We report the upgrade of the epifluorescence microscope of the GSI heavy-ion microprobe with a galvo-scanned, 488 nm laser diode. The laser is focussed into the object plane by the water-immersion objective resulting in a focal spot size of about 1 μm. To increase temporal and spatial resolution a water-immersion objective with a high numerical aperture is integrated into the custom-build microscope. The upgraded system can now be used to bleach GFP-tagged proteins recruited to DNA damage induced by targeted single-ion irradiation. The system is demonstrated on NIH 3T3 cells with Ku80-GFP ion-targeted in heterochromatic and euchromatic DNA. Fluorescence recovery after photobleaching (FRAP) is shown to be significantly slower in heterochromatin.

The heavy-ion microbeam is used for targeted irradiation of living cells (in culture) with ions from the UNILAC linear accelerator of GSI [1]. A custom-made epifluorescence microscope is available for beam targeting as well as for observation of proteins (tagged by fluorescent dyes) in sub-compartments of interest. In order to gain better insight into DNA damage repair processes, we intend to employ a focused laser beam to locally bleach these fluorescent markers, thus unmarking sub-sets of molecules and analysing their dynamics and kinetics (FRAP, FLIM) [2]. Here, the development of a dedicated laser system for such bleaching experiments at the bio-endstation of the GSI microbeam facility is described.

When people plan to adapt their ion microprobe for the targeted irradiation of biological cells, they often claim that they expect a targeting accuracy in the range of their beam spot diameter, because they assume that reaching a sub-mu m beam spot is the most difficult part of the job.Although many microprobes have now a beam spot diameter of some hundred nano-meters or less, nobody reached a targeting accuracy below 1 mu m. Besides obvious reasons, like mechanical or thermal instabilities, there is a more difficult problem to overcome: one still needs a light microscope to locate both the microbeam and the cells to be irradiated, and there are various light-optical effects, which can give misleading information about the position of the beam and the cells. (C) 2009 Elsevier B.V. All rights reserved.

In collaboration with GSI Biophysics, the microbeam, operated with ions from the UNILAC linear accelerator of GSI, is routinely employed for targeted irradiation of living cells in culture [1]. A custom-made fluorescence microscope is used both for beam targeting and for cell observation via fluorescent dyes that specifically highlight the biological agent of interest. Green Fluorescent Protein (GFP) is often used to study dynamics of repair proteins shortly after broad-beam ion irradiation in a beamline microscope [2, 3]. A combination of the high-quality in-situ GFP microscopy, currently available at the X6 beamline microscope, with the accurate targeting of single ions performed at the X0 microbeam would greatly widen the possibilities of the microbeam for radiobiological studies. Here, we report on technical improvements of the in-situ GFP microscopy setup at the microbeam facility.

In collaboration with GSI Biophysics, the microbeam, operated with ions from the UNILAC linear accelerator of GSI, is routinely employed for targeted irradiation of livi ng cells in culture [1]. An online fluorescence microscope is used both for beam targeting [1] and for cell observation after irradiation via fluorescent dyes [2]. Fluorescence microscopy requires an intense light source at a set of specific wavelengths that excite the fluorescent dyes. For this purpose, we used to employ a highpressure mercury arc lamp with a number of bandpass filters on a mechanical filter changer along with a mechanical shutter.

For a single ion hit facility built to irradiate specific targets inside biological cells, it is necessary to prove that the ions hit the selected targets reliably as the ion hits usually cannot be seen. That ability is traditionally tested either indirectly by aiming at preetched tracks in a nuclear track detector, or directly by making the ion tracks inside cells visible by a stain coupled to special proteins produced in response to ion hits. However, both methods are time consuming and hits can be verified only after the experiment. That means targeting errors in the experiment cannot be corrected during the experiment. Therefore, we developed a fast online hit verification method that measures the targeting accuracy electronically with a spatial resolution of ±1μm before cell irradiation takes place.

Biological effects in unirradiated cells located near to irradiated cells are called bystander effects. Targeted irradiation of single cells by means of the heavy ion microprobe developed at GSI [1] was used to improve our understanding of these effects on a molecular and cellular basis. The protein of our interest was the cyclindependent kinase inhibitor CDKN1A (p21). Based on previous results obtained by broadbeam irradiation, we investigated the spatial distribution of the induction of p21 protein on the level of single bystander cells. We targeted single cells (3%) of a confluent monolayer of normal human fibroblasts (AG1522C) with 5 particles of carbon ions in 2 crossed lines per nucleus. Three hours after exposure we performed immunofluorescence (IF) staining for CDKN1A and 53BP1 protein. 53BP protein is recruited to sites of DNA damage immediately after irradiation by forming foci and is used to identify the targeted cells unambiguously (see figure 1a: the irradiated cell indicated with an arrow and enlarged in the box).

