Metabolites such as fumarate are not widely quantitated in tissue due to high cost of current methods as well as a lack of effective methodology. Metabolites are often highly polar, non-volatile and have poor detectability making analysis difficult. Chan et al. used high resolution magic angle spinning-nuclear magnetic resonance (NMR) and gas chromatography coupled to mass spectrometry (GC-MS) to compare the metabolic profile of biopsied CRC tumours and matched normal tissue from 31 patients384. They identified a reduction of fumarate in CRC compared to normal tissue. A similar result was found by Denkert et al. using GC time-of-flight MS (GC-TOFMS) for 15 paired CRC and normal tissue patient samples385. Both of these studies showed a reduction in TCA cycle metabolites which links well with Otto Warburg’s theory of reduced mitochondrial metabolism386. Hirayama et al. used capillary electrophoresis TOFMS (CE-TOFMS), which they argue is more suited to metabolism analysis due to its high resolution and ability to simultaneously quantify charged low-molecular weight compounds335. They identified an increase in fumarate in 16 CRC tissue samples compared to matched normal.
My findings are in accord with Hirayama and colleagues finding of elevated fumarate in CRC tissue compared to matched normal335. In a cohort of 83 matched normal and CRC tissue, for the first time an elevation in succination, as measured by 2-SC staining, is shown (Figure 3.2.A). My work has also shown elevated succination in the early stage of the cancer, an adenoma, compared to matched normal tissue (Figure 3.3.A). These data suggest that elevated fumarate is an early event in CRC tumorigenesis and that protein succination could be a factor influencing tumour development.
Succination in CRC tissue is further elevated by the presence of T2D. This is another novel finding. The trend of succination in adenomas is still to be confirmed. In this study, the number of adenomas with known T2D status was limited. From the small cohort used, succination was lower in tissue from T2D patients compared to non-diabetic patients, however, this was not significant.
Interestingly, succination was not significantly elevated in normal tissue from T2D compared to non-diabetics, although it was raised in T2D compared to non-diabetic patients. This suggests that the elevated fumarate is a consequence of an oncogenic event that has led to a change in cellular metabolism, which is then exacerbated by the presence of T2D and could be one reason why T2D patients with CRC have a poorer outcome and more invasive cancer124.
T2D is a metabolic disorder where succination has already been confirmed elevated in certain tissues. Adipocytes from db/db and ob/ob mice342 which are mouse models of human diabetes, and skeletal muscle from streptozotocin-induced diabetic rats333 were found to have elevated succination as well as adipocytes treated with 30mM glucose versus 5mM glucose339 in vitro. Interestingly, undifferentiated fibroblasts cultured in the same high glucose media did not exhibit an increase in succination339 whereas differentiated 3T3 fibroblasts do exhibit increased succination under 30mM glucose310. Frizzell et al. 2012 showed that 3T3 cells treated with 30mM glucose compared to 5mM also showed significant increases in cellular ATP/adenosine diphosphate (ADP), NADH/NAD+, mitochondrial membrane potential and cellular fumarate concentration. This was postulated to an increase in NADH/NAD+ ratio resulting in the inhibition of NAD+-dependent dehydrogenases. Cellular fumarate and succination was decreased upon addition of chemical uncouplers which reduce the NADH/NAD+ ratio and upon addition of metformin (an inhibitor of complex I in the ETC) to high glucose culture media. The conclusion of the study was that elevated succination in diabetic conditions was a result of glucotoxicity-driven mitochondrial stress310,343.
Mitochondrial function is known to be deregulated in cancer387 and that increased glycolysis is a feature of cancer. Therefore, it could be suggested that glucotoxicity-driven mitochondrial stress exists within CRC which leads to elevated fumarate and, therefore, elevated succination. In turn, the glucotoxicity-driven mitochondrial stress is further elevated in patients with T2D, leading to higher levels of succination. Unfortunately, I was unable to determine the diabetic treatment for the majority of the CRC patients, as hospital records are incomplete. Nevertheless, it would be interesting to study the succination levels in T2D CRC patients with and without metformin treatment; I would hypothesise that succination would be lower in patients on metformin.
Surprisingly, FH was elevated in CRC tissue compared to matched normal (Figure 3.5). This suggests that FH is upregulated in conjunction with, or as a consequence of elevated fumarate, but is unable to remove the excess fumarate. FH may be dysfunctional. However, somatic FH mutations are rare and restricted to HLRCC patients388; No FH mutations have been found in CRC patients388. More work to determine FH expression in this set of patients will be completed so that more concrete conclusions can be made. FH is a tetrameric enzyme, composed of four identical subunits of 50kDa each389. There are 3 active sites (site A) and one lower affinity site (site B). A report by Mescam et al. suggests that at low fumarate concentrations (<1mM) the enzyme shows Michaelis- Menten kinetics; at 0.001-0.033M allosteric activation of the enzyme by binding to site B is observed; at 0.1M and above fumarate actually inhibits FH390. Therefore, it could be that fumarate itself is inhibiting the action of FH, leading to further increases in fumarate. Hirayama et al. showed that the fumarate concentration in colon tumour was around an average of 50nmol/g which is equivalent to 50mM and, therefore, close to allosteric inhibition. This is an average concentration and local concentrations maybe much higher leading to inhibition.
