RELATED APPLICATIONS 0001 This application is a continuation of U.S. application Ser. No. 09/074,936, filed May 7, 1998, to Franz Hillenkamp, entitled IR-MALDI Mass Spectrometry of Nucleic Acids Using Liquid Matrices. The subject matter this application is herein incorporated by reference in its entirety.

1. BACKGROUND OF THE INVENTION 0002 Mass spectrometry provides a means of weighing individual molecules by ionizing the molecules in vacuo and making them fly by volatilization. Under the influence of combinations of electric and magnetic fields, the ions follow trajectories depending on their individual mass (m) and charge (z). For molecules of low molecular weight, mass spectrometry has long been part of the routine physical-organic repertoire for analysis and characterization of organic molecules by the determination of the mass of the parent molecular ion. In addition, by arranging collisions of this parent molecular ion with other particles (e.g., argon atoms), the molecular ion is fragmented, forming secondary ions by the so-called collision induced dissociation (CID). The fragmentation pattern/pathway very often allows the derivation of detailed structural information. 0003 During the last decade, mass spectrometry (MS) has become an important analytical tool in the analysis of biological macromolecules. This is due at least in part to introduction of the so-called soft ionization methods, namely Matrix-Assisted Laser Desorption/Ionization (MALDI) and ElectroSpray Ionization (ESI), which allow intact ionization, detection and exact mass determination of large molecules, i.e. well exceeding 300 kDa in mass (Fenn, J. B., et al., (1989) Science 246, 64-71; Karas M. & Hillenkamp F. (1988) Anal. Chem. 60, 2299-3001). 0004 MALDI-MS (reviewed in (Nordhoff E., et al., (1997) Mass Spectrom. Rev. 15: 67-138.) and ESI-MS have been used to analyze nucleic acids. However, since nucleic acids are very polar biomolecules, that are difficult to volatize, there has been an upper mass limit for clear and accurate resolution. ESI would seem to be superior to MALDI for the intact desorption of large nucleic acids even in the MDa mass range (Fuerstenau S. D. & Benner W. H. (1995). Rapid Commun. Mass Spectrom. 9, 15281538; Chen R., Cheng X., Mitchell et al., (1995). Anal. Chem. 67, 1159-1163.). However, mass assignment is very poor and only possible with an uncertainty of around 10%. The largest nucleic acids that have been accurately mass determined by ESI-MS, so far, are a 114 base pair double stranded PCR product (Muddiman D. C., Wunschel D. S., Lis C., Pas-Tolic L., Fox K. F., Fox A., Anderson G. A. & Smith R. D. (1996) Anal. Chem. 68, 3705-3712) of about 65 kDa in mass and a 120 nucleotide E. coli 5S rRNA of about 39 kDa in mass (Limbach, P. A. Crain, P. F., McCloskey, J. A., (1995) J. Am. Soc. Mass Spectrom. 6:27-39.). ESI furthermore requires extensive sample purification. A few reports on the MALDI-MS of large DNA molecules with lasers emitting in the ultraviolet (UV) have been reported (Ross P. L. & P. Belgrader (1997) Anal. Chem. 69: 3966-3972; Tang K., et al., (1994) Rapid Commun. Mass Spectrum. 8: 727-730; Bai J., et al., (1995) Rapid Commun. Mass Spectrum. 9: 1172-1176; Liu Y-H-, et al., (1995) Anal. Chem. 67: 3482-3490 and Siegert C. W., et al., (1997) Anal. Biochem. 243, 55-65. However, based on these reports it is clear that analysis of nucleic acids exceeding 30 kDa in mass (i.e. ca. a 100mer) by UV-MALDI-MS gets increasingly difficult with a current upper mass limit of about 90 kDa (Ross P. L. & P. Belgrader (1997) Anal. Chem. 69: 3966-3972). The inferior quality of the DNA UV-MALDI-spectra has been attributed to a combination of ion fragmentation and multiple salt formation of the phosphate backbone. Since RNA is considerably more stable than DNA under UV-MALDI conditions, the accessible mass range for RNA is up to about 150 kDa (Kirpekar F., et al., (1994). Nucleic Acids Res. 22, 3866-3870). The analysis of nucleic acids by IR-MALDI with solid matrices (mostly succinic acid and, to a lesser extent, urea and nicotinic acid) has been described (Nordhoff, E. et al., (1992) Rapid Commun. Mass Spectrom. 6: 771-776; Nordhoff, E. et al., (1993) Nucleic Acids Res. 21: 3347-3357; and Nordhoff, E. et al., (1995) J. Mass Spec. 30: 99-112). The 1992 Nordhoff et al., paper reports that a 20-mer of DNA and an 80-mer of RNA were about the uppermost limit for resolution. The 1993 Nordhoff et al. paper, however, provides a distinct spectra for a 26-mer of DNA and a 104-mer of tRNA. The 1995 Nordhoff et al., paper shows a substantially better spectra for the analysis of a 40-mer by UV-MALDI with the solid matrix, 3-hydroxy picolinic acid, than by IR-MALDI with succinic acid (See FIGS. 1(d) and 1(e)). In fact the 1995 paper reports that IR-MALDI resulted in a substantial degree of prompt fragmentation. 0005 Nucleic acid analysis can be useful, for example, for diagnosing the existence of a genetic disease or chromosomal abnormality; a predisposition to a disease or condition, infection by a pathogenic organism or to provide information relating to identity, heredity or compatibility. Since mass spectrometry can be performed relatively quickly and is amenable to automation, improved methods for obtaining accurate mass spectra for larger nucleic acid molecules (e.g. larger than about 90 kDa of DNA and 150 kDa of RNA) are clearly needed.

2. SUMMARY OF THE INVENTION 0006 In one aspect, the invention features processes for rapidly and accurately determining the mass of nucleic acids (e.g. DNA or RNA) using infrared matrix assisted laser desorption ionization (IR-MALDI) mass spectrometry and a liquid matrix. 0007 In a preferred embodiment, a solution containing the nucleic acid and a liquid matrix is deposited onto a substrate to form a homogeneous, transparent thin layer of nucleic acid solution, which is then illuminated with infrared radiation, so that the nucleic acid is desorbed and ionized, thereby emitting ion particles, which are then analyzed using a mass analyzer to determine the identity of the nucleic acid. Preferably, sample preparation and deposition is performed using an automated device. 0008 Preferred liquid matrices for use herein have a sufficient absorption at the wavelength of the laser to be used in performing desorption and ionization and are a liquid (not a solid or a gas) at room temperature (20 C.). For absorption purposes, the liquid matrix can contain at least one chromophore or functional group that strongly absorbs infrared radiation. Preferred functional groups include: nitro, sulfonyl, sulfonic acid, sulfonamide, nitrile or cyanide, carbonyl, aldehyde, carboxylic acid, amide, ester, anhydride, ketone, amine, hydroxyl, aromatic rings, dienes and other conjugated systems. 0009 Particularly preferred liquid matrices are substituted or unsubstituted: (1) alcohols, including: glycerol, 1,2- or 1,3-propane diol, 1,2-, 1,3- or 1,4-butane diol, triethanolamine; (2) carboxylic acids including: formic acid, lactic acid, acetic acid, propionic acid, butanoic acid, pentanoic acid, hexanoic acid and esters thereof; (3) primary or secondary amides including acetamide, propanamide, butanamide, pentanamide and hexanamide, whether branched or unbranched; (4) primary or secondary amines, including propylamine, butylamine, pentylamine, hexylamine, heptylamine, diethylamine and dipropylamine; (5) nitrites, hydrazine and hydrazide. 0010 Preferably, a liquid matrix for use herein, is miscible with a nucleic acid compatible solvent. It is also preferable that the liquid matrix is vacuum stable, i.e. has a low vapor pressure, so that the sample does not evaporate quickly in the mass analyzer. Preferably the liquid should also be of an appropriate viscosity to facilitate dispensing of micro- to nano-liter volumes of matrix alone or mixed with a nucleic acid compatible solvent. Mixtures of different liquid matrices and additives to such matrices may be desirable to confer one or more of the above-described properties. 0011 Once prepared, the nucleic acid/matrix solution is deposited as a thin layer on a substrate, which is preferably contained within a vacuum chamber. Preferred substrates for holding the nucleic acid/matrix solution are selected from the group consisting of: beads, capillaries, flat supports, pins and wafers, with or without filter plates. Preferably the temperature of the substrate can be regulated to cool the nucleic acid/matrix solution to a temperature that is below room temperature. 0012 Preferred infrared radiation is in the mid-IR wavelength region from about 2.5 m to about 12 m. Particularly preferred sources of radiation include CO, CO2 and Er lasers. In certain embodiments, the laser can be an optic fiber or the laser radiation can be coupled to the mass spectrometer by fiber optics. 0013 In a further preferred embodiment, the ion particles generated from the analyte are extracted for analysis by the mass analyzer in a delayed fashion prior to separation and detection in a mass analyzer. Preferred separation formats include linear or reflector (with linear and nonlinear fields, e.g. curved field reflectron) time-of-flight (TOF), single or multiple quadrupole, single or multiple magnetic sector, Fourier transform ion cyclotron resonance (FTICR) or ion trap mass spectrometers. 0014 Using the processes reported herein, accurate (i.e at least about 1% accurate) masses of sample DNA can be obtained for at least about 2000-mers of DNA (i.e. masses of at least about 650 kDa) and at least about 1200-mers of RNA (i.e. masses of at least about 400 kDa). In addition, signals of single stranded as well as double stranded nucleic acids can be obtained in the spectra. 0015 The improved accuracy for measuring the mass of deoxyribonucleic acids (DNA) by IR-MALDI mass spectrometry (accuracy of at least about 1%) is far superior to that provided by standard agarose gel sizing of nucleic acids (accuracy of about 5%). Mass determination of ribonucleic acids (RNA) by IR-MALDI mass spectrometry (accuracy of at least about 0.5%) is even more significant, since an accurate size determination of RNA by gel analysis is difficult if not impossible, in part because of the absence of suitable size markers and of a really well-suited gel matrix. 0016 As important as the extension in mass range is the dramatic decrease in the amount of analyte needed for preparation, down to the low femtomole (fmol) and even the attomole (amol) range even with an essentially simple preparation method. In addition, by using a liquid rather than a solid matrix, the ion signals generated have been found to be more reproducible from shot to shot. Use of a liquid matrix also facilitates sample dispensation, for example to various fields of a chip array. Further, by using a liquid matrix in conjunction with IR-MALDI mass spectrometry, essentially all sample left on the target after IR-MALDI analysis can be retrieved for further use. 0017 Other features and advantages of the invention will be apparent from the following detailed description and claims.

