Knowledge of mammalian CDC73 has essentially been driven by focused efforts to map, and then characterize, a gene responsible for familial primary hyperparathyroidism. Primary hyperparathyroidism (HPT) is estimated to affect three in every 1000 people. Primary HPT refers to hypersecretion of parathyroid hormone (PTH) by chief cells in the parathyroid in a setting of disrupted calcium homeostasis. Sporadic HPT is caused by the presence of a single parathyroid adenoma in the majority of patients (80–85%) or hyperplasia that affects more than one gland in 15–20% of patients. Sporadic parathyroid carcinoma is an exceedingly rare tumor, responsible for <1% of cases of primary HPT. Parathyroid abnormalities can also present as components of the familial syndrome Hyperparathyroidism Jaw Tumor syndrome (HPT-JT). Abnormalities that are part of the HPT-JT phenotype include parathyroid carcinoma (10–15%), fibro-osseous tumors of the mandible or maxilla, renal hamartomas and/or cystic kidney disease, and an increased risk of Wilm’s tumors (reviewed in (Marsh et al. 2007). Benign parathyroid tumors that occur within HPT-JT are generally believed to have malignant potential (Howell et al. 2003). An increased risk of developing uterine tumors has also been reported. While parathyroid tumors have been found in children who are members of HPT-JT families, the majority (~80%) of affected individuals in these families will develop HPT due to the presence of a parathyroid tumor by 40 years of age (Marx et al. 2002). Within the context of HPT-JT, parathyroid tumors are aggressive and often have a cystic histology.
Using traditional linkage strategies and a candidate gene approach, members of numerous HPT-JT families were studied to identify a gene mapping to chromosome 1q25 as the causative gene for HPT-JT (Carpten et al. 2002). The HPT-JT gene, called HRPT2 (hyperparathyroidism 2), had previously been known as C1orf28 (chromosome 1 open reading frame 28) (Sood et al. 2001). This gene consists of 1596 nucleotides organized into 17 exons encoding the ubiquitously expressed 531 amino acid protein parafibromin (named for parathyroid disease and fibro-osseous lesions). In recent years, it has become more common to refer to this gene as CDC73 (Cell Division Cycle 73) named for its budding yeast homologue Cdc73. The C-terminus of CDC73 shows strong homology to yeast Cdc73; however, does not have known functional domains. The Drosophila homologue of CDC73 is Hyrax (Mosimann et al. 2006).
Mutation and Allelic Loss of CDC73
The majority of CDC73 mutations are nonsense or frameshift mutatations, predicted to lead to truncation of the full length protein in a manner consistent with loss of wild-type function. Few missense mutations have been reported (reviewed in (Marsh et al. 2007)). Around 80% of CDC73 mutations occur in exons 1, 2, and 7, marking these exons as mutational hotspots. Approximately 60% of HPT-JT families have an associated germline mutation in CDC73, including reports of a number of founder mutations. Around 70% of sporadic parathyroid carcinomas have a somatic mutation in CDC73. CDC73 mutations have also been identified in renal tumors. Further evidence that CDC73 functions as a tumor suppressor is the fulfillment of Knudson’s 2-hit hypothesis, given that mutations have been identified on both alleles, as well as the presence of a mutation in conjunction with loss of the wild-type allele (Howell et al. 2003, Shattuck et al. 2003). Hypermethylation of the CDC73 promoter is not seen in parathyroid tumors (Hahn et al. 2010).
Cellular Localization of CDC73
CDC73 is a nuclear protein with an evolutionarily conserved bipartite nuclear localization signal at residues 125–139 (Hahn & Marsh 2005, Lin et al. 2007). Its nuclear localization is consistent with its role in transcription discussed below. Further, CDC73 also localizes within the nucelolus, with three nucleolar localization signals identified at resides 76–92, 192–194, and 393–409 (Hahn & Marsh 2007). Loss of nuclear CDC73 as the result of a CDC73 mutation(s), at times combined with allelic loss at this locus, has been harnessed as an immunohistochemical diagnostic marker of parathyroid carcinoma (Gill et al. 2006). Furthermore, while mechanism(s) has not been elucidated, tissue array studies of CDC73 protein in primary tumors have reported decreased nuclear CDC73 staining in breast, gastric, and colorectal tumors.
