1. Introduction
2. Factors to check during optimization
3. Factors to optimize
4. The normal chemistry of the PCR with regard to specificity control
5. Special substances for control of specificity in PCR
6. Blocking reagents added to the PCR mix
   1. BSA
      1. Milky white PCR liquid in DIAPOPS after thermal cycling
   2. Blocking Reagent (BR)
7. Conclusion

1. Introduction
When a new PCR ( ) is developed, the first action is to select suitable primers ( ). This requires a suitable target sequence ( ). The selection of PCR primers is in most cases made with help of a computer program. These programs are designed to indicate conditions concerning salt concentration and annealing temperature, at which the PCR will perform well. However, in most cases these suggestions are not optimal, and further optimization is necessary.

2. Factors to check during optimization
There are three major factors to check during the optimization of the PCR:

• How many distinct DNA sequences, besides the correct one, can generate a positive signal (specificity). These are called false positives.
• How many copies of the DNA target sequence are necessary to generate a positive signal (limit of detection or sometimes called sensitivity).

If the target sequence is present in different subtypes of the organism being analyzed:

• How many strains or types of the organisms, which should be positive, are positive (sensitivity or selectivity).

In this text, the term "specificity" is used to indicate the likelihood of a PCR amplifying a sequence homologous to the true target sequence. If the specificity is high, only the correct sequence will be amplified. If the specificity is low, sequences not in total homology with the target sequence may also be amplified.

3. Factors to optimize
In addition to the selection of primers ( ) optimal amplification depends on several factors including temperature profile, and the concentration of reagents in the buffer. The most straightforward way of optimizing a PCR with a given primer pair is to change the concentration of MgCl2 or the annealing temperature. The PCR program is a very important factor to optimize. This is mostly done by changing the annealing temperature. However, the number of cycles, the period of time ( ), and the temperatures in the PCR program are also very important.

4. The normal chemistry of the PCR with regard to specificity control
In most reports, the concentrations of the single compounds in the PCR buffer mix are basically the same as described in the first published report of amplification using the thermostable Taq-polymerase (Saiki et al., 1988). Briefly: 50 mM KCl, 10 mM Tris [pH 8.4], 2,5 mM MgCl2, 1 µM of each primer, 200 µM of each mononucleotide, 200 µg/ml gelatin and 2 units/100 µl of Taq-polymerase. Subsequently, the gelatin was shown to have no effect (Blanchard et al. 1993), but is nevertheless still used routinely in many laboratories. Modest concentrations of KCl stimulate the synthesis rate of Taq-polymerase with an apparent optimum at 50 mM. Higher KCl concentrations are increasingly inhibitory (Gelfand, 1989). The Mg²⁺-ion binds tightly to the phosphate sugar backbone of nucleotides and nucleic acids, and variation in the MgCl2 concentration has strong and complex effects on experiments involving nucleic acid interactions. Variations of the Mg²⁺ concentration below 4 mM, can improve the performance of PCR by affecting the specificity (lower concentrations raise specificity, higher concentrations lower the specificity) (Blanchard et al. 1993). In general, each single assay must be optimized with regard to the concentration of Mg²⁺. The effect of variations in the dNTP concentration is closely correlated to the Mg²⁺ concentration, due to the interaction between mononucleotides and the Mg²⁺-ion. A higher concentration of Mg²⁺ allows amplification with a higher concentration of dNTP's, which is not seen at lower Mg²⁺ concentrations (Blanchard et al. 1993). Taq-polymerase in concentrations of 2-2.5 units/100 µl reaction is normally used. Concentrations higher than 4 units/100 µl can generate non-specific products and may reduce the yield of the desired product (Saiki, 1989). Increasing the concentration of TRIS in the buffer is reported to decrease the specificity (Blanchard et al. 1993), and therefore, the TRIS concentration can also be used to optimize the PCR (Rasmussen et al. 1996).

5. Special substances for control of specificity in PCR
Many substances not necessary for the basic PCR DNA amplification may influence the reaction either by enhancing the specificity by inhibiting non-homologous hybridization, or by increasing the product yield by preventing inhibition of the polymerase. Table 1 shows a list of some chemicals or conditions which can be used to improve or affect the performance of the PCR.

Table 1: Factors and compounds affecting the PCR and the way in which they influence the reaction.

6. Blocking reagents added to the PCR mix

6. a) BSA

Figure 1: Showing a photo of two NucleoLink Strips containing the liquid in the wells after thermal cycling. In the Strip on the left, 10 mg/ml of BSA was added to the PCR mix before cycling, and as seen, the liquid is opaque. In the Strip on the right only 1 mg/ml BSA was added, and the liquid is clear after thermal cycling. The DIAPOPS signals from these two NucleoLink Strips can be seen in Figure 3 ( ), and the agarose gel analysis can be seen in Figure 2 ( ).

