Even when using advanced electron microscopical techniques, it is not always possible to clearly distinguish primitive life forms like bacteria and archaea from nonliving mineral particles (Thomas-Keprta et al., 2002). In the context of detecting life on Mars and other foreign worlds, the degree of difficulty is exacerbated by device vibration throughout the voyage, temperature sensitivity, and, amongst other challenges, a higher necessary threshold of operator expertise. Nevertheless, every planetary mission has used optical equipment for imaging, but the focus has primarily been on geological studies rather than the direct observation of potential microorganisms (Nadeau et al., 2018). To date, no direct microscopic observations have been made on Mars with sufficient resolution to detect bacteria or other microscopic life forms (Nadeau et al., 2016). There is, therefore, a need for the development of life-detection instruments, which may involve machine learning algorithms, that can be employed with minimal resources yet still retain the capacity to sufficiently detect microbial life and differentiate it from the surrounding environment (Riekeles et al., 2021).

Microbial motility, the directed motion of microbes under their own propulsion, can be clearly distinguished from random Brownian movement by microscopic techniques and is, therefore, a prominent biosignature of life (Thomas-Keprta et al., 2002; Lindensmith et al., 2016; Nadeau et al., 2016; Nadeau et al., 2018; Riekeles et al., 2021; Riekeles et al., 2024b). Given that it has evolved independently multiple times on Earth (Miyata et al., 2020), motility might also be a fundamental trait of extraterrestrial life that can be exploited for its detection in a resource-depleted setting (Neveu et al., 2018). Motility, a characteristic present across all three domains of life—bacteria, archaea, and eukaryotes—can be derived via several molecular mechanisms in response to various environmental cues, such as avoiding harmful substances or locating new resources and nutrients (Allen, 1981). The run-and-tumble method is one such motility pattern in several prokaryotes (Madigan et al., 2018, pp. 92–100). The clockwise or counterclockwise rotation of filamentous appendages, which are capable of rotating at speeds of up to 1,000 revolutions per second, allows bacteria (Madigan et al., 2018, pp. 92–100) to alternate between swimming phases (i.e., runs) and tumbling phases (i.e., changes in direction). The swimming phase propels the cell forward at speeds up to 60 cell lengths per second, whereas the tumbling phase involves reorientations of swimming direction. The Mot proteins (MotA and MotB) located in the cytoplasmic membrane selectively conduct protons, thereby providing a proton motive force (PMF) that can serve as the necessary energy source for rotation (Bardy et al., 2003).

Other common bacterial motility forms include twitching motility, sliding motility, and flagellar propulsion. Twitching motility involves a cyclic pulling mechanism via the extension, fastening, and retraction of cellular attachments referred to as type IV pili (Merz et al., 2000; Skerker and Berg, 2001; Mattick, 2002; Craig et al., 2019). Sliding motility, instead, relies on the expansive forces of cell growth within a colony (Hölscher and Kovács, 2017). Alternatively, spirochetes have evolved to utilize flagella stemming from the periplasmic space (between their outer membrane and cell wall) at opposing ends of the cell. The coordinated rotation of these flagella provides movement when rotating symmetrically and restricts movement when rotating asymmetrically (Bardy et al., 2003).

Following the concept of the ubiquitous nature of motility, many archaea also use an appendage that is functionally similar but structurally distinct from the bacterial flagellum (Albers and Jarrell, 2018). The archaellum consists of a helical filament attached to a membrane-embedded motor complex. This filament is composed of proteins called archaellins, which are synthesized as pre-proteins in a manner similar to that of the type IV pili in bacteria (Jarrell et al., 2021). Unlike bacterial flagella, which make use of electrochemical gradients, the archaellum presumably uses ATP as its energy source (Nuno de Sousa Machado et al., 2022). Both the archaellum’s preserved ability to rotate clockwise and counterclockwise and the large percentage of amino acid sequence overlap with flagellin, a key protein in the filament region of bacterial flagella, suggest that flagellar-like motility has deep evolutionary roots in the history of life. Furthermore, archaea appear to have acquired their chemosensory systems from bacteria and, subsequently, have successfully integrated these systems with the archaellum (Li et al., 2019). Despite these seemingly ancient similarities, structural comparison of the flagellar proteins from the archaeal archaellum and bacterial flagellum imply that flagellar motility evolved independently in these two groups, likely diverging over 3.5 billion years ago (Madigan et al., 2018).

