Wednesday, April 30, 2003

Strategies For Microarray Analysis of Limiting Amounts of RNA - Brief Funct Genomic Proteomic.

Summary:
This review evaluates current signal and sample amplification technologies, including those that can be used to generate labelled cDNA populations for array analysis from as little as a single cell. Options for expression profiling are to increase cDNA labelling and hybridisation efficiency, or to use an amplification strategy to generate enough RNA/cDNA for use with a standard labelling method. Sample amplification approaches must preserve the representation of the relative abundances of the different RNAs within the starting population and must also be highly reproducible. (Briefings in Functional Genomics and Proteomics, Vol 2, No 1, 31-36, April 2003)


Notes:

Introduction

  • microarrays
    • have become a standard technology for measuring relative and absolute levels of gene expression
    • interest in increasing resolving power of this technology has grown
      • particularly in terms of input material required to generate robust data
    • drive for this
      • partly technical
      • partly motivated by biological and clinical concerns
        • main goal is to use defined populations of cells or small pieces of complex tissue (e.g. clinical biopsy) for expression profiling
      • associated with a reduction in the amount of cells that can be harvested
        • e.g. laser capture microdissection: possible to collect defined cells from fesh and fixed tissue sections
  • ultimate aim of this increase in resolution:
    • to enable reproducible expression profiling at the level of single cells
      • several reports that this is currently feasible
      • likely to be in general use in the near future
  • current methods
    • require microgram amounts of total RNA for generating labelled cDNA populations for microarray analysis
    • equivalent of over 1 million cells
  • efforts to reduce this requirement focus on two complementary approaches:
    • signal amplification and detection
      • allowing the use of smaller amounts of input RNA
    • RNA sample amplification
      • to generate enough material for standard labelled cDNA synthesis, hybridisation and detection
    • combination of both
Signal versus sample amplification: theory
  • ideally:
    • extract the RNA from a single cell, directly label that RNA and hybridise it to some form of microarray
  • many practical issues
    • from: difficulty of harvesting picogram quantities of RNA contained in a typical cell
    • to: hybridisation kinetics for very small numbers of molecules at relatively low concentrations
  • mRNA abundance: three classes (tissue based estimates from brain cDNA libraries)
    1. high abundance transcripts
      • ~1/6th of the mass of mRNA
      • represents 100 different transcripts
    2. medium abundance
      • ~45% of mass of mRNA
      • 2,000 different transcripts
    3. low abundance
      • ~40% of mass of mRNA
      • 45,000 different transcripts
  • inherent technical challenge in labelling all of these low abundance transcripts for microarray hybridisation under any circumstances and developing hybridisation conditions that would ensure that all molecules hybridise in a reasonable timeframe
  • under conditions where the input RNA and the corresponding absolute numbers of each low abundance transcript are low, these problems become more accute, with less room for errors in each step of the generation of labelled cDNA population
  • a final technical hurdle is the detection of the extremely small numbers of molecules harvested from single cells
Signal amplification
  • currently: two main methods for generating labelled cDNA populations for array analysis
    1. direct incorporation of fluorescent label-conjugated nucleotides
    2. incorporation of modified nucleotides followed by dye coupling to those modified nucleotides
      • amino-allyl labeling method
      • introduced for
        • relative cost reasons
        • reduce the biases in incorporation rates of different fluorophore-labelled nucleotides
    • in widespread use and commercial kits available
  • novel strategies
    • labelling cDNA populations as well as amplifying that label such that smaller numbers of hybridised molecules can be reproducibly detected and quantified
    • e.g.
      • enzymatic amplification
        • e.g. tyramide signal amplification
      • use of dendrimers
        • increases amount of label per nucleotide and thus per labelled cDNA molecule
        • several hundred fluorescent tags per dendrimer
        • input amounts of RNA down to 0.5ug
          • still considerable amount
        • hybridisation takes far longer than with standard methods
          • due to size of molecules
          • typically of order of several days
      • alternative detection methods, e.g.
        • quantum dots
        • rolling circle amplification
Sample amplification
  • amplification of the input RNA to generate enough material for standard labelled cDNA synthesis
    • alternative to signal amplification
  • currently: two approaches
    1. PCR-based or exponential amplification
    2. linear amplification
  • linear amplification
    • first described by Eberwine et al. as a method for single cell analysis; now common method
    • antisense RNA synthesis from a population of double-stranded cDNA molecules, all carrying a standard recognition site for T7 RNA polymerase
    • used in Affymetrix system
    • curretnly, amplification of nanogram quantities of total RNA (equivalent of 50 - 1000 cells) requires two rounds of T7 linear amplification
    • feature: shortening of the amplified transcripts, compared to their parent mRNA population, with the associated 3'-bias in the amplified material
    • disadvantages:
      • labour intensive
        • requires synthesis and purification of double-stranded cDNA from the starting RNA, followed by at least one round of RNA synthesis and amplification
        • this RNA is in turn used to synthesise double-stranded cDNA, followed by a second round of RNA synthesis
        • typical time taken to generate amplified RNA from picogram quantities of input total RNA is of the order of 3-5 days
  • PCR-based amplification
    • general principle: introduction of PCR-priming sites at either end of each reverse-transcribed cDNA molecule, followed by global amplification of the entire population of molecules
    • potential pitfalls (sources of sampling, non-representative amplification):
      1. during each step
        • failure to introduce priming sites to the ends of every RNA/cDNA molecule in the starting population will introduce sampling into the amplification process with under-representation and possible amplification of those molecules
      2. during the oligo-dT primed reverse transcription steps
      3. during the PCR itself
        • when the exponential nature of the process amplifies any variations in the amplification efficiency of particular templates
        • most significant source of error during PCR based amplification
    • Clontech's SMART system
      • has been succesfully used for generating labelled cDNA for array analysis from limiting amounts of RNA
      • has been shown to preserve the relative abundance of RNA molecules in the amplified population
    • advantages of PCR
      • rapid (exponential) amplification of cDNA population: less than 1 day
      • short, relatively simple protocols
      • particularly useful in medium- and high-throughput situations where many smaples are to be studied
  • amplified material can be labelled to generate labelled cDNA populations for array analysis in a number of different ways
    • RNA generated by linear amplification can be labelled using standard direct and indirect labelling methods, or with signal amplification methods
    • amplified cDNA can be labelled by random primer-mediated incorporation of either directly or indirectly labelled nucleotides
Pushing the system: the challenge of single cell expression profiling
  • even with current labelling technologies, generating enough cDNA from a single cell for a single microarray hybridisation requires around 10^6-fold amplification of th emRNA content of that cell
    • total degree of amplification depends on the cell type used, given the wide range of total RNA content in different cell types
      • from as little as 1pg to as much as 50pg
      • only 1-5% of this mass of RNA is composed of mRNA
        • containing an estimated total of 100,000 - 300,000 molecules of mRNA
    • amplifying 300,000 molecules of different abundances to generate this mass of material represents a considerable challenge
  • the particular acute problems for amplifying single-cell material are
    • the efficiency of priming the intial RT and
    • the efficiency of the subsequent steps to prepare the cDNA for amplification
      • be they the introduction of a second priming site for PCR amplification or production of dsDNA from the single stranded material
    • failure of either step for a sub-population of the cDNA will result in the absence of detection of low abundance transcripts
  • assuming that all amplification methods introduce some degree of error over the million-fold amplification procedure, it is likely that amplification from single cells is an inherently noisy procedure

Expand notes