Functional Non-Coding DNA Part I Non-coding genes and non-coding elements of coding genes BNFO 602/691 Biological Sequence Analysis Mark Reimers, VIPBG

What Does ‘Functional Non-Coding DNA’ Mean? DNA whose sequence affects transcripts made from DNA in some way Could affect transcription levels, splicing or sequestering of RNA Three main ways to identify functional non- coding elements – Sequence characteristics – favored bases – Genomic conservation – Epigenetic marks and open chromatin especially outside of genes

Types of Non-Coding Elements Non-coding RNAs – miRNAs, lncRNAs, etc Non-coding gene elements – UTRs, splice sites, poly-adenylation sites, splice sites and regulating element, RNA-binding sites DNA elements outside genes – our main focus – Promoters – Enhancers/Silencers – Insulators

Types of Non-Coding RNA microRNAs Silencing RNAs Small nuclear/nucleolar RNAs Piwi-Interacting RNAs Long Non-Coding RNAs Circular RNAs Still other RNAs??? Comprehensive data base at

Micro-RNAs Micro-RNAs are small non-coding RNA molecules, about 21– 25 nucleotides in length They are processed from much longer genes, or from introns within mRNA, by several molecular pathways Micro-RNAs base-pair with complementary sequences within mRNA molecules, often in 3’ or 5’ UTR. miRNA binding usually results in gene repression either via translational stalling or by triggering mRNA degradation Image by Charles Mallery, U of Miami

Micro-RNAs The human genome encodes over 1500 miRNAs, which are believed to affect more than half of human genes miRNAs are abundant in many cell types – Thousands of copies per cell of some miRNAs – Those within gene introns share regulation miRNAs are well-conserved across vertebrates – No orthologs between plant and animal miRNAs – miRBase is the comprehensive repository of micro- RNAs

Other Short RNAs: siRNA Small interfering RNAs are double-stranded with an overhang They are processed by some of the same machinery as miRNAs and have some of the same effects

Other Short RNAs: piRNA Piwi-Interacting RNAs are longer base single-stranded RNAs – PIWI (P-element Induces Wimpy Testis) protein Over 50,000 sequences known in mouse – They are the largest class of nc-RNA They seem to play an ancient role in defense against retro-viruses and transposons

Other Short RNAs: snRNAs & snoRNAs Small nuclear RNAs (snRNAs) are typically ~ 150 bases long, and associate with protein – Many conserved copies of each snRNA gene – U1-U6 snRNAs key parts of splicing machinery Small nucleolar RNAs (snoRNAs) – Guide chemical modifications of other RNAs – Prader-Willi syndrome results from deletion of region containing 29 copies of SNORD116 on chr 15q11 U6 snRNA

Long Non-Coding RNAs Many long (>200bp) stretches of genome are transcribed and have epigenetic marks like those of protein-coding genes Most of these are spliced RNAs with two (or more) exons GENCODE v15 has 13.5K lncRNA See also – Derrien et al, Genome Research 2012 – Lee, Science 2012 From Derrien et al Genome Res 2012

Many lncRNAs Induce Silencing Coat nearby gene(s) and silence them Xist binds to gene clusters first Xist binds disparate parts of chromosome Many lncRNA are antisense to genes Some lncRNAs maintain pluripotency of stem cells From Jeannie Lee lab (Harvard) website

Long Non-Coding RNAs - 2 Most lncRNAs are expressed in only a few tissues Most human lncRNAs are specific to the primate lineage From Derrien et al Genome Res 2012

Circular RNAs Several thousand non-coding RNAs apparently form circular structures Many form complexes with AGO and seem to absorb attached miRNAs, blocking processing CDR1 has 70 conserved binding sites for mir7

Functional Pseudo-Genes Pseudo-genes are copies of genes that are decaying and rarely (never) make proteins Some pseudo-genes act to absorb negative regulators of the original gene – eg. SRGAP2B

How to Identify Non-Coding RNAs? Short (and long) RNA transcriptomes Promoter chromatin marks for independent (non-embedded) miRNAs and lncRNAs

DEMO: Display HOTAIR & XIST Tracks in UCSC Browser

Non-Coding Elements of Genes TSS 5' UTRs Introns Splicing regulation sites 3' UTRs Termination/Poly-adenylation sites

Transcription Start Sites Transcription of most genes may initiate at several distinct clusters of locations with distinct promoters for each TSS Two major types of metazoan TSS: CG-rich broad TSS, and narrow (often tissue-specific) TSS

Transcription Start Sites Transcription often starts at CG within promoter

5’ Untranslated Regions First exon often contains dozens to thousands of bases before Start codon (median 150) Sometimes contains regulatory sequences, e.g. binding sites for RNA binding proteins, and translation initiators

Splice Regulatory Sites Splicing is achieved through binding of spliceosome to recognition sequences on nascent RNA molecule

Splice Regulatory Sites Tissue-specific splice regulatory sites are highly conserved From Merkin et al Science 2012

Splicing Patterns Evolve in All Tissues Except Brain From Merkin et al Science 2012

Non-Coding Elements in Coding Exons Many regulatory sites occur within coding exons, esp. toward 5’ end These constrain some codons as much as protein sequence Many human SNPs break TFBS but have little effect on protein (AFAWK) From Stergachis et al Science 2013

3’ Untranslated Regions Longest exon is usually 3’UTR (>1000 nt) Typically 1/3 – 1/2 of a gene is in 5’ & 3’ UTRs 3’UTR has binding sites for miRNAs and RNA binding proteins AU-rich elements (AREs) stabilize mRNA Proteins recognize complex secondary structure GRIK4 3’UTR secondary structure is conserved

RNA Binding-Protein Sites mRNAs are usually further processed (e.g. transported or sequestered) RNA binding proteins recognize specific motifs within secondary structure of 3’ or 5’ UTR These sites are often highly conserved From Ray et al Nature 2013

Poly-adenylation/Termination Sites Transcripts can be terminated and poly-adenylated at sites with specific sequences Most genes have alternate poly-adenylation sites Median lengths of 3’UTR are 250 & 1773 bp (mouse)

Poly-adenylation/Termination Sites Rapidly proliferating cells express gene isoforms with short 3’ UTRs Neurons typically have longer 3’ UTRs Elkon et al, NRG 2013Types of alternate poly-adenylation

DEMO: GAPDH and GABRA1 in UCSC Browser