3. BRIEF DESCRIPTION OF THE FIGURES 0018 FIG. 1 shows the following mass spectra of a synthetic DNA 70-mer: FIG. 1(a) ultraviolet matrix assisted laser desorption ionization (UV-MALDI) and detection by a linear time-of-flight (TOF) instrument using delayed extraction and a 3 hydroxypicolinic acid (3HPA) matrix (sum of 20 single shot mass spectra); FIG. 1(b) UV-MALDI-reflectron (ret) TOF spectrum, using delayed extraction and a 3HPA matrix (sum of 25 single shot mass spectra); and FIG. 1(c) IR-MALDI-retTOF spectrum, using delayed extraction and a glycerol matrix, (sum of 15 single shot mass spectra). 0019 FIG. 2 shows the following IR-MALDI RetTOF mass spectra using a 2.94 m wavelength and a glycerol matrix: FIG. 2(a) a synthetic DNA 21 mer (sum of 10 single shot spectra); FIG. 2(b) a DNA mixture comprising a restriction enzyme products of a 280mer (87 kDa), a 360mer (112 kDa), a 920mer (285 kDa) and a 1400mer (433 kDa) (sum of 10 single shot spectra); FIG. 2(c) DNA mixture; restriction enzyme products of a 130mer (ca. 40 kDa), a 640mer (198 kDa) and a 2180mer (674 kDa) (sum of 20 single shot spectra); and (d) an RNA 1206mer (ca.387 kDa), (sum of 15 single shot spectra). Ordinate scaling is intercomparible. 0020 FIG. 3 shows the spectra of a 515-mer double stranded PCR DNA product. The following different total amounts of sample were loaded: 3(a) 300 fmol, single shot spectrum; 3(b) 3 fmol, sum of single shot spectra; 3(c) 300 amol, sum of 25 single shot spectra obtained using an IR-MALDI RetTOF, wherein the laser emitted at a wavelength of 2.94 m using a glycerol matrix.

4. DETAILED DESCRIPTION OF THE INVENTION 0021 The instant invention is based, at least in part, on the surprising finding that high resolution mass spectra of large nucleic acid molecules (DNA and RNA) can be obtained by desorbing and ionizing the nucleic acids in a liquid matrix using a laser that emits in the infrared electromagnetic wavelength. 0022 The invention therefore features a process for performing matrix assisted laser desorption ionization (MALDI) mass spectrometry, comprising mixing a nucleic acid solution with a liquid matrix to form a matrix/nucleic acid solution and depositing the solution onto a substrate to form a homogeneous, thin layer of matrix/nucleic acid solution. The nucleic acid containing substrate can then be illuminated with infrared radiation of an appropriate wavelength to be absorbed by the matrix, so that the nucleic acid is desorbed and ionized, thereby emitting ion particles that can be extracted (separated) and analyzed by a mass analyzer to determine the mass of the nucleic acid. 0023 Nucleic acids to be analyzed according to the processes of the invention can include any single stranded or double stranded polynucleotide, such as deoxyribonucleic acid (DNA), including genomic or cDNA, ribonucleic acid (RNA) or an analog of RNA or DNA, as well as nucleotides or nucleosides and any derivative thereof. Nucleic acids can be of any size ranging from single nucleotides or nucleosides to tens of thousands of base pairs (-mers). For analysis herein, preferred nucleic acids are thousand-mers or less. 0024 Nucleic acids may be obtained from a biological sample (i.e. any material obtained from any living source (e.g. human, animal, plant, bacteria, fungi, protist, virus) using any of a number of procedures, which are well-known in the art. The particular isolation procedure chosen being appropriate for the particular biological sample. For example, freeze-thaw and alkaline lysis procedures can be useful for obtaining nucleic acid molecules from solid materials; heat and alkaline lysis procedures can be useful for obtaining nucleic acid molecules from urine; and proteinase K extraction can be used to obtain nucleic acids from blood (Rolff, A et al., PCR: Clinical Diagnostics and Research, Springer (1994)). Prior to being mixed with a liquid matrix, the particular nucleic acid may require further processing to yield a relatively pure nucleic acid sample. For example, a standard ethanol precipitation may be performed on restriction enzyme digested DNA. Alternatively, PCR products may require primer removal prior to analysis. Likewise, RNA strands can be separated from the molar excess of premature termination products always present in in vitro transcription reactions. 0025 As used herein, the term liquid matrix is meant to refer to a matrix that has a sufficient absorption at the wavelength of the laser to be used in performing desorption and ionization and that is a liquid (not a solid or a gas) at room temperature (about 20 C.). 0026 For absorption purposes, the liquid matrix can contain at least one chromophore or functional group that strongly absorbs infrared radiation. Examples of appropriate functional groups include: nitro, sulfonyl, sulfonic acid, sulfonamide, nitrile or cyanide, carbonyl, aldehyde, carboxylic acid, amide, ester, anhydride, ketone, amine, hydroxyl, aromatic rings, dienes and other conjugated systems. 0027 Preferred liquid matrices are substituted or unsubstituted: (1) alcohols, including: glycerol, 1,2- or 1,3-propane diol, 1,2-, 1,3- or 1,4-butane diol, triethanolamine; (2) carboxylic acids including: formic acid, lactic acid, acetic acid, propionic acid, butanoic acid, pentanoic acid, hexanoic acid and esters thereof; (3) primary or secondary amides including acetamide, propanamide, butanamide, pentanamide and hexanamide, whether branched or unbranched; (4) primary or secondary amines, including propylamine, butylamine, pentylamine, hexylamine, heptylamine, diethylamine and dipropylamine; (5) nitrites, hydrazine and hydrazide. Particularly preferred compounds are comprised of eight or fewer carbon atoms. For example, particularly preferred carboxylic acids and amides are comprised of six or fewer carbon atoms, preferred amines are comprised of about three to about seven carbons and preferred nitrites are comprised of eight or fewer carbons. However, compounds that are unsaturated to any degree may be comprised of a larger number of carbons, since unsaturation confers liquid properties on a compound. Although the particular compound used as a liquid matrix must contain a functional group, the matrix is preferably not so reactive that it may fragment or otherwise damage the nucleic acid to be analyzed. 0028 An appropriate liquid should also be miscible with a nucleic acid compatible solvent. Preferably the liquid should also be of an appropriate viscosity, e.g. typically less than or equal to about 1.5 Ns/m2, (the viscosity of glycerol at room temperature) to facilitate dispensing of micro- to nano-liter volumes of matrix alone or mixed with a nucleic acid compatible solvent. For use herein, a liquid matrix should also have an appropriate survival time in the vacuum of the analyzer (typically having a pressure in the range of about 105 to about 1010 mbars) to allow the analysis to be completed. Liquids having an appropriate survival time are vacuum stable, a property, which is strictly a function of the vapor pressure of the matrix, which in turn is strongly dependent on the sample temperature. Preferred matrices have a low vapor pressure at room temperature, so that less than about fifty percent of the sample in a mass analyzer having a back pressure, which is less than or equal to 105 mbars, evaporates in the time needed for the analysis of all samples introduced (e.g. from about 10 min to about 2 hrs.). For example, for a single sample, the analysis may be performed in minutes. However, for multiple samples, the analysis may require hours for completion. For example, glycerol can be used as a matrix at room temperature in a vacuum for about 10 to 15 minutes. However, if glycerol is to be used for analyzing multiple samples in a single vacuum, the vacuum may need to be cooled to maintain the sample at a temperature, which is in the range of about 50 C. to about 100 C. (173 K to about 223K) for the time required to complete the analysis. Triethanolamine, in contrast, has a much lower vapor pressure than glycerol and can survive in a vacuum for at least about one hour even at room temperature. 