1. BioSci 203 blumberg lecture 3 page 1 © copyright Bruce Blumberg 2001-2005. All rights reserved Bio Sci 203 Lecture 3 - cDNA library screening Bruce Blumberg (blumberg@uci.edu) –office – 2113E McGaugh Hall –824-8573 –lab (x46873, x43116) –office hours MWF 11-12. http://blumberg-serv.bio.uci.edu/bio203-w2005/index.htm http://blumberg.bio.uci.edu/bio203-w2005/index.htm Link is also main class web site Today –cDNA library construction –cDNA library screening, a.k.a. how do I clone my gene? NO CLASS ON MONDAY 3-7 due to Gottlieb Seminar –Senior stem cell job candidate, attendance is encouraged

2. BioSci 203 blumberg lecture 3 page 2 © copyright Bruce Blumberg 2001-2005. All rights reserved What do I need to know about E. coli genetics? (contd) suppressors –supE - inserts glutamine at UAG (amber) codons –supF - inserts tyrosine at UAG (amber) codons many older phages have S100 am which can only be suppressed by supF –λZAP, λgt11, λZipLOX

3. BioSci 203 blumberg lecture 3 page 3 © copyright Bruce Blumberg 2001-2005. All rights reserved Construction of cDNA libraries What is a cDNA library? What are they good for? –Collection of DNA copies representing the expressed mRNA population of a cell, tissue, organ or embryo –Identifying and isolating expressed mRNAs –functional identification of gene products –cataloging expression patterns for a particular tissue EST sequencing and microarray analysis

4. BioSci 203 blumberg lecture 3 page 4 © copyright Bruce Blumberg 2001-2005. All rights reserved Determinants of library quality What constitutes a full-length cDNA? –Strictly it is an exact copy of the mRNA –full-length protein coding sequence considered acceptable for most purposes mRNA –full-length, capped mRNAs are critical to making full-length libraries –cytoplasmic mRNAs are best 1st strand synthesis –complete first strand needs to be synthesized –issues about enzymes 2nd strand synthesis –thought to be less important than 1st strand (probably not) choice of vector –plasmids are best for EST sequencing –phages are best for manual screening how will library quality be evaluated –test with 2, 4, 6, 8 kb probes to ensure that these are well represented

5. BioSci 203 blumberg lecture 3 page 5 © copyright Bruce Blumberg 2001-2005. All rights reserved cDNA synthesis Scheme –mRNA is isolated from source of interest –1-2 ug is denatured and annealed to primer containing d(T) n –reverse transcriptase copies mRNA into cDNA –DNA polymerase I and Rnase H convert remaining mRNA into DNA –cDNA is rendered blunt ended –linkers or adapters are added for cloning –cDNA is ligated into a suitable vector –vector is introduced into bacteria Caveats –there is lots of bad information out there much is derived from vendors who want to increase sales of their enzymes or kits –all manufacturers do not make equal quality enzymes –most kits are optimized for speed at the expense of quality –small points can make a big difference in the final outcome

6. BioSci 203 blumberg lecture 3 page 6 © copyright Bruce Blumberg 2001-2005. All rights reserved cDNA synthesis (contd) Preparation of mRNA –want minimum of non poly A+ mRNAs –affinity chromatography oligo d(T) or (U) –Oligo d(T) 30 latex (Nippon Roche) works best overall (a.k.a. OligoTex Qiagen) –2 successive runs gives ~90% pure A+ mRNA denaturation of mRNA –critical step –most protocols use heat denaturation –CH 3 HgOH is method of choice for best libraries First strand synthesis - lots of misinformation about enzymes –reverse transcriptase contains 2 subunits polymerase RNase H - critical for processivity of the enzyme! –Manufacturers prefer to sell MMLV RNase H- RT – cloned and cheap

7. BioSci 203 blumberg lecture 3 page 7 © copyright Bruce Blumberg 2001-2005. All rights reserved cDNA synthesis (contd) Example of comparisons between enzymes and buffers –Mfg supplied buffers NOT optimal –Literature references not optimal either –Enzymes vary a lot AMVSuperscript II sigma R B T both First strand synthesis (contdbest enzyme for 1 st strand synthesis is AMV RT from Seikagaku America But not best overall –thought that 1st strand is main failure point in cDNA synthesis - NOT –great improvement in 1st strand synthesis is addition of 0.6 M trehalose to reaction allows rxns to run at ~60° C

