Use of Paraffin Wax Film in MALDI-TOF Analysis of DNA - American

that a probe tip with a paraffin wax film (Parafilm) surface improves the MALDI performance in DNA analysis com- pared to the commonly used metal surf...
0 downloads 0 Views 172KB Size
Download PDF
Anal. Chem. 1998, 70, 3088-3093

Use of Paraffin Wax Film in MALDI-TOF Analysis of DNA K. C. Hung, H. Rashidzadeh, Y. Wang, and Baochuan Guo*

Department of Chemistry, Cleveland State University, Cleveland, Ohio 44115

Poor detection limits and strong salt effects are two of the main problems encountered in the matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometric analysis of DNA. This work demonstrates that a probe tip with a paraffin wax film (Parafilm) surface improves the MALDI performance in DNA analysis compared to the commonly used metal surface. First, the use of Parafilm increases the detection sensitivity. It was found that the detection limit achieved with Parafilm was 5 times better than that obtained using stainless steel for a 85mer. More importantly, the Parafilm method could improve detection of larger DNA components in the presence of a large excess of a smaller DNA component or in a DNA mixture. This feature is important to analyses of PCR and sequencing products. Second, we found that the use of Parafilm increased the salt tolerance limits for the 17-, 41-, and 85mers studied in this work and that the salt effect was less sensitive to the DNA size. Third, this method offers other analytical benefits, including producing a more homogeneous coverage of matrix/DNA, adding no extra cost and time to sample preparation, and eliminating the commonly required step for cleaning the probe after analysis. In this paper, we will also present our perspectives on why the use of Parafilm can improve the MALDI-TOF performance in DNA analysis.

Matrix-assisted laser desorption/ionization time-of-flight (MALDITOF) mass spectrometry is a potential method for low-cost, highaccuracy DNA analysis. During the past several years, progress has been made in the MALDI-TOF analysis of DNA. Becker and co-workers discovered a 3-hydroxypicolinic acid (3-HPA) matrix,1 which was used for the detection of DNA containing 89 bases in subsequent work.2 Chen’s group demonstrated the possibility of detecting DNA consisting of over 500 bases using a mixture of picolinic acid and 3-HPA acid as the matrix.3 More recently, Becker and co-workers reported successful detection of DNA of over 1000 bases.4 The sequencing feasibility of MALDI-TOF was first demonstrated by Smith and co-workers, who analyzed (1) Wu, K. J.; Steding, A.; Becker, C. H. Rapid Commun. Mass Spectrom. 1993, 7, 142. (2) Wu, K. J.; Shaler, T. A.; Becker, C. H. Anal. Chem. 1994, 66, 1637. (3) Tang, K.; Taranenko, N. I.; Allman, S. L.; Chang, Y. L.; Chen, C. H. Rapid Commun. Mass Spectrom. 1994, 8, 727. (4) Hunter, J.; Hua, L.; Becker, C. H. Presented at the 45th ASMS Conference on Mass Spectrometry and Allied Topics, Palm Springs, CA, June 1-5, 1997.

