Spatiotemporal Tracking of Near‐Infrared Fluorescent Single‐Walled Carbon Nanotubes in C. Elegans Nematodes Confined in a Microfluidics Platform

Caenorhabditis elegans (C. elegans) nematodes are a powerful model organism for diverse biological and biomedical studies, benefiting from their genetic similarities to humans, small size, and transparency. However, fluorescence imaging of C. elegans can be challenging due to the strong autofluorescence in the visible range, which obscures the signal of common fluorescent proteins or dyes. Single‐walled carbon nanotubes (SWCNTs) fluoresce in the near‐infrared (NIR) range, where there is no autofluorescence background. Herein, a platform is developed for in vivo NIR imaging of C. elegans gastrointestinal tract using biocompatible SWCNTs functionalized by single‐stranded DNA or phospholipid‐polyethylene glycol (PEG). The SWCNTs serve as fluorescent tracking probes within the worm gut, following internalization along with food intake. A microfluidic confinement device is employed to ensure an anesthetics‐free feeding environment, allowing spatiotemporal control over the SWCNTs intake imaging. Furthermore, improvements in spectral colocalization, real‐time detection of intracorporeal SWCNT dynamics, and digestive trajectory tracking are demonstrated. Owing to the unique optical properties of SWCNTs and the confinement of the worms in the microfluidics system, the proposed platform facilitates advanced in vivo imaging of C. elegans in both the visible and NIR regions, opening numerous avenues for advancing research of C. elegans and other microscopic model organisms.


Introduction
A widely used in vivo model organism, the nematode Caenorhabditis elegans (C.elegans) is a transparent, non-parasitic soildwelling roundworm that has become a well-established signaling. [18]The C. elegans nematode is also becoming a popular platform for a deeper understanding of the underlying mechanism of anti-parasitic agents and various human drugs, [19,20] as well as for discovering new bioactive compounds. [21]maging of C. elegans internal processes, certain proteins, cellular compartments, or tissues, are normally performed by labeling targets with fluorescent markers.Fluorescent markers are used to indicate defects in function, gene expression, development, or protein interactions in vivo. [22,23]For example, the green fluorescent protein (GFP) reporter is commonly used in transgenic C. elegans worms, [24,25] and the 40,60-diamidino-2phenylindole hydrochloride (DAPI) is most commonly used as DNA dye for worm cell compartments visualization. [26]][29][30][31] Owing to the worm's transparency, strong autofluorescence in C. elegans can be detected throughout the visible spectral range and is caused by a few major sources.For example, in the GFP wavelength range (excitation 470-490 nm and emission 500-550 nm), the source is primarily the presence of flavin-containing molecules such as riboflavin and flavin adenine dinucleotide (FAD) in the intracellular lysosome-derived granules, mitochondria or extracellular collagen present in the cuticle, uterus and intestine of the worm, [22,32] whereas in the DAPI wavelength range (excitation 325-375 nm and emission 435-485 nm), the autofluorescence is mainly due to the presence of lipofuscin, a fluorescent pigment that accumulates in aging cells and which is located in the gut granules. [33,34][44][45] Assessing the processes and fate of ingested particles in the worm's intestine is therefore of great interest.50] Another common example that uses the feeding mechanism in C. elegans model is the study of host-microbiota interactions and bacterial infections related to human pathogens. [51]Typically, these bacteria express a fluorescent protein and are introduced into the worms through food intake.54][55] Many research studies have explored intestinal imaging in C. elegans by employing conventional visible fluorescent proteins or particles, [56] such as glucose uptake [57] or abnormal pharyngeal structure in response to phm-2 gene mutation. [58]However, the majority of frequently used fluorescent dyes emit fluorescence in the visible wavelength range, which overlaps with the autofluorescence exhibited by the worms so that the fluorescent agents can unambiguously label only proteins or compartments that are spatially separated from the intestines and the uterus of the worm.To address this challenge, a few approaches were introduced, such as administering RNA interference (RNAi) to the worms through feeding and thus reducing the autofluorescence of the intestine; however, it could also cause changes in gene functions. [22,59]Another possible approach to address the fluorescence crosstalk is to separate the spectral bands of the fluorescence proteins emission from the worm's autofluorescence with a triple band filter set; however, it can also block some of the fluorescence emission of the dye. [60]63][64] NIR imaging techniques, which take advantage of the unique properties of the NIR spectral range, have emerged as a powerful tool for non-invasive in vivo imaging. [61,65,66]In particular, singlewalled carbon nanotubes (SWCNTs) have been found to exhibit fluorescence in the near-infrared II (NIR-II) region, mainly between 900 and 1400 nm. [67]This is due to their unique semiconducting electronic band structure, which allows them to absorb and emit light at specific wavelengths.Different chiralities, which are labeled by two indices (n,m) that describe the structure of the nanotube, can yield a variety of optical transitions. [68]77][78][79][80][81][82] Being highly hydrophobic nanostructures, the pristine SWC-NTs tend to clump in an aqueous environment.However, their properties can be altered to suit many applications by surface functionalization.[114][115][116][117][118][119][120][121][122][123] Indeed, imaging C. elegans using the fluorescence emission of SWCNTs can be advantageous owing to the spectral distinguishability between the SWCNTs' NIR fluorescence and the autofluorescence of the worm in the visible range, thereby improving quantitative analysis as well.Hence, the utilization of SWCNTs as fluorescence probes can be extended beyond specific areas or organs.126][127] Long-term in vivo imaging of SWCNTs within C. elegans requires a well-controlled spatiotemporal environment for restricting the worms in an optically accessible setting and avoiding movement-related artifacts, while allowing for continuous feeding and washing procedures.[130] Worms immobilization techniques such as introducing anesthetics or using polystyrene nanoparticles on agarose pads, [131][132][133] were either ill-controlled or contained compounds that were often not compatible with the physiological processes being studied and usually precluded long-term imaging. [134,135]ne popular method for C. elegans immobilization is using microfluidic devices, owing to the precision of control, minimal animal stress, compatibility with imaging settings, high throughput, reusability, and versatility.152][153] Despite the significant potential of using NIR imaging with functionalized SWCNTs in C. elegans confined within a microfluidic device, thorough investigations of such a platform that can utilize long-term NIR imaging for in vivo applications are lacking.For example, the suggested platform would enable deeper insights into the worm's digestive tract physiology and could have significant implications for understanding gastrointestinal disorders and drug development.
In this work, we introduce a platform for in vivo imaging of SWCNTs as NIR fluorescent probes within C. elegans nematodes, facilitated with a microfluidic device for spatial and temporal control that enables whole organism imaging (Scheme 1).We use the nutritional source for C. elegans worms, namely, the E. coli bacteria, as a conduit to introduce SWCNTs into the worms through the food intake process.We demonstrate multispectral fluorescence imaging and colocalization of visible and NIR imaging channels, where the NIR fluorescent SWCNT signals are spectrally distinct from the worm autofluorescence in the visible range.We utilize and examine two types of functionalized SWCNTs, namely, with single-stranded DNA (ssDNA) and phospholipid-polyethylene glycol (PEG), where both are validated as ingestible NIR fluorescent probes in C. elegans.Finally, we compare different feeding profiles showcasing the functionalized SWCNTs as NIR fluorescent real-time tracking agents inside the digestive tract of C. elegans.This work reveals new possibilities for the use of NIR fluorescent probes and nanosensors based on functionalized SWCNTs in C. elegans and other model organisms for a wide variety of bioimaging applications.

