Two-photon ablation with 1278nm laser radiation
P Fischer, A McWilliam, L Paterson, C T A Brown, W Sibbett, K Dholakia and M P MacDonald
SUPA, J.F.Allen Research Labs, School of Physics and Astronomy, University of St Andrews, Fife, KY16 9SS
Abstract. We report on precise two-photon ablation with 110fs laser pulses at 1278 nm, emitted from a Cr:Forsterite laser. Selective two-photon ablation of Muntjac deer chromosomes is demonstrated. The two-photon absorption at 639 nm was enhanced by using methylene blue dye as a photosensitiser. This stain has a strong absorption in the region around 650 nm but 100% transmission around 1300 nm, allowing increased specificity: material that has absorbed the dye is ablated but undyed material is left unaffected. The low group velocity dispersion in glass at 1278 nm led to negligible pulse broadening in the focussing objective where the 100fs pulses stretched to 110 fs. This contrasts to the 100fs pulses at 780 nm that were measured to stretch to 300 fs under the same conditions.
1. Introduction
The interaction of light with biological matter such as tissue, cells and intracellular bodies like organelles and chromosomes is highly wavelength dependent. It is also often necessary to use complex optics to deliver the light and when, as is becoming more and more common, short pulses are used, there is an additional wavelength dependence added by the dispersion of the optical system itself [1]. Precise intracellular microsurgery and in-depth penetration of light into tissue are both examples of processes that are made possible by utilizing the multi-photon interaction of short-pulse laser radiation with matter. The first can be used to accurately incise or ablate intra-cellular components, hence affecting the manner in which the cell functions [2], and the second to image deep within tissue [3]. In both cases it is important that as little absorption as possible occurs out with the focal volume and this can be done by keeping single-photon absorption to a minimum. Hence, it is of great importance to select the wavelength of the laser with care, taking into account single and multi-photon cross sections plus the dispersion characteristics of the optical system.
In multi-photon interactions, longer wavelength radiation requires more incident photons for ionisation [4] such that the absorption cross section is lower than at shorter wavelengths and the localisation of the 2-photon process is enhanced. Furthermore, the multi-photon absorption cross section has been found to decrease in many biological tissues with longer wavelength [5] and single photon absorption has a window at 1300 nm, allowing high specificity when using a photosensitiser, as only sensitised material will interact with the laser.
In this paper we report on two-photon ablation of Muntjac deer chromosomes using femtosecond radiation in the spectral region around 1300 nm in combination with suitable staining. In addition to its intrinsic advantages for reducing out of focus damage and for light delivery into tissue compared to light at 780nm, the spectral range around 1300 nm has very low group velocity dispersion (GVD) in a wide variety of glass materials [6].
Short pulses are central to many processes, not least multi-photon absorption, yet when delivering pulses to a sample it is common to use a high numerical aperture microscope objective (which typically contains a large optical path length). When using the most common ultra-short pulse wavelength at 780 nm, GVD is high and the pulse that is delivered to the tissue is significantly stretched compared to the pulse length emitted by the laser. Hence, 1300 nm ultrashort pulse radiation has two significant advantages over many other wavelengths: low GVD allowing pulse lengths to be maintained (even through cheap telecommunications fibre) and very high specificity, with the aid of a membrane permeable organelle/protein specific dye.
2. Materials and Methods
Laser induced chromosome cutting was demonstrated by Liang et Al. in 1993 and 1994, where nanosecond pulses from a frequency doubled Nd:YAG laser were used [7, 8]. Two photon chromosome cutting was demonstrated in 2001 by Koenig et al. using femtosecond pulses from a Ti:Sapphire laser (l = 800 nm, repetition rate 80 MHz femtosecond laser, pulse duration 170 fs pulse average power 15-100 mW, with and without Giemsa staining) [9]. In this work, as a demonstration of the ability of pulses in the 1300nm spectral region to ablate intracellular material via a two-photon interaction, selective ablation of Muntjac deer chromosomes was performed. By staining the metaphase spread, efficient 2-photon absorption is achieved where, unlike when using radiation around 800nm, only photosensitised material is capable of interacting with the laser radiation reducing the chances of damaging material other than the target material
2.1. Sample preparation
The absorption peak of DNA is in the ultraviolet at 260 nm [10]. To obtain absorption at our wavelength, chromosomes from Muntjac cells (MJ) were photosensitised using a methylene blue dye. This dye has a strong absorption in the range of 650 nm but 100% transmission at 1300 nm, as shown in Figure 1(a) and Figure 1(b), respectively. As well as having suitable absorption characteristics, Methylene blue has been used as a drug to treat methemoglobinemias and has also been used as an antimalarial drug [11]. The toxicity of Methylene blue has been investigated in two mammalian test systems and it was found to be mutagenic in cultured mammalian cells, however this genotoxicity is not expressed in vivo [12].