Approximately 35-40% of human CRCs have an activating missense mutation in KRAS380– 383. These affect hotspots in codons 12 and 13 which lock KRAS in an active GTP-bound conformation, constitutively presenting a docking surface for RAF kinases391. BRAF mutations are found in 10% of CRCs, and are most likely to be a V600E amino acid substitution, although other mutations at codon 600 or neighbouring positions are documented29. The presence of these mutations in our cohort is in accord with these results and were selected for examination as they are amongst the most prevalent mutations found.
Although the use of KRAS and BRAF mutation status as a predictive biomarker of response to anti-EGFR therapy is well supported, there are inconsistencies within the literature regarding the association between KRAS and BRAF mutations and CRC survival. There are discrepancies related to specific mutations, age, stage at diagnosis and treatment. The
collaborative database called RASCAL (The Kirsten ras in-colorectal-cancer collaborative group)392. They identified that only one mutation of codon 12, the KRAS p.G12V mutation, found in 8.6% of patients had a statistically significant impact on outcome, but only among patients with Dukes’ C CRC392. BRAF mutation is associated with poorer survival of CRC patients29,393. Unfortunately, the specific KRAS or BRAF mutation is not reported in patient notes, therefore, it was not possible to distinguish the effect of specific KRAS and BRAF mutations on 2-SC score in this study. In vitro data identified that HCT116 and DLD1 CRC cell lines with KRAS mutations increase GLUT1 expression after 4 days of culture in 0.5mM glucose, leading to increased glucose uptake and increased lactate production, although mitochondrial function and oxidative respiration were not affected394. It was also shown that the WT KRAS CRC cells which survived 4 days of culture with 0.5mM glucose increased their mutation rate of KRAS to increase GLUT1 expression and, therefore, glucose uptake394. Another study which involved 8 KRAS WT and 8 KRAS mutant human colon tumours identified an association between a 2-fold increase in expression of glycolytic and glutamine metabolic proteins and KRAS mutation status376. KRAS mutation in CRC has also been found to affect amino acid metabolism. Human CRC cell lines and clinical specimens with KRAS mutation were found to have an increase in asparagine synthetase (ASNS) which was induced via the PI3K-AKT-mTOR pathway395. Subsequent knock down of ASNS in KRAS mutant CRC cell lines led to growth suppression. It was also found that asparagine addition prevented cell death from glutamine depletion. Weinberg et al. reported that the pentose phosphate pathway (PPP), not glycolysis was essential for KRAS mutant CRC cell growth396. Miyo et al. reported that in KRAS mutant CRC cell lines resistance to glucose-deprived conditions is associated with increased levels of both GLUD1 and SLC25A13 (a mitochondrial aspartate-glutamate carrier)397.
From the work detailed above it is clear that KRAS mutations influence cancer cell metabolism in a cell and tissue dependent manner. It has also been reported that KRAS driven NSCLC cells in vitro use nutrients differently compared to human lung tumours;
KRAS mutant tumours were less dependent on glutaminase that NSCLC cells in vitro398.
This highlights the importance of studying cancer metabolism in a physiological context. More work is needed to determine the exact role of KRAS in cancer.
Currently, there is no literature which link KRAS or BRAF mutation to an increase in succination. This study hypothesised that presence of KRAS or BRAF mutation would increase succination via increased use of the TCA cycle. The data show that there is no significant effect of KRAS or BRAF mutation status on 2-SC score. However, there is an increase in 2-SC score in samples from patients with T2D and KRAS or BRAF mutation combined compared to samples from non-diabetic KRAS or BRAF WT patients. This suggests an interaction between T2D and KRAS or BRAF mutation which could be attributed to greater glucose availability and increased ability to uptake glucose leading to more succination. I would, therefore, hypothesise that KRAS mutant CRC cells would channel glucose-derived metabolites into the TCA cycle to a greater extent than KRAS WT CRC cells and, therefore, lead to increased fumarate and increased succination. This chapter presented evidence that fumarate is elevated in the early stage of CRC, an adenoma. There is increased succination in both adenoma and CRC tissue compared to match normal tissue. Succination in CRC is further increased in the presence of T2D. Additionally, the 2-SC staining was non-uniform across tumours, underlining the metabolic heterogeneity of CRC tissue. T2D patients with KRAS/BRAF mutation exhibit a further increase in succination compared to T2D patients with WT KRAS/BRAF. The possibility of a link between 2-SC score and patient prognosis requires further investigation and would be an interesting avenue to pursue. The functional effects of succination are likely to be wide-spread and would take more time to investigate through collaboration. Additionally, the role of FH in CRC patients should be further investigated (Chapter 8, Section 8.1). Here, I found FH is present in CRC tissue at an elevated level, although it remains to be determined if FH is fully active or inhibited by high levels of fumarate. Loss of function at the genetic levels is less likely, as there are no known mutations in FH in CRC patients.