0029 Mixtures of different liquid matrices and additives to such matrices may be desirable to confer one or more of the above-described properties. For example, an appropriate liquid matrix could be comprised of a small amount of an IR chromophore containing solution and a greater amount of an IR invisible (i.e. nonabsorbing) material, in which, for example, the nucleic acid is soluble. It may also be useful to use a matrix, which is doped with a small amount of a compound or compounds having a high extinction coefficient (E) at the laser wavelength used for desorption and ionization, e.g. dinitrobenzenes, polyenes, etc. An additive that acidifies the liquid matrix may also be added to dissociate double stranded nucleic acids or to denature secondary structures of nucleic acids, such as that of t-RNA. Additional additives may be useful for minimizing salt formation between the matrix and the phosphate backbone of the nucleic acid. For example, the additive can comprise an ammonium salt or ammonium-loaded ion exchange bead, which removes alkali ions from the matrix. Alternatively, the liquid matrix can be distilled prior to mixture with the nucleic acid solution, to minimize salt formation between the matrix and the phosphate backbone of the nucleic acid. The liquid matrix can also be mixed with an appropriate volume of water or other liquid to control sample viscosity and rate of evaporation. Since literally all of the water is evaporated during mass analysis, an easily manipulated volume (e.g. 1 l) can be used for sample preparation and transfer, but still result in a very small volume of liquid matrix. As a result, only small volumes of nucleic acid are required to yield about 1016 moles to about 1012 moles (about 100 attomoles to about 1 picomole) of nucleic acid in the final liquid matrix droplet. 0030 As shown in the following examples, when glycerol is used as a matrix, the final analyte-to-glycerol molar ratio (concentration) should be in the range of about 104 to 109 depending on the mass of the nucleic acid and the total amount of nucleic acid available. For example, for the sensitivity test described in the following examples, the relatively high concentration of nucleic acid used was measured by standard UV-spectrophotometry. Practically speaking, one typically knows the approximate amount of nucleic acid generated, e.g. from a PCR or transcription reaction. The large range specified indicates that the actual amount of nucleic acid analyzed is not very critical. Typically, a greater amount of nucleic acid results in a better spectra. However, there may be instances where the nucleic acid sample requires dilution. 0031 Preferably, nucleic acid samples are prepared and deposited as a thin layer (e.g. a monolayer to about a 100 m layer, preferably between about 1 m to 10 m) onto a substrate using an automated device, so that multiple samples can be prepared and analyzed on a single sample support plate with only one transfer into the vacuum of the analyzer and requiring only a relatively short period of time for analysis. Appropriate automated sample handling systems for use in the instant process are described, for example, in U.S. Pat. Nos. 5,705,813, 5,716,825 and 5,498,545. Any substrate on which the nucleic acid/liquid matrix can be deposited and retained for desorption and ionization of the nucleic acid can be used in the process of the instant invention. Preferred substrates are beads (e.g. silica gel, controlled pore glass, magnetic, Sephadex/Sepharose, cellulose), capillaries, flat supports (e.g. filters, plates or membranes made of glass, metal surfaces (steel, gold, silver, aluminum, copper or silicon) or plastic (e.g. polyethylene, polypropylene, polyamide, polyvinylidenefluoride), pins (e.g. arrays of pins suitable for combinatorial synthesis or analysis of beads in pits of flat surfaces such as wafers, with or without filter plates). 0032 The sample containing substrate can then be analyzed in a vacuum chamber of a mass analyzer to identify the nucleic acid. Preferably, the mass analyzer can maintain the temperature of a sample at a preselected value, e.g. a temperature in the range of at least about 100 C. to about 80 C., during sample preparation, disposition or analysis. For example, improved spectra may be obtained, in some instances, by cooling the sample to a temperature below room temperature (i.e. below 20 C.) during sample preparation and/or mass analysis. Further, as described above, the vacuum stability of a matrix may be increased by cooling. Alternatively, it may be useful to heat a sample to denature double stranded nucleic acids into single strands or to decrease the viscosity during sample preparation. 0033 Desorption and ionization of the sample is then performed in the mass analyzer using infrared radiation. Infrared radiation or infrared wavelength refers to electromagnetic wavelengths, which are longer than those of red light in the visible spectrum and shorter than radar waves. Preferred infrared wavelengths for use in the instant invention are in the mid-IR wavelength region (i.e. from about 2.5 m to about 12 m). Preferred sources of infrared radiation are CO lasers, which emit at about 6 m, CO2 lasers, which emit at about 9.2-11, Er lasers with any of a variety of crystals (e.g. YAG or YILF) emitting at wavelengths around 3 m and and optical paramagnetic oscillator lasers emitting in the range of about 2.5 m to about 12 m The pulse duration and/or size of the irradiated area (spot size) can be varied to generate multiple charged ions. A preferred pulse duration is in the range of about 100 picoseconds (ps) to about 500 nanoseconds (ns). A preferred spot size is in the range of about 50 m in diameter to about 1 mm. 0034 IR-MALDI can be matched with an appropriate mass analyzer, including linear (lin) or reflector (ret) (with linear and nonlinear fields, e.g. curved field reflectron) time-of-flight (TOF), single or multiple quadrupole, single or multiple magnetic sector, Fourier transform ion cyclotron resonance (FTICR) or ion trap. Preferably detection is performed using a linear or reflectron mode TOF instrument in positive or negative ion mode, so that the ions are accelerated through a total potential difference of about 3-30 kV in the split extraction source using either static or delayed ion extraction (DE). Time-of-flight (TOF) mass spectrometers separate ions according to their mass-to-charge ratio by measuring the time it takes generated ions to travel to a detector. The technology behind TOF mass spectrometers is described for example in U.S. Pat. Nos. 5,627,369, 5,625,184, 5,498,545, 5,160,840 and 5,045, 694, the teachings of which are each specifically incorporated herein by reference. 0035 Delayed extraction with delay times ranging from about 50 ns to about 5 s may improve the mass resolution of some nucleic acids (e.g. nucleic acids in the mass range of from about 30 kDa to about 50 kDa using either a liquid or solid matrix). 0036 The improved processes for detecting nucleic acids by mass spectromety can be useful, for example, for diagnosing the existence of any one of the more than 3000 known genetic diseases (Human Genome Mutations, D. N. Cooper and M. Krawczak, BIOS Publishers, 1993) including hemophilias, thalassemias, Duchenne Muscular Dystrophy (DMD), Huntington's Disease (HD), Alzheimer's Disease and Cystic Fibrosis (CF) or genetic diseases to be identified. In addition, the processes can be useful for diagnosing certain birth defects, which are the result of chromosomal abnormalities such as Trisomy 21 (Down's Syndrome), Trisomy 13 (Patau Syndrome), Trisomy 18 (Edward's Syndrome), Monosomy X (Turner's Syndrome) and other sex chromosome aneuploidies such as Kleinfelter's Syndrome (XXY). The processes can also be used to detect certain DNA sequences that may predispose an individual to any of a number of diseases, such as diabetes, arteriosclerosis, obesity, various autoimmune diseases and cancer (e.g. colorectal, breast, ovarian, prostate, lung) or that render an individual suitable or unsuitable for a particular medical treatment. 0037 Alternatively, the processes can be used to detect nucleic acids that are characteristic of viruses, bacteria, fungi and other infectious organisms, which are different from the sequences contained in the host cell. Finally, the processes can be used to detect characteristic nucleic acid sequences that provide information relating to identity, heredity or compatibility. 