Of the various analytical methods developed in microscopy during the latter half of the century, SIMS imaging is probably one of the most powerful and sophisticated. Originally introduced in the early sixties by Castaing & Slodzian,¹ this technique is based on the mass spectral analysis of secondary ions extracted from the surface of a solid sample under the impact of an energetic beam of primary ions.

This "ion microscopy" has been widely applied to surface analysis in geology, metallurgy and semi-conductor research and is now a central analytical tool in each of these fields. In spite of sporadic application of this technique in biology, starting 30 years ago,² SIMS microscopy has been considered only as a marginal method for solving problems in the life sciences, due mainly to poor lateral resolution (1-0.5 µm).

To deal with this issue, a concerted collaborative effort between the University of Paris-Sud, Orsay (Pr G.Slodzian), the French Space Agency (ONERA) (B.Daigne), and CAMECA, a high-tech company developing EPMA and SIMS instruments (http://www.cameca.fr/), has been established to design a new ion microprobe with high technical standards in both mass and lateral resolution, as well as in sensitivity. A machine possessing these characteristics is expected to open the way to spectacular new applications in cell biology, pharmacology and medicine.

A prototype of this new microprobe was installed at Harvard Medical School in February 1999, and commercialization of this instrument, called "Nanosims 50", began this year. Three units are now in operation. Two of them will be used to study cosmic dust from meteorites (Washington University, MI, USA (http://presolar.wustl.edu/nanosims/index.html) and Max Planck Institute, Mainz, Germany), the third instrument, to be devoted to biological studies, will be installed at the end of this year in our biophysics group (INSERM U350) in the Raymond Latarjet laboratory of the Curie Institut, which is located on the campus of the University of Paris-Sud in Orsay.

Isolated MCF-7 cell treated by pazelliptine (10-5M) for 1h. Field 100x100 µm - A : optical image (x400) of the cell after SIMS analysis (note the strong erosion of the cell surface) - B & C : SIMS images (IMS 4F) at m = 26 (12C14N-) and m = 27 (12C15N-), Mass resolution (M/DM) = 5000 - D : subtraction of calculated 12C15N- values from image C. from Jean-Luc GUERQUIN-KERN, Maïté COPPEY, Danièle CARREZ, Anne-Christine BRUNET, Chi Hung NGUYEN, Christian RIVALLE, Georges SLODZIAN, Alain CROISY. Complementary advantages of fluorescence and SIMS microscopies in the study of cellular localization of two new antitumor drugs. Microsc Res Techn, 1997, (36) :287-295

---

Background

Upon the impact of primary ions of several KeV energies on the surface of a solid sample, most chemical bonds are broken and atoms or polyatomic fragments are ejected from the very superficial atomic layers of the specimen (1-2 nm), either as neutral or charged particles (ions).

In an ion microscope, the sample is maintained at a potential of 3 to 4.5 KV. Thus, secondary ions resulting from the impact of the primary beam, and bearing a charge of the same sign as the sample, can be readily extracted through a first electromagnetic lens called an immersion lens. This secondary beam is then focused and guided to the entrance of a mass spectrometer using several transfer lenses. An electrostatic sector allows an energy sorting of the secondary ions. The beam is refocalized before entering the magnetic sector, which can be adjusted for selection of specific ions on the basis of their mass to charge ratio (m/z). Projection lenses guide the selected ions either to a Faraday cylinder for total ion current measurement or to a visualization screen for imaging purpose. The whole system is maintained under ultra-high vacuum (10-8-10-9 Torr). In the first SIMS imaging systems, the primary beam diameter was 25 to 250 µm with an energy of 10 KeV and an intensity varying from 10 to 100 pA, depending on the nature of the primary ions used. Such conditions were highly destructive and most biological preparations were completely vaporized after a single analysis, especially in the case of trace analysis. Lateral resolution was only around 1 to 0.5 µm.

Synopsis of an IMS-3F ion microscope

The mass spectrometer part of the first commercial system (SMI 300, CAMECA) was a simple combination of a magnetic prism and an electrostatic mirror with a mass resolution (M/DM) of only 300, a value far below the required resolution for biological samples.
Improvement in the design of the mass spectrometer led to the CAMECA IMS 3F series, with high mass resolution allowing the separation of secondary ions such as 12C14N- (m = 26.045) and 12C21H2- (m = 26.024). However, such high resolution power was strongly detrimental to sensitivity.

Indeed, the sensitivity of an ion microscope is essentially dependent upon two main characteristics :

the ionization rate of the considered atom (or molecule), t(A+/-), which is the ratio between the number of ions formed, n(A+/-), and the number of atoms, n(A), within the studied volume :

t(A+/-) = n(A+/-) / n(A)

This value is strongly related to the properties of the target atom (ionization potential, electron affinity) and its chemical environment, as well as to the nature of the primary ions.

The transmission of the system, h, which can be roughly considered as the ratio between the number of ions formed and the number of ions detected. At high mass resolution, the overall transmission of most ion microscopes was limited to 10-15% thus strongly affecting the sensitivity. In practice, only the effective ionic yield, tu, which is a function of both t(A+/-) and h, has to be considered

tu = t(A+/-) * h

From these relations, one can assume that, for a specific SIMS microscope, tu would be influenced only by the choice of the primary ions. Thus, for elements such as Na, K, Ca..., the best choice is to select imaging of positive ions using O2- from a duoplasmatron, which is still the best available source of primary negative ions. On the other hand, for organic matter, better sensitivity will be obtained by collecting negative secondary ions ( C-, CN-, P-...) produced with a beam of positive primary ions such as Cs+.

