Factors Influencing Genetic Transcription Initiation And Termination

Eunice Kennedy Shriver National Institute Of Child Health & Human Development

Investigators

Linked publications & trials

Abstract

Rare human disorders often result from mutations in genes that are critical for cell growth and maintenance. Such genes are sometimes called Housekeeping genes. We began studies on ribonucleases H (RNase H) from a biochemical and structural approach long before their importance in human disorders became apparent. Studies on retroviruses described a viral RNase H in many mammalian viruses including HIV-AIDs virus. The AIDs type viruses utilize their RNases H for viral replication in which the viral RNA genome is copied to DNA whereby duplex RNA/DNA hybrid generates the molecule recognized the viral RNase H. We have described the basic structure of both RNase H1 and H2 and are now aiming to study their impact in vivo. Ribonucleases H are enzymes that recognize RNA/DNA hybrids and digest the RNA. Two types of cellular RNases H are known RNase H1 and RNaseH2. RNases H1 comprise a single protein whereas RNases H2 has three distinct proteins. Early embryonic lethality is observed with loss of any one of these four genes, indicating the essential nature of these two RNases H. RNA/DNA hybrids are found in nuclear DNA and the small DNA in mitochondria. RNase H1 is present in mitochondria while RNase H2 is not. Thus, in mitochondria, RNase H substrates are restricted to RNase H1. We found that mitochondria fail to replicate when RNase H1 is absent suggesting that lethality of RNase H1 null embryos is due to failure to replicate mitochondrial DNA. This is a bit surprising since the RNase H1 in mitochondria accounts for only 5% of total RNase H1. However, human patients with mutated RNase H1 exhibit mitochondrial myopathies, supporting the major action of RNase H1 is its mitochondria RNA/DNA. Eukaryotic RNases H2 are much more abundant than RNase H1 and are capable of hydrolyzing most non-mitochondrial RNA/DNA hybrids possibly even those normally degraded by RNase H1. More than 1 million ribonucleotides (rNMPs) are found per mouse cell when RNase H2 is absent. These rNMPs are generated when RNA primers incorporate the first DNA moiety by the DNA polymerase (RRRD) and when rNMPs are incorporated by DNA polymerases due to incomplete exclusion of uses of riboNTPs during DNA replication and repair. Many of the abundant rNMPs in DNA are thought to exist as single rNMPs embedded in DNA (DDDRDD). RNase H1 is incapable of carrying out these reactions. Aicardi Goutières Syndrome (AGS) is a severe human disorder caused by mutations in more than five genes, three of which encode subunits of RNase H2. We generated a mouse model of AGS with a mutant form of RNase H2 where a Glycine residue near the catalytic center is replaced by Serine. The mutated RNase H2 (G37S) has significant loss of RNA/DNA hydrolysis and incision at DDRDD. Mice homozygous for the G37S mutation are perinatal lethal, which already tells us that this mouse does not mimic the human disorder in which AGS patients are born with a Type I interferonopathy phenotype. Another group has studied a mouse model with a mutation that is found most frequently in AGS patients, finding the mouse exhibited no phenotype at all. After several generations the G37S mice started to produce viable G37S homozygous mice with phenotypes similar to other mice that are models of the human disorder Diamond Blackfan Anemia (DBA). The mice are small and have white ventral spotting. Human patients and mice with DBA mainly have only one copy of a gene (haploinsufficiency) encoding any of several ribosomal protein genes. Involvement of ribosomes is a disorder are classified as ribosomopathies (ribosomal-opathies) which result in decreased protein synthesis. We used a chemical treatment and a genetic mutation to prevent the decrease in protein synthesis neither of which affected the size of the mice nor the white belly spot. DBA patients are often smaller in size than normal which is caused by altering the major nutrient pathway. A few papers report that DBA patients can have their need for blood transfusions by treatment Leucine, an amino acid that can overcome nutritional deficiencies. We provided Leucine in drinking water for the mice prior to breeding and for pups after birth. We found that leucine-treated mice remained small but had decreases or absence is white belly spotting. The white belly spots result from defects in abundance/and or the nutritional status of melanocytes. Improvement of nutrition in the neural crest by feeding Leucine diminishes the white spotting. We have not uncovered the cause of perinatal lethality nor explained the changes to produce live mice for our mouse model of AGS which morphed into DBA in mic. . The appearance of live mice enabled us to continue breeding them over a long time period. We very rarely see Mendelian frequencies of offspring nor any changes in DNA sequences accounting for the observed small size and white belly spotting. The consistent DNA sequence and the variation in unusual distribution and properties of the mice could be due to 1) inability of RNA-seq results to detect long non-coding RNAs or small nucleolar RNAs 2) epigenetic difference and 3) alternative splicing. Our current hypotheses to explain our findings are 1) R-loops are most abundant in rDNA causing dsDNA breaks activating the cGAS-Sting pathway and 2) Failure to obtain normal frequencies of homozygous G37S homozygous mice is related to differences in alternative splicing in more than one gene. We are currently testing