CDC73 Is a Member of the Human PAF1 Transcriptional Complex
Cellular Function of CDC73
CDC73 Affects Chromatin Remodeling
CDC73 has been shown to influence cancer-associated chromatin remodeling. It is a binding partner of the E3 RING finger ubiquitin ligases RNF20 and RNF40 (Hahn et al. 2012) (Fig. 3). These two ubiquitin ligases function in a complex to add a single ubiquitin molecule to histone H2B at lysine (K) 120 (H2Bub1). This posttranslational histone modification works to decondense chromatin, actively pushing apart chromatin strands, making it more accessible to DNA repair factors, as well as factors involved in gene transcription (reviewed in (Cole et al. 2015). Parathyroid tumors that have a CDC73 mutation show loss of nuclear CDC73 and loss of H2Bub1 detected by immunohistochemistry (Hahn et al. 2012). Downregulation of CDC73 in cancer cell line models using small interfering (si)RNA also shows loss of H2Bub1 (Hahn et al. 2012). Like CDC73, H2Bub1 has known roles in transcriptional elongation, its presence important in creating an open chromatin structure that allows RNA polymerase II to move through and transcribe genes (Fig. 1). It also has roles in 3′-end processing and other key cellular functions. Cancer-associated mutations in CDC73 and the key regulatory epigenomic modification of H2Bub1 are clearly intimately entwined at a physical and functional level.
In summary, CDC73 is a protein with tumor suppressive roles that functions as a member of the RNA polymerase II-PAF1 complex. To date, no role for CDC73 has been reported independently of this complex. While there are no known functional domains, CDC73 has homologues in budding yeast and Drosophila, both model systems having shined light on the functions of this tumor suppressor. Knowledge of the interaction partners of CDC73, including β-catenin and the RING finger E3 ligases RNF20 and RNF40, has provided clues as to some of the major cellular roles of this tumor suppressor, including in chromatin remodeling and Wnt signaling. CDC73 is the only member of the PAF1 complex to harbor cancer-associated mutations, with mutations reported in both familial and sporadic parathyroid tumors, as well as in renal cancers. Loss of nuclear CDC73 has been developed as a diagnostic immunohistochemical test for the clinical confirmation of parathyroid carcinoma.
- Gill AJ, Clarkson A, Gimm O, Keil J, Dralle H, Howell VM, Marsh DJ. Loss of nuclear expression of parafibromin distinguishes parathyroid carcinomas and hyperparathyroidism-jaw tumor (HPT-JT) syndrome-related adenomas from sporadic parathyroid adenomas and hyperplasias. Am J Surg Pathol. 2006;30:1140–9.CrossRefPubMedGoogle Scholar
- Hahn MA, Howell VM, Gill AJ, Clarkson A, Weaire-Buchanan G, Robinson BG, Delbridge L, Gimm O, Schmitt WD, Teh BT, et al. CDC73/HRPT2 CpG island hypermethylation and mutation of 5′-untranslated sequence are uncommon mechanisms of silencing parafibromin in parathyroid tumors. Endocr Relat Cancer. 2010;17:273–82.CrossRefPubMedGoogle Scholar
- Sood R, Bonner TI, Makalowska I, Stephan DA, Robbins CM, Connors TD, Morgenbesser SD, Su K, Faruque MU, Pinkett H, et al. Cloning and characterization of 13 novel transcripts and the human RGS8 gene from the 1q25 region encompassing the hereditary prostate cancer (HPC1) locus. Genomics. 2001;73:211–22.CrossRefPubMedGoogle Scholar