Platform for Imaging Internalized SWCNTs in C. Elegans
In order to image SWCNTs internalized by C. elegans nematodes, we first established the microfluidics platform and optimized the working conditions.For our study, we chose two different functionalization classes for the SWCNTs to be used as model NIR fluorescent probes, namely, a singlestranded DNA functionalized SWCNT, specifically (GT) 15 -SWCNTs (Figure S1, Supporting Information), owing to the ease of sample preparation, [123,154] and phospholipid-PEG function-Scheme 1.Schematic illustration of the experimental setup and procedures.a) C. elegans loading into the microfluidic platform: 10-20 synchronized adult hermaphrodite worms are handpicked and incubated with SWCNTs and E. coli as a nutrition source, followed by manual loading into the microfluidic cell inlet tube, later to be used for imaging.b) Real-time feedthrough system setup: The microfluidic device, pre-conditioned and pre-loaded with C. elegans worms, is placed on the microscope stage and infused with a mixture of SWCNTs and E. coli as a nutrition source for scrutinized incubation and imaging.Objective and syringe pump illustrations were adapted from grabcad.com.
[157] Both the DNA and PEGylated-lipid SWCNT functionalizations had previously shown no adverse impact on the viability, well-being, and compatibility with C. elegans, [50] cells [158,159] and mice, [75,160] rendering them compelling imaging and sensing agents.
The working concentrations of the chosen SWCNTs had to be considered carefully, as low concentrations might result in weak signals below the detection limit of the imaging system, whereas high concentrations could give rise to substantial background fluorescence from SWCNT binding to the surface of the worms or to the microfluidic device and might affect the biocompatibility.Based on previous research, [50,161] the concentration for both types of functionalized SWCNTs was chosen to be 0.5 mg L −1 , for optimal signal-to-noise ratio in the NIR images and confirmed biocompatibility.For SWCNT fluorescence excitation, we used a 730 nm CW laser source, which resonates with the absorption of the predominant (10,2), (9,4), (8,6), and (8,7) chiralities (Figures S1c and S2c, Supporting Information), all of which emit in the NIR-II, specifically, at 1070 nm, 1120 nm, 1192 nm, and 1278 nm, respectively.Since the C. elegans worms are known to be photosensitive, [162,163] we had to evaluate the trade-off between obtaining high-quality data with sufficient excitation power for the SWCNT, and ensuring the safety and well-being of the C. elegans worms.To ensure a wide experimental framework consisting of two different types of functionalized SWCNTs with different excitation-emission responses and colloidal manifestation, multiple optical setups, and short and long imaging durations, we optimized our system to an excitation intensity of 570 mW (≈10 W mm −2 ) at the sample plane.The worm's safety was ensured first by using a NIR-I excitation wavelength (730 nm), which overlaps with the biological transparency window, so only a small amount of the energy is absorbed by the biological sample to generate heat or free radicals.Furthermore, we applied short exposure pulses (150-200 msec) with a duty cycle of ≈30% onto-off time, a convection cooling mechanism provided by the microfluidic device, and real-time behavior monitoring.
The SWCNTs were internalized by the worms via food-intake.The C. elegans worms ingest food through their mouth into their Pharynx by a peristaltic-like pumping action of the surrounding liquid containing E. coli bacteria, [164] where it passes through the Corpus to the terminal bulb.At the worm's terminal bulb, the grinder physically breaks the food, and it passes through the pharyngeal-intestinal valve into the lumen of the anterior intestines (Figure 1).For internalization, the C. elegans worms can be pre-incubated with the functionalized SWCNTs in a feeding medium consisting of M9 buffer and E. coli bacteria, prior to live imaging.
Inspired by the original design of the "Worm Spa" microfluidics, and to enable worm confinement, we fabricated a modified version of the microfluidic device suitable for our purposes, and subsequently refined the process of worm loading.The Polydimethylsiloxane (PDMS) microfluidic devices were cast from a prefabricated silicon wafer master mold and adhered to a microscope cover glass via a plasma process, thus providing an imaging accessible area of ≈2.5 mm × 10 mm, covering 32 wormconfining channels, each of which was 50 ± 10 μm in width and height.To prevent clogging of worms in the fluidic channels during the loading process, we loaded no more than 20 adult hermaphrodite C. elegans worms in one experimental sequence that were individually handpicked (Figure S4 and Movie S1, Supporting Information) from a pre-synchronized N2-Bristol strain colony (Scheme 1a) and maintained in the loading medium.The worms were then loaded into the microfluidic device inlet and pushed into the confinement channels using a syringe pump system infusing M9 buffer (Scheme 1b; Figure S5a and Movie S2, Supporting Information).The loading and infusion protocols were optimized to prevent leaks and blockages, and to smoothly guide individual worms into several of the 32 confinement channels.When the optimal positioning of the worms within the central part of the confinement channel was achieved (Figure S5b and Movie S3, Supporting Information), the worms were ready for imaging.This procedure was successfully reproduced throughout our experiments, allowing us to choose suitable candidates from several confined worms in each experimental session.
Given the controlled spatiotemporal microfluidics environment we established, in addition to pre-incubation of the worms with SWCNTs for internalization, our platform can also enable real-time feeding during the live imaging, while the worms are confined within the channels (Scheme 1b).In this setup, we were able to repeatedly follow several confined worms in low magnifications, leaping from one worm to another until we detected NIR fluorescent signals within the worm.Once zooming-in on a preferable worm, we continued monitoring its digestion process under high magnification throughout the feeding process, allowing us to scrutinize one worm per feeding session.
For fluorescence microscopy, we used a wide range of optical objectives for imaging ranging from 4× up to 100×, to cover as many imaging scales and configurations as possible.Lower magnifications, such as the 4× and 10×, were normally used to scan the areas of interest and identify candidate worms, whereas the higher magnifications served to focus on the optical plane of interest and subsequently image the internal organs with the internalized NIR fluorescent SWCNTs.For the C. elegans autofluorescence imaging, we chose to demonstrate two commonly used excitation wavelengths, corresponding to the green fluorescent protein (GFP) channel, excited at 480 nm, and 4′,6-diamidino-2-phenylindole (DAPI) channel, excited at 365 nm, having emission central wavelengths of 530 nm and 430 nm, respectively.The autofluorescence in the visible range was excited using an LED illumination system, and its emission was imaged with an EM-CCD camera mounted on an inverted fluorescence microscope which was also used to record a transmitted brightfield image.We used a separate InGaAs camera mounted on the same microscope for the NIR fluorescence imaging, where all the different images were later overlaid.
Lastly, we optimized the live imaging frame rates and exposure times for optimal signal-to-noise ratio (SNR) in each experimental configuration.Frame rates varied in the range between 2 frames per second for high temporal resolution imaging, and a frame every 10 s for stress-free long time-lapse imaging.