the splicing possibility using long read of RNA to be able to capture alternative splicing. Separation of Function for RNases H1 and H2 RNase H1 Transcription of the Rnaseh1 gene produces a single mRNA that is translated by a leaky-scanning process in which translation initiates at the first in-frame AUG produces an N-terminal protein which targets the protein to mitochondria. Upon entering the mitochondria, the N-terminal tag is removed. Skipping the first AUG allows production of a protein initiating at a second AUG that enters the nucleus. Modifying the AUG should eliminate the major RNase H1 activity in the cell. We recently have viable mice which don’t have the second AUG codon. The RNase H1 activity is greatly decreased. We now have to characterize the mice for cellular location of the protein. We will also be producing Embryonic Stem Cells (ESCs) from these mice and examine for sensitivities to DNA damaging agents; what changes might occur in cellular RNA/DNA hybrids and what effects might result from loss of RNase H1 interacting with RNase H1. Our studies describe in Biorxiv doi:https://doi.org/10.1101/2025.04.30.651504 on loss of RNase H1 in early B cell development describe activation of the mitochondrial Unfolded Protein Response without affecting the nuclear R-loops resulting in early embryonic arrest. Patients with mutations in RNase H1 exhibit symptoms of a mitochondrial myopathy. Taken together, it seems that the major role of RNase H1 is its role in mitochondria. Is there a role for RNase H1 in other parts of the cell? RNase H2 We attempted to uncover non-mitochondria substrates by eliminating non-mitochondrial RNase H1 while retaining the mitochondrial form. We obtained viable fertile mice in which the totality of RNase H1 is in mitochondria. We named this RNase H1mt. Remarkably, only 11 transcripts showed significant changes when compared to wt: 7 increased 4 decreased and mostly attributable to differences in males. Perhaps RNase H2 can resolve any substrates normally handled by RNase H1. We have developed a Hybrid Defective RNase H2 mouse that retains only 5% of RNA/DNA activity but has about 70% of the activity for initiating removal of RNMP embedded in DNA – we named this RNase H2-HD. HD mice are viable and fertile when homozygous and even when the second copy of RNase H2 produces no active enzyme. The strain with homozygous RNase H1mt and RNase H2-HD are viable and fertile. One interpretation of our result is that the importance of each enzyme is its unique function: mtDNA replication (H1) and RER (H2). We have identified amino acids in the catalytic subunit of RNase H2 which can be modified to drastically decrease the ability of RNase H2 to cleave at the DDRDD sites. A mouse with modified RNase H2 is embryonic lethal at the same stage of development as complete absence of RNase H2 as in RNase H2-null mice. This result supports the original suggestion that the abundant rNMPs in DNA are the cause embryonic lethality. We have modified amino acid in the catalytic subunit of RNase H2 in attempts to eliminate or significantly reduce the RNA/DNA degradative activity of the enzyme while retaining the DDRDD activity. We produced a mouse with 5% RNA/DNA and 26% of DDRDD activity. RNA exosome: RNA transcripts can hybridize with the template DNA strand co- or post-transcriptionally, forming RNA/DNA hybrids and displacing a single DNA strand to create R-loops. These structures can have both physiological and pathological roles, and cells employ multiple mechanisms to prevent or resolve their formation. Here, we investigate how RNases H and the RNA exosome cooperate in budding yeast to avert unscheduled R-loop accumulation. Eukaryotic RNase H1 and H2 (RNases H) degrade the RNA within RNA/DNA hybrids, eliminating R-loops. The RNA exosome, a major 3’-5’ RNA degradation complex, processes RNA precursors and removes cryptic or defective transcripts, preventing or resolving R-loops. Drop-test growth assays revealed that yeast strains defective for both nuclear RNA exosome and RNases H show hypersensitivity to hydroxyurea, which reduces dNTP pools and inhibits DNA synthesis. This indicates that RNases H and exosome activities suppress R-loop formation that would otherwise exacerbate replicative stress under low dNTP conditions. To dissect their contributions to R-loop control, we performed high-resolution, strand-specific mapping of RNA/DNA hybrids using CRAC (cross-linking and analysis of cDNAs) in exosome-proficient and -deficient strains. Overexpression of a catalytically inactive RNase H1 “bait” in exosome-proficient cells showed widespread hybrid accumulation particularly at highly transcribed loci. Interestingly, exosome-deficient strains showed hybrid accumulation at sites typically targeted by the exosome, such as cryptic unstable transcripts and the 3′ ends of small nucleolar RNA genes, highlighting the exosome’s role in preventing R-loops from these RNA species. Accumulation of R-loops can result in double-stranded DNA breaks, which are repaired by homologous recombination (HR). To further investigate the relationship between RNases H and the exosome in maintaining genome stability, we employed a genetic system that links R-loops to HR (Kim and Jinks-Robertson, DNA Repair, 2011). In mutants defective for both exosome and RNase H activities, we observed a simultaneous increase in RNA accumulation and R-loop formation at the construct, which correlated with a marked rise in recombination rates. In conclusion, RNases H and the RNA exosome act together to maintain genome stability by suppressing the detrimental effects of R-loops.