(a) /
(b)
Figure 1. Absorption (a) and transmission (b) curves of methylene blue. The absorption spectrum was obtained using a dedicated absorption spectrometer and includes corrections for the water that was used as the solvent for the dye. The transmission was measured with an ellipsometer. (The noise in the transmission spectrum just above 900 nm is due to a change of lamp in the spectrometer.) The vertical lines mark the single photon (solid) and two-photon wavelength (dashed).
Metaphase spreads of Muntjac chromosomes were prepared as follows: Muntjac cells were grown in T125 flasks (in Minimal Essential Medium (MEM) supplemented with 10% foetal calf serum) until approximately 80% confluent. Colcemid (a chemical that inhibits mitotic spindle formation) was added to the medium at a final concentration of 0.1 mg/ml and cells were left for 3 hours. Colcemid halts cells at the metaphase stage of the cell cycle. Mitotic cells were harvested, treated with hypotonic solution (KCl: H2O, 1:1) to make them swell and then washed in fixative solution (methanol: acetic acid, 3:1). Twenty microlitres of metaphase cells in fixative solution were dropped onto ice cold slides coated in ethanol, then left to dry. Once dry, the slides were stained in methylene blue solution (2g methylene blue, 0.5g NaCl, 100ml H2O) for five minutes, then rinsed in tap water and left to dry. The metaphase chromosomes fixed to these slides were then exposed to the laser.
2.2. Experimental set-up for two-photon interaction
The set-up shown in figure 2 was used to achieve two-photon ablation of the chromosomes. The radiation from the laser was focused to a spot size of w = 1.9 mm using an oil immersion microscope objective (Olympus Ach, 100x, NA = 1.25).
The sample was held on a xyz-stage with computer controlled actuators (Newport CMA12-PP). A program scanned the sample stage in the x-direction with a scan speed of 5 mm/s while changing the z direction in 200nm steps. The laser parameters as described above result in a total energy per spot of 28 mJ deposited in 6.7x107 pulses. The peak power was 3.5 kW and the peak intensity was approximately 31 GW/cm2. Imaging was obtained using a long working distance microscope objective (100x, N.A. 0.7, working distance 6 mm) and a camera (Soliton, TC-2912) whose sensitivity extends into the 1300nm spectral range.
Figure 2. Experimental set-up used for the two-photon interaction. Illumination of the sample was achieved through the focusing objective using a Koehler arrangement.2.3. Delivery of ultrashort pulses
An in-house built Cr:Forsterite laser (l = 1278 nm), either continuous wave or modelocked with a pulse duration of 100 fs, average power 150 mW and a repetition rate of 180 MHz [13] was used for the experiments. The autocorrelation trace and the spectrum of the laser radiation are displayed in figure 3(a) and figure 3 (b), respectively.
The wavelength of the Cr:Forsterite laser is very close to the zero-dispersion wavelength at 1300 nm so the pulse duration remains essentially unchanged - even though no special optics and coatings were used. We measured an average power delivered to the sample plane of 75 mW and a pulse duration of 110 fs.
Both the transmission and pulse broadening were measured using a system of two identical microscope objectives (the second one to collect the light focused down by the first objective), allowing collimated light to be used for reliable power and pulse length measurements.
Figure 3. Autocorrelation trace (left) and emission spectrum (right) of the used Cr:Forsterite laser.The initial pulse duration t0, as well as the broadened pulse length through two microscope objectives t2, were measured using an autocorrelator. Assuming a pulse shape described by EQ 1.