0038 Examples of disease causing viruses that infect humans and animals and which may be detected by the disclosed processes include: Retroviridae (e.g., human immunodeficiency viruses, such as HIV-1 (also referred to as HTLV-III, LAV or HTLV-III/LAV, See Ratner, L. et al., Nature, Vol. 313, Pp. 227-284 (1985); Wain Hobson, S. et al., Cell, Vol. 40: Pp. 9-17 (1985)); HIV-2 (See Guyader et al., Nature, Vol. 328, Pp. 662-669 (1987); European Patent Application No. 0 269 520; Chakraborti et al., Nature, Vol. 328, Pp. 543-547 (1987); and European Patent Application No. 0 655 501); and other isolates, such as HIV-LP (International Publication No. WO 94/00562 entitled A Novel Human Immunodeficiency Virus; Picornaviridae (e.g., polio viruses, hepatitis A virus, (Gust, I. D., et al., Intervirology, Vol. 20, Pp. 1-7 (1983); entero viruses, human coxasackie viruses, rhinoviruses, echoviruses); Calciviridae (e.g., strains that cause gastroenteritis); Togaviridae (e.g., equine encephalitis viruses, rubella viruses); Flaviridae (e.g., dengue viruses, encephalitis viruses, yellow fever viruses); Coronaviridae (e.g. coronaviruses); Rhabdoviridae (e.g., vesicular stomatitis viruses, rabies viruses); Filoviridae (e.g., ebola viruses); Paramyxoviridae (e.g., parainfluenza viruses, mumps virus, measles virus, respiratory syncytial virus); Orthomyxoviridae (e.g., influenza viruses); Bungaviridae (e.g., Hantaan viruses, bunga viruses, phleboviruses and Nairo viruses); Arena viridae (hemorrhagic fever viruses); Reoviridae (e.g., reoviruses orbiviruses and rotaviruses); Birnaviridae; Hepadnaviridae (Hepatitis B virus); Parvoviridae (parvoviruses); Papovaviridae (papilloma viruses, polyoma viruses); Adenoviridae (most adenoviruses); Herpesviridae (herpes simplex virus (HSV) 1 and 2, varicella zoster virus, cytomegalovirus (CMV), herpes viruses'); Poxviridae (variola viruses, vaccinia viruses, pox viruses); and Iridoviridae (e.g., African swine fever virus); and unclassified viruses (e.g., the etiological agents of Spongiform encephalopathies, the agent of delta hepatities (thought to be a defective satellite of hepatitis B virus), the agents of non-A, non-B hepatitis (class 1internally transmitted; class 2parentally transmitted (i.e., Hepatitis C); Norwalk and related viruses, astroviruses). 0039 Examples of infectious bacteria include: Helicobacter pyloris, Borelia burgdorferi, Legionella pneumophilia, Mycobacteria sps (e.g. M. tuberculosis, M. avium, M. intracellulare, M. kansaii, M. gordonae), Staphylococcus aureus, Neisseria gonorrhoeae, Neisseria meningitidis, Listeria monocytogenes, Streptococcus pyogenes (Group A Streptococcus), Streptococcus agalactiae (Group B Streptococcus), Streptococcus (viridans group), Streptococcus faecalis, Streptococcus bovis, Streptococcus (anaerobic sps.), Streptococcus pneumoniae, pathogenic Campylobacter sp., Enterococcus sp., Haemophilus influenzae, Bacillus antracis, corynebacterium diphtheriae, corynebacterium sp., Erysipelothrix rhusiopathiae, Clostridium perfringers, Clostridium tetani, Enterobacter aerogenes, Klebsiella pneumoniae, Pasturella multocida, Bacteroides sp., Fusobacterium nucleatum, Streptobacillus moniliformis, Treponema pallidium, Treponema pertenue, Leptospira, and Actinomyces isrealli. 0040 Examples of infectious fungi include: Cryptococcus neoformans, Histoplasma capsulatum, Coccidioides immitis, Blastomyces dermatidis, Chlamydia trachomatis, Candida albicans. Other infectious organisms (i.e., protists) include: Plasmodium falciparum and Tocxoplasma gondii. The present invention is further illustrated by the following examples which should not be construed as limiting in any way. The contents of all cited references (including literature references, issued patents, published patent applications as cited throughout this application are hereby expressly incorporated by reference. The practice of the present invention will employ, unless otherwise indicated, conventional techniques which are within the skill of the art. Such techniques are explained fully in the literature. See, for example, Molecular Cloning A Laboratory Manual, 2nd Ed., ed. by Sambrook, Fritsch and Maniatis (Cold Spring Harbor Laboratory Press: 1989); DNA Cloning, Volumes I and II (D. N. Glover ed., 1985); Oligonucleotide Synthesis (M. J. Gait ed., 1984); Mullis et al. U.S. Pat. No. 4,683,195; Nucleic Acid Hybridization (B. D. Hames & S. J. Higgins eds. 1984); Transcription And Translation (B. D. Hames & S. J. Higgins eds. 1984); B. Perbal, A Practical Guide To Molecular Cloning (1984); the treatise, Methods In Enzymology (Academic Press, Inc., N.Y.).

EXAMPLE 1 0041 MALDI Mass Spectrometry of Nucleic Acids Ranging From a 70-mer to a 2180-mer 0042 Materials and Methods 0043 Samples 0044 Synthetic oligodeoxynucleotides were obtained from Pharmacia Biotech (Uppsala, Sweden). The 70-mer was FPLC-purified by the supplier, while the smaller ones were used without additional purification. Plasmid DNA was purified from the E. coli strain DH5a by use of the Qiagen midiprep kit (QIAGEN GmbH, Hilden, Germany) according to the manufactures recommendations. Restriction enzymes were obtained from New England Biolabs GmbH (Schwalbach/Taunus, Germany); restriction enzyme digests of plasmid DNA were performed according to the supplier's protocols. Samples intended for MALDI-MS analysis were adjusted to 10 mM EDTA and 2 M NH4-acetate, and precipitated with 2 volumes of ethanol. The pellet was washed once with 70% ethanol and dissolved in water to an approximate concentration of 0.5 pmol/L. 0045 The 1206 nucleotide in vitro transcript was synthesised and ethanol precipitated according to standard procedures (Kirkepar, F. et al., (1994) Nucleic Acids Res. 22: 3866-3870), using the restriction enzyme Scal digested plasmid pBluescript KS as template for the T3 RNA polymerase (MBI Fermentas, Vilnius, Lithuania). A 10 L Poros 50 R2 (PerSeptive Biosystems, Framingham, Mass.) reverse phase column was prepared and equilibrated with 3% acetonitrile/10 mM triethyl ammoniumacetate (TEAA) as described elsewhere (Kussman, M. et al., (1997) J. Mass. Spectrom. 32: 593-6010. The RNA sample was adjusted to 0.3 M TEAA and loaded onto the column. The column was washed with 200 L 3% acetonitrile/10 mM TEAA, and the sample was eluted with 10 uL 25% acetonitrile/10 mM TEAA. Subsequent to lyophilisation, the eluate was dissolved in 5 L water; the estimated sample concentration was 1 pmol/L. 0046 A crude DNA preparation from mycoplasma-infected HELA cells was made, and a PCR performed essentially as described (Hopert, A. et al., (1993) J. Immunol. Methods 164: 91-100) using the primers 5-CGC CTG AGT AGT ACG TTC GC-3 (SEQ ID NO. 1) and 5-GCG GTG TGT ACA AGA CCC GA-3 (SEQ ID NO. 2), and recombinant Taq DNA polymerase (MBI Fermentas, Vilnius, Lithuania). The PCR results in an approximate 515 bp DNA fragment originating from the 16S rRNA gene of mycoplasma (Hopert, A. et al., (1993) J. Immunol. Methods 164: 91-100); the precise length of the PCR product cannot be predicted because the species of the mycoplasma is unknown. A reamplification by PCR was performed under identical conditions using the same primer set, and the final product was adjusted to 4 mM EDTA/2M NH4-acetate, and precipitated as described for the restriction enzyme digests. The pellet was dissolved in 200 L water and purified over a Microcon-100 (Amicon GmbH, Witten, Germany) microconcentrator, by three successive diafiltrations with 100 L water as recommended by the manufacturer. The retenate was lyophilised and re-dissolved in water to a concentration of 0.6 pmol/L as determined by UV spectrophotometry. 0047 Sample Preparation 0048 For IR-MALDI, glycerol was used as the matrix. The glycerol was incubated with an equal volume of a H-cation exchange bead suspension (Dowex 50W-X8. Biorad AG, Munich, Germany) in order to reduce subsequent alkali salt formation of the nucleic acid backbone phosphates. Typically 0.5-1 L of glycerol were mixed with an equal amount of an aqueous analyte solution on the target to give a final analyte-to-glycerol molar ratio of the sample of about 104-107, depending, on the mass of the analyte. The mixture was smeared out evenly over an area of ca.1-2 mm2 to form a homogeneous, transparent thin layer on the stainless steel substrate. The water was evaporated off at a pressure of typically 102-1 Pa, before sample introduction into the mass spectrometer. 0049 Samples for UV-MALDI-MS were prepared by on-target mixing of 1 L of a 105 to 106 M aqueous analyte solution with 0.7 L of a 50 g/l 3-hydroxypicolinic acid (3HPA) solution in 20% acetonitrile. About ten ammonium-loaded cation exchange beads were added to the samples before drying in a cool stream of air (Nordshoff, E. et al., (1992) Rapid Commun. Mass Spectrom. 60: 771-776). 0050 Instrumental 0051 The experiments were carried out with an in-house built MALDI single stage reflectron time-of-flight (TOF) mass spectrometer of 3.5 m equivalent flight length (Berkenkamp, S. et al., (1997) Rapid Commun. Mass Spectrom. 