HOW DO CHANGES IN DNA CAUSE CHANGES IN FUMARASE?
If your genetic test results show that you have an alteration in your fumarase gene, you may have questions about what the letters, numbers, and symbols that describe the alteration mean. You may also wonder how changes in the DNA of your fumarase gene interfere with the making of fumarase in your cells.
How DNA Codes for Proteins
Fumarase is an enzyme, meaning it is a protein that catalyzes chemical reactions. DNA is the code for making proteins. Proteins are like strings of beads in which the beads are amino acids. Proteins have 20 different kinds of beads--they are the 20 amino acids. DNA provides the code for putting the amino acid "beads" in order. DNA is found in the chromosomes. In the pictures we see of chromosomes, they look a bit like worms that have been pinched in the middle. That is because the picture is taken at a time when the DNA (and the proteins surrounding it) is all coiled up. At other times it is stretched out and is like a very long filament. The DNA filament can be thought of as a double-stranded string of beads. The two strands only have 4 kinds of beads--they are the nucleotide bases, symbolized by the letters A, T, C and G (adenine, thymine, cytosine and guanine). Along the length of this double strand, the beads are paired up so that A of one strand is always connected to T of the other strand, and C is always connected to G.
DNA is used as a template to make messenger RNA (mRNA), which is more directly involved in the assembly of protein. mRNA can also be thought of as a "string of beads". Like DNA it has 4 kinds of beads. Three of them are the same as in DNA (A, C, and G), but in place of thymine (T), mRNA contains uracil (U). mRNA is single-stranded, so the beads are not paired. However, in the process of making mRNA from DNA, the attraction and pairing of the beads is important. Wherever there is an A on DNA, a U will be placed in the mRNA strand that is being assembled on it. Likewise, a T codes for an A, C codes for G, and G codes for C. The process of making mRNA from DNA is called transcription. During transcription, the DNA strands separate at the point where transcription is taking place, and only one strand of the DNA is used. The corresponding mRNA beads are lined up and joined together on that strand of DNA. Later, the newly formed strand of mRNA is separated from the DNA template, and the two DNA strands reunite. The new strand of mRNA is complementary to the DNA strand that made it. For example, a DNA sequence of ATCGTTACC would result in an mRNA sequence of UAGCAAUGG. mRNA is the direct code for proteins. It dictates the sequence of amino acid beads that will be lined up and connected to make a protein in a process that is called translation. Every 3 RNA beads codes for an amino acid. This nucleotide triplet is called a codon. Since there are three beads in a codon, and there are 4 different kinds of beads in mRNA, there are 64 different 3-letter combinations to code for 20 amino acids. (See the charts at the end of this section.) Therefore some amino acids are coded for by more than one 3-letter combination. Three codons are nonsense codons, meaning they code for nothing. They are used as stop signals. One codon, which codes for methionine, is the start codon. It signals where translation is to begin.
So you can see that if there is a change in the sequence of DNA beads (mutation) there will be a change in the sequence of mRNA beads. Sometimes the change will make no difference in the protein that is made. For example, if UAU on the mRNA strand is changed to UAC, there will be no change because UAU and UAC both code for the amino acid tyrosine. However, if UAU on the mRNA strand is changed to CAU, then histidine will be put where tyrosine is supposed to be. This is called a missense mutation. A missense mutation is one that causes a different amino acid to be put in place of the correct one. Sometimes, instead of a substitution, a mutation can cause protein making to come to a halt. For example, if UAU is changed to UAG, it will code for nothing because it is a stop codon. This is called a nonsense mutation. Enzymes that have had only one or just a few of their amino acids changed can sometimes still function, depending on whether the amino acids are located in a position on the enzyme that is important or not. Nonsense mutations, which stop the protein making process and lead to truncated proteins, are more likely to cause problems.
Many details were left out of the above explanation so that the understanding of how DNA codes for mRNA, and how mRNA, in turn, codes for the sequence of amino acids that makes up a protein would be easier to comprehend. Just a few of these details will be added here. There are actually several different kinds of RNA. In addition to mRNA there is another type of RNA, called transfer RNA (tRNA), which brings the amino acids to the mRNA during the assembly process. There are many different forms of tRNA, each one with a particular amino acid attached to it. Each tRNA has a binding site that can attach to the mRNA. That binding site is called the anti-codon and consists of three nucleotides that are complementary to the mRNA nucleotides. For example, if a section of mRNA is being used to make protein, and the next three nucleotides are UAC, the tRNA with the anti-codon AUG will attach to those three nucleotides and that tRNA will have the amino acid tyrosine attached to it. Tyrosine will then be added to the growing chain of amino acids. There is a third kind of RNA called ribosomal RNA (rRNA) which, with the addition of some proteins, forms a ribosome. A ribosome attaches to a strand of mRNA at the point at which protein making will begin, and then moves from codon to codon providing a framework for the amino acid assembly. Transfer RNAs reach the mRNA by entering the ribosome, and then the ribosome attaches the amino acids to each other. The ribosome moves down the mRNA strand much like a car on a roller coaster, and, as it does, a growing chain of amino acids, called a peptide, emerges from it.