There is a lack of consensus regarding the overall percentage of (known) motile prokaryotes, partially due to many studies focusing solely on using the genome to predict structure and function rather than making direct cellular observations. In 1996, approximately 80% of the bacteria described were motile by means of flagella (Aizawa, 1996). Utilizing more recent data from a BacDive database sample of 14,309 different bacterial strains, it was determined that 40% were motile (Reimer et al., 2022). The website is available at https://bacdive.dsmz.de/advsearch (accessed on 27 November 2024). Overall, the proportion of motile bacteria can vary widely in natural environments at any given sampling time, with reports indicating a range between 5% and 80% (Grossart et al., 2001; Fenchel and Thar, 2004). For example, a study on bacterial swimming behavior in the ocean near the Scripps pier in San Diego demonstrated that motility can vary seasonally (Soutourina et al., 2001).

Light stimuli (phototaxis), orientation in magnetic fields (magnetotaxis), and movement in response to chemical gradients, known as chemotaxis—the primary focus of this study—are among the most common stimulants of microbial motility. Numerous studies have demonstrated that bacteria predominantly utilize chemotaxis to migrate toward environmental conditions that optimize their growth, such as being attracted to amino acids and sugars (Colin et al., 2021). In certain bacterial species, the most potent chemoattractants align with the molecules that are preferentially consumed by the cells, as exemplified by Escherichia coli’s response to specific L-amino acids (Yang et al., 2015; Colin et al., 2021). However, this strict correlation between the magnitude of chemotaxis and the direct metabolic or physiological advantages conferred by the chemoattractant does not always hold true in other bacterial species (Colin et al., 2021). For instance, the chemotactic system in B. subtilis, which was determined to have a heightened sensitivity to environmental cues (Yang et al., 2015), was found to be attracted to ethanol despite it not being a direct metabolic substrate nor providing seemingly any physiological benefit (Tohidifar et al., 2020). The authors proposed that ethanol conferred an indirect benefit to Bacillus subtilis, which could use the molecule to locate and prey on ethanol-fermenting microorganisms (Tohidifar et al., 2020). Rather than the chemoattractant leading directly to a beneficial metabolic or physiological source, this suggests a multistep process where bacteria utilize chemoattractants to locate particular environments that are more likely to contain their preferred nutrient sources (Colin et al., 2021).

Amino acids, particularly proteinogenic amino acids, have long been recognized as potent chemoattractants, presumably due to their essential role in biological processes (Mesibov and Adler, 1972; Vuppula et al., 2010). Some of the oldest and most critical biochemical pathways, such as carbon fixation methods, have evolved around the need to synthesize such key amino acids (Braakman and Smith, 2012). Correspondingly, the high degree of chemotactic sensitivity for these molecules is best exemplified in some organisms by an abundance of proteinogenic amino acid receptors in their membranes. One such proteinogenic amino acid, L-serine, has been determined to be among the most potent chemoattractants for B. subtilis and various other microbes (Ordal and Gibson, 1977). This finding may stem from the fact that L-serine is particularly unique due to its hydroxyl group, which is only found in two proteinogenic amino acids and is a common site for regulatory phosphorylation and dephosphorylation by protein kinases and phosphatases, respectively (Oxenrider and Kennelly, 1993). Concerning its astrobiological relevance, L-serine has been found has been found in a meteorite and in the 162173 Ryugu asteroid, indicating its presence even before the formation of our Solar System (Koga and Naraoka, 2017; Naraoka et al., 2023). Given that early Earth and Mars were bombarded by carbonaceous asteroids, L-serine likely exists on Mars (Pizzarello and Shock, 2010; Zhu et al., 2022). If life developed on Mars with a similar biochemistry to known life on Earth, it seems plausible that L-serine could also be a potent chemoattractant for hypothetical Martian microbes. In the field of astrobiology, chemotaxis has been discussed and investigated in (Nadeau et al., 2016; Acres et al., 2021). Furthermore, in the development of a microscopic system for life detection, L-serine has already been proposed to stimulate microbial motility (Lindensmith et al., 2016).