11:1399-1406, which is set forth as Example 2). The mass spectrometer can also be used in the linear TOF mode. Unless specifically mentioned, the experiments reported here have been carried out in reflectronand positive ion mode. Ions are accelerated through a total potential difference of about 16-25 kV in the split extraction source using either static or delayed ion extraction (DE). A Venetian blind secondary electron multiplier (EMI 9643) with a conversion dynode, mounted 10 mm in front of the cathode (ion impact energy of about 20-40 kV, depending on ion mass) or a Chevron Micro-Channel plate (Galileo Co., Sturbridge, Mass., USA) are used for ion detection. For high mass ions, the potential between the conversion dynode and the electron multiplier cathode is set to several thousand volts in order to increase the ion signal by making efficient use of the secondary ions. If maximum mass resolution is sought in the mass range up to several thousand Daltons, the potential between the two electrodes is kept below about 500 V in order to detect secondary electrons only and thereby avoid the time (and mass) dispersion of the secondary ions (see e.g. FIG. 2a). Signals are processed by a transient recorder with a time resolution of about 0.5 ns (LeCroy 9350). The digitized data are transferred to a PC for storage and further evaluation. For IR-MALDI experiments, an Er-YAG-Laser emitting at 2.94 m (Spectrum GmbH, Berlin, Germany; T80-90 ns, energy stability ca. 2-4% from shot to shot) was used. A frequency tripled Nd-YAG laser, emitting in the UV at 355 nm (TZ6 ns) was used for direct comparison between IR- and UV-MALDI. Single laser pulses are focused to a spot diameter of ca. 150 m (IR) and 100 m (UV) on the sample under an angle of 45. Samples are observed in situ with a CCD camera of about 5 m resolution. 0052 Results 0053 UV-MALDI spectra of DNA with at least about 50 nucleotides and with a reasonable quality could be obtained only in the linTOF, delayed-ion-extraction (DE) mode. FIGS. 1a and 1b demonstrate the striking differences in spectra quality for the two modes of operation for a synthetic DNA 70-mer (ca. 21.5 kDa) and a 3HPA matrix (355 nm). The quality of the spectrum of FIG. 1b, obtained in reflectron mode is quite inferior to that of FIG. 1a in several respects. Signal intensity as well as signal-to-noise ratio are considerably degraded as is the mass resolution, down to 15 (M/m; FWHM) from 65 in the spectrum of FIG. 1a. The saturated signal in the mass range below approximately 2000 Da in FIG. 2b reflects the increased laser fluence necessary to obtain analyte signals of the intensity shown. The loss in mass resolution is, for the most part, a result of the sloping low mass edge of the peaks, signaling abundant metastable small neutral losses. Exact mass determination is severely compromised by the loss of spectral quality. The IR-MALDI spectrum (refTOF, DE mode) of the same DNA 70-mer with glycerol as matrix is shown in FIG. 1c. The quality of this spectrum is comparable to UV-MALDI analysis obtained in the linear mode with respect to signal intensity and mass resolution (FIG. 1a). The base peak has a steeply rising low mass edge, demonstrating an essential absence of any metastable small neutral loss. This behaviour was consistently observed for IR-MALDI of nucleic acid with glycerol as a matrix, qualifying it as a very gentle desorption method forming ions of nucleic acids of high ion stability. This contrasts strikingly to the IR-MALDI spectra of nucleic acids obtained with succinic acid as matrix (Nordhoff, E. et al., (1993) Nucl. Acids Res. 21: 3347-3357; FIGS. 1(d) and 1(c)). The absence of literally all metastable neutral loss for the glycerol matrix, therefore, was a very unexpected result not anticipated based on prior experience (See the Background of the Invention). 0054 This leads to a broad mass range for the analysis of nucleic acids, from small oligonucleotides up to more than 2000 nt. as demonstrated in FIG. 2. A refTOF mass spectrum of a synthetic DNA 21-mer is shown in FIG. 2a. With delayed ion extraction a mass resolution of 1050 (FWHM) is obtained, comparable to the resolution obtained with the instrument for proteins in this mass range. Several poorly resolved peaks on the high mass side of the analyte peak appear in the spectrum. They are detection artefacts of residual secondary ions generated at the conversion dynode operated here in a mode to preferentially detect only secondary electrons in order to not degrade mass resolution by the ion detection system. FIG. 2b demonstrates the high mass range with a restriction enzyme digest of a plasmid (pBluescript-KS digested with BgIl and Rsal), yielding four fragments of 280 bp, 360 bp, 920 bp, and 1,400 bp. All four signals represent single strands and are the composite signal of the two complementary strands. Very weak, if any signals of the double stranded oligomers are apparent in this spectrum. Tentatively, the dissociation of the double strands in samples prepared with purified glycerol is attributed to an acidification by the H ion exchange resin. Not enough experience has, however, been accumulated so far to identify all additional parameters determining double strand dissociation under IR-MALDI conditions. The mass resolution of all high mass ion signals is about 50 (FWHM) and appears to be relatively independent of the ion mass. The IR-MALDI mass spectrum of FIG. 1c shows the upper mass limit measured so far for a restriction enzyme digest (130 bp, 640 bp, and 2,180 bp). The signal of the 2,180 nt ss-fragment was obtained only after heating the restriction digest to a temperature of 95 C. for 5 minutes. Such large DNA fragments apparently do not get separated into single strands under the conditions used, in contrast to the samples up to 1400 bp. The relatively poor mass resolution of ca. 30 for the 2,180 nt fragment in this spectrum and the strong background signals indicate an upper mass limit for IR-MALDI-MS of nucleic acids of approximately 700 kDa under the current conditions. Accordingly, the double stranded 2,180 nt fragment was not observable. 0055 IR-MALDI-MS of large RNA is also possible as shown in FIG. 2d for an RNA 1206 nt in vitro transcript. The increased ion-stability for RNA compared to DNA, well documented for UV-MALDI, was not observed in IR-MALDI within the mass range tested in these experiments. Large DNA ions as well as large RNA appeared to be of comparable stability, stable enough even for TOF analysis in the reflectron mode. The large hump, centered at about 50 kDa is believed to reflect impurities of the sample rather than metastable fragments. The comparably steep rise of the peak at the low mass side also testifies to a very limited loss of small neutrals such as single bases. 0056 One advantage of glycerol as matrix is the superior shot to shot reproducibility and mass precision (200-400 ppm) (Nordhoff, E. et al., (1993) Nucleic Acids Res. 21: 3347-3357. These values, originally determined for proteins, are also valid for the analysis of smaller oligonucleotides. However, mass accuracy was found to be mass dependent. Using, an external 2 point calibration with angiotensin II (1047 Da) and bovine insulin (5743 Da) the mass of the 21 mer (6398 Da) in FIG. 2a was determined to within 2 Da of the known mass, i.e. an accuracy of 0.03%. The molecular mass of the 70mer (theoretical mass: 21517 Da) was determined to within 25 Da i.e. a mass accuracy of 0.1% from the spectrum of FIG. 1c, calibrated with cytochrome C oligomers. (M, 2M, 3M). 0057 For all of the ten different samples of high mass DNA analyzed, the measured mass was within less than about 1% of the theoretical mass derived from the sequence (see e g. FIGS. 2b and 2c). The average mass of the two single strands was used as the theoretical mass in the case of DNA restriction enzyme fragments. The masses of the two single strands never differed by more than about 1%. Only one large mass RNA has been measured so far (FIG. 2d). The measured mass of this RNA is 388,270 Da, whereas the mass calculated from the gene sequence is 386,606 Da. Given that the sample most likely is a heterogeneous mixture of the species expected from the gene sequence with less abundant products extended by one to three extra nucleotides (Melton, D. A. et al., (1984) Nucleic Acids Res. 12: 7035-7056), the actual mass of the RNA sample is probably about 500 Da larger than the one calculated from the sequence. It would therefore appear as though a mass accuracy of at least about 1% as observed for DNA, can also be achieved for RNA. 0058 For external 4 point calibrations of large DNA/RNA with molecular masses between 100-400 kDa, either clusters of cytochrome C (e.g. 10M, 20M, 30M, 40M) or multimers of an IgG monoclonal antibody (e.g. 2M3M, 4M) were used. For analytes exceeding 500 kDa the calibration with IgG monoclonal antibody was found to be most exact. Mass calibration of unknown DNA fragments using DNA or RNA calibrants may be more desirable, resulting in a more accurate mass determination. 0059 Experiments to evaluate the sensitivity of IR-MALDI-MS of large nucleic acids with glycerol as matrix have been carried out with a PCR-product of approximately 515 nt and unknown sequence. Its mass was measured to 318,480 Da. For these measurements, glycerol, not subjected to ion exchange purification, was used. The spectra show dominant signals of the double stranded moiety. Tentatively the dissociation of the double strands in samples prepared with purified glycerol is attributed to an acidification of the glycerol by the protons exchanged for the cations. Although additional parameters may be involved in the double strand dissociation under IR-MALDI conditions. The starting concentration for the dilution experiment was 0.6 pmol/L as determined by UV spectrophotometry. The mass spectra in FIG. 3 were obtained by loading different amounts of sample onto the target. For the single shot mass spectrum in FIG. 3a, 300 fmol of the PCR-product had been loaded. The quality of this spectrum with an S/N-ratio better than 100 and a mass resolution of 65 (FWHM) for the double-strand indicates that the analyte to matrix ratio (A/M) of about 107 is well suited for an analyte of this size (about 320 kDa). 