3088 Analytical Chemistry, Vol. 70, No. 14, July 15, 1998

mixtures that mimicked DNA sequencing ladders.5 Shaler and co-workers first reported sequencing a 45mer DNA with an enzymatic preparation of DNA ladders.6 Roskey et al. reported sequencing a synthetic DNA template of 50 bases by using a delayed ion extraction technique.7 They also showed that delayed extraction offered a better mass resolution and detection limit when small DNA molecules were analyzed. Smith and co-workers reported sequencing reactions of bacteriophage M13 and succeeded in sequencing DNA up to 35 bases.8 Recently, Chen and co-workers reported complete sequencing a DNA with 50 bases by using both the conventional Sanger and cycle sequencing methods for DNA ladder preparation.9 However, the currently used MALDI-TOF method could not sequence a large DNA (over 100 bases) containing all four bases. Two of the main reasons for this inability are strong salt effects and poor detection limits. The salt effects are due mainly to the presence of a variety of small molecules introduced into the PCR or sequencing reactions. The presence of these molecules leads to poor mass resolution and detection limits. Becker and co-workers investigated the effect of many molecules present in the sequencing products on the MALDI-TOF analysis of DNA and found that metal salts have the strongest effect on this analysis.10 Protocols for Sanger sequencing generally produce an average of a few femtomoles of each termination product. Although a detection limit of a few femtomoles of DNA was demonstrated for MALDI-TOF,6 a low detection limit could be achieved only for highly pure, small DNA molecules.11,12 It has been shown that an important characteristic of MALDI is that detection limits become poorer with increasing mass and that detection limits are typically in the high femtomole or low picomole ranges for large DNA.7 Another notable feature of MALDI is that much weaker intensities are obtained for these larger DNA components when (5) Fitzgerald, M.; Parr, G.; Smith, L. M. Rapid Commun. Mass Spectrom. 1993, 7, 895. (6) Shaler, T. A.; Tan, Y.; Wickham, J.; Wu, K. J.; Christopher, C. H. Rapid Commun. Mass Spectrom. 1995, 9, 942. (7) Roskey, M. Y.; Juhasz, P.; Smirnov, I. P.; Takach, E. J.; Martin, S. A.; Haff, L. A. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 4724. (8) Mouradian, S.; Rank, D.; Smith, L. M. Rapid Commun. Mass Spectrom. 1996, 10, 1475. (9) Taranenko, N. I.; Chung, C. N.; Zhu, F.; Allman, S. L.; Golovle, V. V.; Isola, N. R.; Martin, S. A.; Haff, L. A.; Chen, C. H. Rapid Commun. Mass Spectrom. 1997, 11, 386. (10) Shaler, T.; Wickham, J.; Sannes, K.; Wu, K.; Becker, C. Anal. Chem. 1996, 68, 576. (11) Zhu, Y. F.; Chung, C. N.; Taranenko, N. I.; Allman, S. L.; Martin, S. A.; Haff, L.; Chen, C. H. Rapid Commun. Mass Spectrom. 1996, 10, 383. (12) Zhu, Y. F.; Taranenko, N. I.; Allman, S. L.; Martin, S. A.; Haff, L.; Chen, C. H. Rapid Commun. Mass Spectrom. 1996, 10, 1591. S0003-2700(98)00090-0 CCC: $15.00