Internalized SWCNTs Detection and Colocalization
We set to establish a reliable system for detecting, visualizing, and monitoring the SWCNT particles that were internalized by the worms.It was particularly crucial to devise a technique for differentiating intra-corporeal NIR fluorescent SWCNTs from those that were outside the worms.Such a platform enables further exploration of the different optical and spatiotemporal aspects of SWCNT internalization in C. elegans worms.
The C. elegans is a typical roundworm whose digestive tract is a simple tube with an asymmetrical "twist" in the intestine between anterior and posterior segments (Figure S6, Supporting Information). [165]In many cases, such as in our work, the main region of interest and scrutiny is the anterior part of the intestinal lumen, since it is the most dynamic section and holds most of the feeding mechanism.Within the anterior segment of the worm, where the pharynx, pharyngeal-intestinal valve, and a portion of the intestinal lumen are located, the intestine tube is mainly situated along its central longitudinal axis (Figure 1).Therefore, we expected to find, in a captured image of the longitudinal center cross-section, fluorescent SWCNTs that are exclusively located inside the worm's body.Any other NIR fluorescent signals, as we realized, were barely discernible.Hence, we assumed them to be outside the focal plane as external entities.
The proposed technique is straightforward.Initially, we opted for a conspicuous and easily identifiable morphological feature of the worm, such as the terminal bulb grinder.This component is situated symmetrically in the center of the pharynx's vertical axis (in-plane), thereby enabling qualitative z-positioning, which brings the focal plane approximately to the central vertical axis.Based on this, we can deduce that the radial edges of the worm lie within a range of ≈± 40 μm, considering that an average adult hermaphrodite measures up to 80 μm in diameter at its broadest point, [166] and the worm is confined in a squared microfluidic channel of 50 ± 10 μm (Figure 1).
For the next step, we set this central axis plane as z = 0 and took a stack of 11 images at different z positions of the worm, ranging from + 50 μm to − 50 μm relative to the z = 0 plane in 10 μm intervals, as illustrated for 20× magnification with (GT) 15 -SWCNTs (Figure 2a; Figure S7 and Movie S4, Supporting Information).Subsequently, we identified a region of interest (ROI) showing SWCNTs fluorescence in the NIR channel, and evaluated the intensity of a vertical line that intersected the fluorescent region (Figure 2b) for every z-position image (Figure 2c).Then, for each slice, we established and assessed the full width at half-maximum (FWHM) of the primary intensity profile, which served as a marker for determining the z-position of the fluorescence image that was "in-focus".Choosing the minimum FWHM value from the entire stack of images (Figure 2c,d), we obtained the optimal image, which in this case was at z = − 10 μm plane.A similar process was reproduced for a different imaging setting with the 60× objective, in which case the optimal image was found to be at the z = − 20 μm plane (Figure 2e).Considering that we qualitatively set our z = 0 plane based on visual assessment of the terminal bulb grinder, which essentially may yield a certain spatial tolerance, we could still ascertain the presence of the SWCNTs inside the body of the worm based on this technique even with a ± 20 μm offset.
To further reinforce the level of confidence in detecting SWC-NTs within the worm, a complementary measurement was conducted, based on the assumption that internalized SWCNTs have similar spatial dynamics as the worm, whereas the position of the extra-corporeal SWCNTs is independent with respect to the worm motion.A worm with candidate SWCNTs was temporally imaged for approximately two minutes, and the position of the labeled particles was tracked relative to a fixed point in the microfluidic space and to the worm's terminal bulb grinder (Figure 2f,g; Movie S5, Supporting Information).As expected, it can be clearly seen that the internalized SWCNT candidates, marked in blue, have an almost identical motion as the worm's grinder, marked in red, and are entirely independent with respect to the green mark, which labels external SWCNTs that appear static.Slight differences can be observed between the internal SWCNTs and the grinder due to the worm's body contractions, elongations, and bends.These observations were consistent in every experiment of real-time feeding and reassured a proper monitoring of internalized SWCNTs.Using the two complementary methods, we could relatively easily detect and colocalize the SWCNTs' NIR fluorescence within the worm's digestive tract.

In Vivo SWCNT Imaging
Having established a method for detecting and confirming SWC-NTs internalized within C. elegans, we proceeded to optimize the optical imaging configuration.We examined typical images of the internalized SWCNTs captured at various magnifications and compared the performance of the two different SWCNT functionalizations, namely (GT) 15 -SWCNTs (Figure 3a) and DPPE-PEG-SWCNTs (Figure 3b).One immediate observation was that the DNA-functionalized SWCNTs have a larger tendency to cluster within the worm compared to the PEGylated SWCNTs.This effect is attributed to the relatively larger hydrophilic layer of the PEG functionalization around the SWCNT, which prevents individual SWCNTs from coming in close proximity with the surrounding ones.This fact is also evident in the feeding medium (Figure S8, Supporting Information), where comparing the two classes of functionalized SWCNTs at the same concentrations, they exhibit distinct morphological features in their dispersion.Since ssDNA-suspended SWCNTs are prone to clustering, as can be seen consistently in multiple cases throughout the experimental work, the ability to optically detect them with smaller magnifications increases.Indeed, we managed to detect (GT) 15 -SWCNTs clusters of 10 − 20 μm in size, with a magnification as low as 10× (Figure 3a).[169] Detecting and imaging the DPPE-PEG-SWCNTs presented a distinct challenge.Their intake by the worms was less often detectable, since they remained as individual particles or as very small bundles in the order of 1 μm, and could only be identified with the higher magnification setups.On rare  occasions, small DPPE-PEG-SWCNTs bundles could be seen at 60× magnification, as shown in the worm's Metacorpus (Figure 3b).However, for most cases, the detection and imaging of DPPE-PEG-SWCNTs was made possible only through the 100× setup (Figure 3b).Still, the DPPE-PEG-SWCNTs benefit from higher fluorescence intensity compared to their (GT) 15 -SWCNTs counterparts and have better single-particle dynamics.