Direct imaging required strong primary ion intensities, thus the sample was quickly destroyed, and the sequential acquisition of several images from different ions did not allow a good superposition of the various images since each of them were not coming from the same volume.
A first major breakthrough was made when the direct imaging system was replaced by a microprobe which scans the surface of the sample. Using this process, Levi-Setti obtained a strong improvement in lateral resolution using a liquid gallium source (20nm diameter, 1.6pA intensity).3 Another improvement was the introduction of a low energy electron beam for compensation of positive charge accumulation at the sample level in order to prevent burst of insulating samples.4 This device, together with a cesium ion microprobe, has been successfully tested with a commercial SIMS apparatus (IMS 4F, CAMECA).5 These advances ultimately led to the design of a new SIMS microprobe with high resolution in terms of both lateral definition and mass separation 6 : the NanoSims 50.This machine has decisive capabilities which allow biological applications of SIMS microscopy at a level up till now inaccessible.

---

Description of NanoSims 50

As for previous SIMS microscopes, the NanoSims 50 contains a source(s) of primary ions, a mass spectrometer and a detection device, its main characteristics are :

High lateral resolution (<50 nm in cesium, <150 nm in oxygen)

Capability to measure up to 5 masses (ions) in parallel, coming from the same microvolume and ensuring perfect isotopic ratio from the same small volume or perfect image superimposition.

Very good transmission even at high mass resolution (60% at M/DM = 5000)

Direct optical observation of the sample within the ionization chamber allowing fast selection of the interesting areas

Charge compensation for insulating sample (in secondary negative ions mode).

No change has been made in the ion sources, but the primary beam path has been strongly modified to become co-axial with the secondary beam within the objective column. The immersion lens is used for both primary ions focalization and secondary ions collection. This configuration allows a strong shortening of the distance between the lens and the sample. Thus the primary beam can be focused to a very thin probe. This also improves the effective yield of collection of secondary ions, as well as a decrease in the aberration of the system. The mass spectrometer is double focusing, with an electrostatic filter and a magnetic sector in the Mattauch-Herzog configuration. The mass spectra is displayed along the focal plane of the magnet and up to five detectors can be moved along this plane, thus allowing the collection of up to five ion images coming from the same microvolume. The mass resolution is generally around 4000 to 5000, with a 60% transmission for the C- ions of graphite. The transmission is still 40% at M/DM = 7500 and 10% for M/DM = 15000. Slodzian has shown that, with this new microprobe, only 200 carbon atoms are required to obtain a C- analytical signal.6 This means that, in this case, tu has a value around 5x10-3. However, this yield can vary strongly with the nature of the observed atom and its environment.

Synopsis Nanosims 50

Lateral resolution is essentially dependent on the size of the microprobe and the number of image points (pixels). A maximum area of 200µm x 200µm can be explored and the pixel number can be adjusted between 16 x 16 to 2048 x 2048. The probe has routinely a 0.1 µm diameter but can be decreased to 40 µm (Cs+).

---

Future Applications

Several founding projects have been identified through various collaborations involving the Curie Institute and other laboratories engaged in cancer research (Orsay University, INSERM, National Museum of Natural History, Aventis). These projects are essentially concern with cancerology through three main areas of application:

Antitumor pharmacology

Intracellular localization of new antitumor drugs
(in collaboration with C. Monneret & Ph.Maillard, UMR176 CNRS/Institut Curie and H. de Thé, UPR 9051 CNRS, Hôpital St-Louis).

Intracellular trafic of plasmidic DNA designed for gene therapy
(in collaboration with D. Scherman & P. Wils, UMR133 CNRS/Aventis).

Intracellular localization of antisens oligonucleotides
(in collaboration with T. Le Doan, CNRS ERS 0571, UPS & N. Chi Hung, UMR176 CNRS/Institut Curie).


Cytogenetic studies of cancer cells

Analysis of the distribution of Mg, Ca and Zn ions and of Br, F and I atoms in metaphasic chromosomes
(In collaboration with B.Dutrillaux, UMR 147 CNRS/Institut Curie).

Nuclear medicine and radiotoxicology

Targetting of the melanic cell, application to diagnostic and treatment of melanoma
(In collaboration with J.C.Madelmont, Unité INSERM 484, CHR Clermont-Ferrand).

R&D of new radiolabelled oligonucleotides targetted to MDR cancer cells
(In collaboration with D.Fagret, Equipe INSERM 00.08, Nuclear Medicine Department, CHU Grenoble).

This laboratory is open to the biological scientific community, on the basis of programs selected by a committee of experts.

For further informations mail to : Alain.Croisy@curie.u-psud.fr - Jean-Luc.Guerquin-Kern@curie.u-psud.fr

---

References

1 : R. Castaing et G. Slodzian (1962), J. Microsc., 1, 395-410
2 : P.Galle (1970), Ann.Phys.Biol.Med.,42, 83-94
3 : R. Levi-Setti (1985), Scan.Electr.Microsc., 2, 535-551
4 : G.Slodzian (1987), Optik, 77, 148-155
5 : H.N.Migeon et al. (1990), Surface & Interface Anal., 16, 9-13
6 : G.Slodzian et al.(1990), C.R.Acad.Sc., 311(série II), 57-64
G.Slodzian et al. (1992), Biol.Cell, 74, 43-50