Introduction
Mars may have provided in its early history an environment conducive for an independent origin of life, but the surface conditions today are no longer hospitable. Low atmospheric pressure and temperature result in an environment where liquid water is not stable, and the lack of a thick atmosphere and significant ozone layer results in a solar ultraviolet (UV) radiation environment that extends down to biologically damaging UV wavelengths of ∼200 nm (e.g. Cockell et al. Reference Cockell, Catling, Davis, Snook, Kepner, Lee and Mckay2000; Patel et al. Reference Patel, Bérces, Kerékgyárto, Rontó, Lammer and Zarnecki2004). In addition, the ionizing radiation field produced by the unshielded flux of energetic cosmic rays is also destructive to the complex molecules of life (Baumstark-Khan & Facius Reference Baumstark-Khan and Facius2001; Dartnell Reference Dartnell2011) and extends metres into the Martian subsurface, far deeper than UV penetration (Pavlov et al. Reference Pavlov, Blinov and Konstantinov2002; Dartnell et al. Reference Dartnell, Desorgher, Ward and Coates2007a, Reference Dartnell, Desorgher, Ward and Coatesb). These factors, among others, mean that the surface of Mars presents a challenging environment for life to persist. In spite of the harsh exposed surface environment, potential microniche refuges for microbial life, offering protection against the unfiltered solar UV radiation and oxidizing conditions, include cryptoendolith communities within fractures or pore spaces of surface rocks (Friedmann Reference Friedmann1986; Wynn-Williams & Edwards Reference Wynn-Williams and Edwards2000), polar ice deposits (Cordoba-Jabonero et al. Reference Cordoba-Jabonero, Zorzano, Selsis, Patel and Cockell2005) or deeper in the near-surface regolith. Only a few millimetres of dust coverage is required to screen the damaging UV irradiation to survivable levels (e.g. Cockell & Raven Reference Cockell and Raven2004) and numerous studies have been performed on the viability of various microorganisms in the Martian surface environment (see Olsson-Francis & Cockell (Reference Olsson-Francis and Cockell2010) for a review), but once exposed to the unshielded UV flux photolytic destruction can be rapid. This study extends this body of work by investigating the detection window for specific biosignatures once previously shielded samples have become exposed to solar UV on the Martian surface during probe operations.
Instruments designed to detect organic molecules or other potential tracers of life on Mars include gas chromatography mass spectrometry (GCMS), as installed aboard NASA's Mars Science Laboratory (Mahaffy et al. Reference Mahaffy2012) and proposed for the ESA-Roscosmos ExoMars rover (Evans-Nguyen et al. Reference Evans-Nguyen, Becker, Doroshenko and Cotter2008; Goesmann et al. Reference Goesmann, Becker and Raulin2009); Raman spectroscopy (Ellery & Wynn-Williams Reference Ellery and Wynn-Williams2003; Jorge Villar & Edwards Reference Jorge Villar and Edwards2006; Marshall & Olcott Marshall Reference Marshall and Olcott Marshall2010; Rull et al. Reference Rull2010), and antibody-based systems such as the Life Marker Chip (Sims et al. Reference Sims, Cullen, Bannister, Grant, Henry, Jones, McKnight, Thompson and Wilson2005). Such instrumentation can offer very high sensitivity and discriminatory power, but often requires sample preparation, long exposure times or limited resources or reagents.
It has been proposed therefore to use a complementary instrument based on exciting native autofluorescence from Martian organic compounds, or biomolecules within microorganisms, to rapidly survey potential target regions or triage samples before scrutiny with more discriminatory instruments (Griffiths et al. Reference Griffiths, Coates, Muller, Storrie-Lombardi, Jaumann, Josset, Paar and Barnes2008; Storrie-Lombardi et al. 2008, Reference Storrie-Lombardi, Muller, Fisk, Cousins, Sattler, Griffiths and Coates2009; Weinstein et al. Reference Weinstein2008; Muller et al. Reference Muller, Storrie-Lombardi and Fisk2009; Dartnell et al. Reference Dartnell, Storrie-Lombardi and Ward2010, Reference Dartnell, Storrie-Lombardi, Mullineaux, Ruban, Wright, Griffiths, Muller and Ward2011, Reference Dartnell, Patel, Storrie-Lombardi, Ward and Muller2012). Such an instrument could employ either UV laser diodes for remote detection or UV LEDs with a close-up imager (such as the Mars hand lens imager, MAHLI, aboard Mars Science Laboratory, which has two 365 nm LEDs for UV illumination at night; Edgett et al. Reference Edgett, Ravine and Caplinger2009).