DNA is a double-stranded helix. When it is used as a template to make mRNA, the two strands are first separated at the point where transcription will begin, and then just one strand is "read" or transcribed to make the mRNA. This strand is called the template strand. One might think that the template strand would be the one used when DNA sequences are reported, but it is not. Instead it is the conventional practice to report the sequence of the other strand of DNA, which is called the "coding DNA". The reason for this is that, since coding DNA and mRNA are both complementary to the template DNA, their sequences are almost identical. The only difference is that wherever there is a T for thymine in DNA, there is a U for uracil in mRNA. Researchers find it less confusing to use DNA sequences and mRNA sequences that are nearly alike. The "c." that precedes numbers in the mutation database stands for "coding DNA sequence". This is so that it will not be confused with the other DNA strand which, in fact, actually serves as the template for mRNA. (There are other sequences with which it could be confused as well.) In the charts that follow, you will see the coding DNA and template DNA sequences that code for the mRNA codons. The mRNA codons code for the amino acids. This is accomplished by transfer RNAs that each have a particular amino acid attached. Only the transfer RNA with a nucleotide triplet (anti-codon) that is complementary to the next three nucleotides on the mRNA strand (codon) will be brought in and temporarily attached to the mRNA. The amino acid that it brings with it will be added to the growing chain of amino acids which form a peptide.
The Code for Amino Acids
The Fumarase Gene
Proteins are not always just simple peptides. Often, after peptides are made they are joined together and twisted into shape to make protein subunits. The subunits combine to make the finished protein. Fumarase is made out of four identical subunits that are put together (a homotetramer). The subunits are coded for by the fumarase gene, which is located on chromosome 1. (The precise location is 1q42.1, meaning that it is in sub-band 1 of band 42 on the long arm (q) of chromosome 1.) The fumarase gene consists of more than 22,000 nucleotide base pairs. Not all of these base pairs are used to code for the amino acid sequence of fumarase. There are 10 sections of the gene that are used, and they are separated by much larger intervening sequences that are not used.
The parts that are used are called exons, and the parts in between are called introns. After the messenger RNA is made from the DNA, the intron portions of the mRNA are cut off and the exon portions are joined (spliced) together to form one long chain that is then used to make fumarase. The exons of fumarase are numbered 1 through 10. Each of the introns is given the number of the exon that precedes it.
Changes or mutations found in introns that are located close to exons may sometimes cause incorrect splicing of the mRNA, which leads to incorrect proteins being made. Most of the mutations of the fumarase gene that are known to cause HLRCC are located on one of the 10 exons. However, there is one splice site mutation which is located on intron 2 that is recorded on the TCA Cycle Gene Mutation Database. There are also 5 more splice site mutations reported in a recently published paper from France. Most genetic screening only looks for mutations in the exons plus the portions of the introns that are located near the exons.
Types of Mutations
When cells divide, the chromosomes containing the DNA divide so that half the DNA goes to each of the two new cells. Later, each cell replicates the DNA it has received. During this process of cell division and DNA replication, mistakes can be made. Mistakes can also occur when DNA is damaged by radiation, sunlight and chemicals. Often the cells can repair damaged DNA, but as we get older our repair mechanisms don't work as well and mutations can accumulate.
There are many different types of mutations that can occur. Sometimes a different nucleotide is substituted for the correct nucleotide. This is called a substitution mutation and is denoted with the symbol >. For example, if the nucleotide thymine was put in the sequence where cytosine was supposed to be, it would be written as C>T. Sometimes a nucleotide may be inserted between two nucleotides. This is called an insertion and is indicated by "ins". Sometimes a nucleotide is accidentally removed from a sequence. This is called a deletion and is indicated with "del". These three types of mutations are called point mutations because they are caused by changes in a single nucleotide. Some mutations involve more than one nucleotide. For example, deletions may involve the omission of several nucleotides. In some cases, an entire exon or even a whole gene may be deleted. Another type of mutation that can involve many nucleotides is a duplication, in which a segment of nucleotides is duplicated resulting in a repetition of the same sequence. This is indicated with "dupl". Another type of error is an inversion, in which a segment of DNA is reversed, and the sequence of nucleotides is backwards. This is indicated with "inv". Some of these mutations can be frameshift mutations, that is, they shift the reading frame of the mRNA by one or two nucleotides. If this happens, the nucleotides are no longer read in the correct group of three. This causes the wrong amino acids to be put in place after the mutation point and also, since accidental stop codons can be created, usually causes the resulting peptide to be much shorter.