In-situ life detection instruments for the autonomous detection of motility require technical sophistication that includes classical or advanced computer vision or machine learning and needs considerable computational resources, which is challenging to accommodate for life detection missions in space (Bedrossian et al., 2018; Riekeles et al., 2021; Riekeles et al., 2024a; Wronkiewicz et al., 2024). Furthermore, the Martian sediment, which is among the most likely habitats for putative microorganisms, must be physically or (bio-)chemically separated from any hypothetical microorganisms to enhance microbial detection. To address these challenges, we suggest a simplified method (Figure 1). The methodology is based on the work of (Elgamoudi and Ketley, 2016), inducing chemotactic motility and detecting it through the usage of μ-slides and subsequent blob detection. The μ-slide devices consist of two opposing chambers separated by a semipermeable membrane, where, in principle, one of the chambers contains a chemoattractant (e.g., L-serine) and is sterile at the starting point, while the other is inoculated with putative microbes. The semipermeable membrane divides the two sections, preventing sediment particles from traversing (which we tested in a pre-experiment) and yet allowing motile organisms to pass through (i.e., traveling up the chemoattractive gradient) (Figure 2). Particles observed to have crossed the partition towards the stimulant are likely to have done so actively, indicating they are living organisms. Exploiting the principles of chemotaxis and utilizing small and relatively inexpensive μ-slides could facilitate both sediment-microbe separation, which would allow for easier imaging and computing, but, more importantly, would allow for active microbial motility detection that requires only low-computational blob detection algorithms to indicate living organisms.

We have tested the setup for the Gram-positive bacterium B. subtilis, the Gram-negative bacterium P. haloplanktis, and the archaeon H. volcanii. These organisms were chosen for two main reasons. Firstly, given that no extraterrestrial microbe has yet been identified, the traits necessary to survive such a harsh environment are not well understood, and, therefore, it is challenging to choose representative organisms for examination. However, we can speculate that, for survival in cold, arid, extraterrestrial climates like that of Mars, specific characteristics, such as spore-forming capacity and psychrophilic and halophilic aptitude, would likely be advantageous for survival. Each selected organism exhibits one or more of these traits, resulting in a better extraterrestrial-exemplary organism. Secondly, previous studies have found B. subtilis to be attracted to L-serine, allowing it to act as a positive control, and P. haloplanktis to be weakly repelled by L-serine, allowing us to conceptualize it as a negative control (Ordal and Gibson, 1977; Garrity and Ordal, 1995; Barbara and Mitchell, 2003). The inclusion of H. volcanii, a well-characterized halophilic archaeon, broadens the study’s taxonomic diversity and enables insights into its chemotactic responses.

The results show that all three organisms demonstrated significant chemotactic behavior in response to L-serine, further supporting the basis of chemotaxis and, more broadly, motility as an important microbial biosignature. That said, each organism’s particular chemotactic response varied with respect to L-serine concentration and time of exposure. As expected, B. subtilis’ chemotactic response to L-serine, which had been previously firmly established allowing it to act as a positive control (Ordal and Gibson, 1977; Garrity and Ordal, 1995), was reaffirmed across all tested concentrations and time points. The response was significant across both the 10 mM and 20 mM concentrations at the 1-hour timepoint, though it was particularly strong under the 20 mM conditions, suggesting that, at this time, higher concentrations of attractant translated to a greater chemotactic response. Interestingly, as time progressed from 1 to 3 h, there was an apparent decrease in affinity for the 20 mM concentration and no change in the affinity for the 10 mM concentration. Though the mean number of blob counts within the 10 mM and 20 mM chemoattractant chambers was greater at the 3-hour time point than the 1-hour by 360% and 215%, respectively, it should be noted that the control chamber cell density also experienced a 325% increase in cell density. This was an isolated event not observed in either of the other organism’s control chambers, which remained constant throughout the experiments, suggesting that the control sample of B. subtilis motility increases between the 1 h and 3 h timepoint. This might be a survival mechanism due to nutrient depletion, which has also been observed for E. coli bacteria grown in media with a single carbon source of low-nutritional value (Ni et al., 2020).