In the DIAPOPS procedure the coated wells are blocked with BSA before addition of the PCR mix. This is carried out to prevent the polymerase enzyme from attaching to the wall and thereby being inactivated. BSA can also be added to the PCR mix in a concentration of 1 mg/ml without affecting the reaction. If the concentration of BSA is raised to 10 mg/ml in the PCR, the liquid turns milky white after the thermal cycling. This is illustrated in Figure 1.

In spite of the milky white appearance after thermal cycling of the PCR mix containing 10 mg/ml BSA seen in Figure 1, the liquid phase DNA amplification is not totally inhibited. This is illustrated in Figure 2. This Figure shows an analysis on agarose gels of the liquid phase PCR products shown in Figure 1. With 10 mg/ml BSA the DNA amplification, shown on the right agarose gel, is not as efficient as when the concentration of BSA in the PCR-mix is only 1 mg/ml, which is shown on the left agarose gel in Figure 2. It is obvious that the concentration of the liquid phase PCR products is lower with the high BSA concentration, and the limit of detection has been decreased compared to the left hand agarose gel in Figure 2.

Figure 2

Figure 2: Showing liquid phase PCR product analyzed on an agarose gel. The products are from a DIAPOPS analysis of a dilution series of BLV plasmid DNA ( ). A concentration of 1 mg/ml BSA was added to the PCR mix analyzed in the gel on the left, and a concentration of 10 mg/ml to the PCR mix is analyzed in the gel on the right. The PCR mix analyzed in the gel on the right was milky white after the thermal cycling as illustrated in Figure 1 ( ). Even though the liquid phase amplification is functioning, the solid phase DNA amplification is not working, which is illustrated in Figure 3 ( ).

A different experiment with identical dilution series performed on the same day is presented elsewhere ( ). In this experiment, BSA was not added to the PCR mix before amplification, but one of the coated NucleoLink Strips was blocked with BSA before addition of the PCR mix as described in the procedure ( ), while the other was used directly without BSA blocking. The agarose gels from the analysis of the liquid phase PCR product clearly indicates that the DNA amplification in the Strips blocked with BSA are more reproducible compared to the Strips without BSA ( ).

Figure 3: Showing the DIAPOPS analysis of the solid phase products in the NucleoLink Strips where the liquid phase DNA amplification illustrated in Figure 2 was performed. The reactions containing 1 mg/ml BSA have created a high concentration of solid phase products, while almost no solid phase product can be detected in the Strips of the reactions containing 10 mg/ml BSA, even though both series of reaction showed high concentrations of liquid phase PCR products.

6 a i) Milky white PCR liquid in DIAPOPS after thermal cycling
The milky white PCR liquid illustrated in Figure 1 has also been observed when using the normal BSA blocking procedure before addition of the PCR mix ( ). This is most likely because the blocking liquid has not been entirely removed causing the subsequent concentration of BSA to be too high. It is very important to empty the wells of the NucleoLink Strips entirely, using the method described elsewhere ( ). When the NucleoLink Strips have not been entirely emptied after blocking, and the PCR liquid is milky after thermal cycling, it is expected that a liquid phase DNA amplification has been successful as illustrated in Figure 2 right. However, no solid phase DNA amplification is expected and there will be no, or only very low, DIAPOPS signals, as observed in Figure 3.

6 b) Blocking Reagent (BR)
Another blocking reagent, skimmed milk powder, purchased from Boehringer Mannheim as Blocking Reagent (BR) ( ), was also examined by addition to the PCR. Addition of this protein in a concentration of 0.5% totally obstructed the reaction. For this reason, BR is only used after the DNA amplification ( ).

7. Conclusion
During the first experiments with a new PCR system, an optimization is necessary in most cases. This optimization is normally performed by changing the annealing temperature or the concentration of MgCl2. Other substances for controlling the specificity can also be added to the PCR. A number of these are presented in table 1 ( ).

It is possible to add BSA to the PCR mix at a concentration of 1 mg/ml, without losing any efficiency of the DNA amplification. If the concentration of BSA is 10 mg/ml, the PCR liquid turns milky white after the thermal cycling. The liquid phase DNA amplification is not totally inhibited, but the solid phase DNA amplification will be absent, and if the PCR liquid is milky white after the thermal cycling, no DIAPOPS signal will be observed.