0060 The mass spectrum in FIG. 3b was obtained using a 3 fmol total load (A/M about 2109). A strong background signal now dominates the low mass range. Total signal intensity, mass resolution (of about 25 FWHM for the ds-ion signal) and S/N-ratio are significantly degraded compared to FIG. 3a. However, mass determination is still possible with an accuracy of about 1%. The spectrum in FIG. 3c has been obtained from a very small sample volume forming an approximately 270 m diameter sample spot on the target and a total sample load of only 300 amol (A/M about 81010). Such small sample volumes can be realized by either dispensing the small volumes by micropipettes as described in the literature (See e.g. Little, D. P., (1997) Anal. Chem. 69: 4540-4546) or by preparing the analyte in a standard microliter volume of a suitable glycerol/water mixture. In the latter case, the water is then evaporated off prior to or upon insertion of the sample into the vacuum. The poor mass resolution of only about 10 classifies this amount of analyte as the limit for the given instrument and detection system for a mass accuracy of better than about 3%. Compared to values reported for UV-MALDI-MS (Tang, K. et al., (1994) Rapid Commun. Mass Spectrom. 8: 727-730; and FIGS. 5 and 6), the sensitivity demonstrated here for IR-MALDI-MS demonstrates an improvement of at least about 2-3 orders of magnitude for nucleic acids of this size.

EXAMPLE 2 0061 Performance of IR Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry 0062 The performance characteristics of two lasers emitting in the mid infrared, an Er-YAG (2.94 m wavelength, 80-90 ms pulse width), and an Er-YSGG infrared laser (2.79 m wavelength, 80 ns pulse width), in matrix-assisted laser desorption/lionization mass spectrometry (IR-MALDI-MS) of biological macromolecules, was studied. Glycerol and succinic acid were used as matrices. In IR-MALDI sample consumption per laser shot typically exceeds that of UV-MALDI by about two orders of magnitude. Using glycerol as matrix, the reproducibility of the ion signals from shot to shot is comparable to the best values achieved in UV-MALDI. The same holds true for the precision and accuracy of the mass determination. For succinic acid all these values are significantly worse, due to the strong sample heterogeneity as typically found in dried droplet preparations. Metastable fragmentation is comparable for UV- and IR-MALDI in the law mass range, but is significantly less for the IR in the mass range above ca. 20 kDA, leading to an improved mass resolution and an extended high mass limit for IR-MALDI. 0063 In this Example (reproduced from Berkenkamp, S. et al., (1997) Rapid Commun. Mass Spectrom. 11:1399-1406 and incorporated by reference herein and in copending U.S. application Ser. No. 09/074,936), results and performance data for IR-MALDI analysis obtained with ER lasers emitting at 2.94 m and 2.79 m, and the applicability of delayed extraction for an improved mass resolution, is presented. In particular, is demonstrated that the extent of metastable fragmentation is different for the mass resolution of high molecular mass analytes. 0064 Experimental 0065 The experiments were carried out with an in-house built, single stage reflectron TOF mass spectrometer of 3.5 m equivalent flight length. The mass spectrometer can also be used in linear mode. Unless specifically mentioned, the experiments reported here have been carried out in reflectron mode. In the split extraction source, ions are accelerated through total potential differences of 12-20 kV using either static or delayed extraction. In the delayed mode extraction, a maximum potential difference of 8 kV can be switched; the minimum delay for ion extraction is 120 ns. No arcing was observed under these operation conditions in the positive or negative ion modes for any of the matrices used. A venetian blind secondary electron multiplier. (EMI R2362) with a conversion dynode, mounted 10 mm in front of the cathode (ion impact energy 20-27 kV, depending on ion mass), or a Chevron microchannel plat detector (Galileo Co., Sturbridge, Mass., USA), are used for ion detection. Signals were processed by a transient recorder with a time resolution of 2.5 ns (LeCroy 9450A) in the majority of the experiments. For the high mass-resolution experiments a LeCroy 9348La recorder with a time resolution of 0.5 ns was used. The data are transferred to a PC for storage and further evaluation. The instrument is equipped with two infrared lasers, one emitting at 2.94 m (Er-YAG:Fa.Spectrum GmbH, Berlin, Germany: T80-90 ns, energy stability ca. 2-4% from shot to shot) and a second radiating at 2.79 m (Er-YSGG: Schwartz Electro Optics. USA: T80 ns; energy stability ca. 2%). A frequency tripled Nd-YAG laser, emitting in the UV at 355 nm (T16 ns) is used for direct comparison between IR- and UV-MALDI. Single laser-pulses are focused to a spot diameter of ca. 200 m (IR) and 100 m (UV) on the sample under an angle of 45. Samples are observed in situ with a CCD camera of ca. 5 m resolution. The stainless steel substrate can be cooled with liquid nitrogen to a temperature of ca. 150-170 K. Its temperature is monitored by a thermo-couple with an accuracy of 5 K. 0066 Sample Preparation 0067 A wide variety of small molecules can be used as matrices in IR-MALDI as described previously. Succinic acid (solid matrix) and glycerol (fluid or frozen, solid) were prepferred. DHBs (2,5-dihydroxybenzoic acid mixed with 10% 2-hydroxy-5-methoxybenzoic acid), a common matrix in UV-MALDI also functions in IR-MALDI and was used for comparison in some cases. Additionally, mixtures of compounds, e.g. succinic acid/DBHs or succinic acid/TRIS (Tris-hydroxymethylaminomethane) were found suitable. Solid matrix samples were prepared using the standard dried droplet method by mixing ca. 1-3 l of a 104 105 M aqueous analyte solution with 1-5 L of a 30 g/L matrix solution on the target and subsequently drying in a stream of cool air. 0068 For glycerol, analytes were either dissolved directly at a concentration of 0.5-10 g/L, or the glycerol was mixed with an aqueous analyte solution to a final analyte-to-glycerol molar ration of 2104-1105, depending on the mass of the analyte. A volume of typically 1 L is applied to the stainless steel substrate and smeared out evenly over an area of ca. 3-4 mm2 to form a homogeneous, transparent thin layer. If an aqueous analyte solution is mixed with the glycerol, the water must be evaporated off before sample introduction into the mass spectrometer, usually at a pressure of 102-1 Pa. Samples are either inserted directly into the mass spectrometer or are cooled down to a temperature of ca. 150-170 K in liquid nitrogen before insertion. 0069 For the matrices investigated, IR-MALDI was found to be quite tolerant with respect to salts and buffers. Spectra of samples containing NaCl at concentrations of up to 200 mM, saccharose up to 20% (w/v) or Tris/HCl buffer up to 100 mM have, for example, been obtained without significant loss in spectral quality with succinic acid as well as glycerol. 0070 Results 0071 Sample Consumption and Analytical Sensitivity 0072 The analytical sensitivity of MALDI can be limited by either the minimal concentration of the analyte solution used for the analysis or the total amount of analyte available. For the actual measurement the total sample volume used for the preparation and the analyte-to-matrix ratio in the sample can be adjusted within certain limits to accommodate a given situation. In this section typical, as well as limiting, values for these quantities in IR-MALDI is presented and compared to the corresponding UV-MALDI values. As an introduction a few general differences between IR- and UV-MALDI will be discussed. 0073 Given comparable laser spot sizes on the sample, 100-1000-times more material is desorbed by the IR as compared to the UV laser beam, because of the correspondingly smaller absorption coefficient and higher penetration depth of the radiation into the sample. Under the experimental conditions described here the single shot ion signals in the IR and UV typically were of comparable intensity. With that much more material ablated per laser pulse in IR-MALDI, the material is either primarily removed as larger clusters and small particles, as actually observed experimentally, and/or the ion yield is smaller by two to three orders of magnitude. 