This web page was produced as an assignment for Genetics 677, an undergraduate course at UW-Madison
Conclusions
The purpose of my research was to use bioinformatics programs to analyze the LDLR gene and protein in order to better understand familial hypercholesterolemia. Initial homology and phylogeny analysis for both the mRNA and amino acid sequence of LDLR yielded very similar results. Pan troglodytes has the most similar sequence and is evolutionarily closest to Homo sapiens, followed by Bos taurus and Mus musculus. These results were expected, given the importance of cholesterol metabolism in higher organisms. One finding that was somewhat surprising was the fact that as the homologs grew more divergent their DNA and mRNA sequences decreased in size while the amino acid sequence remained fairly constant. For instance, Dan rerio, the most divergent homolog, has DNA and mRNA sequences of about half the length of Homo sapiens, while it has slightly more amino acids. This can be viewed on the Homology page. Further analysis also found conserved LDLR-like genes in Drosophila melanogaster and Caenorhabditis elegans, indicating a deep evolutionary conservation of cholesterol metabolism.
I also used gene ontology to gain a more in depth understanding of all the functions and localization of LDLR. The results were expected given my reading of the literature and analysis of the LDLR domain functions. Most of the ontology terms supported LDLR's involvement in lipid metabolism, cholesterol homeostasis, endocytosis, and lipoprotein particle binding. Protein domain analysis supported these ontology terms as well. LDLa domains bind LDL particles via interactions with ApoB. LDLb domains form a "six-bladed" beta propeller that displaces the LDL particle once the receptor is endocytosed. The complex biochemistry that LDLR undergoes in order to bind and internalize LDL particles before then being recycled back to the plasma membrane helped me realize why so many different mutations cause an FH phenotype.
Because most of the protein is made up of functionally important domains, it is not surprise a slight amino acid change could cause dysfunction of LDLR. However, I did want to see if there was a certain area within the protein where more mutations occurred compared to other areas. I went through the LDLR mutation database and counted the number of different mutations that cause amino acid changes. I then created a map of the mutation numbers across the LDLR protein, which can be seen in figure 1. It can be seen that 893 different mutations have been found within the LDLR gene that cause amino acid changes. There are a few mutations in the database within the gene promoter and as well as within exons that do not cause amino acid changes but still give an FH phenotype. These mutations were not included in figure 1. Overall, the distribution of mutations is fairly consistent throughout the entire protein. Although it may seem like there are significantly more mutations within the LDLa domains compared to the LDLb domains, this is actually a reflection of the number of amino acids contained within each area of the protein. The 401 different mutations found within the LDLa domains is over a span of 315 amino acids, while the 275 mutations within the LDLb domains are over a span of only 220 amino acids. This comes out to a ratio of 1.27 mutations per amino acid across the LDLa domains compared to 1.25 mutations per amino acid across the LDLb domains. This makes sense because of the complex biochemistry involved with LDLR, as stated earlier.
Because most of the protein is made up of functionally important domains, it is not surprise a slight amino acid change could cause dysfunction of LDLR. However, I did want to see if there was a certain area within the protein where more mutations occurred compared to other areas. I went through the LDLR mutation database and counted the number of different mutations that cause amino acid changes. I then created a map of the mutation numbers across the LDLR protein, which can be seen in figure 1. It can be seen that 893 different mutations have been found within the LDLR gene that cause amino acid changes. There are a few mutations in the database within the gene promoter and as well as within exons that do not cause amino acid changes but still give an FH phenotype. These mutations were not included in figure 1. Overall, the distribution of mutations is fairly consistent throughout the entire protein. Although it may seem like there are significantly more mutations within the LDLa domains compared to the LDLb domains, this is actually a reflection of the number of amino acids contained within each area of the protein. The 401 different mutations found within the LDLa domains is over a span of 315 amino acids, while the 275 mutations within the LDLb domains are over a span of only 220 amino acids. This comes out to a ratio of 1.27 mutations per amino acid across the LDLa domains compared to 1.25 mutations per amino acid across the LDLb domains. This makes sense because of the complex biochemistry involved with LDLR, as stated earlier.
Figure 1-Schematic showing protein domains and the number of different mutations that cause amino acid changes within LDLR. Domains are not to scale, however the numbers below the domains are based on SMART amino acid domain boundaries. Image was created using PowerPoint. Click to enlarge.
LDLR has been shown to interact, either directly or indirectly with many proteins. Some of the most important proteins shown to interact directly with the LDLR protein include APOB, APOE, PCSK9, and LDLRAP1. Explanations of the functions of these proteins can be found on the Interactions page. Indirectly interacting proteins, either via transcriptional regulation or other mechanisms, include SREBF1, SREBF2, HMGCS1, and NR1H3. SREBF1 and 2 are transcription factors that upregulate LDLR gene expression when cellular concentrations of cholesterol are low (1). A diagram of the SREBF pathway can be found here. HMGCS1 is the gene that encodes for HMG-CoA synthase, an important enzyme in the cholesterol biosynthesis pathway (2). This gene is also upregulated by SREBF1 and 2. Finally, NR1H3 encodes for a protein known as Liver X Receptor (LXR). This transcription factor is activated by high cholesterol levels causing it to upregulate multiple genes involved in cholesterol efflux (1). A diagram of this pathway can be found here.
As for the microarray analysis on LDLR, I was very disappointed. Most of the microarray studies were unrelated to FH and only showed LDLR expression levels in response to individuals with other diseases or after certain chemical treatments. I did find one study that did a comparison of normal monocytes to FH monocytes using mircoarrays (3). They found substantial differential gene expression in the FH cells, however the dataset is too large to download (>180Mb). I think that microarray analysis is an important tool that needs to be utilized more frequently in studying and treating FH. There are multiple cell types that should be compared between normal and FH individuals in order to establish expression patterns that could be used as a diagnostic tool. I will further explain the usefulness of microarrays on the Future Directions page.
References
1. Weber, L. W., Boll, M., & Stampfl, A. (2004). Maintaining cholesterol homeostasis: sterol regulatory element-binding proteins. World Journal of Gastroenterology, 10(21), 3081-7. PMID: 15457548
2. Vock, C., Doring, F., & Nitz, I. (2008). Transcriptional regulation of HMG-CoA synthase and HMG-CoA reductase genes by human ACBP. Cellular Physiology and Biochemisty, 22(5-6), 515-24. PMID: 19088433
3. Mosig, S., Rennert, K. Buttner, P., Krause, S., Lutjohann, D., Soufl, M., Heller, R., Funke, H. (2008). Monocytes of patients with familial hypercholesterolemia show alterations in cholesterol metabolism. BMC Medical Genomics, 1, 60. doi:10.1186/1755-8794-1-60
2. Vock, C., Doring, F., & Nitz, I. (2008). Transcriptional regulation of HMG-CoA synthase and HMG-CoA reductase genes by human ACBP. Cellular Physiology and Biochemisty, 22(5-6), 515-24. PMID: 19088433
3. Mosig, S., Rennert, K. Buttner, P., Krause, S., Lutjohann, D., Soufl, M., Heller, R., Funke, H. (2008). Monocytes of patients with familial hypercholesterolemia show alterations in cholesterol metabolism. BMC Medical Genomics, 1, 60. doi:10.1186/1755-8794-1-60
