  Indeed, treatment of cells for up to 4h with CL67 caused more suppression of HIF pathway genes compared to non-HIF pathway genes (Figure 1).  Some suppression of non-HIF pathway genes by CL67 was observed (particularly by 4h) but this may be due to indirect suppression of these genes via HIF pathway genes.  This data suggests that CL67 is acting mainly via HIF-1α and HIF-2α, giving us increased confidence that CL67 should not be expected to cause excessive toxicity.

Figure 1 – CL67 decreases expression of HIF pathway genes to a greater extent than non-HIF pathway genes. * denotes a significant difference compared to non-HIF pathway genes treated with CL67 for 2h.  ** denotes a statistically significant difference compared to non-HIF pathway genes treated with CL67 for 4h.


Walsh Report Figure 1

Aim 2:  To determine the effect of CL67 on kidney cancer cell growth in vivo

Although we know that CL67 inhibits the growth of kidney cancer cells growing in dishes, we do not know if it inhibits kidney cancer cells in animals.  Toxicity studies to determine the correct dose of CL67 to use in animals were, interestingly, negative meaning that despite high doses of CL67 no toxicity was observed.  Normally, a dose is chosen to test for anti-tumour activity which is just below the ‘maximal tolerated dose’ ie. the maximum dose that can be tolerated by the mice without causing undue toxicity.  However, no maximal dose was found as the mice did not display signs of distress or toxicity at any dose tested.  This may be further evidence that CL67 acts selectively via HIF-1α and HIF-2α as, in theory, the HIF pathway should not be activated in normal physiology.  Therefore, a dose of 40mg/kg of CL67 was chosen to test against kidney cancer cells grown in mice for this project, as significant inhibition of the growth of sarcoma cells (a type of bone cancer) in mice was seen at 40mg/kg by my collaborators at UCL suggesting that CL67 is effective in mice at this dose, at least in sarcoma (a tumour-type which is normally very resistant to cancer treatment, and is known to also have high levels of the HIF pathway).

In the kidney cancer experiments 3 different cell lines were chosen which have different expression levels of the various HIF proteins.  We wanted to use cell lines with different genetic backgrounds to see whether CL67 was active in different tumours.  A498 and 786-0 cell lines (common to the majority of kidney cancers) do not express HIF-1α protein significantly but express high levels of HIF-2α, whereas ACHN cells express both HIF-1α and HIF-2α.  Therefore, male nude mice (aged 6-8 weeks) were implanted with 1×107 renal cancer cells (per mouse) on one flank.  When tumours grew to approximately 95mm3 (74.9-111.1mm3) the mice were randomly assigned to the either receive placebo (vehicle only) or CL67 (40mg/kg).  Mice were treated three times per week for 3 weeks, intravenously, and were monitored daily for signs of distress and tumour growth.  There were no reports of any adverse health effects during the study.   Although there was a little variation in bodyweight during the study (Figure 2), mean bodyweight did not decrease below 95% of treatment start weight during the study showing that the mice tolerated treatment well.

Figure 2 –  Body weights of mice treated with CL67 for 3 weeks.  Values shown are mean ±SD; n=8 for all groups.  CL67 treatment did not affect body weight significantly over the course of treatment.

Walsh Report Figure 2


Tumours in vehicle-treated animals grew steadily during the study (Figure 3A-C). At the end of the dosing period, tumours from animals treated with CL67 40mg/kg were significantly smaller than vehicle controls in all 3 cell lines (786-0:  p=0.0019; A498:  p<0.0001; ACHN:  p=0.0043; 2-tailed T-test).   At the end of the study period CL67-treated tumours were 24.0% (for 786-0 tumours), 47.6% (for A498 tumours), and 64.9% (for ACHN tumours) of the size of vehicle-treated tumours.  Significant anti-cancer activity is defined by the National Cancer Institute (NCI), which states that any compound which has a value of less than or equal to 42% has demonstrated significant anti-cancer activity. Thus, in this model, CL67 showed anti-cancer activity in 786-0 cells, but although CL67 has resulted in control of tumour growth in A498 and ACHN cells, it has not demonstrated anti-cancer activity as per NCI definition.  I await the final results of immunohistochemistry experiments which are investigating the levels of inhibition of HIF-1α and HIF-2α, in addition to the downstream targets VEGF, GLUT-1, and CA-9 whose results may indicate whether the level of inhibition of tumour growth correlates with target inhibition in these different models.

Figure 3A – Volume of 786-0 tumours implanted subcutaneously in male nude mice treated with CL67. Values shown are mean ±SD; n=8 for all groups.

Walsh Report Figure 3A

Figure 3B – Volume of A498 tumours implanted subcutaneously in male nude mice treated with CL67. Values shown are mean ±SD; n=8 for all groups.

Walsh Report Figure 3B

Figure 3C – Volume of ACHN tumours implanted subcutaneously in male nude mice treated with CL67. Values shown are mean ±SD; n=8 for all groups.

Walsh Report Figure 3C


The generous granting of my TUF scholarship has enabled significant progress to be made towards exploiting the HIF pathway as an anti-cancer strategy in kidney cancer.  Evidence from kidney cancer cells grown in the laboratory in dishes, as well cancer cells grown in animals has shown that CL67 inhibits the HIF pathway in a selective manner, which suggests that it is likely to be a safe anti-cancer strategy in kidney cancer.  Experiments in animal models have also confirmed that this is also potentially an effective anti-cancer strategy in kidney cancer.  Remaining outstanding immunohistochemistry experiments will investigate whether the effectiveness of CL67 is determined by the levels of inhibition of target proteins in the HIF pathway.  UCL Technology are assessing whether to extend the patent and take CL67 forward to other animal models and more extensive testing prior to human studies, based on the animal work to date.


“This work would not have been possible without the support of The Urology Foundation.”

Find out more