© 1998 American Chemical Society Published on Web 05/28/1998

they are analyzed in a DNA mixture sample than when individual DNA molecules are analyzed.5 The conventional MALDI sample preparation method uses a metal sample surface. Recently, several groups have used membrane materials to replace metals for protein analysis.13-16 The aim of these works was to explore the possibility of interfacing MALDI-TOF to SDS-PAGE separation. More recently, several groups have also used dielectric surfaces to remove salts on the sample probe.17,18 During the course of a recent study of MALDI sample preparation, we discovered that the use of paraffin wax film (Parafilm), a commonly used laboratory material, to replace the metal probe surface could considerably improve the performance of MALDI-TOF in DNA analysis. Reporting this new sample preparation method and its applications to DNA analysis constitute the main focus of this article. EXPERIMENTAL SECTION MADLI-TOF Instrument. The experiments were performed using a homemade linear MALDI-TOF instrument that was operated on the continuous extraction mode. A detailed description of this instrument has been given previously.19 Briefly, the sample probe tip was inserted through a vacuum lock and into the acceleration region such that the front end of the metal surface was flush with the first acceleration plate. A pulsed Nd:YAG laser producing a wavelength of 355 nm was used for MALDI. The two-stage acceleration voltages were set at 30 and 15 kV, respectively. A pulsed electric field of 1500 V/cm was applied to the deflection plates to remove low-mass ions. An electron multiplier was used to detect ions. Thereafter, the ion signal was recorded using a Tektronix 520 digital oscilloscope and subsequently transferred to a 486 PC computer for processing. All mass spectra were produced in the positive ion mode and averaged by more than 10 shots. A Grams/32 (Galactic, Salem, NH) program was used to analyze the recorded data. Materials. Synthetic DNA samples of 17-, 41-, 62-, and 85mers were obtained from National Biosciences, Inc. (Plymouth, MN). The matrix 3-HPA, diammonium citrate, and the metal salts of NaCl and MgCl2 were obtained from Aldrich (Milwaukee, WI). All samples were used without further purification. Parafilm was obtained from American National Can Co. Saturated solutions of the matrix 3-HPA (∼0.5 M) were prepared in a 50 mM diammonium citrate solution containing 15% (by volume) acetonitrile. Thereafter, this solution was diluted by a factor of 2 for sample preparation. Stock DNA solutions were prepared in water. The concentration of the DNA solution was determined through UV absorption measurements at 260 nm. The DNA mixture solutions were prepared prior to loading them on the sample tip. For the metal salt effect studies, the solution of a particular salt in water was prepared first and then mixed with DNA solutions at a specified concentration ratio. Sample Preparation. The procedure to prepare a sample tip with Parafilm is rather simple. This material is highly flexible (13) Vestling, M. M.; Fenselau, C. Anal. Chem. 1994, 66, 471. (14) Strupat, K.; Karas, M.; Hillenkamp, F.; Eckerskorn, C.; Lottspeich, F. Anal. Chem. 1994, 66, 464. (15) Loo, R. R. O.; Stevenson, T. I.; Mitchell, C.; Loo, J. A.; Andrews, P. C. Anal. Chem. 1996, 68, 1910. (16) Blackledge, A.; Alexander, A. J. Anal. Chem. 1995, 67, 843. (17) Worrall, T. A.; Cotter, R. J.; Woods, A. S. Anal. Chem. 1998, 70, 750. (18) Brockman, A. H.; Dodd, B. S.; Orlando, R. Anal. Chem. 1997, 69, 4716. (19) Chen, H.; Guo, B. C. Anal. Chem. 1997, 69, 4399.

and can be stretched into thin films. First, we cut many small pieces of Parafilm from a large roll and stretched one of them into a thin film. Next, the thin film was stretched further on a metal probe tip and attached to the tip without using any adhesive materials. 3-HPA/DNA film preparation is different from the commonly used method, in which a mixture of DNA and 3-HPA was loaded on the probe tip. In this work, 0.4-1.0 µL of the 3-HPA solution was first loaded onto Parafilm and allowed to dry in the air at room temperature. The DNA solution was added then to the dried 3-HPA film and air-dried before analysis. RESULTS AND DISCUSSION General Observation. Parafilm is a blend of olefins and thus is hydrophobic in nature. When the aqueous 3-HPA solution was spotted on the Parafilm surface, the solution tended to form a droplet instead of quickly spreading over a large area, as is the case on the metal surface. It should be noted that the size of the droplet depends also on the amount of acetonitrile present in the matrix solution, since acetonitrile tends to spread the droplet. The droplet shrinks as the solvent evaporates. Thereafter, the matrix 3-HPA participates and forms a small, thick film. Under the conditions used in this work, we found that the matrix film areas formed on Parafilm were about 5-10 times smaller than those formed on a metal surface using the same amount of the 3-HPA solution. When a small amount of the DNA solution was spotted on the 3-HPA film, the DNA solution quickly covered the entire surface of the film and did not spread to the Parafilm surface. Hence, the DNA sample size on the sample tip was actually dictated by the matrix film area. It should be noted that the crystal formation rate on Parafilm was significantly slower than that occurring on the metal surface because a smaller droplet was formed. Figure 1 shows the MALDI-TOF mass spectrum of a DNA 17mer obtained by loading the matrix solution onto Parafilm, followed by adding the 17mer solution to the dried matrix film. The total amount of DNA loaded was about 1 pmol. The typical resolution of the 17mer obtained on Parafilm was 40, similar to that obtained on a metal surface. The better resolution spectra obtained using DE-MALDI-TOF revealed that the poor resolution observed was due to metal attachment.20 We found that the desorption threshold with Parafilm was similar to that observed with a metal surface, but, as will be shown later, the use of Parafilm improves detection sensitivity. Compared to a metal surface probe, a Parafilm surface probe greatly improved the shot-to-shot and sample-to-sample reproducibility of the DNA ion signal, especially when small amounts of DNA were loaded. It was easy to locate sweet spots, even without the assistance of a video camera. These findings suggest that the use of Parafilm produces a more homogeneous coverage of matrix/DNA over the sample surface. The Parafilm method offers other analytical benefits in terms of cost, time, and ease of preparing the MALDI sample. First, this method adds virtually no extra cost to DNA analysis, since Parafilm is inexpensive. Second, it is easy and takes only a few seconds to make a good Parafilm film on a metal probe. Therefore, this method adds virtually no extra time to the MALDI sample preparation. Third, (20) This study was performed in PerSpective Biosystems, Inc. using VoyagerDE RP System.

