1999 "Rigi" Meeting

Structure Mapping of Nucleic Acids by Tandem Mass Spectrometry

Pamela F. Crain
Depart. of Medicinal Chemistry, University of Utah
30 So. 2000 East, Rm. 311A
Salt Lake City, UT 84112, USA

Prior to the introduction of ESI and MALDI, mass spectrometric analysis of nucleic acids and their naturally occurring modified constituents was a challenging task owing to the polarity of the polyphosphate backbone. Prohibitively large sample amounts and/or digestion to monomers or very short, uninformative oligonucleotides were often required. Using these new ionization methods, it has now become possible to mass analyze these polyanions as polynucleotides of up to several hundred nucleotides (nt) in length.

Our laboratory is focused on the structure analysis of natural RNAs: what modified nucleosides among the 95 that are known (1) are present, and what are their sequence locations? Given such a large number of known modifications, mass spectrometry has become an ideal technique for unambiguous structure assignment compared with traditional chromatography-based methods. Analysis of natural RNA presents a special problem, however, unlike sequence analysis of DNA (which is ideally suited for MALDI-TOF). Small amounts of DNA can be readily amplified in vitro with high fidelity using the polymerase chain reaction (PCR), but the naturally occurring nucleoside modifications that are of interest will be lost if RNA is PCR-amplified. RNAs must therefore be analyzed as cell isolates, and high sensitivity is required for their analysis.

We use ESI and reversed phase LC/MS with a 1,1,1,3,3,3-hexafluoroisopropanol-based solvent system (2) for the analysis of mixtures of oligonucleotides generated from RNase-treated RNA (3,4). Using a triple quadrupole instrument, scan functions for product ion, precursor ion and multiple reaction monitoring, as required for the particular structure problem, can all be executed for intact oligonucleotides, and for fragment ions generated by in-source fragmentation at high cone (nozzle-skimmer) voltage. Examples to illustrate these approaches will be selected from current structure work on tRNA (75-85 nt) and 16S rRNA (~1500 nt) from thermophilic microorganisms, although they can be equally applied to DNA structure studies.

References
1. J. Rozenski, P.F. Crain & J.A. McCloskey.(1999) Nucleic Acids Res. 27, 196-197.
2. A. Apffel, J.A. Chakel, S. Fisher, K. Lichtenwalter & W.S. Hancock (1997) Anal. Chem. 69, 1320-1325.
3. B. Felden, K. Hanawa, J.F. Atkins, H. Himeno, A. Muto, R.F. Gesteland, J.A. McCloskey & P.F. Crain (1998) EMBO J. 17, 3188-3196.
4. P.F. Crain, D.E. Ruffner, Y. Ho, F. Qiu, J. Rozenski & J.A. McCloskey. In Mass Spectrometry in Biology and Medicine. A.L. Burlingame, S.A. Carr & M.A. Baldwin, eds. Humana Press, Totowa, NJ, 1999, pp. 531-555.

A new generation of accelerator mass spectrometers - smaller, simpler and cheaper

M. Suter, ETH Zürich

14C concentrations at natural levels were detected with accelerator mass spectrometry (AMS) for the first time in summer 1977. The method was based on accelerator technology developed for research in nuclear and atomic physics. Meanwhile dedicated facilities are available which are successfully applied in studies of various long-lived radionuclides in earth and climate research. In recent years, AMS has also been applied routinely for biomedical radiotracer studies (14C, 26Al, 41Ca). The main disadvantage of AMS is that the instruments are large and require special installations.

In recent years significant advances have been made towards AMS at sub-MeV energies, which allows to reduce the size of the installation significantly. This has become possible by applying a different technique to eliminate the interfering molecules. The new facilities can be placed in a regular size laboratory and their operation is much simpler. At ETH/PSI a prototype of such a system has been built. It demonstrates that radiocarbon dating can be performed with high performance. Other instruments specially designed for biomedical applications are also being developed by various accelerator companies.

A review of the new developments is given and the potential for application will be discussed.

MALDI Mass Spectrometry in Milk Science

Pietro Traldi
CNR, Area della Ricerca, Corso Stati Uniti 4, I-35100 Padova (Italy)

Milk is a highly-nourishing food owing to its contents of fats (mainly triglycerides contained in globules), sugars (essentially lactose), vitamins, mineral salts and proteins (1). In particular, proteins are very important from the nutritional point of view and a wide number of analytical methods have been applied for their characterization (1). By these approaches two main groups of protein have been identified in milk:

• caseins, which are the most abundant of the protein fraction and are present as aggregates. The main classes of caseins are: alpha s1, alpha s2, beta, k and gamma;
• whey proteins, such as alpha-lactalbumin, beta-lactoglobulin, serum albumin and proteoso peptones.

Protein composition varies during lactation (2), from one animal species to another (3) and can also be varied by thermal treatment employed for preserving the milk quality.