Since the mid 1990's, an increasing number of charged particle microbeams have been designed to deliver a preset number of ions to individual living cells with the micron resolution.These tools provide a powerful technique to investigate the cellular response to low doses of radiations.During the last years, the single ion hit facility in operation on the GSI microbeam since 1987 has been upgraded for the irradiation of individual living cells in vitro.This setup presents two main peculiarities compared to the microbeams used up to now for cell irradiation.First, the beam's micrometric size is obtained by magnetic focusing and not by a simple collimation.This allows obtaining a smaller beam spot, a better defined linear energy transfer, and a high irradiation throughput.Then, the GSI microbeam is able to focus ions from carbon to uranium with energies between 1.4 MeV/u to 11.4 MeV/u.The range of accessible linear energy transfer is thus considerably extended compared to light ions microbeam in operation today.The design of the GSI microbeam is described, including the beam control, the online cell localisation, the cell dish designed specifically for microbeam irradiation, and the cell irradiation procedures.Experimental tests performed to check the global aiming accuracy as well as the first cellular irradiations are presented.

Quantitative charge collection measurements have been performed on SOI capacitors using a 58 N 33+ micro beam at 3.6 MeV/u. Collection efficiencies as well as charge sharing effects on adjacent devices through coincidence measurements are reported and discussed in the frame of comparisons between Synopsys code simulations and experimental results.

The possibilities of coincident elastic recoil detection for mass analysis around mass 64 have been investigated at the GSI heavy-ion microprobe. Partial separation between the stable isotopes 63Cu and 63Cu has been obtained using a pair of position-sensitive detectors accepting both scattered projectiles and recoils between 38° and 52°. Total energy spectra gated by the peaks in the mass spectrum show small differences. For the surface nearest to the detectors depth resolution is better than 25 nm.

The technical concept of the scanning ion microscope (SIM) at GSI, which has now reached a resolution of 0.5 μ, will be presented together with some details of general interest like diagnostics of very weak microbeams, beam scanning and switching. Results of past experiments concerning secondary electron imaging, energy loss imaging and the use of the microbeam for microlithography will be reported as well as the special problems of operating a SIM at a pulsed accelerator. Advantages of a SIM compared to other microscopes will be discussed.

An antireflection treatment based on etched nuclear tracks is presented. The optical surface to be treated is irradiated by about 1011–1012 ions/cm2 and etched until the latent tracks develop into etch pits which have a depth of 1/2 of the wavelength of visible light and a largest diameter of no more than 1000A˚. A surface treated in this way exhibits the familiar antireflection effect of rough surfaces but still looks fairly clear. The method can be applied to all materials in which the track etching speed is at least 5 times the etching speed in the bulk material, i.e. to most plastics and some optical glasses. The result of an exploratory experiment, using 750 keV argon ions and the plastic material CR-39, is described and possible improvements are suggested.

A shadow cast technique is presented which generates relief like replicas of submicroscopic objects using irradiation by accelerated heavy ions followed by an etching step (nuclear track etch technique). Some applications to the investigation of biological objects and to the technology of microelectronics are demonstrated. Distinctive advantages over existing X-ray technique are 1.1) Small accelerators/implanters can be successfully employed.2.2) The technique offers high density contrast due to the well defined ion range.3.3) A nearly unlimited range of materials can be directly structured by the technique.

A digital processor has been built, which calculates within 2.5 μs position and total energy loss of a particle in a position sensitive detector of the charge division type. The processor has been designed to accept signals from all common types of ADCs with up to 14-bit resolution. Its digital output of maximum 14 bits can be accepted by all common multichannel analyzers.

An automatic scanner has been built to count tracks in nuclear emulsions exposed in magnetic spectrographs. It scans up to 200 times faster than human scanners. Counting accuracy for high contrast plates is comparable to that of human scanners except in extremely low or high density regions. The device described differs from other automatic scanners in that it has a fiber optic illumination system, a specially designed parallel processing computer for track recognition and a relatively low price. Special attention is given to the problem of optical preprocessing of track information to facilitate the recognition of tracks.