Analytical Chemistry, Vol. 70, No. 14, July 15, 1998

3089

Figure 1. MALDI-TOF mass spectrum of a 17mer obtained using Parafilm. The total amount of the 17mer loaded on the probe was about 1 pmol. The unlabeled peak corresponds to the doubly charged 17mer.

although thin films gave the best results, the exact thickness of Parafilm was not critical. Fourth, because of its inert nature, Parafilm does not react with the matrix or DNA solutions. Fifth, the commonly required step of cleaning the probe tip after each analysis is eliminated, since it is easy to replace the used Parafilm. Detection Limit. Detection limits are important to the MALDI-TOF analysis of DNA. The first issue explored was the detection limit in a sample containing a single DNA component. We found that the Parafilm method could improve the detection of DNA, especially the detection of larger DNA. Figure 2 displays the MALDI-TOF spectrum of an 85mer using Parafilm. The total amount of the 85mer loaded was 100 fmol. Although we were able to detect less than 100 fmol of this 85mer on Parafilm, the signal-to-noise and reproducibility became poorer. The detection limit achieved on Parafilm for the 85mer was about 5 times better than that obtained with stainless steel. It should be noted that, since detection limits also depend on the sample purity and the experimental conditions, this comparison was based on our evaluation of both Parafilm and stainless steel using the same DNA sample and instrumental conditions. With regard to smaller DNA, we were able to detect as low as 20 fmol of 41mer using the Parafilm method. The second issue explored was the detection limit for a larger DNA component in the presence of a large excess of a smaller DNA component. For PCR and sequencing reactions, a large excess of the primer DNA may be present in the reaction products. Recently, Buchanan and co-workers have investigated this issue and found that a 28-fold molar excess of a 20mer (8.3 pmol loaded) 3090 Analytical Chemistry, Vol. 70, No. 14, July 15, 1998

Figure 2. MALDI-TOF mass spectrum of an 85mer. The total amount of the 85mer loaded on the probe was 100 fmol. The unlabeled peak corresponds to the doubly charged 85mer.

eliminated any signal from a 72mer (0.3 pmol loaded).21 Hence, the removal of a large excess of untreated primers may be crucial to the success of detecting larger DNA components. In this work, we revisited this issue by using the Parafilm method. Figure 3 displays a typical MALDI-TOF spectrum of 62mers (300 fmol loaded) in the presence of a 28-fold excess of 17mer (8.4 pmol). In general, the absolute signal level of the 62mer decreased in the presence of an excess of 17mer, but we were still able to routinely observe good 62mer signals under the conditions used. This indicated that the use of Parafilm also improved the detection of a larger DNA component in the presence of a large excess of a smaller DNA component. The third issue explored was the signal intensities of larger DNA components in a DNA mixture, since the sequencing products contain many DNA components and MALDI-TOF should be able to detect all of the DNA components in such a mixture. Smith and co-workers reported that a notable feature in the MALDI analysis of a mixture of DNA was the decrease in peak intensities for the larger DNA components in a DNA mixture.5 In the spectra reported in that paper, the peak intensity of the larger 40mer component was considerably weaker than that of the smaller 18mer component. In a recent work, Little and co-workers found that equimolar samples of a 15- and 36mer, when analyzed as a mixture, resulted in approximately 5-fold higher signal intensities for the smaller 15mer.22 We have also examined this issue using the Parafilm method and found that this method (21) Hurst, G. B.; Doktycz, M. J.; Vass, A. A.; Buchanan, M. V. Rapid Commun. Mass Spectrom. 1996, 10, 377. (22) Little, D. P.; Cornish, T. J.; O’Donnell, M. J.; Braun, A.; Cotter, R. J.; Koster, H. Anal. Chem. 1997, 69, 4540.