Immobilization Effect on Colocalization Mismatch of Autofluorescence and NIR Channels
Multi-spectral imaging in C. elegans worms may become a challenge in an uncontrolled spatial environment since free-roaming worm locomotion in the presence of food or other chemotaxis cues is hectic.In order to exemplify the role of worm restriction by the microfluidic device, we compared the locomotion behavior of C. elegans in the spatially controlled and non-controlled environments using (GT) 15 -SWCNTs as internalized probes.Indeed, the free-roaming worm covered a larger area within a field of view (FOV) compared to the worm that was confined within a microfluidic channel (Figure 4a,b).In many cases, the free-roaming worms exited the static FOV within a short period of time, even when observed using a relatively small magnification of 20×, rendering the worm's tracking and overlay imaging a challenging task.In contrast, the worms inside the microfluidic device were confined to a small area and were successfully imaged for at least 5 min (Figure 4c,d).
Rapid movement of free-roaming worms poses an additional challenge on multi-spectral imaging (Figure 5).Specifically, physically rotating a filter turret within a microscope can take up to several seconds, and during the acquisition of consecutive frames of brightfield, visible fluorescence, and NIR fluorescence, the worm can substantially change its position.As a result, when overlaying the entire multi-spectral stack (Figure 5a), the combined image unravels the colocalization mismatch challenge (Figure 5b).In contrast, the limited degree of freedom of a confined worm only allowed for minor bends, which resulted in a minor mismatch between the fluorescent channels, mainly the GFP and DAPI, to the brightfield and NIR channels (Figure 5c).Although this undesirable effect was still observed in the confined worm, it was negligible compared to the freeroaming worm imaged under similar conditions (Figure 5b,c), which exhibited significant mismatches that made colocalization impossible.Moreover, the confinement in the microfluidics platform allowed us to use a much higher magnifica- tion for multi-spectral overlaid imaging, such as 100×, which otherwise was not possible in a non-anesthetized free-roaming worm.
The NIR fluorescent signal of internalized SWCNTs could now be detected in the worm's terminal bulb grinder and right pass the pharyngeal-intestinal valve, and could be overlaid with the worm's GFP and DAPI autofluorescence channels and the bright-field image (Figure 6).The FWHM of the NIR fluorescence clusters, in this case, was as small as 1 μm.This result highlights the capability to image functionalized SWCNTs in the NIR range, through the strong autofluorescence interference of the worm's intestinal granules.

Feeding Profiles Effect on SWCNT Internalization
The preliminary strategy adopted to internalize SWCNTs within C. Elegans worms relied on feeding methods used in earlier studies. [50,170,171]Specifically, worms were pre-incubated for a few hours in a feeding medium that contained E. Coli and SWC-NTs, before loading them into the microfluidics channels and acquiring images over time.We performed pre-incubation cycles with either (GT) 15 -SWCNTs or DPPE-PEG-SWCNTs for at least 2 h.Although in both cases we observed a successful internalization of the SWCNTs, the (GT) 15 -SWCNTs were more easily detectable as they were present as larger clusters deep within the intestinal bulb, in the intestinal "twist" and posterior section, as seen repeatedly in multiple worms (Figure 7a; Figures S6 and S9, Supporting Information), whereas the DPPE-PEG-SWCNTs were found as smaller clusters in the corpus and terminal bulb parts of the worm's pharynx (Figure 7a).The (GT) 15 -SWCNTs pre-incubation results were similar for longer periods of pre-incubation, namely, 3 and 4 h (Figure 7b), showing accumulation in the pharyngeal-intestinal valve and the anterior part of the intestinal lumen as well as other internal sections (Figure S9 and Movie S6, Supporting Information).The clusters of the (GT) 15 -SWCNTs that appeared following a few hours of pre-incubation and were observed irrespective of the feeding duration can be attributed to the morphological characteristics of the ssDNA-suspended SWCNTs as were first evident in the feeding medium itself (Figure S8, Supporting Information) and could hint on similar binding affinities within the C. elegans digestive tract.Another reason can be linked to the gut microenvironment, consisting of a rich presence of proteins and other biomolecules within the intestinal lumen, such as C-type Lectin or mucins. [172,173]This can explain the differences we observed between the pre-incubation results of the (GT) 15 -SWCNTs and the DPPE-PEG-SWCNTs, since for the latter, the PEG corona can prevent nonspecific adsorption, [174] and therefore cluster formations are less pronounced.
In addition to pre-incubation with SWCNTs, we used a different internalization approach, in which we first loaded C. Elegans worms into the microfluidics device and exposed them to a feeding medium containing E. Coli and SWCNTs, while monitoring in real-time the intake and digestive progression of the SWCNTs for 2 h (Figure 7c).The SWCNTs' fluorescence could be detected in the meta-corpus and terminalbulb as early as 15 min after the introduction of the SWC-NTs, for both the (GT) 15 -SWCNTs and the DPPE-PEG-SWCNTs, where fluorescence spots as small as 1 μm could be visualized.Throughout the imaging duration, we continuously observed intake of SWCNTs passing through the worm's pharynx and throughout the pharyngeal pathway, without large clustering in the worm's corpus, isthmus or terminal bulb, probably owing to the mechanical and structural properties of the pharynx where muscular walls create strong contractions during pumping. [164]Moreover, the mucus secreted in the pharynx provides a barrier to the wall's lining and may contribute to the prevention of such aggregations. [175,176]While the DPPE-PEG-SWCNTs did not form large clusters, rendering their tracking somewhat more challenging, the (GT) 15 -SWCNTs began to accumulate and cluster in various internal segments of the worm's intestinal lumen ≈90 min following their introduction, in agreement with the results of the pre-incubation internalization.Compared to the pre-incubation approach, in which the SWCNTs intake process cannot be captured, realtime imaging of the confined worms enabled us to monitor the intake dynamics of SWCNTs.This platform thus provided us a new opportunity for scrutinizing and tracking the feeding process in real-time, in a controlled spatiotemporal setting.

Worm's Digestive Path Tracking
One of the advantages of using the microfluidics device for worm confinement and live imaging is the ability to identify and monitor a trajectory of functionalized SWCNTs through the digestive tract of the worm.Specifically, we successfully monitored the translocation of (GT) 15 -SWCNTs through the anterior segment of the intestinal lumen (Figure 8a; Movie S8, Supporting Information).During a period of 15 min, we observed a small cluster of SWCNTs, of ≈5 μm in diameter (blue circle), separated and displaced from a source cluster of ≈10 μm in diameter (red circle).The SWCNTs then progressed through the intestine lumen toward the posterior section of the worm's intestinal lumen.The source cluster remained relatively stationary within the worm frame of reference despite the clusters' motion, resulting from the worm's motion back and forth along the fluidic channel.We measured the distance (Figure 8b) between the centers of the source cluster and the moving cluster as a function of time (green line) in the frames in which both were considered to be in focus according to our previously established criterion (Figure 2).In addition, we determined the distance between both the source cluster (red) and the moving cluster (blue) from a stationary reference point on the microfluidic device (yellow).The distance between the moving cluster and the source cluster exhibited a steady and continuous increase up to 120 μm.This distance is roughly equivalent to 13% of the average length of the adult hermaphrodite intestine and corresponds to a digestion speed of 8 μm min −1 .Studies have shown that the defecation cycle time of an adult C. elegans is 45-50 s. [177,178] However, transport and defecation times through an average adult C. elegans intestine depends on several factors, such as the type and quantity of food consumed, the age, health of the worm, and the environmental conditions.Specifically, when the food medium contains non-food particles, and the uptake is exhausted, particles start to accumulate within regions of the intestinal lumen, and the defecation process slows down to time scales of 15 min and higher. [179]These previous studies are in good agreement with the time scale measured in this work.
While the (GT) 15 -SWCNTs showed a higher tendency to cluster compared to the DPPE-PEG-SWCNTs, they enable detection and tracking in lower magnifications imaging, benefiting from a larger FOV.The DPPE-PEG-SWCNTs, on the other hand, could only be detected and imaged using the highest magnification 100x objective.This resulted in a relatively small FOV, of ≈115 μm in width.Following 90 min of C. elegans feeding with DPPE-PEG-SWCNTs, we observed several NIR fluorescent signals, of ≈1 μm in diameter, and monitored their position (Figure 8c).Despite their strong fluorescence signal, accurately measuring the displacement of the NIR fluorescent cues was challenging due to their small size, the limited focal depth, and the small FOV.Nevertheless, we were able to qualitatively capture the trajectory of the DPPE-PEG-SWCNTs over a 25-minute interval.We observed distinct organelles within the worm, such as a portion of the anterior gonad sheath (indicated by a blue dashed line) and two oocytes (indicated by a green dashed line).We also marked two points of interest (labeled as I and II) and monitored the movement of the NIR signals relative to these reference points.We could clearly see that the DPPE-PEG-SWCNT probes move deeper within the intestinal lumen.
We demonstrated that both the (GT) 15 -SWCNTs and the DPPE-PEG-SWCNTs could be used for tracking within the digestive tract of the C. elegans worm confined in a microfluidics device.Although the DPPE-PEG-SWCNTs have higher fluorescence intensity, the (GT) 15 -SWCNTs were preferable for the detection and monitoring of food trajectory owing to larger FOV given the ability to use lower magnification for fluorescence imaging.In terms of temporal resolution, our platform provided a satisfactory temporal resolution for the purpose of tracking SWCNT probes within the intestine.However, for specific requirements such as monitoring pharyngeal pumping behaviors where short temporal dynamics dominate, or tracking SWCNT probes in conjunction with other fluorescent dyes where heavy multispectral overlaying is needed and colocalization is challenging, the imaging setup should be adapted accordingly by, for example, modifications to the microscope filter turret and dichroic mirrors, optical parallelization or even a more sensitive camera sensor which allows shorter exposure times.