0074 The intensity of low mass signals, the signal-to-noise ratio and mass resolution are the main criteria for the quality of recorded mass spectra. The optimal molar analyte-to-matrix ration on the target for the signal intensity and quality of spectra depends on the molecular mass of the analyte. For analytes with molecular masses below ca. 50 kDa a ratio of 2104-2105 was found to be optimal. A ratio of 105 was found to work best for analytes with masses exceeding 50 kDa. Hence, in particular in the low mass range, the optimal analyte concentration in IR-MALDI exceeds that typically used in UV-MALDI by approximately one to two orders of magnitude. Spectra of reasonable quality can be obtained for analyte-to-matrix ratios down to 2106-107, depending on molecular mass of the analyte. This corresponds to a sample consumption per laser shot of 1-50 fmol. In routine applications, spectra of IgG monoclonal antibodies (ca. 150 kDa) have been obtained with reasonable quality using 107 M aqueous solutions using either succinic acid or glycerol as matrix. This compares to values of only a few attomole in UV-MALDI. 0075 The results demonstrated the attainable sensitivity for a lysozyme/glycerol preparation. Here, 5.5103 L of a 1.2107 mol/L frozen glycerol solution, corresponding to a total amount of prepared lysozyme of ca. 0.7 fmol, were used for the preparation. This corresponds to a molar A/M-ratio of ca. 108. The frozen sample on the target had a diameter of about 1.5 times that of the laser spot on the sample. After the ten laser exposures summed for the spectrum some sample was still left on the target. Consequently, sample consumption of average was less than 70 attomol per single spectrum. However, when using such low A/M ratios the mass resolution is down at ml/m10-20. The signal-to-noise (S/N) ratio has also degraded substantially and the low mass background signals become excessive. For UV-MALDI a sample consumption in the high zeptomole range has been reported (see, Jespersen et al. (1996) p. 217 in Mass Spectrom. in Biol. Sci, Burlingame, Ed.). 0076 In IR-MALDI there is also a pronounced dependence on the A/M ratio of the yield of (non-specific) analyte oligomers or multiply charged ions. These tendencies are more pronounced in IR- as compared to UV-MALDI. A mass spectrum of hen egg lysozyme from a preparation with a A/M-ration of 2104 was obtained. Homo-oligomers of lysozyme up to the 25th mer (ca. 500 kDA) were identified in this spectrum. Conversely, signals of multiply charged analyte ions become dominant in the spectra for A/M-ratios below a value of ca. 106 whereas oligomer signals decrease to values below the noise level. These trends have also been observed with succinic acid and the water of hydration as matrices ( ) Sadeghi (1997) Rapid Commun. Mass. Spectrom. 11:393), and are particularly pronounced for analytes with molecular masses exceeding 50 kDa. Given the observed high yield of analyte homo-oligomers, the details of the distribution in particular the most abundant oligomer signal, can be influenced significantly by changes in the ion extraction conditions, e.g. by using low (soft or mild) or high (hard or harsh) ion extraction fields in a two stage ion source. Low extraction fields will shift the distribution towards a higher degree of oligomerization. This observation is a strong indication of gas phase processes in the expanding desorption plume besides possibly reflecting differences in the acceptance of the spectrometer under different extraction conditions. 0077 Reproducibility 0078 For Er-YAG lasers presently used, the pulse shot-to-shot stability is ca. 2%, a value quite comparable to that of the best UV lasers, and, thus, does not contribute substantially to signal variation. Laser energy fluctuations from shot to shot, therefore, play only a minor role with these lasers. Another source of signal variation from shot-to-shot is the sample homogeneity, which depends on the matrix and the preparation technique. Solid matrices such as succinic acid typically form heterogeneous microcrystalline patterns. High quality mas spectra can only be achieved from sweet spots. For some matrices like DHB (2,5-dihydroxybenzoic acid mixed with 10% 2-hydroxy-5-methoxybenzoic acid) this is also true for UV-MALDI. In UV-MALDI 50-100 spectra can be obtained from any sweet spot, in contrast to only 3-4 spectra from a given sweet spot in IR-MALDI, because of the much larger sample consumption per exposure. 0079 Liquid matrices such as glycerol form very homogeneous layers and spectra of comparable quality can be obtained from all locations across the sample. Also, the surface of these liquid samples recovers after every laser shot and more than 500 spectra of almost identical quality can be obtained from the same spot of a typical penetration. The spectrum of hen egg lysozyme was obtained after more that 250 shots on the same location at a laser repetition rate of 2 Hz. No significant differences in signal intensity, mass resolution, or S/N ratio as well as oligomer distribution, were observed between the early and late exposures. As a result, reproducibility of IR-MALDI spectra was found to be comparable to, if not better than, that for UV-MALDI preparations if glycerol is used as matrix. For solid matrices such as succinic acid reproducibility of ion signals from shot to shot may become a problem and considerable experience of the operator is often required for good results. 0080 Fragmentation 0081 Fragmentation is also an important parameter in MALDI-MS. Generally the yield of so call prompt fragments, generated during the desorption and ionization process on a time scale short compared to the ion extraction time, is very low in UV-as well as IR-MALDI. These fragments are detected at their true (fragment) mass in linear as well as reflecting time-of-flight (TOF) spectrometers. This is also true under delayed extraction with delay times of ca. 1 s or less. In IR-MALDI, a somewhat increased yield for such prompt fragments of oligonucleotides has been observed but the overall yield was very low nonetheless. Metastable ion decay, on the microsecond to hundreds of microseconds time scale in the field free region of the TOF mass spectrometer is much more important. On the one hand it degrades mass resolution in reflectron (reTOF) instruments, but on the other hand it can be used for structural analysis in the post-source decay (PSD) mode. No significant differences between UV- and IR-MALDI have so far been found for the metastable fragmentation of peptides and proteins in the mass range up to ca. 20 kDa. For analytes with a molecular mass above 20 kDa a markedly different metastable fragmentation has been observed. reTOF spectra of an IgG monoclonal antibody (mouse, MW ca. 150 kDa) obtained with UV-(matrix: DHBs) and IR-MALDI (matrix: succinic acid; matrix: glycerol) were obtained. The base peak of the parent ion had a rather symmetrical shape in both IR-MALDI spectra, whereas it showed a strong tail on the low mass side in the UV-MALDI spectrum, testifying to a significant amount of metastable decay. The peak width full width at half maximum (FWHM) decreases from a value of 2000 Da in the UV spectrum to values of 1000 Da in the IR spectrum with succinic acid as matrix, and merely 700 Da if glycerol is used as matrix. It was also noticeable that the anlayte signals ride on a substantially elevated baseline in the UV-MALDI spectrum which results from delayed fragmentations somewhere within the ion source. No such baseline distortions were observed in the IR-MALDI spectra if the desorption fluence remained within a range of ca. 1-1.5 times ion detection threshold fluence. 0082 It has been reported in the literature that in UV-MALDI the degree of metastable decay increases significantly with degrading source back pressure. Two spectra of an IgG monoclonal antibody (mouse) weree obtained with (a) UV- and (b) IR-MALDI at a source back pressure of 4103 Pa as compared to a back pressure of 4104 Pa used to obtain the spectra described above. The UV-MALDI spectrum exhibited signals with a substantially increased tailing to the low mass side, particularly visible for the oligomers 2M and 3M2; whereas no such tailing was seen in the IR-MALDI spectrum. 