Fluorescence-based detection systems have proved themselves to be sensitive and discriminatory, and have been used in many terrestrial applications. These include the assessment of pollution (JiJi et al. Reference JiJi, Cooper and Booksh1999; Ko et al. Reference Ko, Lee, Kim and Kim2003; Alberts & Takács Reference Alberts and Takács2004; Cory and McKnight Reference Cory and McKnight2005) and identification of potentially pathogenic or toxic microorganisms in environmental water or food preparation (Patra & Mishra Reference Patra and Mishra2001; Hua et al. Reference Hua, Dolan, Mcghee, Clevenger and Deng2007; Sohn et al. Reference Sohn, Himmelsbach, Barton and Fedorka-Cray2009; Ziegmann et al. Reference Ziegmann, Abert, Müller and Frimmel2010). Of most direct relevance here, fluorescence systems have successfully detected trace biomolecules or microbial life in the Atacama desert (Weinstein et al. Reference Weinstein2008), Antarctic sandstone (Nadeau et al. Reference Nadeau, Perreault, Niederberger, Whyte, Sun and Leon2008), and glacial (Rohde & Price Reference Rohde and Price2007) and Antarctic ice (Storrie-Lombardi & Sattler Reference Storrie-Lombardi and Sattler2009).
Excitation-emission matrices
The fluorescence response from a target can be quantitatively described by the generation of an excitation–emission matrix (EEM). Briefly, an EEM is created by recording the fluorescent emission spectrum from an experimental sample across a range of excitation wavelengths, and plotting the fluorescence intensity produced at each excitation–emission (ex–em) wavelength combination as a contour map.
Figure 1 displays the EEMs generated for the two model organisms considered here. Deinococcus radiodurans (left) is selected as a polyextremophile able to tolerate high ionizing radiation doses (such as from the cosmic ray flux on to the unshielded Martian surface), and Synechocystis sp. PCC 6803 (right) is taken as a model cyanobacterium exhibiting the same cellular fluorophores as cryptoendolith organisms inhabiting UV-shielded microniches. Such representations of the full fluorescence response of microorganisms clearly exhibit different fluorophores within the cell, appearing as fluorescence features at particular ex–em locations within the parameter space. The intense feature was seen at UV wavelengths, with a peak ex–em pairing of about 280–340 nm and labelled (1) in Fig. 1, is due to the amino acid tryptophan bound within cellular proteins (e.g. Ammor Reference Ammor2007).
Excitation–emission matrices (EEMs) representing the full fluorescence response across a broad range of wavelengths for both model organisms studied here, the polyextremophile Deinococcus radiodurans (left) and the cyanobacterium Synechocystis sp. PCC 6803 (right). Fluorescence intensity is in arbitrary units and colour-coded on a log-10 scale as indicated. Horizontal black bars indicate the spectral ranges analysed to quantify the degradation of signal intensity of different cellular fluorophores with exposure to UV. Cellular fluorophores are labelled as follows: (1) tryptophan; (2) pool of small cellular fluorophores including NAD(P)H; (3) photosynthetic pigments, including chlorophyll and phycocyanin.
Both Deinococcus and Synechocystis also exhibit weaker fluorescence at an ex–em pairing of about 340–450 nm, labelled (2) in the EEMs, which is generally attributed to a mixed pool of cellular fluorophores including the metabolic coenzymes NADH and NADPH (Ammor Reference Ammor2007). In addition, Synechocystis exhibits two broad, intense fluorescence features with a peak emission at about 660 nm, labelled (3). This is fluorescence of the cyanobacterial photosynthetic system for efficiently gathering light and transducing this energy into growth. The photopigment chlorophyll and phycocyanin, a component of the phycobilisome molecular complex that aids light-gathering, are both intrinsically highly fluorescent due to their biological function (Keränen et al. Reference Keränen, Aro and Tyystjärvi1999).
The horizontal black lines marked in Fig. 1 indicate the spectral regions of interest for following the degradation of these cellular fluorophores by exposure to UV radiation. These are the excitation wavelengths selected for the most sensitive detection of each fluorescent feature, and the width of each bar indicates the emission wavelength range of the spectra plotted in Fig. 4.
