PCR Primer Design

wupfzhongguo

[fontMolecular Biology Today 2001. 2(2): 27-32.

PCR Primer Design
Vinay K. Singh and Anil Kumar
  
Bioinformatics Sub-centre, School of Biotechnology, Devi Ahilya University, Khandwa Road, Indore 452 017 MP, India
  
  
Abstract
  
To make PCR a specific, efficient and cost effective tool for researchers and clinicians the most important aspect is oligonucleotide primer design. This review discusses various aspects of primer design. Advice is provided for optimal design and the role of bioinformatic tools is highlighted. The authors discuss theoretical considerations and compare computational and experimental studies.
  
Introduction
  
Bioinformatics is a newly-emerged inter-disciplinary research area spanning a range of specialties that include molecular biology, biophysics, computer science, mathematics and statistics. It makes use of scientific and technological advances in the areas of computer science, information technology and communication technology to solve complex problems in life sciences, particularly problems in biotechnology. Bioinformatics comprises of the development and application of algorithms for the analysis and interpretation of data, for the design and construction of vital databases, and for the design of experiments.
Bioinformatics is used interchangeably with the terms biocomputing and computational biology. However, biocomputing is more correctly defined as the systematic development and application of computing systems and computational solution techniques to model biological phenomena. Polymerase chain reaction (PCR) is one such phenomenon. PCR is used for the in vitro amplification of DNA at the logarithmic scale. Various components of the PCR reaction such as Taq DNA polymerase, assay buffer, deoxynucleoside triphosphates, stabilizing agents, and primers make it possible for the DNA template to be amplified sufficiently in vitro to attain detectable quantities. PCR can be used for various purposes such as the amplification of human specific DNA sequences, differentiation of species, sub-species and strains, DNA sequencing, detection of mutations, monitoring cancer therapy, detection of bacterial and viral infections, pre-determination of sex, linkage analysis using single sperm cells, ascertaining recombinant clones and studying molecular evolution. PCR is a sensitive technique and therefore highly susceptible to contamination which may result in false positivity. To make PCR a specific, efficient and cost effective tool for researchers and clinicians the most important component of the PCR is the oligonucleotide primers. Literature searches indicate that insufficient experimental work has been done in the field of bioinformatics especially in the field of nucleic acid sequence analyses. Inadequate experimental data is available (at least in the public domain) for the establishment of primer design strategies. In this review the authors aim to establish various aspects and types of PCR and primer design theory, supported by computational and experimental data.
  
PCR Primer Design
  
Selective amplification of nucleic acid molecules, that are initially present in minute quantities, provides a powerful tool for analyzing nucleic acids (Saiki et al., 1985; Mullis et al., 1987). The polymerase chain reaction is an enzymatic reaction, which follows relatively simple, predictable and well understood mathematical principles. However the scientist often relies on intuition to optimise the reaction. To make PCR an efficient and cost effective tool, some components of PCR such as Taq DNA polymerase, assay buffer, deoxynucleoside triphosphates (dNTPs), stabilizing agents (Sarkar et al., 1990), DNA Template and oligonucleotide primers must be considered in greater detail (Linz et al., 1990). Efficacy and sensitivity of PCR largely depend on the efficiency of primers (He et al., 1994). The ability for an oligonucleotide to serve as a primer for PCR is dependent on several factors including: a) the kinetics of association and dissociation of primer-template duplexes at the annealing and extension temperatures; b) duplex stability of mismatched nucleotides and their location; and c) the efficiency with which the polymerase can recognize and extend a mismatched duplex. The primers which are unique for the target sequence to be amplified should fulfill certain criteria such as primer length, GC%, annealing and melting temperature, 5' end stability, 3' end specificity etc (Dieffenbach et al., 1993).
DNA template quality or purity is not particularly significant for amplification. Provided DNA does not contain any inhibitor of Taq DNA polymerase, it can be isolated by almost any method (Murray and Thompson, 1980; Sambrook et al., 1989; Kaneko et al., 1989; Mercier et al., 1990; Kawasaki 1990a; Green et al., 1991; Keller et al., 1993; Klebe et al., 1996; Singh and Naik, 2000).
Taq DNA polymerase also plays an important role (Drummond and Gelfand, 1989). Taq DNA polymerase from different suppliers may behave differently because of the different formulations, assay conditions and/or unit definitions. Recommended concentration ranges between 1-2.5 Units/50-100 ml reaction (Lawyer et al.,1989) when other parameters are optimal.
Most of the reviews on PCR optimization (Erlich et al., 1991; Dieffenbach 1993; Roux 1995) consider different parameters of PCR but generally do not discuss basic concepts of PCR primer design. Because of the requirements for different strategies of PCR, more effective PCR studies would be attainable by considering the basic concepts of PCR primer design.
  