Surprisingly, P. haloplanktis, which was expected to act as a negative control based on a previous study describing a repulsive effect between the microbe and L-serine (Barbara and Mitchell, 2003), also exhibited a statistically significant attractive chemotactic response to both the 10 mM and 20 mM L-serine concentrations at all observed time points. However, given that the prior studies had observed the repulsive effect at significantly lower L-serine concentrations (0.6 mM) and duration of exposure (less than 1 min), these findings do not directly contradict the earlier study (Barbara and Mitchell, 2003). A biphasic response has been observed in E. coli when leucine is used as a signaling molecule and was either an attractant or repellent, depending on its concentration (Khan and Trentham, 2004). P. haloplanktis might exhibit a similar response when exposed to L-serine. This highlights the importance of revisiting and reassessing established biological understandings, especially under different experimental conditions or environmental contexts, as a particular chemoattractant’s repulsive or attractive nature can vary with concentration, duration of exposure, and likely other situational factors. Moreover, the transition from 1 to 3 h was a point of interest as the results diverged for the 10 mM and 20mM, where, despite the 20 mM cell density deviating insignificantly from the 1-hour timepoint, the 10 mM chamber cell density was significantly less than both its 1-hour density and the 3-h 20 mM cell density. Synthesizing these results with the previously reported study, it appears that P. haloplanktis’ chemotactic response towards L-serine can be conceptualized as a scaled transition from repulsive at lower chemotactic agent concentrations and durations, such as 0.6 mM and less than a minute, towards attractive at higher concentrations and longer durations of exposure, such as 20 mM at 1 h. The caveat being, however, that intermediary concentrations still capable of inducing a significant attractive response, such as 10 mM, trend more quickly towards being weaker, indifferent, or possibly even repulsive stimuli as time progresses and the attractant gradient disperses than higher concentrations, such as 20 mM, which were able to maintain their affinity across the 3-h time period. Further work will be needed to address the concentrations and durations of exposure at which there is a transition in the chemotactic response from repulsive to attractive and vice versa.

Research on the chemotactic behavior of archaea has been relatively limited until recent years. However, a growing focus on H. volcanii within the scientific community has significantly deepened our understanding (Kokoeva et al., 2002; Quax et al., 2018; Li et al., 2019). The question of whether L-serine induces chemotaxis in H. volcanii had not been addressed before this study. The results presented here suggest, for the first time, an increased attraction of the archaeon toward both the 10 mM and 20 mM concentrations of L-serine. When observed after 1 h, there was no apparent difference in the microbe’s affinity for the differing chemoattractant concentrations. However, after 3 h there was a significant contrast in chemotactic responses between the two concentrations with an increasing affinity for the higher concentration whereas the affinity for the 10 mM remained approximately the same throughout the study together suggesting that as time progresses H. volcanii’s chemotactic response towards higher concentrations, such as 20 mM, becomes more attractive. In contrast, the response towards lower concentrations, such as 10 mM, remains relatively constant.

With respect to the presented results for all three organisms, it should be noted that observed increases in chemoattractant chamber cell density across the 3-h time period were attributed to a flux of individual microbes chemotactically traversing the semipermeable-membrane up the L-serine gradient rather than cells metabolically utilizing L-serine as a carbon source permitting cell growth beyond what would normally be feasible within the provided minimal medium. This conclusion was supported by growth curves for all three organisms across the entirety of the 3-h experiment following the PBS washing procedure.