Figure 3. MALDI-TOF mass spectra of a 62mer (300 fmol loaded) obtained in the presence of a 28-fold excess of a 17mer. The unlabeled peaks correspond to the multiply charged DNA or DNA cluster ions.

Figure 4. MALDI-TOF mass spectrum of a mixture of 17-, 41-, 62-, and 82mers. A total of 500 fmol of each of the mixture components was loaded onto Parafilm. The unlabeled peaks correspond to either the multiply charged DNA or DNA cluster ions.

improved detection of the larger DNA components in a DNA mixture sample. Figure 4 displays a typical spectrum of a mixture sample of 17-, 41-, 62-, and 85mers obtained using Parafilm. A total of 0.5 pmol of each of the mixture components was loaded, which was similar to the concentration level used in Smith and co-workers’ work.5 The peaks corresponding to the monomer DNA ions are labeled in Figure 4, and the unlabeled peaks correspond to either multiply charged DNA or DNA cluster ions. As can be seen from this figure, the peak height of the 17mer was about 1.6 times higher than that of the 41mer, but the total ion intensity (peak area) of the 41mer is very similar to that of the 17mer. This suggests that there is no significant drop in sensitivity with increasing DNA size from 17- to 41mer. Even for the much larger 62- and 85mers, the total signal levels were still better than 40% and 20% of the signal level of the 17mer, respectively. Salt Effects. The original aim of this work was to develop a new sample preparation method that improves detection limits and produces a more homogeneous sample film. But, to our surprise, the use of Parafilm could also reduce the effect of metal salts. Figure 5 displays the MALDI-TOF mass spectra of a 17mer and insulin. As will be discussed later, the use of dielectric materials such as Parafilm could lead to the peak position variation from sample to sample and from spot to spot. This made it difficult to utilize external calibration. Hence, insulin was added to the 17mer solution for the purpose of internal calibration and evaluation of mass accuracy. It was found that mass accuracy was better than 0.3%, similar to that obtained with stainless steel. The relatively poor mass accuracy is due to poor mass resolution associated with using the 3-HPA matrix.

Figure 5a was obtained in the absence of salt, while parts b and c of Figure 5 were produced in the presence of 100 mM NaCl and 0.5 mM MgCl2 in the 17mer solution, respectively. The amounts of 17mer and insulin loaded onto the tip were 2 and 1 pmol, respectively. The poor insulin mass resolution shown in Figure 5 resulted from the use of the matrix 3-HPA. It was seen that, even at the levels of 100 mM NaCl and 1 mM MgCl2, the17mer spectra did not degrade appreciably. Figure 6 displays the MALDI-TOF spectra of a 41mer. Figure 6a was obtained in the absence of salt, while parts b and c of Figure 6 were produced in the presence of 100 mM NaCl and 1 mM MgCl2 in the 41mer solution, respectively. The total amount of 41mer loaded on the tip was 1 pmol. It can be seen from Figure 6, at the level of 100 mM NaCl and 1 mM MgCl2, that the 41mer spectra did not degrade significantly. Moreover, no laser power increase was needed to generate Figure 6a,b. Recently, Becker and co-workers conducted an investigation into the salt effect on DNA analysis and found that the salt tolerance limits were only 5 mM NaCl and 0.0 2mM MgSO4 for a 41mer DNA.10 Obviously, our results show that the DNA tolerance toward metal salts are increased when Parafilm was used. Becker and co-workers also found that the salt tolerance significantly dropped with increasing size of DNA. As shown above, when Parafilm was used, the tolerance toward NaCl and MgCl2 did not change significantly with increasing size of DNA. To further examine this interesting result, we studied the salt effect on the analysis of a 85mer DNA. Figure 7 displays the MALDI-TOF spectra of the 85mer. Figure 7a was obtained in the absence of salt, while parts b and c of Figure 7 were produced in the presence of 100 mM NaCl and 1 Analytical Chemistry, Vol. 70, No. 14, July 15, 1998