Conclusion
Biological and biomedical research, among many others, has long aimed to visualize and analyze the physiological processes of living organisms in real-time.NIR imaging techniques have recently become a valuable non-invasive tool for in vivo imaging, due to the distinct characteristics of NIR light, including low tissue autofluorescence and the ability to penetrate deep into tissues.Although using NIR fluorescent materials also has its challenges, such as specialized imaging equipment, and possibly higher costs, with careful evaluation of the research needs, one can benefit extensively from the unique properties of NIR imaging.][182] Due to its ease of genetic manipulation and a significant level of molecular and cellular similarity with humans, this organism is highly advantageous for research purposes. [1,2]Although the strong autofluorescence sourcing from the C. elegans internal organs, such as the uterus and intestine, in the entire visible spectrum is extensively harnessed for many research fields, it covers most of the worm's body and in many experimental cases interferes spatially and spectrally with external fluorescent markers.However, C. elegans show no autofluorescence in the NIR spectral range (>900 nm), thus allowing the use of NIR fluorescent SWCNTs as efficient biomarkers, sensors, and tracking agents within the worm's entire volume.
The internalization of SWCNTs inside C. elegans organisms has been explored in the past for various research fields, yet in a very ill-controlled spatiotemporal setting.Rapid locomotion of the worms presented colocalization artifacts, and the worm's anesthetic immobilization affected the physiological parameters.
In our work, we assembled a spatially and temporally controlled NIR imaging platform that utilizes the advantages of using the wild-type C. elegans nematodes as an in vivo organism model with functionalized SWCNTs as NIR fluorescent markers, all in a microfluidic device used for worm's stress-free immobilization that does not require anesthesia.Owing to their biocompatibility, we chose two widely used types of functionalized SWC-NTs as our NIR fluorescent markers, single-stranded DNA SWC-NTs ((GT) 15 -SWCNTs), and phospholipid PEGylated SWCNTS ((DPPE)-PEG (5 kDa)-SWCNTs).We internalized the functionalized SWCNTs in the C. elegans worms through food intake with E. coli bacteria (OP50) broth, following their immobilization inside microfluidic device confinement channels.First, we showed a method to correctly detect and label candidate SWCNT particles within the C. elegans digestive tract.The proposed method utilized Z-stacking NIR imaging to measure "in-focus" fluorescent signals.Furthermore, we measured the SWCNTs' synchronized motion with the worm's body to reassure the intra-corporeal presence of the SWCNTs.Live NIR and brightfield overlaid images of the worm's pharyngeal corpus and terminal bulb grinder were used to set the focal plane for SWCNTs intake and tracking.Once we were able to reproduce the detection of fluorescent SWCNTs within the worm's pharynx and intestine, we showed the ability to image the SWCNTs with various optical settings.We identified large cluster formations of (GT) 15 -SWCNTs in low magnifications (10×), in contrast to small cluster formation of (DPPE)-PEG (5 kDa)-SWCNTs which could only be detected in higher magnifications (> 60×).
While spectral separation of the SWCNTs' NIR fluorescence and the worm's visible autofluorescence (at the GFP and DAPI channels) was previously proven, [50] it was never before tested without the use of anesthetic means of immobilization.In our study, we utilized a microfluidic device and created a stress-free environment for the free-roaming C. elegans worms, allowing us to successfully image both the worm's visible autofluorescence and the SWCNTs' NIR fluorescence spectral ranges and colocalize them with minimal misalignment.In contrast, visible and NIR dual channel imaging of un-anesthetized worms on agar-treated glass slide suffered from significant misalignment between NIR, DAPI and GFP channels, and colocalization was nearly impossible in most cases.
Another new and important functionality we demonstrated in our work is real-time NIR imaging and tracking of SWCNT intake in un-anesthetized and yet immobilized C. elegans worms.Initially, we observed and characterized the intake of ssDNA SWCNTs and PEGylated SWCNTs during live imaging, and were able to identify the fluorescent SWCNT particles in the worm's corpus and terminal bulb from the first minutes of the feeding process.Our results from the 2-hour real-time feeding imaging were in good agreement with the intake picture we observed in worms that were pre-incubated with SWCNTs for 2-4 h.Most notably, after ≈2 h of feeding (in both methods), the (GT) 15 -SWCNTs started accumulating in the anterior parts of the worm's intestinal lumen, creating distinct trackable clusters.The DPPE-PEG-SWCNTs, on the other hand, showed little to no clustering throughout the real-time feeding and were detectable only with the high magnification (100×), which, again, matched our pre-incubation (>2 h) observations, where distinct clusters were hardly noticeable.Further, we also succeeded in tracking and measuring the (GT) 15 -SWCNTs trajectory through an anterior part of the intestine owing to the easily detectable ssDNA SWC-NTs clusters in lower magnifications with larger FOV.The DPPE-PEG-SWCNTs intestinal trajectories were indeed recognized due to their higher fluorescence signal; however, their small spot sizes limited our tracking capabilities to higher magnifications and smaller FOV.
In summary, we were able to demonstrate a method to image, detect, and track two types of functionalized NIR fluorescent SWCNTs that were internalized via food intake into wild-type C. elegans worms confined within a microfluidic device.We showed improved colocalization of multi-spectral fluorescence imaging and revalidated the spatial freedom of NIR fluorescence probes within a highly visible autofluorescence environment of the C. elegans.Finally, we showcased a spatiotemporally controlled platform and a method to track and characterize, with real-time feeding, different intake behaviors of fluorescent nano-sensors using two types of SWCNTs, the (GT) 15 -SWCNTs and the DPPE-PEG-SWCNTs, each expressed different uptake, accumulation, detectability and dynamics throughout the feeding process owing to their specific molecular complexes and optical characteristics.
Our work opens new experimental and research possibilities previously limited or impeded, in the fields of biomedi-cal NIR imaging of live organisms such as C. elegans, specifically with functionalized single-walled carbon nanotubes as fluorescent agents.Other types of functionalized SWCNTs can be further explored for feeding behaviors related research, or expanded to other fields, such as developmental or neurological studies where SWCNTs are introduced in different methods.Moreover, the microfluidic devices and the imaging setups can be tailored and rendered fit to other relevant needs with increased performance and reproducibility, such as conducting multiple parallel worms imaging with large group quantitative statistics, by customizing the microscopy setup and programming a motorized state such that consecutive images can be captures from several regions of interest including different areas of the same worm, or interesting areas in different worms in adjacent channels.Further, the temporal resolution can also be increased by modifications to the microscope filter turret, dichroic mirrors, optical parallelization, or even a higher sensitivity grade of the camera sensor, which allows shorter exposure times.This will support experiments with fast dynamics analysis and heavy multispectral overlaying with challenging colocalization.