8. BioSci 203 blumberg lecture 3 page 8 © copyright Bruce Blumberg 2001-2005. All rights reserved cDNA synthesis (contd) 2nd strand –must remove mRNA –best way is with RNAse H so that fragments serve as primers for DNA pol I –Gubler and Hoffman (1983) Gene 25, 263 –in my experience, 2nd strand synthesis is the point of failure in cDNA virtually all kits shortcut this step (1-2 hrs) should be overnight recent improvement is to use thermostabile RNAse H, DNA ligase and DNA polymerase to maximize production of 2nd strand.

9. BioSci 203 blumberg lecture 3 page 9 © copyright Bruce Blumberg 2001-2005. All rights reserved cDNA synthesis (contd)

10. BioSci 203 blumberg lecture 3 page 10 © copyright Bruce Blumberg 2001-2005. All rights reserved cDNA synthesis (contd) Cloning –after 2nd strand is made, the ends must be blunted and linkers or adapters added usually T4 DNA polymerase WHY? –perfect cDNAs will retain 2-20 bp of RNA at the 5’ end. Linkers can not be added to this by any DNA ligase! But T4 RNA ligase can ligate DNA-RNA and stimulates blunt end ligation 10x no commercial products use T4 RNA ligase so it is no wonder that full- length cDNAs are lost –if internal restriction sites have not been protected, they need to be methylated now before linkers are added. Most methylase preps are not clean

11. BioSci 203 blumberg lecture 3 page 11 © copyright Bruce Blumberg 2001-2005. All rights reserved Full-length mRNA isolation and cDNA synthesis Ways to capture cap structures and presumably full-length mRNAs –affinity chromatography with eIF-4E (cap binding protein a.k.a. Capture –selection with antibody to cap structure –oligo capping –biotinylated cap trapper 5’ oligo capping - Maruyama, K., and Sugano, S. (1994). Gene 138, 171-4. –uncapped mRNAs are dephosphorylated so that they cannot be ligated –cap structure is removed, only previously capped mRNAs have 5’ PO 4 –RNA ligase can ligate a 5’-OH oligo to the 5’ end of the mRNA –This can be used to prime 2nd strand synthesis

12. BioSci 203 blumberg lecture 3 page 12 © copyright Bruce Blumberg 2001-2005. All rights reserved Full-length mRNA isolation and cDNA synthesis (contd) 5’ oligo capping (contd) –advantages very simple no homopolymeric regions to worry about can put arbitrary sequence at 5’ end. –Enables custom vector construction –also enables PCR to make driver for normalization –disadvantages cap trapper paper claims this method only gives 70% full-length cDNAs high quality TAP is not easy to find original paper used PCR between 5’ and 3’ primer to make cDNAs –PCR => bias!

13. BioSci 203 blumberg lecture 3 page 13 © copyright Bruce Blumberg 2001-2005. All rights reserved Full-length mRNA isolation and cDNA synthesis (contd) Cap trapping Carninci, P. et al. (1996) Genomics 37: 327- 336. –Principle is that a biotin residue is chemically added to the cap structure –approach 1st strand cDNA is synthesized treatment with RNAse I cuts any cDNA:mRNA duplexes which are not absolutely complete complete cDNAs are isolated by streptavidin chromatography RNA is hydrolyzed cDNA is tailed with dG –What are pitfalls of this? 2nd strand synthesis is primed with dC adapter added cloned advantages –claimed to give 90% recovery of full-length cDNAs –lots of history at RIKEN disadvantages –homopolymeric region –many steps -> points of failure

14. BioSci 203 blumberg lecture 3 page 14 © copyright Bruce Blumberg 2001-2005. All rights reserved Full-length mRNA isolation and cDNA synthesis (contd)

15. BioSci 203 blumberg lecture 3 page 15 © copyright Bruce Blumberg 2001-2005. All rights reserved Full-length mRNA isolation and cDNA synthesis (contd) Cloning of cDNAs –most methods require linker or adapter addition followed by restriction digestion –relies on methylation to protect internal sites or use of rare cutters –A new alternative is ExoIII-mediated subcloning no methylation no restriction digestion no ligation no multimerization of vector or inserts 100% oriented