Primer Length: a Hard Core Factor
  
Length of a primer is a critical parameter (Wu et al., 1991). The rule-of-thumb is to use a primer with a minimal length that ensures a denaturation temperature of 55-56°C. This greatly enhances specificity and efficiency. For general studies, primers of typically 17-34 nucleotides in length are the best. Primer >16 nucleotides in length are not generally annealed specifically to non-target DNA sequence (e.g. human DNA in an assay for bacterial infection). For example, a short primer sequence, such as a 12 bp oligonucleotide, binds to 200 specific annealing sites in the human genome. (The genome consists of 3x109 nucleotides: 3 x 109/412=200). In contrast, a 20 mer sequence is expected to randomly exist only once every 420 nucleotides and as such, has only a 1 in 400 probability of existing by chance in the human genome. Primers, 18-24 mer are accepted as best in being sequence specific if the annealing temperature of the PCR reactions is set within 5°C of the primer Td (dissociation temperature of the primer/template duplex) (Dieffenbach, 1993). Primers work exceptionally well for the sequence with least intra-strand secondary structure. This is because secondary structure impedes primer annealing and extension. Longer primers (28-35 mer) are required only to discriminate homologous genes within different species or when a perfect complementary sequence to all the template is not expected. They could also be used when extra sequence information e.g. a motif binding site, restriction endonuclease site or GC clamp is attached to 5' end. Such extensions do not generally alter annealing to the sequence specific portion of the primer (Sheffield et al., 1989).
Although the following formula is generally used for determining melting temperature (Tm):
Tm = 4 (G+C) + 2(A+T)
  
Frier et al. (1986) showed that the nearest-neighbor calculation is better for calculating the melting temperature of longer primers because this also takes account of thermodynamic parameters. Using improved nearest-neighbor thermodynamic values given by John SantaLucia (1995), a good estimate of melting temperature can be obtained for oligonucleotide analysis.
  
Terminal Nucleotides Make a Difference
  
Both the terminals of the primer are of vital importance for a successful amplification. The 3'-end position in the primer affects mispriming. However, for certain reactions, such as amplification refractory mutation system (ARMS), this mispriming is required (Newton et al., 1989; Old, 1991; Tan et al., 1994). Runs (3 or more) of C's or G's at the 3' end of the primer should be avoided as G + C rich sequence leads to mispriming. Complementarity at the 3' end of the primer elevates mispriming as this promotes the formation of a primer dimer artifact and reduces the yield of the desired product (Huang et al., 1992). The stability of the primer is determined by its false priming efficiency; ideally it should have a stable 5' end and an unstable 3' end. If the primer has a stable 3' end, it will bind to a site which is complementary to the sequence rather than the target site and may lead to secondary bands. It is adequate to have G or C in last 3 bases at 5' termini for the efficient binding of the primer to the target site. This GC clamp reduces spurious secondary bands (Sheffield et al., 1989).
  
GC Content, Tm and Ta are Interrelated
  
GC content, melting temperature and annealing temperature are strictly dependent on one another (Rychlik et al., 1990). GC% is an important characteristic of DNA and provides information about the strength of annealing. A GC of 50-60% is recommended. The value recommended by Dieffenbach (1993) is 45-55%.
  
Secondary Structure
  
An important factor to consider when designing a primer is the presence of secondary structures. This greatly reduces the number of primer molecules available for bonding in the reaction. The presence of hairpin loops reduces the efficiency by limiting the ability to bind to the target site (Singh et al., 2000). It is well established that under a given set of conditions, the relative stability of a DNA duplex structure depends on its nucleotide sequence (Cantor and Schimmel, 1980). More specifically, the stability of a DNA duplex appears to depend primarily on the identity of the nearest-neighbor nucleotides. The overall stability and the melting behavior of any DNA duplex structure can be predicted from its primary sequence if the relative stability (DG0) and the temperature dependent behavior (DH0, DCp0) of each DNA's nearest-neighbor interaction is known (Marky and Breslauer, 1982). Tinoco et al., (1971, 1973) and Uhlenbeck et al., (1973) have predicted stability and melting behavior of RNA molecules for which they and others have determined the appropriate thermodynamic data. But, to the best of our knowledge, no experimental data is available to support the prediction of the thermodynamic properties of hairpin structures, an important factor to consider when designing a primer. Single stranded nucleic acid sequences may have secondary structures due to the presence of complementary sequences within the primer length e.g. hairpin loops and primer-dimer structures. We have recently shown experimentally that hairpin loops, if present, can greatly reduce the efficiency of the reaction by limiting primer availability and the ability to bind to the target site (Singh et al., 2000). The effect of primer-template mismatches on the PCR has been studied earlier in a Human Immunodeficiency Virus (HIV) model (Kwok et al., 1990). Studies have also been performed for the characterization of hairpins (Marky et al., 1983, 1985), cruciforms (Marky et al., 1985), bulge and interior loops (Patel et al., 1982 , 1983).
  
Dimers and False Priming Cause Misleading Results
  
Annealing between the 3' end of one primer molecule and the 5' end of another primer molecule and subsequent extension results in a sharp background product known as primer dimer. Its subsequent amplification product can compete with the amplification of the larger target. If the primer binds anywhere else than the target site, the amplification specificity is reduced significantly (Breslauer et al., 1986). This leads to a weak output or a smear. This occurs again when some bases at 3' end of the primer bind to target sequence and achieve favorable chances of extension (Chou et al., 1992).
To minimize the possibility of dimers and false priming, PCR is generally performed at high temperature (>50°C), but primers may be extended non-specifically prior to thermal cycling if the sample is completely mixed at room temperature (RT) (Hung et al., 1990). To prevent this occurring the Hot Start