Rationale
Here, we expand on recent work (Storrie-Lombardi et al. 2008, Reference Storrie-Lombardi, Muller, Fisk, Cousins, Sattler, Griffiths and Coates2009; Muller et al. Reference Muller, Storrie-Lombardi and Fisk2009; Dartnell et al. Reference Dartnell, Storrie-Lombardi and Ward2010, Reference Dartnell, Storrie-Lombardi, Mullineaux, Ruban, Wright, Griffiths, Muller and Ward2011, Reference Dartnell, Patel, Storrie-Lombardi, Ward and Muller2012) on the utilization of fluorescence for surveying for Martian organics or organisms. These recent studies have demonstrated the feasibility of detecting autofluorescence of polycyclic aromatic hydrocarbons (PAHs) on mineral grains using 365 nm LED (Storrie-Lombardi et al. 2008) or 375 nm laser diode (Muller et al. Reference Muller, Storrie-Lombardi and Fisk2009) excitation sources and calculated the detection limit to be about 20–145 ppm using the PanCam filter wheel (Storrie-Lombardi et al. Reference Storrie-Lombardi, Muller, Fisk, Cousins, Sattler, Griffiths and Coates2009). The photolytic degradation half-life of these representative PAHs, once uncovered by probe operations, has been experimentally determined to between 25 and 60 h of noontime summer irradiation (Dartnell et al. Reference Dartnell, Patel, Storrie-Lombardi, Ward and Muller2012). Biological work has focused on characterizing the complete fluorescence response from different extremophilic and photosynthetic microorganisms to identify the best excitation and emission wavelengths for an instrument (Dartnell et al. Reference Dartnell, Storrie-Lombardi and Ward2010) and the rate of loss of the fluorescence biosignature from cyanobacteria exposed to ionizing radiation, such as from unshielded cosmic rays on the Martian surface (Dartnell et al. Reference Dartnell, Storrie-Lombardi, Mullineaux, Ruban, Wright, Griffiths, Muller and Ward2011). The present study builds on this existing body of knowledge to now look at the rate of degradation of fluorescence signal from model microorganisms exposed unfiltered solar UV radiation.
In order to have survived until the present day, localizations of Martian organics or organisms must have been protected from solar UV and so a Mars exploration probe would need to examine freshly exposed material. A rock abrasion tool (RAT) such as carried by the Mars Exploration Rovers Spirit and Opportunity (Gorevan et al. Reference Gorevan2003), the rock coring drill on Mars Science Laboratory (Okon Reference Okon2010) and the 2 m subsurface drill proposed for ExoMars (Vago et al. Reference Vago, Gardini, Kminek, Baglioni, Gianfiglio, Santovincenzo, Bayon and van Winnendael2006) are all appropriate devices for this. Once exposed to the unfiltered solar UV spectrum on the Martian surface, however, any potential fluorophores will be readily photodegraded and their fluorescence signal lost.
So the key question for the functionality of such a fluorescence-based technique is: on what timescale are the emission signals of different fluorescent targets destroyed by solar UV once uncovered? What is the window of opportunity for detection by instrumentation aboard a rover? In the longer term, the fluorescence-based detection of ice-bound organisms may have significance for the rapid testing of ice profiles for cryoperserved microbial life in the Europan ice shell.
In this study, we expose frozen thin samples of two model microorganisms, the polyextremophile D. radiodurans and the cyanobacterium Synechocystis sp. PCC 6803, to broad-spectrum UV radiation in a Mars simulation chamber. The changes to the cellular fluorescence response are quantified after total fluences calibrated to between 0.25 and 64 h of equatorial noontime Martian sunlight, and thus the window of opportunity for detection for a fluorescence-based instrument is assessed.
Methods
Cell cultures
Deinococcus radiodurans
The 200 ml of Tryptone digest, Glucose, Yeast extract (TGY) broth in a 1 l conical flask was inoculated with starter culture and incubated at 30 °C with constant agitation at 200 rpm for 44 h to culture until mid-late log phase. The nutrient growth medium TGY is itself fluorescent and so the cell culture was washed by centrifugation and resuspension of the pellet in phosphate buffer solution (PBS; Difco) which is non-fluorescent (Dartnell et al. Reference Dartnell, Storrie-Lombardi and Ward2010).