The study’s experimental setup diverged from earlier protocols, in which the μ-slides were used for long-term chemotactic assays lasting up to 24 h (Elgamoudi and Ketley, 2016). Previous methodologies generally filled both distal chambers of the μ-slide with a concentration of agarose gel up to 1% (Elgamoudi and Ketley, 2016). Alternatively, this experiment focused efforts on a much shorter duration for chemotactic observation by electing to remove the motile-resistive agarose in both the right chemoattractant chamber as well as in the left organismal chamber, effectively restricting the presence of agarose to only the central semipermeable membrane. This modification allows for less resistive movement within each chamber, yet still retains the capacity for interchamber selective permeability. The approach proved viable and can be utilized in the future as a more time-effective alternative for studies of microbial motility, such as on in-situ life detection missions in which longer-duration cultivation may not be possible. Like in previous space-related research (Lindensmith et al., 2016), the use of L-serine in these experiments proved successful, justifying further investigation.

As with any other technique intended to investigate the existence of extraterrestrial life, methodological assumptions must be made and addressed as limitations of the approach to specifying the approach’s breadth of utility. Three such limiting assumptions made in this study are 1) that extraterrestrial life would have the capacity for chemotaxis, in this case in response to L-serine at a particular concentration and time point, and 2) that not only is the semipermeable membrane capable of resisting non-cellular interchamber translocation of sediment particles at various agar concentrations but also that 3) researchers have prior knowledge of the desired isolated organism’s motility characteristics so as to properly adjust the membrane’s density. At present, these assumptions are likely too restrictive for this approach to be put into practice during life detection missions on Mars or other planetary bodies, where such prior knowledge is unknown.

Further studies could address these limitations and, thereby, make this approach more practical by incorporating a more comprehensive variety of chemoattractants including other amino acids or glucose, which, despite L-serine’s chemotactic potency across all three domains of known life, would provide an even more extensive chemotactic stimulus reaching organisms that would not otherwise be chemotactically sensitive. As previously stated, organic compounds with biological relevance, such as racemic mixtures of various proteinogenic and non-proteinogenic amino acids, have been found in extraterrestrial samples such as the Murchison meteorite and asteroid (162173) Ryugu (Koga and Naraoka, 2017; Naraoka et al., 2023). Moreover, there is evidence suggesting the existence of ribose and other essential sugars within primitive meteorites (Furukawa et al., 2019). On Earth, many of these detected molecules have demonstrated the ability to induce chemotactic responses in species including E. coli’s sensitivity towards ribose, L-alanine, and various other amino acids and B. subtilis’ response to glycine and cysteine, amongst other stimulants (Ordal and Gibson, 1977; Kondoh et al., 1979; Yang et al., 2015). Given their existence in space and capacity for chemotactic stimulation, the methodology presented here could be adapted to include these attractants individually or in unique combinations in an effort to induce a greater response magnitude and a broader spectrum of reactivity. In future studies, the limitations of the semipermeable membrane should be further addressed through the development of an agnostic semipermeable membrane; meaning a membrane that maintains its ability to restrict noncellular translocation yet also, without additional density manipulation, permits the passage of a broader spectrum of potential microbes with various sizes, shapes, and biochemical properties.

Chemotaxis remains a broad and only partially understood field that holds much promise for both living microbe isolation and identification that would be of interest to a variety of disciplines. In the context of astrobiology, ongoing research is required to facilitate the most resource-effective methodology that could be used for in-situ life detection missions. When synthesized with previous findings, this study’s results effectively demonstrate the viability of μ-slides for chemotactic observation of prokaryotes across shorter (several hours) and longer (up to 24 h) timeframes. Though further adaptation is required for its real-life feasibility, the minimal necessary technical and resource requirements, as well as the reduced need for continuous observation, make their usage an attractive and feasible alternative for space missions, where power, operator handling, data storage, and complex data processing capabilities are limited.