Another type of error is the accidental creation of a splice site where it should not occur. There are certain sequences of nucleotides that signal where splicing is supposed to be. Sometimes a mutation creates a false splice site signal that leads to a mistake in the protein product.
The gene mutations that have been discovered for fumarase are listed in the LOVD TCA Cycle Gene Mutation Database. Actually, the preferred term for a change in a gene is "sequence variant", because not all changes cause disorders, as the word mutation might imply. Some are just harmless variations that can be found in normal people. The database lists the alterations in the usual sequence of nucleotides, and it indicates whether the person was normal or had a disorder such as HLRCC, MCUL, FH deficiency, etc. Each listing of a sequence variant starts with the number of the exon on which it was located (or if it was on an intron, the number of the nearest exon.) Then, in most cases, there is "c." This stands for "coding DNA sequence". This means that it is based on the sequence of nucleotides in the coding DNA strand and not, for example, the template DNA strand, an RNA strand, or mitochondrial DNA. In this system, for each exon of the fumarase gene, the normal sequence of DNA nucleotides for that exon has been recorded in a databank. The nucleotides are numbered, and the numbers continue and do not start over with each exon. The nucleotides of the introns are not numbered. (However, in "genomic" sequences, all of the nucleotides in both introns and exons are included.) The number that follows is the number of the nucleotide (or nucleotides) in which there was a change. This is followed by a symbol or abbreviation that indicates the type of change. When the location of the change is on an intron, the number of the closest nucleotide that is on an exon is given, and then there is a plus or minus sign followed by the number of nucleotides away from that nucleotide that the change is located. For example, suppose a mutation is located between exons 3 and 4, but is closer to exon 3, and is located 24 nucleotides away from exon 3. The number of the last nucleotide in exon 3 would be given followed by +24. If instead, the mutation is only 16 nucleotides away from exon 4, then the location would be given as the number of the first nucleotide in exon 4 followed by "-16". Sometimes genetic testing laboratories don't use this nomenclature. In the above example, they might use "IVS3" instead of the number of the closest nucleotide. This was the old way of indicating a location on intron 3. IVS stands for intervening sequence, which means intron.
If you have received a genetic report indicating a fumarase mutation, you may wish to see the part of the coding DNA sequence in which your mutation is located. You may do this by going to the LOVD FH Homepage and selecting "Genomic Reference Sequence", which shows the entire sequence of nucleotides in the fumarase gene, with the amino acids coded for indicated by letters underneath.
Examples of HLRCC Mutations from the LOVD TCA Cycle Gene Mutation Database
02 c.157G>T Nonsense This mutation is located in exon 2 of the fumarase gene. Nucleotide 157 was changed from guanine to thymine. This resulted in the creation of a stop codon where there should not have been one, which caused a shorter protein than normal to be made.
03 c.302G>C CGA>CCA Missense This mutation is located in exon 3 of the fumarase gene. Nucleotide 302 was changed from guanine to cytosine. This resulted in the amino acid proline being put in where arginine was supposed to be.
05 c.671_672delAG Frameshift This mutation is located in exon 5 of the fumarase gene. Nucleotides 671 and 672 were lost leading to a frameshift in reading which caused subsequent codons to be wrong.
How Do Mutations in Fumarase Cause Cancer?
Fumarase is believed to be a tumor suppressor, that is, its presence prevents the formation of tumors. There are several theories about how fumarase is a tumor suppressor and how lack of fumarase in cells can cause them to become cancerous. HLRCC tumors are often atypical (unusual) in their appearance with large misshapen nucleoli. A distinguishing feature of HLRCC is the presence of large eosinophilic nucleoli surrounded by clear halos.
Fumarase is known to be a very important enzyme for energy production in the mitochondria. In the mitochondria there is a series of biochemical reactions that is called the Krebs cycle. (See diagram below.) In the cycle, succinate is converted to fumarate and fumarate is converted to L-malate. The conversion of fumarate to malate occurs because of the action of the enzyme fumarase. If fumarase is missing, then an excess of fumarate accumulates, and the energy producing cycle is blocked.
The Krebs Cycle was discovered by Hans A. Krebs. It takes place in the innermost space of the mitochondria and requires the presence of oxygen. Together with the electron transport chain, it produces most of the cells' energy.