3091

Figure 5. MALDI-TOF mass spectra of a 17mer and insulin obtained (a) in the absence of salts and (b and c) in the presence of 50 mM NaCl and 1 mM MgCl2, respectively. The total amounts of 17mer and insulin loaded were 2 and 1 pmol, respectively.

mM MgCl2, respectively. The total amount of 85mer loaded was 1 pmol. Once again, the 85mer spectra did not degrade appreciably at these salt levels. Clearly, this suggested that the salt effect was less sensitive to the DNA size when Parafilm was used. It should be noted that, although the use of Parafilm suppressed the salt effect, it could not completely eliminate it. The better resolution spectra of small DNA reveal that metals still attach to DNA to a certain degree if metal salts are present.20 Hence, further investigation into the desalting mechanism that occurs on hydrophobic surfaces is warranted, since a better understanding of this process will provide guidance for developing better sample preparation methods for eliminating salt effects in DNA analysis. Further Discussion. As mentioned above, a problem associated with the Parafilm method is variation of the peak position from sample to sample and from spot to spot. A similar problem was also observed with the use of other dielectric materials.23 This could be due to the fact that the electric field near the desorption spot may be no longer uniform when dielectric materials are attached to the metal tip. In addition, the film prepared on Parafilm was thicker than that produced on a metal surface. Uneven surface levels could also cause this variation. Currently, we are examining several different methods to reduce the thickness of the sample film prepared on Parafilm. It should also be noted that this variation is a problem only if external calibration is used. The use of internal calibration obviates this problem. (23) Loo, R. O. Presented at the 45th ASMS Conference on Mass Spectrometry and Allied Topics, Palm Springs, CA, June 1-5, 1997.

3092 Analytical Chemistry, Vol. 70, No. 14, July 15, 1998

Figure 6. MALDI-TOF mass spectra of a 41mer obtained (a) in the absence of salts and (b and c) in the presence of 100 mM NaCl and 1 mM MgCl2, respectively. The total amount of the 41mer loaded was 1 pmol.

Primer and template DNA are present in the products of sequencing and PCR reactions. One should be able to use the primer and template as internal standards to calibrate the resulting DNA spectra. Moreover, this variation problem may occur only under continuous extraction conditions. It has been reported that the use of delayed extraction minimized the peak variation problem associated with the use of dielectric materials.23 Next, we elaborate on the detection limit issues. In general, the use of Parafilm improves the detection of DNA. This could be due to the fact the sample size prepared on Parafilm was smaller than that obtained with metals, and thereby the DNA concentration on a desorption spot of Parafilm might be higher when the same amount of DNA was loaded. More interestingly, this new method worked better for a larger DNA and enhanced the detection of larger DNA in the presence of a large excess of a smaller DNA or in a DNA mixture. This feature is important to sequencing applications and PCR product analyses. We also found that we could generate over 50 good DNA spectra with a DNA sample of 0.5 mm (100 fmol loaded). This suggests that the sample size that we prepared on Parafilm might still be too large and that further improvement in detection can be obtained by reducing the sample size to 0.1 mm. In fact, Little and co-workers have reported the generation of the DNA sample size of 0.1 mm on a silicon substrate and attained a detection limit for a 36mer less than 1 fmol.22 We are presently working on a project to further reduce the sample size prepared on Parafilm to achieve low femtomole detection limits. Now, we consider the salt effect. Lubman and co-workers reported that the use of Nafion and nitrocellulose (NC) substrates reduced the metal salt effect.24-26 Both Nafion and NC are