0083 Another observation is the dependence of metastable fragmentation on the analyte-to-matrix ratio. In reTOF UV-MALDI, too high a A/M-ratio will usually result in a degraded S/N ratio and a loss of mass resolution. This was shown in with spectra for cytochrome c and DHB as the matrix. A higher A/M-ratio of 103 used to obtain the spectrum, resulted in a strong, low-mass tail of the peaks, again stronger for the dimer as compared to the parent ion peak, but not seen in the spectrum obtained from a sample with an A/M ratio of 104. The IR-MALDI spectrum shown of the same sample used showed no such tailing, again indicating substantially less metastable fragmentation. For IR-MALDI of cytochrome c using the water of hydration as intrinsic matrix the molar A/M ratio is even higher (ca. 5103), yet no significant metastable decay was observed (Berkenkamp et al. (1996) Proc. Natl. Acad. Sci. U.S.A. 93:7003), 0084 It is a generally held notion that collisions of ions with matrix neutrals in the plume and with residual gas molecules in the spectrometer are the major cause of metastable fragmentation (see, e.g., Spengler et al. (1992) J. Phys. Chem. 96:9678). Considering that much more material is desorbed in IR versus UV-MALDI, resulting presumably in a more extended plume, and that in addition proportionally more of the absorbed laser energy goes into the analyte molecule in the IR, the finding of much less metastable fragmentation in IR-MALDI under all the different conditions presented above was not expected. Contrary to intuition, IR-MALDI seems to be a milder method than UV-MALDI. 0085 Accessible Mass Range 0086 The lower degree of fragmentation gives IR-MALDI an advantage over UV-MALDI for the analysis of very high mass analytes, particularly when an ion mirror is used. Not only does this lead to stronger signals of large parent molecular ions, it also, and more importantly, allows the use of higher laser fluences up to about twice the ion detection threshold fluence without deterioration in spectral quality as would be the case in UV-MALDI under such conditions. This increases the high mass signals even further. 0087 A spectrum of gramicidin-S-synthetase of mass 510 kDa, prepared in glycerol matrix from an aqueous solution containing 50 mM Tris/HCL, 18% (w/v) saccharose and 5 mM dithiothreitol, with a quite acceptable S/N ratio and mass resolution of m/m50 was obtained. No signals of this analyte could be obtained with UV-MALDI under a variety of conditions tried. 0088 A mass spectrum of an IgG monoclonal antibody (mouse) demonstrated that IR-MALDI in combination with a TOF mass analyser can be used for the analysis of biomolecules with molecular weights exceeding 1 MDa. Multiply charged ions of the 13-mer homo-oligomer of ca. 2 MDa mass could unambiguously be identified in the spectrum and signals of ions of other oligomers with m/z values as high as 900 000 were also clearly identified in the spectrum. 0089 Mass Resolution 0090 Delayed ion extraction (DE) is used for enhanced mass resolution in UV-MALDI-TOF MS. It was not immediately clear whether DE would be as advantageous in IR-MALDI as well, considering the much higher desorbed sample amount and the possibly substantially different plume expansion dynamics. In fact, for peptides in the 1000 Da mass range, the mass resolution is only about 200 (FWHM), down from about 1000 for UV-MALDI under otherwise comparable conditions, indicating difference in the ion generation process. Nonetheless DE gave equal mass resolution for IR- and UV-MALDI within the accuracy of the measurement. This was demostrated with a reTOF spectrum (sodiated gramicidin-s, MW 1164.5 Da) obtained with the Er-YAG laser at 2.94 m and 80 ns pulse width and succinic acid as matrix. The mass resolution in this spectrum was 1000, corresponding to a width of the individual peaks of 3.5 ns, limited by the time resolution of the dual MCP detector (3.0 ns). Using a reTOF for Mellitin (2846 Da) a mass resolution of 9500 and one of 1500 for cytochrome c (12360 Da) were obtained. Thus an enhancement in resolution by factors of ca. 50 for peptides and of 4 for cytochrome c was achieved. 0091 In the high mass range mass resolution in the linear TOF with static ion extraction at 20 kV is limited to a value of ca. 50, equal for IR- and UV-MALDI. In both cases the mass resolution is determined by the distribution of initial ion velocities and kinetic energies. Using DE the mass resolution could be improved by a factor of 3 to equal values of ca. 150 for IR- as well as UV-MALDI. 0092 For analytes exceeding 50 kDa mass resolution in IR-MALDI can, however, be improved even more with a reTOF analyser, in contrast to UV-MALDI. As demonstrated by the spectra discussed above (reTOF spectra of an IgG monoclonal antibody) the strongly decreased metastble fragmentation in IR-MALDI resulted in a peak width of only 700 Da for the parent ion peak of a monoclonal antibody desorbed with the IR-laser out of a glycerol matrix, compared to a peak width of ca. 2000 Da for the UV-MALDI spectrum. The peak width of 700 Da, corresponding to a resolution of about 200, was the best obtained for all experimental conditions tested. Using either a matrix other than glycerol, or an Er-YAG laser with a longer pulse width (120 ns instead of 90 ns) or switching to the Er-YSGG laser at a wavelength of 2.79 m and 80 ns pulse width, led to slightly larger peak widths. No significant improvement in mass resolution was observed for the IR-MALDI reTOF spectra using DE for large masses, in agreement with observations made in UV-MALDI. Electrospray mass spectra of an IgG monoclonal antibody10 show that the peak width of 700 Da reflects the peak envelope of the various glycosylation states of the molecule. The instrumental mass resolution has, therefore, been even somewhat higher. This assumption was supported by a spectrum of chondroitinase in which the peak of the singly charged dimer was observed at mass 224 kDa with a peak width of 700 Da, corresponding to a mass resolution of 300. 0093 Mass Accuracy and Precision 0094 Similar to the reproducibility of the intensity of ion signals in IR-MALDI, precision of the mass determination as given by the standard deviation of sequential measurements depends on the matrix (sample morphology) as well as on the shot-to-shot variation of the laser pulse energy. 0095 For prompt ion extraction and solid matrices, such as succinic acid and dried droplet preparations, mass precision is typically 400-500 ppm for molecular weights up to 150 kDa. It is limited by the strong heterogeneity of the sample morphology and the need for a frequent change of the desorption location on the sample as described above. For prompt extraction and liquid matrices such as glycerol, forming very homogeneous layers, variations in total flight time resulting from variations of the laser pulse energy determines the precision of the mass determination. If, for example, the laser energy is raised intentionally from threshold (Io) to 1.51 Io, an increase in flight time of ca. 0.1% for cytochrome c (total flight time ca. 253 s) was observed. Thus for glycerol the precision of mass determination depends on the stability of laser output energy. For IR lasers of current design with glycerol as matrix, a precision of the mass determination of 200-400 ppm can be achieved up to a mass of approximately 150 kDa. For analytes below 30 kDa this precision is lower by about one order of magnitude than the values typically obtained in prompt extraction UV-MALDI. In the high mass regime, precision in IR-MALDI was found to be better by at least a factor of 2, most likely due to the enhanced mass resolution of IR-MALDI in this mass range. 0096 The mass accuracy for prompt extraction IR-MALDI was determined by external calibration with 3 well known standards. (Low mass range: angiotensin (human), mellitin, bovine insulin; high mass range: cytochrome c (horse heart), apo-myoglobin (horse), subtilisin Carlsberg (bacillus subtilis)). In the mass range up to 30 kDa, 5 sum spectra of 15 single shots each were used to obtain the calibration factors from the calibration spectrum and the mass of the unknown in the second spectrum. For both matrices, succinic acid and glycerol, the absolute mass accuracy has been found to be 102-5103 ppm depending on molecular mass. For proteins up to 40 kDa the mass accuracy of 100-500 ppm is in good agreement with the previously described numbers for UV-MALDI (static extraction).20 For analytes exceeding 40 kDA accuracy is 1-5103 ppm.

CONCLUSIONS 0097 As judged by the lesser degree of metastable fragmentation compared to the UV, IR-MALDI appears to be the milder of the two techniques for generating biomolecular ions. Glycerol or an equivalent material is the matrix of choice for many applications because of its superior reproducibility in comparison to solid matrices. Among the two lasers tested in this study, Er-YAG laser performs slightly better than Ef-YSGG laser for glycerol and substantially better for succinic acid as matrix. The lesser metastble fragmentation makes IR-MALDI also particularly well suited for the analysis of high mass analytes in the reTOF mode. Delayed ion extraction works well with IR-MALDI, with results comparable to UV-MALDI.