Theory 1 – Blockage of Krebs Cycle
One theory of cancer in HLRCC is that the blockage of the Krebs Cycle causes chemicals called reactive oxygen species (like hydrogen peroxide) to be formed. The reactive oxygen species can cause cells to form tumors.
Theory 2 – Lack of Fumarase
A second theory is that fumarase has other important functions in cells in addition to its role in the Krebs cycle. For example, Ohad Yogev, et al, reported in 2010 that fumarase was involved in DNA repair. It was already known that two different forms of fumarase are made from the same fumarase gene. One is sent to the mitochondria for its function in the Krebs Cycle. The other, called cytosolic fumarase is sent to the cytoplasm. Up until recently, the function of cytosolic fumarase was not known. By studying yeast cells it was discovered that cytosolic fumarase plays a role in detecting and repairing DNA damage, particularly double-strand DNA breaks. According to this theory, if cells lack the fumarase they need to repair damaged DNA, they are more likely to grow into tumors.
Theory 3 - Pseudohypoxia
A third theory of how loss of fumarase leads to cancer is the pseudohypoxia theory. There are times when the cells in our body don't get enough oxygen. This is called hypoxia. The cells have a way of dealing with this that uses a chemical regulator called hypoxia inducible factor (HIF). HIF has the ability to activate genes that protect cells from low oxygen. These genes prevent cell death, promote the growth of blood vessels, increase energy production, and promote cell growth and replication. Some of these genes are: vascular endothelial growth factor (VEGF), glucose transporter 1 (GLUT-1), platelet-derived growth factor (PDGF), transforming growth factor alpha (TGF-alpha), and erythropoietin (EPO). These same genes which prevent cell death during times of low oxygen are also believed to promote tumor growth.
When there is adequate oxygen in the cells, HIF is not needed. Cells have a mechanism that keeps it from being produced when it is not needed. HIF is made from subunits, one of which is called HIF-1alpha. When this subunit is in adequate supply, HIF can be made. If it is not available, then HIF cannot be made. HIF-1alpha is constantly being produced, but it is also constantly being destroyed by enzymes called prolyl hydroxylases (PHDs). PHDs must have oxygen to do their job. If the oxygen level is too low, PHDs are unable to destroy HIF-1 alpha, and we say that HIF-1alpha has been stabilized. When this happens HIF can be made from its subunits. After HIF is made, it increases the transcription rate of the above-mentioned genes that increase the growth of blood vessels and help cells survive.
In addition to oxygen, PHDs also need to bind to alpha-ketoglutarate (AKG) in order to perform their function of destroying HIF-1alpha. An abundance of fumarate can block the alpha-ketoglutarate from binding with the PHDs. A lack of fumarase for the Krebs cycle can lead to a buildup of excess fumarate, which can prevent AKG from binding to PHDs. This causes the stabilization of HIF-1alpha even when oxygen levels are high. This condition is called pseudohypoxia (false low oxygen). With HIF-1alpha stabilized, HIF becomes abundant and the genes that promote cell and blood vessel growth are perpetually turned on. According to the pseudohypoxia theory, this is a major factor leading to tumor formation.
The Most Recent Research
Although stabilization of HIF-1alpha does occur in HLRCC, researchers wondered why HLRCC is so different from other cancer syndromes in which HIF is believed to play the primary role, such as VHL and SDH deficiency. VHL and SDH deficiency both cause paragangliomas, pheochromocytomas, and clear cell renal cell carcinoma. But HLRCC has a different clinical picture that includes uterine and cutaneous leiomyomas and papillary renal cell cancer. Another puzzle was that HIF causes upregulation of vascular endothelial growth factor (VEGF), and anti-VEGF agents are very effective against clear cell RCC, but they are not very effective against papillary type 2 RCC. For these reasons researchers were not satisfied that increased HIF is the primary mechanism for tumor formation in HLRCC.
A clue to what might be the cause of papillary tumors came from diabetes research. It was discovered that in diabetics, mitochondrial stress could cause accumulation of fumarate and that the excess fumarate could damage proteins. Proteins can be thought of as chains of amino acids that are bent and twisted into precise shapes, and these shapes are stable because of chemical attractions that form between the amino acids in different places along the chains. One of the amino acids is cysteine. When fumarate is present in excess, it can have a special chemical reaction with cysteine. When this happens, the cysteine is converted into S-(2-succinyl)-cysteine, or 2SC for short. This process is referred to as protein succination. When cysteine is succinated, it can change the attraction between the amino acids that help hold the shape of the protein. This can cause a protein to change its shape, and the change in shape can alter its ability to function. In diabetics, it is believed that the enzyme GADP is damaged by succination of its cysteine residues and that this contributes to some of the complications of diabetes.