Experimental Section
SWCNT Functionalization: For DNA functionalization, 1 mg of SWC-NTs (HiPCO, NanoIntegris) was suspended with 2 mg single-stranded (GT) 15 DNA oligonucleotides (Integrated DNA Technologies) in 0.1 m NaCl via bath sonication (Elma P-30H, 80 Hz for 10 min), followed by two cycles of direct tip sonication (QSonica Q125, 3 mm tip, 4 W) for 20 min in an ice bath.Aggregates and impurities were then separated from the individually suspended SWCNTs by centrifuging the sample twice for 90 min at 16 100 rcf.After each centrifugation step, 80% of the supernatant was collected, and the pellet was discarded.
For Phospholipid PEG functionalization, SWCNTs were first suspended with sodium cholate (SC, Sigma Aldrich), which was later removed by dialysis in the presence of phospholipid-poly(ethylene-glycol) (1,2dipalmitoyl-sn-glycero-3-phosphoethanolamine-PEG (5 kDa), Avanti Polar Lipids) using methods similar to those previously published. [183]In brief, SWCNTs (HiPCO, NanoIntegris) were first suspended with 2% wt SC via bath sonication (80 Hz for 10 min), followed by two cycles of direct tip sonication (6 mm tip, 12 W) for 30 min in an ice bath.To remove SWCNT aggregates and impurities, the suspension was ultracentrifuged (OPTIMA XPN-80, Beckman-Coulter, 41,300 rpm for 4 h, 4°C), the top 80% of the supernatant was collected, and the pellet was discarded.Subsequently, a surfactant exchange was performed.To this means, a mixture of SC-SWCNTs (40 mg L −1 ) and 2 mg mL −1 DPPE-PEG (5 kDa) was dialyzed against water using a dialysis membrane (Spectra-Por Float-A-Lyzer G2, Spectrum labs, MWCO: 0.5-1 kDa, 5 mL).Dialysis was performed for 7 days with multiple water exchanges to remove SC and allow the adsorption of the DPPE-PEG onto the SWCNTs.
SWCNT Characterization: Successful suspensions were validated by recording their ultraviolet-visible-near-infrared (UV-vis-NIR) absorption spectra, using a UV-vis-NIR spectrophotometer (Shimadzu UV-3600 PLUS), in a 1 cm path length quartz cuvette (Starna cells Inc.) where sharp distinguishable peaks indicated a successful suspension (Figures S1a and  S2a, Supporting Information).The concentration of (GT) 15 -SWCNT and DPPE-PEG-SWCNTs were determined spectroscopically with an extinction coefficient of  632 nm = 0.036 L mg −1 cm −1 [184,185] A redshift relative to the initial SC-SWCNT suspension of the DPPE-PEG-SWCNTs was also observed, indicating the surfactant exchange. [186]or SWCNT fluorescence characterization, the excitation-emission map was recorded, (Figures S1b and S2b, Supporting Information).Samples of suspended (GT) 15 -SWCNT and DPPE-PEG-SWCNTs were diluted to 0.5 mg L −1 in NaCl 0.1 m and were added to the wells of a 96 well plate.The samples were illuminated with a supercontinuum white-light laser (NKT-photonics, Super-K Extreme) coupled to a tunable bandwidth filter (NKT-photonics, Varia, Δ = 20 nm) scanned between 500 nm to 840 nm with an excitation time of 2 s per wavelength, a 1 nm wavelength step size, and 20 mW (at 730 nm) intensity.Emission spectra were recorded on an inverted fluorescence microscope (Olympus IX73) coupled to a spectrograph and a liquid-nitrogen cooled InGaAs detector (HRS-300SS, and PyLoN-IR 1024-1.7,Princeton Instruments, Teledyne Technologies) or adapted from the excitation-emission map (Figures S1c and S2c, Supporting Information).
C. Elegans Growth and Maintenance: The N2 Bristol strain Caenorhabditis Elegans (C.elegans) was used as a wild-type animal model in this study (kindly received from Prof. Limor Broday, Tel-Aviv University, Israel).The worms were grown and maintained on standard nematode growth medium (NGM) plates seeded with Escherichia coli (E.Coli) strain OP50 as a food source at 22 ± 1°C.For maintaining the stock culture, the worms were transferred to fresh NGM plates every 5-7 days, similarly as described in previous works. [187,188]ge-synchronized worms were obtained using the alkaline hypochlorite method. [49,189]In brief, cultured worms were collected from 2-3 NGM plates by repeating twice a wash-and-collect cycle followed by the addition of diluted bleach (1%), thus, terminating adult worms and leaving behind a population of eggs which were then transferred to fresh NGM plates for hatching and culturing period of 72-96 h at 22 ± 1°C, resulting in synchronized adult hermaphrodites.
Microfluidic Device Fabrication and Assembly: Microfluidic devices (MFD) for longitudinal studies of C. elegans worms were fabricated based on a slightly modified design of the "WormSpa" MFD developed by Kopito and Levine. [150,151]Standard soft-lithography techniques [190] were used to create a master mold for the polydimethylsiloxane (PDMS; Sylgard 184, Dow Corning Corp) structures consisting of 9 individual microfluidic cells, each containing 32 channels.Fabricated at the Tel Aviv University Center for Nanoscience and Nanotechnology facilities, a 4′′ silicon wafer was spin-coated with SU-8 photoresist to obtain a uniform height of ≈50 μm.This layer was exposed to UV light through a patterned 5″ photomask to achieve the microfluidic cell-specific design features.This layer was then etched and cleaned to remove the unexposed SU-8 photoresist. [191]The Silicon master mold was adhered to a 120 mm × 20 mm borosilicate culture plate (SCHOTT) to be used for PDMS casting.
The SU-8 master mold was then used to fabricate the PDMS microfluidic devices.The PDMS was mixed at a ratio of 10:1 (base to curing agent), degassed in a vacuum chamber for 4 h to remove bubbles, and slowly poured onto the master to a height of ≈7 mm.The PDMS layer was cured at 70°C for 2 h and peeled off from the SU-8 mold after cooling.Individual cells were cut out and fluid interconnecting holes were punched in manually with a 1.2 mm rapid core microfluidic punch (PT-T983-12, Darwin microfluidics).The PDMS cells were cleaned with tape and irreversibly bonded to a 24 mm × 50 mm #1.0 microscope cover glass (BN1052431STC, Bar-Naor Ltd.) via low-pressure oxygen plasma treatment at 100 W and 0.5 mBar for 60 s (Atto, Diener electronic).
The fluid interconnectivity to the devices was accomplished by assembly of a 90°stainless steel 18-gauge PDMS coupler (PN-BEN-18G, Darwin microfluidics) that was inserted into the fluid interconnect holes, and each was connected to a 1.02 × 1.78 mm tubing (SA-AAD04127, Tygon ND 100-80, Darwin microfluidics).Lastly, the inlet and outlet ports of the tubing were connected to an 18-gauge Blunt-end Luer Lock Syringe Needle (AE-18G, Darwin microfluidics).