Synechocystis sp. PCC 6803
100 ml of BG11 growth medium (Castenholz Reference Castenholz1988) in a 1 l conical flask was inoculated with starter culture and incubated for 3 days at 25 °C with constant agitation of 130 rpm in a photosynthesis incubator (Innova 4340, New Brunswick Scientific, St. Albans, UK) providing a photon flux of 46 μE m−2 s−1. Although it had already been determined that the BG11 growth medium produces no fluorescent emission of its own (Dartnell et al. Reference Dartnell, Storrie-Lombardi and Ward2010), the culture was centrifuged and resuspended in PBS to allow dilution to the target cell density, as described below.
Sample preparation
The target OD600 for optimal measurement of fluorescence of a cell suspension using the spectrofluorimeter instrument described below had been found previously to be around 0.5 (Dartnell et al. Reference Dartnell, Storrie-Lombardi and Ward2010), representing a compromise between sufficient fluorophore concentration and minimization of scattering or self-shielding by higher cell densities. Cell samples of 1 ml were to be exposed to UV, whereas the quartz spectroflourimeter cuvette is 3 ml, so the Deinococcus and Synechocystis master cell suspensions were both made to three times the final target OD600 of 0.5.
Approximately 1 ml samples of the prepared cell suspension were pipetted into 35 mm diameter (× 10 mm depth) tissue culture dishes. This volume and dish diameter selection results in a sample depth of only 1 mm, avoiding the problem of self-shielding during UV irradiation. The absorption coefficient of water over the wavelength range 185–400 nm is negligible (Warren Reference Warren1984) and so the UV exposure is uniform through the 1 mm depth. Cell samples were frozen to −80 °C before transportation on dry ice to the Mars simulation chamber. Control samples were prepared and transported identically to the test samples, but not exposed to UV radiation in the Mars simulation chamber.
UV irradiation
Cell sample dishes were laid flat on a bed of dry ice to maintain a temperature of −80 °C (representative of the Martian surface) throughout irradiation and arranged in the Mars simulation chamber, as shown in Fig. 2.
Photograph of the setup of the experimental exposure. Very thin layers of frozen cell suspensions within tissue culture dishes rest on a bed of dry ice, to emulate Martian surface temperatures, within the Mars simulation chamber at the Open University. The UV irradiation lamp is not yet switched on in this image.
The xenon lamp used in this study produced a UV emission spectrum closely matching that modelled for the unfiltered sunlight incident on the Martian equatorial surface at noon during northern summer (Ls=90°) and nominal atmospheric dust loading (Patel et al. Reference Patel, Zarnecki and Catling2002). The flux rate is also calibrated to be ten times greater to allow accelerated irradiations, and so to create Martian equivalent exposures of 0.25, 0.5, 1, 2, 4, 8, 32 and 64 h, real-time durations of 1.5 min, 3 min, 6 min, 12 min, 24 min, 48 min, 3 h 12 min, and 6 h 24 min, respectively, were employed. See Dartnell et al. (Reference Dartnell, Patel, Storrie-Lombardi, Ward and Muller2012) for further details on the irradiation spectrum. This distribution of dose levels was designed to allow for detailed observation of the effects through low exposures while also providing results on the response after much greater irradiation. This data coverage over a wide scale proved to be very effective in revealing, for example, the complex behaviour of the 448 nm emission feature of Synechocystis (shown in Fig. 5, solid green line). Thus, the experimental procedure reproduced Martian surface temperature and irradiation spectrum, but was conducted at ambient atmospheric conditions. Photolytic chemistry and thus biomolecule destruction from UV radiation is temperature-dependent, but not affected by pressure per se at the molecular level, and so the Martian low atmospheric pressure was not experimentally recreated in order to remove sublimation of the ice samples as a confounding factor for fluorescence degradation.
Fluorescence analysis
Immediately before analysis, the 1 ml cell samples were thawed, diluted in 2 ml of PBS to the optimum OD600 of 0.5, and pipetted into a 3 ml UV-C quartz cuvette (Fisher, Loughborough, UK) cleansed with 70% ethanol and which yields no background fluorescence of its own.
Fluorescent emission of samples was measured with a Perkin Elmer LS55 fluorescence spectrometer (Perkin Elmer, Cambridge, UK), which consists of a pulsed xenon arc lamp light source, excitation and emission scanning monochromators, a sample compartment with cuvette holder, and a photomultiplier tube detector.