/ (1)where I(t) is the time dependent intensity and I0 the peak value of I(t). The pulse duration t1 after the first microscope objective is described by [14] in terms of the initial pulse length:
/ (2)and similarly, the pulse duration after the second microscope objective is described by
/ (3).The GVD f is the same for each of the objectives and hence can be treated as a constant, allowing the system of equations (2) and (3) to be solved to obtain an expression (EQ 4) for the pulse length t1 after one objective:
/ (4)Pulse broadening was measured for the Cr:Forsterite laser as well as a Ti:Sapphire laser. The pulse durations after one microscope objective, calculated using EQ 4, are displayed in table 1.
Table 1. Pulse durations t1 and group velocity dispersion parameter f after one microscope objective for Cr:Forsterite (l = 1278 nm) and a Ti:Sapphire laser (l = 780 nm) when the initial pulse t 1 is 100 fs .
Cr:Forsterite / Ti:SapphireMicroscope Objective / f [fs2] / t1 [fs] / f [fs2] / t1[fs]
x40 / 2193 / 117 / 8004 / 243
x60 / 2430 / 121 / 9509 / 282
x100 / 1664 / 110 / 10200 / 300
Thus, a 100fs long pulse from the Cr:Forsterite laser, when passed through the microscope objective used for ablation (Olympus ACH, 100x, NA = 1.25), is increased by only 10% to 110 fs, whereas the same microscope objective broadens the pulse of a Ti:Sapphire laser by 300%, leading to a pulse duration of 300 fs.
3. Experimental Results
The 1300nm wavelength region has outstanding properties when propagating through biological tissue [15]. Preliminary experiments have shown that the radiation of the Cr:Forsterite laser interacts weakly with biological tissue in single photon interaction [16].
The use of femtosecond pulses and tight focusing leads to very high peak intensities, therefore allowing a precise two-photon interaction. It is crucial that the pulse duration of the radiation delivered to the sample is kept as short as possible. Here the low GVD of the Cr:Forsterite radiation in the optical system is highly advantageous.
A dye crystal was used to determine the area of the beam that is effective for the two-photon interaction and the criticality of this focal position. The beam and hole profiles are pictured in figure 4. The hole diameter has a 1mm waist radius whereas the beam diameter has a radius of 1.87 mm. No hole was generated at the same power level when the laser was running in continuous wave mode. This result showed that the two-photon interaction is sensitive to the focal position to within less than 1 mm. The z-position of the translation stage was therefore changed in 200 nm steps.
Figure 4. Beam (broader curve) and hole (narrower curve) profiles. The hole was obtained in two photon interaction of the radiation with a dye crystal.
The same set-up was used for the cutting of mammal (Muntjac) chromosomes, prepared as described above. As mentioned above, the z position of the translation stage was changed in 200 nm steps while the area of interest was scanned in the x-y plane. Figure 5 shows a single chromosome from a metaphase spread before and after interaction with the ultrashort-pulsed Cr:Forsterite laser.
Figure 5. False colour rendering from a micrograph of a Muntjac chromosome before (a) and after (b) two photon ablation with a Cr:Forsterite laser. (The horizontal length of each image is approximately 4.5 mm).
4. Conclusion
We demonstrate for the first time, two-photon ablation of chromosomes using a Cr:Forsterite laser emitting 100 fs pulses at a central wavelength of 1278 nm. The combination of significantly suppressed interaction of this wavelength in the “out of focus” region compared to shorter wavelengths with the outstanding pulse duration maintaining properties due to very low group velocity dispersion in standard optical components, such as lenses and microscope objectives, makes this wavelength a promising candidate for further applications in femtosecond biophotonics.
In contrast to the common ultrashort-pulse sources with outputs around 780 nm where the high GVD prohibits delivery via conventional fibre optics, ultrashort pulse radiation in the range of 1300 nm can be propagated in readily available telecommunications fibre without appreciable pulse broadening.
Acknowledgments
P Fischer acknowledges funding from the Swiss National Science Foundation. We also acknowledge P Bryant for providing the cell line and A E Vasdekis for measuring the spectrum of the methylene blue dye. We thank the UK EPSRC for funding. M P MacDonald acknowledges the support of an EPSRC Advanced Research Fellowship.
References
[1] Cannone F, Chirico G, Baldini G and Diaspro A 2003 Journal of Microscopy-Oxford, pp 149-57
[2] Berns M W, Wang Z F, Dunn A, Wallace V and Venugopalan V 2000 Proceedings of the National Academy of Sciences of the United States of America, pp 9504-7