Microfluidic Device Treatment and Handling: Prior to loading the microfluidic device with worms, it was surface treated and primed. [192]A freshly assembled MFD was connected to a syringe pump system (NE-4000X, New-Era syringe pump systems Inc.).Then, a solution of 5% w/w Pluronic F-127 (Sigma) in ddH 2 O was injected into the MFD at 250 μL min −1 for 20 min and allowed to be adsorbed onto the PDMS surfaces of the micro-chamber.The MFD was then washed with ddH 2 O at 500-700 μL min −1 for few minutes.The MFD was primed with M9 buffer (3 g KH 2 PO 4 , 6 g Na 2 HPO 4 , 5 g NaCl, 1 mL 1 m MgSO 4 , ddH 2 O to 1 liter, sterilize by autoclaving) at 500 μL min −1 for 10 min.Using the syringe pump, the M9 buffer was aspirated and emptied from the micro-chamber and inlet tube, freeing the path for worms loading fluid.
Each Microfluidic device was re-used 1-2 times.A bleach solution (3-5%) was injected into the MFD after completing the experiment, using the same setup as aforementioned, the bleach was infused in both directions of the MFD (inlet and outlet) for 5 min respectively at a flow rate of 500 μL min −1 .A thorough wash with ddH 2 O was then applied, again, in both directions for 5 min each at the same flow rate.
Feeding Experimental Setup: Feedthrough experiments were performed such that the pre-loaded MFD was scrutinized under the imaging microscope while feeding and incubating the worms with the functionalized SWCNTs occurred in real-time.A culture of E. coli OP50 grown in LB media was mixed with SWCNTs at a 1:1 ratio.A syringe pump system (New-Era) infused the mixture into the microfluidic device with a flow rate of 250 μL min −1 for 60 s, followed by a constant flow of 5 μL min −1 all throughout the feeding and incubation time.In order to remove eggs from the vicinity of the worms into the egg-collection area, a 15 s pulse of 250 μL min −1 was initiated every 30 min.Once the SWCNTs were identified within the worm's digestive tract, a wash medium consisting of only M9 buffer was dispensed into the MFD with a flow rate of 400 μL min −1 for 2 min to wash away any undesired SWCNTs that were outside the worm, followed by a constant flow of 20 μL min −1 throughout the imaging sequence.
Another type of experimental setup used in this work utilizes worms that were already incubated with SWCNTs prior to their imaging sequence.Adult hermaphrodite worms (≈10-20 subjects) were handpicked from a synchronized culture NGM plate (Figure S4, Supporting Information) and placed inside a 200 μL loading medium consisting of a 50% E. coli and 50% 1 mg L −1 SWCNT suspension (either (GT) 15 -SWCNT or DPPE-PEG-SWCNT).The loading medium with the worms was then placed in a shaker for incubation between 2-4 h at 25°C, resulting in pre-incubated worms ready to be loaded into the microfluidic chamber.
Worms Loading and Mounting: For free-roaming worms fluorescence imaging, worms were mounted on a 3% agarose gel pad on a glass slide, covered with an 18 mm × 18 mm #1.0 microscope cover glass (Marienfeld), and sealed with wax.As for the binocular imaging, worms were used in their original culture NGM plate.
For loading the worms into the microfluidic device, the 200 μL of incubated worms in the loading medium were dispensed into the microfluidic device inlet tube with a pipette.The inlet tube was then reconnected to the pump system, and M9 buffer was infused into the microfluidic device at a rate of 450 μL min −1 and was supervised through the binocular microscope camera.On the occasion that worms were entangled or clustered, a few iterations of withdrawal and infusion were performed until satisfactory loading was achieved (Figure S5, Supporting Information).The microfluidic device with the loaded worms was transferred to the fluorescence microscope setup for live imaging.
Imaging: The Microfluidic device handling and worms loading procedure was aided by a Binocular reverse microscope (Nikon SMZ800N, Objective Nikon plan 1x WD78) with a digital CMOS camera (PL-D752, Pixelink) and Pixelink CAPTURE software.
For the real-time fluorescence imaging, images were taken via an inverted fluorescence microscope (Olympus IX83) using five different objectives: UPLFLN4X/0.13,UPLFLN10X/0.3,LUCPLFLN20X/0.45,LUC-PLFLN60X/0.7 and UPLFLN100X/1.3.Visible autofluorescence was excited with a LED illumination system (CoolLED, pE4000), choosing 2 different channels covering the visible excitation wavelengths range for DAPI and GFP (365 nm and 460 nm, respectively).Autofluorescence was imaged using two different filter cubes, covering the visible wavelength range of DAPI (Chroma, 49 000), and GFP (Chroma, 49 002).Fluorescence in the visible wavelength range was detected with an EMCCD camera (Andor, iXon Ultra 888).The SWCNT-fluorescence was excited by a 730 nm CW laser (MDL-MD-730-1.5 W, Changchun New Industries) with a neutral density filter (ND02B, 63% transmission, Thorlabs) and an excitation power that was measured at the sample plane as 540, 570, 570, 570, 465 mW for the 4×, 10×, 20×, 60×, and 100× objectives, respectively.The laser excitation light was directed to the sample with a dichroic mir-ror (T900LPXXRXT, Chroma), and the NIR emission of the SWCNTs was detected after an additional 900 nm long-pass emission filter (Chroma, ET900lp) with an InGaAs-camera (Raptor, Ninox 640 VIS-NIR).Videos were taken at frame rates ranging from 2 frames per second down to 0.1 frames per second, depending on the experiment.Exposure times and gains varied between the different wavelengths and objectives used in each experiment for the best SNR and image quality.
Image Processing: All images were processed by Fiji (ImageJ), and MATLAB (R2021b).In general, NIR images were subtracted by a background out-of-focus image, or de-speckled and adjusted for contrast and brightness, where DAPI, GFP, and BF images were slightly adjusted for brightness and contrast.The overlay of the images from the EMCCD and InGaAs cameras was done by adjusting the pixel sizes and the orientation, where overlay parameters of the two images were determined via maximization of the 2D autocorrelation of an identical mark frame taken with both cameras.The images were then cropped to the desired size, and the scale was calculated accordingly.SWCNTs tracking in the worms was performed by Fiji-TrackMate utility using the differences between Gaussian (DoG) and threshold methods.