Emission spectra were recorded between 300 and 800 nm, with an excitation slit width of 8 nm and emission slit width of 5 nm, and data points logged every 0.5 nm at a scan speed of 1000 nm min−1. The excitation wavelength was incremented in 15 nm steps between 240 and 705 nm, which produced a total of 32 emission spectra for each sample. Thus, the complete fluorescence dataset collected for each sample represents a three-dimensional datacube of emission intensity measurements at 32 000 ex–em wavelength combinations. To smooth the recorded emission spectra and reduce the computational demands required for subsequent processing and plotting of the complete EEM, emission spectra were averaged into bins 10 nm wide.
Excitation–emission matrices were generated from the acquired datacube by data transformation and analysis computer code written by the lead author in Mathematica 9.0 (Wolfram Research, Champaign, USA). Each fluorescence profile (for example, in Figs. 1 and 3) is presented on a log-10 intensity scale and displayed with rainbow colour coding and 20 contour lines. The appropriate range for colour coding of the fluorescence intensity was determined from the emission intensity histogram, as explained in Dartnell et al. (Reference Dartnell, Storrie-Lombardi and Ward2010). The error in determining the peak height of the selected fluorescence features from the emission spectra is estimated to be no more than 5%.
The complete EEM dataset acquired on the fluorescence response of the two model organisms after increasing UV irradiation. Emission intensity is in arbitrary units and colour-coded on a log-10 scale as indicated. The total UV doses used are calibrated to the equivalent number of hours of exposure on the Martian surface, calculated for equatorial noontime conditions (Patel et al. Reference Patel, Zarnecki and Catling2002): 0.25, 0.5, 1, 2, 4, 8 and a maximum of 32 h Mars-equivalent exposure for Deinococcus and 64 h for Synechocystis. Minimal degradation of fluorescence signal is observed for Deinococcus, but the cyanobacterium exhibits noticeable diminishment of photosystem fluorescence (peaking at an excitation–emission wavelength pairing of about 615–660 nm) and a stark increase in the fluorescent feature at an ex–em pairing of about 340–440 nm.
Results
Figure 3 displays the complete dataset acquired for the full fluorescence response of the two microorganisms studied here, D. radiodurans and Synechocystis sp. PCC 6803. All EEMs are plotted over the same spectral region (excitation wavelength between 240 and 700 nm; emission wavelength between 300 and 800 nm) and using the same colour-coded log-scale of emission intensity, spanning two orders of magnitude.
Complete EEMs were generated for each UV exposure, rather than only measuring spectra covering the fluorescent features of interest (as indicated in Fig. 1 with horizontal black lines) to enable the detection of any shifting of peak excitation or emission wavelength or appearance of new fluorescent features with irradiation.
While there are no immediately obvious changes in the EEMs generated for D. radiodurans exposed to increasing doses of UV radiation, differences are apparent in the irradiated Synechocystis cells. The height of the intense fluorescent peak centred on an ex–em pairing of about 600–660 nm (labelled (3) in Fig. 1), produced by chlorophyll and phycocyanin associated with the cyanobacterial thylakoid membrane for photosynthesis, can be seen to diminish with increasing exposure. Even more notable is the stark increase in fluorescence of the feature at an ex–em pairing of about 340–450 nm (labelled (2) in Fig. 1).
These fluorescent features are shown in greater spectral detail in Fig. 4. For each fluorescent feature, the optimal excitation wavelength is selected (280 nm, 340 or 345 nm and 615 nm) and the corresponding emission spectrum plotted over the wavelength range indicated in Fig. 1. Fluorescence spectra recorded after Martian equivalent UV exposures of 0 (the control samples), 0.25, 0.5, 1, 2, 4, 8 and 32 h (in the case of Deinococcus) or 64 h (in the case of Synechocystis) are stacked on the same axes and colour-coded from light-grey to black to indicate longer duration of exposure, to enable comparison of the remaining emission intensity after exposure to increasing UV irradiations. The peak of fluorescent emission determined in these spectra is indicated with a green line, and labelled with the wavelength for clarity.