In view of the high power of matrix-assisted laser desorption/ionization (MALDI) mass spectrometry in the mapping of complex protein mixtures, we firstly applied this analytical method to the evaluation of protein content in milk after thermal treatments (4,5), from different origins and in the quality control of cheeses.

The MALDI spectrum of a raw bovine milk allows the detection of the most of milk proteins:

the protonated proteoso peptone pp.8.l leads to the ion at m/z 9170, while gamma 2 and gamma 3 caseins give rise to signals at m/z 11852 and 11595 respectively. Further abundant peaks are due to alpha-lactalbumin (m/z 14212), beta-lactoglobulin (m/z 14394), alpha s1 casein (m/z 23690) and beta casein (m/z 24081). Less abundant peaks are due to k casein (m/z 19122), gamma 1 casein (m/z 20085) and alpha s2 casein (m/z 25248).

In order to preserve the quality of milk, two kinds of thermal treatments are used: pasteurization and sterilization. Pasteurization is a thermal process which kills some of the pathogenic microorganisms and a temperature range of 72°-90° C for 10-30s is usually employed. Ultra-high temperature (UHT) sterilization of milk can be "direct" or "indirect": in the first case a vapour stream is directly mixed with milk, while in the latter case milk is sterilized through an indirect heat exchanger. The typical temperature employed in UHT treatments is in the range 138°-140° C for 2-5s.

The MALDI spectra of three pasteurized bovine milks of different producers were found quite different from the raw milk, demonstrating that the thermal treatment leads, to some extent, to a change in the protein content of milk. These changes are enhanced in the case of UHT milks.

In some cases the thermal treatment has been carried out so severely that caseins are the only proteins still detectable and has led to the denaturation of soluble whey proteins and to the decomposition of proteoso peptones.

Further studies (5) on milk samples pasteurized at a constant homogenization pressure ( 100 bar) and temperatures of 72°, 76°, 80°, 85° and 95° C show that: i) alpha s1 casein monomer is generated by thermal decomposition of a higher-molecular weight casein and ii) alpha-lactalbumin and beta-lactoglobulin are more stable than the other proteins up to 85° C, after which their relative abundances decrease. MALDI spectra obtained from milk samples pasteurized at constant temperature (76° C) but at homogeneization pressure of 0, 50, 100, 150 and 200 bar show that the increased pressure leads to severe degradation of lower molecular weight proteins and to an increase in the relative abundances of caseins.

Moreover, MALDI spectra of milk samples sterilized by different arrangements of equipment show that the milk protein content does not change in the "direct" system, but is severely degraded in the "indirect" system: beta-lactoglobulin practically disappeared in the latter case and there is a clear change in casein content (5).

We thought of interest to analyse milks originating from different mammalians, such as goat, mare, ewe and water buffalo, by MALDI technique.

Typical protein profiles have been obtained for each milk sample, differing from that of cow milk in the relative abundances and in the molecular weight of proteins. This is well in agreement with the genetic polymorphism showed by milk proteins.

Interestingly a specific protein (protein X) at m/z 15791 is detected only in the MALDI spectrum of water buffalo milk. This species has been already detected by other analytical procedures (6) and from HPLC data it was suggested to belong to the beta-lactoglobulins family. Taking into account this peculiarity, we thought of interest to analyse water buffalo mozzarella cheese in order to evaluate possible fraudulence in the mozzarella production, consisting in the addition of cow milk. The MALDI data resulted highly effective, allowing to determine the quantity of cow milk added during the cheese production.

Considering the interesting results obtained by MALDI in the analysis of mozzarella cheeses, we undertook an investigation on other kinds of cheese at different stage of ageing: "Pecorino" cheese, originating from ewe milk, and "Asiago" cheese, coming from bovine milk. The protein profiles of both kind of cheeses are completely different from those of original milks and, by ageing, a clear decrease in caseins content is observed. Thus, it has been possible to obtain an evaluation of the adulteration in "Pecorino" cheese production by adding bovine milk.

More recently the same analytical approach has been employed in the studies of:

1. bacterial protein digestion in yoghurt production;
2. protein profile of human milk;
3. protein profile of infant formulae.

Even in these cases MALDI mass spectrometry has led to interesting results, proving to be a valuable tool in milk science.

References
1. C.Alais, Scienza del Latte, Tecniche Nuove, Milan, Italy (1984)
2. E.A.Jones, J.Dairy Res. 36, 5 (1969)
3. R.A. Shaffenburg, J.Dairy Res. 35, 447 (1968)
4. S.Catinella, P.Traldi, C.Pinelli and E.Dallaturca, Rapid Commun. Mass Spectrom. 10, 1123 (1996)
5. S.Catinella, P.Traldi, C.Pinelli, E.Dallaturca and R.Marsilio, Rapid Commun. Mass Spectrom. 10, 1629 (1996)
6. L.Pellegrino, I.De Toni, A.Tirelli and P.Resmini, Scienza e Tecnica Lattiero-Casearia 42(2), 87 (1991)