Figure 1 .
Figure 1.Longitudinal cross-section brightfield image of a confined C. elegans worm showing the main anterior features: Pharynx (blue), Pharyngeal-intestinal valve (red), and anterior intestinal lumen (green).Width dimensions are explicitly shown on the vertical axis around the longitudinal mid-axis (yellow dashed line set as zero).Vertical and horizontal scaling is 1:1.

Figure 2 .
Figure 2. Validation of SWCNT internalization.a) Z-stack images of a C. elegans worm with (GT) 15 -SWCNTs taken with 20× objective, 10 μm apart, 11 slices in 3D view, where z = 0 marks the middle longitudinal z-slice of the images stack.b) NIR (top) and overlay (bottom) images at z = − 10 μm from the z-stack.A vertical line (yellow) indicates a ROI along which the intensity is analyzed across the different z-slices in the fluorescence channel.The scale bar is 100 μm.The colorbar represents the NIR fluorescence intensity from low (dark red) to high (bright yellow).c) Fluorescence intensity for all the z-slices along the selected ROI, where z = − 10 μm (dark red) intensity profile has the minimal FWHM.d) FWHM of the intensity profiles for every z-slice, highlighting the minimal value at z = − 10 μm.e) Z-stack images of a C. elegans worm with (GT) 15 -SWCNTs taken with 60× objective, 20 μm apart, NIR fluorescence images (top row) and brightfield-NIR overlay images (bottom row).The z = − 20 μm slice (yellow frame) is the slice in focus.The colorbars represent the NIR fluorescence intensity from low (dark red) to high (bright yellow), scale bar is 50 μm.f) Snapshots of a single worm confined within a microfluidic channel, tracking NIR fluorescent cues: suspected internalized (GT) 15 -SWCNTs (blue, A), terminal bulb grinder (red, B), suspected external (GT) 15 -SWCNTs (green, C) and a reference static point (white).The scale bar is 100 μm.g) Distance of the NIR fluorescent cues from the reference point as a function of time, showing a clear movement of the internalized SWCNTs and worm's grinder, whereas the extracorporeal SWCNTs remain static.

Figure 4 .
Figure 4. Tracking pattern comparison of SWCNT probes inside free roaming worms versus confined worms.a) Brightfield and NIR overlay 20× image at t = 0, with tracking ROI (blue) of (GT) 15 -SWCNTs inside a free roaming C. elegans.b) Tracking graph of the SWCNTs inside the ROI in a free roaming C. elegans over a period of 86 s. c) Brightfield and NIR overlay 20× image at t = 0, with tracking ROI (blue) of (GT) 15 -SWCNTs inside a confined C. elegans.d) Tracking graph of the SWCNTs inside the ROI in a confined C. elegans over a period of 288 s.Graphs colorbar is the time evolution in s, Images colorbar represents the NIR fluorescence intensity from low (dark red) to high (bright yellow), scale bars are 100 μm.

Figure 5 .
Figure 5. Worm confinement effect on colocalization mismatch of autofluorescence and NIR channels.a) 3D explosion of separated brightfield, NIR, DAPI, and GFP multichannel layering images acquired with 20× objective of a confined worm with (GT) 15 -SWCNTs, showing the transfer time between filters.b) Overlay image of a free roaming worm with (GT) 15 -SWCNTs, showing (in zoom) a significant mismatch of the worm head (red dashed line) due to the delay of the GFP and DAPI channels after the brightfield image capture.Scale bars are 100 μm.c) Overlay image of a confined worm with (GT) 15 -SWCNTs, showing a red dashed line marking the displacement of the worm during the time lapsed from first to last channel imaging seen in zoom, scale bars are 100 μm.

Figure 6 .
Figure 6.Spectral separation of NIR and visible wavelength images.Detailed images acquired with 100× objective of multichannel fluorescent imaging identifying two NIR fluorescent spots (yellow arrows) emitted by DPPE-PEG-SWCNTs, with the left one visible through both DAPI and GFP autofluorescence channels images.Left (NIR fluorescence), middle (overlay), and right (zoomed ROI) columns show the NIR (top row), NIR and DAPI channels (middle row), and NIR and GFP channels (bottom row) separately.Fluorescence column scale bars are 20 μm, overlay column scale bars are 20 μm, zoom column scale bars are 10 μm.The colorbars on the ROI zoom column images represent the NIR fluorescence intensity from low (dark red) to high (bright yellow).

Figure 7 .
Figure 7. Pre-incubation versus real-time feeding profiles.a) Overlay images (brightfield and NIR) of internalized (GT) 15 -SWCNTs and DPPE-PEG-SWCNTs taken with 60× objective after pre-incubation feeding of 2 h.Scale bars are 50 μm.b) Overlay images (brightfield and NIR) of internalized (GT) 15 -SWCNTs at 2, 3, and 4 h of pre-incubation feeding taken with a 100× objective, showing typical (GT) 15 -SWCNTs clustering past the pharyngealintestinal valve and the front segment of the intestinal lumen.Scale bars are 20 μm.c) Overlay images (brightfield and NIR) of real-time feeding of (GT) 15 -SWCNTs (top) and DPPE-PEG-SWCNTs (bottom) taken with a 100× objective during a 2-hour session showing dynamic ingestion of SWCNTs (yellow arrows) through the pharynx, and inside the intestinal lumen at t = 15, 45, 90, and 120 min.Scale bars are 20 μm.The colorbars represent the NIR fluorescence intensity from low (dark red) to high (bright yellow).

Figure 8 .
Figure 8. Spatiotemporal tracking of internalized SWCNTs.a) Displacement of internalized fluorescent (GT) 15 -SWCNTs over time, imaged with a 60× objective shown in NIR fluorescent channel (left column) and brightfield-NIR overlay image (right column).Markers and labels are for SWCNT source cluster position within the worm (red circle, A), SWCNTs of the displaced cluster position within the intestinal lumen of the worm (blue circle, B), a static reference point (yellow circle, O), and total displacement between the source cluster and the displaced cluster within the worm (green line).Scale bars are 50 μm.b) Total displacement measured over time between the SWCNTs from the source cluster and the displaced SWCNT cluster (green), the SWCNT source cluster relative position to the reference point (red), and the displaced SWCNT cluster relative position to the reference point (blue).c) Overlay of brightfield and NIR fluorescence snapshots taken with a 100× objective showing a right-to-left trajectory of internalized DPPE-PEG-SWCNTs relative to reference features within the C. elegans worm.I is a of reference the gonad sheath (blue dashed curve), and II is a point of reference between the front oocytes (green dashed circles).The pharyngeal-intestinal valve is marked by a red curve.Scale bars are 20 μm.All image colorbars represent the NIR fluorescence intensity from low (dark red) to high (bright yellow).