Stacked fluorescence emission spectra centred on the regions of interest, as indicated in Fig. 1: (1) tryptophan; (2) pool of small cellular fluorophores including NAD(P)H; (3) photosynthetic pigments, including chlorophyll and phycocyanin. Spectra are colour-coded to increasing UV exposure from light grey to black. The peak of each fluorescence feature quantified in this study is indicated by the vertical green line, with the wavelength labelled. The decline in remaining fluorescence from tryptophan (280 nm excitation) and the photosystem (615 nm excitation) of Synechocystis is clearly seen in these plots, and the fluorescence feature elicited by excitation at about 340–345 nm (middle row) in both organisms is revealed to be composed of two distinct peaks.
Both Deinococcus and Synechocystis exhibit a peak in protein-bound tryptophan fluorescence (280 nm excitation; top row) at 342 nm, and the peak emission from the cyanobacterial light-harvesting machinery (bottom pane), which is dominated by chlorophyll and phycocyanin, occurs at 658 nm.
Figure 4 also reveals a spectral characteristic not readily apparent in the full contour-mapped EEMs shown in Fig. 3. The fluorescent feature emitting between about 400 and 500 nm from excitation at about 340 nm seems to be a symmetrical peak in the EEMs (labelled (2) in Fig. 1), but is clearly shown in Fig. 4 (middle row) to be composed of two distinct peaks. These are labelled in Fig. 4 to exhibit emission peaks at 387 and 425 nm for Deinococcus and 394 and 448 nm for Synechocystis. The Synechocystis emission spectra (Fig. 4: middle row, right) are plotted on an extended intensity scale on the y-axis, which clearly shows that this 448 nm fluorescence feature rises to twice (205%) that of the initial intensity after 64 h Mars equivalent hours of UV irradiation.
The peak intensities of each of these fluorescence features, taken at the wavelength indicated, are extracted from these emission spectra and normalized to the initial value as 100% fluorescence signal intensity. Figure 5 plots these results, quantifying the change in fluorescence signal intensity of different cellular fluorophores as a function of increasing UV exposure.
Change in emission intensity of cellular fluorophores as a function of hours of UV exposure on the Martian surface for Deinococcus (dashed) and Synechocystis (solid). Fluorescence profiles are colour-coded to the peak emission wavelength, as labelled on Fig. 4: 342 nm tryptophan fluorescence (purple); 387/394 nm emission of Deinococcus/Synechocystis (blue); 425/448 nm emission of Deinococcus/Synechocystis (green); 658 nm emission from the Synechocystis photosystem (red).
Discussion and conclusions
Degradation of tryptophan and photosystem fluorescence
Figure 5 shows that, while there is experimental scatter of the data points of about ±5%, all three fluorescence features exhibited by Deinococcus (dashed lines) behave similarly in the long-term trend, declining to about 90% of the initial intensity after the maximum exposure of 32 Mars equivalent hours employed in this study. Colonies of Deinococcus are brightly coloured due to their expression of six carotenoid pigments related to β-carotene (Carbonneau et al. Reference Carbonneau, Melin, Perromat and Clerc1989), the dominant of which is deinoxanthin (Lemee et al. Reference Lemee, Peuchant, Clerc, Brunner and Pfander1997). Such carotenoids offer UV photoprotection by quenching oxygen-free radicals produced by short-wavelength irradiation and possibly also by directly screening the UV light (Cockell & Knowland Reference Cockell and Knowland1999), and so the cellular fluorophores within Deinococcus are well shielded from photodegradation.
By comparison, the tryptophan signal intensity in Synechocystis (labelled (1) in Figs. 1 and 4) is much more sensitive to UV photodegradation than the corresponding feature in Deinococcus. Figure 5 shows tryptophan fluorescence (solid purple line) to have been reduced to 80% after 8 h and only 50% after 64 h.
The signal most readily degraded by UV irradiation is fluorescence of the photosynthetic pigments (labelled (3) in Figs. 1 and 4) – diminished to only 35% after 64 h exposure. This can be expected as the biological function of these molecules is to optimize the harvesting of light energy and so they are readily photobleached (an effect also studied by Sinha et al. Reference Sinha, Richter, Faddoul, Braun and Hader2002). UV-B radiation (280–315 nm) is known to damage photosynthetic pigments, and the previous work has shown that the accessory biliproteins in the phycobilisome (such as phycocyanin) are bleached first, followed by carotenoids and chlorophyll damaged last (Sinha et al. Reference Sinha, Kumar, Kumar and Hader1995). Ultraviolet radiation also causes structural destruction of the phycobilisome complex that enhances light collection in photosynthesis. The first step in the breakdown of phycobilisomes is probably the destruction of linker proteins, with degradation products forming as components of the phycobilisome are broken down into hexamers, trimers and then monomers (Banerjee et al. Reference Banerjee, Sinha and Hader1998).