tolerance toward a particular salt should decrease with increase of the DNA size. The work of Becker and co-workers confirmed this point.10 An interesting question is why the metal salt effect in our work is less sensitive to the DNA size. This interesting result could also be explained by the proposed crystallization mechanism. As a result of the crystallization process occurring on Parafilm, there might be a “saturation” effect. In other words, when salts are present in the DNA solution, only small amounts of the salts could be doped into the matrix crystals. Once the salt concentration in the matrix crystals reached a “saturation” level, no additional salts can be doped into the matrix, even when extra amounts of the salts are present in the DNA solution. Hence, the salt effect is less sensitive to the DNA size. If this mechanism is true, one may be able to utilize the crystallization process to remove the salts present in the DNA sample. Furthermore, on the basis of this mechanism, one would expect that the use of other hydrophobic materials as the MALDI sample surface could produce a similar effect. In fact, our preliminary results using Teflon and polyethylene in place of a metal probe surface indicate that these hydrophobic polymer materials also worked well for DNA analyses. Hence, the mechanism proposed in this paper may be helpful in the development of new MALDI sample preparation methods. Figure 7. MALDI-TOF mass spectra of an 85mer obtained (a) in the absence of salts and (b and c) in the presence of 100 mM NaCl and 1 mM MgCl2, respectively. The total amount of the 85mer loaded was 1 pmol.

negatively charged and can effectively bind the metal cations to their “negative” sites and thereby reduce their interference during the MALDI process. Parafilm is a blend of olefins, being neither a polymer nor a polar compound. How, then, does Parafilm reduce the salt effect? Our rationale is that separation of salts from 3-HPA and DNA may occur during crystallization on Parafilm, leading to a reduced salt effect. Less polar molecules interact more strongly with Parafilm given its hydrophobic nature. Less polar 3-HPA and DNA could crystallize first, followed by metal salts. As a result, 3-HPA and DNA could be separated from the salts. This argument may explain an interesting experimental phenomenon, in which we sometimes found that the first few laser shots produced weak or no DNA signals when large amounts of salts, such as 500 mM NaCl, were present in the DNA solution. Thereafter, the DNA signals gradually increased. Sometimes, the DNA signals continued to increase even after 100 laser shots. This could suggest that the top layers of the crystals contained highly concentrated metal salts that were separated from DNA. The number of the negative charges of DNA increases with the size of DNA. Hence, it is reasonable to expect that the (24) Bai, J.; Lubman, D.; Siemieniagw, D. Rapid Commun. Mass Spectrom. 1994, 8, 687. (25) Liu, Y.; Bai, J.; Zhu, Y.; Liang, X.; Siemieniak, D.; Venta, P.; Lubman, D. Rapid Commun. Mass Spectrom. 1995, 9, 735. (26) Bai, J.; Liu, Y.; Cain, T. C.; Lubman, D. M. Anal. Chem. 1994, 66, 3423.

CONCLUSIONS The paper reports a new MALDI sample preparation method involving the use of Parafilm. This method offers the following three major analytical benefits: General Benefits. The Parafilm method (1) produces a more homogeneous coverage of matrix/DNA over the sample surface; (2) adds no extra cost and time to sample preparation; (3) does not react with the molecules present in the matrix or DNA solutions; and (4) eliminates the commonly required step of cleaning the probe tip after each analysis. Detection Limit. This method improves the detection sensitivity. It was found that its detection sensitivity for a 85mer was 5 times better than that obtained with stainless steel. More importantly, it improves the detection of a larger DNA component in the presence of a large excess of a smaller DNA or in a DNA mixture. This is important to the analysis of the PCR and sequencing products. Salt Effects. The Parafilm method improves the DNA tolerance toward metal salts. Moreover, the salt effect was less sensitive to the DNA size. ACKNOWLEDGMENT This research is supported by NIH grant no. HG01437 and by Cleveland State University through an EFFRD grant. Received for review January 29, 1998. Accepted April 24, 1998. AC980090E

Analytical Chemistry, Vol. 70, No. 14, July 15, 1998

3093