16. BioSci 203 blumberg lecture 3 page 16 © copyright Bruce Blumberg 2001-2005. All rights reserved Vectors for cDNA cloning Plasmids vs phage –phage preferred for high density manual screening –plasmids are better for functional screening microinjection transfection panning –phage packaging and infection more efficient than electroporation 10-100x better than best transformation frequency what will the library be used for ? –Consider the intended use as well as other contemplated uses will the library go to an EST project? –Plasmid will it be screened manually –phage or arrayed and screened on high density filters –plasmid will we normalize it? –Probably plasmid

17. BioSci 203 blumberg lecture 3 page 17 © copyright Bruce Blumberg 2001-2005. All rights reserved Vectors for cDNA cloning (contd) Analysis of cDNAs obtained –rate limiting step in clone analysis is getting them into a usable form usually a plasmid –cloning is tedious, particularly if one has many positives some tricks can be used but this is still the bottleneck in about 1985 or so, Stratagene introduced lambda ZAP –phage with an embedded plasmid and M13 packaging signals –plasmid can be automatically excised by adding a helper phage gene II protein replicates plasmid into ss phagemid which is secreted –this was a major advance and many phage libraries today are made in ZAP or its derivatives –early protocols had problems with helper phage but this has been overcome later, others developed a Cre-lox based system –instead of M13 used loxP sites. –When Cre recombinase is added, recombination between the loxP sites excises a plasmid both methods work very well and make analysis of many clones very straightforward

18. BioSci 203 blumberg lecture 3 page 18 © copyright Bruce Blumberg 2001-2005. All rights reserved Vectors for cDNA cloning (contd)

19. BioSci 203 blumberg lecture 3 page 19 © copyright Bruce Blumberg 2001-2005. All rights reserved Vectors for cDNA cloning (contd)

20. BioSci 203 blumberg lecture 3 page 20 © copyright Bruce Blumberg 2001-2005. All rights reserved mRNA frequency and cloning mRNA frequency classes –classic references Bishop et al., 1974 Nature 250, 199-204 Davidson and Britten, 1979 Science 204, 1052-1059 –abundant 10-15 mRNAs that together represent 10-20% of the total RNA mass > 0.2% –intermediate 1,000-2,000 mRNAs together comprising 40-45% of the total 0.05-0.2% abundance –rare 15,000-20,000 mRNAs comprising 40-45% of the total abundance of each is less than 0.05% of the total some of these might only occur at a few copies per cell How does one go about identifying genes that might only occur at a few copies per cell?

21. BioSci 203 blumberg lecture 3 page 21 © copyright Bruce Blumberg 2001-2005. All rights reserved Normalization and subtraction How to identify genes that might only occur at a few copies per cell? –alter the representation of the cDNAs in a library or probe Normalization - process of reducing the frequency of abundant and increasing the frequency of rare mRNAs –Bonaldo et al., 1996 Genome Research 6, 791-806 –normalization is claimed to bring all cDNAs into the same order of magnitude abundance, i.e., within 10 fold of each other rarely works this well. More commonly, abundant genes are reduced 10 fold and rare ones increased 3-10 fold. Intermediate class genes do not change much at all –Approach make a population of cDNAs single stranded –tester hybridize with a large excess of cDNA or mRNA to C o t =5.5 –driver C o t value is critical for success of normalization –5-10 is optimal –higher values are not better

22. BioSci 203 blumberg lecture 3 page 22 © copyright Bruce Blumberg 2001-2005. All rights reserved Normalization and subtraction (contd) –Approach (contd) various approaches to make driver –use mRNA - may not be easy to get –make ssRNA by transcribing library –make ssDNA by ExoIII treating inserts digested from plasmid library –PCR amplification of library experience has demonstrated that the best approach is to use driver derived from the same library by PCR –rapid, simple and effective –other approaches each have various technical difficulties –see the Bonaldo review for details. –What are normalized libraries good for? EST sequencing gene identification –biggest use is to reduce the number of cDNAs that must be screened –good general purpose target to screen »subtracted libraries are useful but limited in utility