The loss of photosystem fluorescence found here, with a decrease of around 20% from the initial signal intensity after 8 h of simulated Martian flux, is less than that reported by a similar study working with dried monolayers of the desiccation-resistant cyanobacterium Chroococcidiopsis sp. 029. Cockell et al. (Reference Cockell, Schuerger, Billi, Friedmann and Panitz2005) found that autofluoresence of the cyanobacteria photosynthetic apparatus (emission>590 nm) was reduced to 14% of the initial intensity after one hour Martian equivalent exposure, and to 5% after 4 h. This greater degradation may be due to irradiation in a monolayer, which may indicate a synergistic destructive effect of cellular dehydration coupled with UV radiation, or a protective effect for our cells irradiated at very low temperature (−80 °C, representative of the Martian surface).
Response of cyanobacterial 448 nm emission feature
The behaviour of the Synechocystis fluorescence observed when exciting at about 345 nm (Fig. 4, middle right) is more complex and interesting. As discussed already, although the feature at an ex–em pairing of about 350–450 nm appears in the EEM representations in Fig. 1 (labelled (2)) and Fig. 3 to be a symmetrical peak; it is revealed by the emission spectra in Fig. 4 to in fact be composed of two distinct peaks.
These two peaks are roughly equivalent in Deinococcus (Fig. 4, middle left), whereas the ratio in Synechocystis between the intensity of the shorter wavelength emission peak (394 nm) and the longer wavelength peak (448 nm) varies substantially with UV exposure. The 394 nm peak is shown in Fig. 5 (solid blue line) to remain relatively unaffected by even the highest UV exposure level and retains a signal intensity closely comparable with the Deinococcus cellular fluorophores (dashed lines). The 448 nm emission peak, however, exhibits a far more complex behaviour in response to UV irradiation (solid green line in Fig. 5). Fluorescence intensity drops rapidly to 60% of the initial value after only brief UV exposure, before steadily rising again. Fluorescence exceeds the initial value (100%) after roughly 12 h of Mars equivalent exposure and continues to increase to twice the original emission after 64 h irradiation.
The spectral characteristics of this fluorescence feature, with an ex–em pairing close to 340–450 nm, and the feature emitting at the same wavelength (located below the label (2) in Fig. 1), with ex–em pairing of 260–450 nm, correspond to those of NADH and NAD+, respectively. Although these are ubiquitous co-enzymes involved in redox metabolism in the cell, they are not thought to be responsible for the shifts in fluorescence seen in the irradiated cyanobacteria. NADH and NAD+ are the reduced and oxidized forms of the same coenzyme and so cannot account for an increase in fluorescence at both of these ex–em pairings simultaneously, as is observed in the irradiated cyanobacteria (Fig. 3).
Instead, the profile of 448 nm emission (Fig. 4, solid green line) is interpreted here as representing an initial rapid photobleaching of the fluorescence of an original pool of cellular fluorophores (up until between half an hour and 1 h of exposure), followed by an accumulation of fluorescent break-down products generated by the photolytic destruction of cellular molecules by the short wavelength UV radiation.
These accumulating fluorescent break-down products are believed to be derived from chlorophyll, a major component of the cyanobacterial cell. The spectral properties of this feature (label (2) in Fig. 1), peaking at an ex–em pairing of 345–448 nm, match that of the fluorescent chlorophyll catabolites (FCCs) that show peak fluorescence at 450 nm after excitation at 350 nm, and are produced by the metabolic degradation pathways within ripening bananas (Moser et al. Reference Moser, Müller, Ebert, Jockusch, Turro and Kräutler2008, Reference Moser, Müller, Holzinger, Lütz, Jockusch, Turro and Kräutler2009), senescent banana leaves (Banala et al. Reference Banala, Moser, Müller, Kreutz, Holzinger, Lütz and Kräutler2010) and those of the peace lily (Kräutler et al. Reference Kräutler, Banala, Moser, Vergeiner, Müller, Lütz and Holzinger2010). The shift in fluorescent emission to shorter wavelengths (blue-shifting), from chlorophyll in healthy cyanobacteria cells at ∼685 nm (label (3) in Fig. 1) to these FCCs at ∼450 nm is caused by changes to the molecular structure of the fluorophore, such as loss of aromatic rings or a decrease in the number of conjugated bonds in a linear chain (Coble Reference Coble1996).