HLRCC researchers who learned about succination of protein in diabetes suspected that there could be similar protein alterations occurring in HLRCC. In HLRCC, conversion of a normal cell to a tumor cell can occur when a cell sustains damage to its one remaining normal FH gene. After this happens, the cell lacks fumarase so the Krebs cycle is stopped at fumarate, and fumarate accumulates instead of being converted to malate.
To understand if succination of proteins by excess fumarate plays a role in HLRCC, researchers focused on a protein called KEAP1. KEAP1 is important because it controls the degradation of another molecule called Nrf2. KEAP1 is shaped a bit like a “V”, and normally the ends of the V attach to both ends of Nrf2 and hold it until it is degraded. If a cell becomes stressed, Nrf2 is released from KEAP1, and it goes to the nucleus where it binds with another protein and then binds to a promoter site on the DNA called ARE (Antioxidant Response Element). ARE is a segment of DNA that activates a group of antioxidant response genes that help protect the cell in times of stress. The researchers suspected that over-activation of some of these anti-oxidant pathways might play a role in the formation of papillary tumors. Researchers learned that loss of the FH gene resulted in the inactivation of KEAP1. They hypothesized that cysteines in the KEAP1 protein were being succinated by fumarate. The researchers discovered that when certain cysteine residues in KEAP1 were succinated, its shape changed so that the arms of the V were more spread out. When this happened, KEAP1 could no longer “grasp” both ends of Nrf2 at the same time, and so it could no longer hold on to it. Without KEAP1 playing its role in the disposal of Nrf2, Nrf2 could accumulate and upregulate the ARE genes. One particular antioxidant whose expression is increased this way is called AKR1B10. Researchers think this gene plays an important role in the formation of papillary tumors.
Patrick Pollard and other researchers at Oxford used a mouse model to investigate the cause of tumor formation in HLRCC. They used different combinations of genetically altered mice, including some with two active FH genes, some with only one active FH gene, and some with no active FH genes combined with active and inactive genes for HIF-1alpha and HIF-2alpha. The mice with inactive FH genes developed cystic changes in their kidneys that are believed to be precursors to papillary tumors because these tumors can arise from cyst walls. The researchers discovered that inactivation of FH led directly to the upregulation of Nrf2 mediated antioxidant pathways and caused the cystic changes believed to be similar to those in HLRCC. Furthermore, they were able to show that the cystic changes were independent of HIF and PHDs. They also demonstrated that the levels of fumarate present were sufficient to succinate KEAP1 and thereby increase Nrf2 activation of ARE genes.
It may turn out that the resistance of HLRCC tumor cells to the usual cancer treatments of chemotherapy and radiation may be due to the protection of the cells by the upregulated ARE genes. For example, inactivation of KEAP1 in non-small-cell lung cancer has been linked to chemotherapy resistance.
The knowledge that fumarate causes proteins to be succinated, led the Oxford researchers to devise staining methods for detecting HLRCC. The staining methods were based on the fact that 2SC is present in HLRCC tumors, but not in other types of kidney tumors. The staining methods involved putting drops of rabbit antibody to 2SC on slides of tumor tissue. This was followed by drops of antibody to rabbit antibody that had either an attached enzyme used for visualization, or an attached fluorescent chemical for use with a fluorescent microscope. When an enzyme was used, a substrate for the enzyme was added next so that areas of tissue containing 2SC became colored. When the researchers tested the stain with a series of patient tumors, the stain predicted FH mutations 100% of the time. There was no staining of non-HLRCC tumors, not even when they tested it on tumors caused by deficiency of SDH, which is another Krebs cycle enzyme. Soon this type of staining will be available to pathology laboratories everywhere so that papillary and collecting duct tumors can be routinely tested. This is so important because many times when there is no family history or other symptoms, HLRCC goes undetected.
This new research has suggested many new possibilities for targeting genes and enzymes in the treatment of HLRCC. One possibility is to inhibit the haem oxygenase gene (also spelled "heme oxygenase”). Haem oxygenase is one of the enzymes that is upregulated by Nrf2/ARE, and it is used by HLRCC tumor cells as an alternate pathway to generate energy, since the Krebs cycle is truncated without fumarase. Another possibility is silencing the AKR1B10 gene with specially made segments of double-stranded RNA that are complementary to the gene. Still another approach is to use drugs that greatly increase the production of reactive oxygen species (ROS). This was tried using a combination of cisplatin and bortezomib. All of these methods have shown promising preliminary results in vitro and in vivo.