The ex–em pairing of these FCCs corresponds very well with that measured here for the anomalously increasing fluorescent feature (labelled (2) in Fig. 1 and seen most prominently in the 64 h EEM in Fig. 3) in UV-exposed cyanobacteria. Thus, it is believed that the stark increase in the ∼450 nm emission features after one hour Mars equivalent UV dose (Fig. 4, solid green line) is due to the accumulation of chlorophyll photolytic products. An identical pattern of growing intensity of a fluorescent feature at ex–em pairing about 350–450 nm has also been reported for the exposure of Synechocysis sp. PCC 6803 to increasing doses of gamma-ray ionizing radiation, and was also interpreted as the accumulation of radiolytically produced chlorophyll break-down products (Dartnell et al. Reference Dartnell, Storrie-Lombardi, Mullineaux, Ruban, Wright, Griffiths, Muller and Ward2011).
Relevance to Mars probe operations
As discussed in the introduction, exploitation of the autofluorescence of aromatic prebiotic compounds or the biomolecules within microorganisms has been proposed for instrumentation suitable for Mars lander probes. Such devices could offer a critical surveying or triaging function for surface operations, identifying the most promising sites or samples for examination with more time, energy or time- and resource-intensive analytical techniques such as GCMS.
For the successful design and operation of such a fluorescence-based instrument, it is important to determine the most effective excitation and emission wavelengths, as well as the response of the fluorescent target molecules once they have been extracted from their protected location beneath the surface and exposed to the intense UV component of the unfiltered sunlight on Mars. The key question is: does the irradiated fluorescence signal persist over the timescales required for robotic probe operations to enable detection?
Recent work has studied the UV photolytic degradation of the fluorescence signal from key prebiotic compounds, PAHs (Dartnell et al. Reference Dartnell, Patel, Storrie-Lombardi, Ward and Muller2012), and here we have extended this work to the fluorescent biosignature of microorganisms.
The native fluorescence of the polyextremophile D. radiodurans, once exposed to the unfiltered Martian sunlight, has been found here to be stable over an appreciable period of time. Figure 4 shows only 10% decrease in the signal strength after an UV exposure equivalent to 32 h of continuous Martian equatorial noon solar flux.
A cyanobacterium, such as Synechocystis sp. PCC 6803, without significant quantities of UV-shielding compounds such as mycosporine-like amino acids or scytonemin, exhibits significant photobleaching of the fluorescence signal from tryptophan and the photosynthetic molecules chlorophyll and phycocyanin by UV irradiation. This study found 20% signal loss from these fluorophores after only 8 h.
These durations represent a conservative estimate as cells emplaced in an environmental setting, embedded within a mineral matrix or partially protected by dust, would receive additional UV shielding. Furthermore, these Mars-equivalent UV fluxes are calculated from a model for equatorial midday sun during northern summer (Ls=90°) and nominal atmospheric dust loading (Patel et al. Reference Patel, Zarnecki and Catling2002), and so represent a relatively high level of dose rate. At increasing latitudes or time in the seasonal cycle, the actual UV flux experienced would be lower. More importantly, the instantaneous UV flux varies during the Martian day as the sun arcs through the sky, and undergoes increasing scattering effects from the atmosphere. Taking this time variation into account and integrating across a complete diurnal cycle reveals that every hour of Mars equivalent exposure used here equates to 0.16 sols (Dartnell et al. Reference Dartnell, Patel, Storrie-Lombardi, Ward and Muller2012).
Thus, Deinococcus autofluorescence remains at roughly 90% of the initial, unexposed intensity, and the Synechocystis tryptophan and photosystem signal at around 65%, over five to six Martian sols. This is a meaningful window of opportunity in the timescale of robotic surface operations and indicates that such fluorescence biosignatures of current life could indeed persist for long enough to be detectable after exposure to the Martian UV radiation.
The detectability of the components of previously protected cyanobacteria would even appear to be enhanced over these timescales due to the increase in fluorescence intensity of the 448 nm emission feature, believed to be due to the accumulation of photolytic breakdown products.
Acknowledgements
LRD was supported by the UCL Institute of Origins Post-Doctoral Research Associateship. MRP was supported by funding from the Science and Technology Facilities Council (STFC).
