A Subset of Circulating Hemocytes Expresses Genes Indicating Neural Precursor Identity
Thanapong Kruangkum, Kenneth Söderhäll, Irene Söderhäll
TL;DR
A small percentage of blood cells in crayfish express genes linked to neural development, suggesting they may act as neural precursors in adult neurogenesis.
Contribution
First demonstration that neural lineage markers are expressed in a distinct population of crustacean hemocytes.
Findings
Around 1% of circulating hemocytes express neural lineage marker transcripts.
Neural marker transcripts co-express with serotonin receptor 1 in some hemocytes.
Treatments like serotonin and brain injury increase the proportion of neural marker-positive hemocytes.
Abstract
Adult neurogenesis in crayfish has been shown to require progenitors from an external source linked with the hematopoietic system, and some hemocyte (blood cell) types are attracted to a neurogenic niche in the brain. By using multiplex RNA-FISH techniques we have for the first time detected specific neural lineage marker transcripts expressed together in a small proportion of the circulating hemocytes (around 1%). This finding agrees with and confirms that there is only a small proportion of hemocytes which can develop further into neurons. Interestingly, these transcripts were co-expressed in the same cell as well as sometimes together with the transcript of the serotonin receptor 1 (5htr1+). Moreover, we could also show that several treatments, including serotonin, astakine, and lipopolysaccharide, as well as an acute brain injury, could induce a greater proportion of such neural…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10- —Uppsala University
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsInvertebrate Immune Response Mechanisms · Neurobiology and Insect Physiology Research · Marine Invertebrate Physiology and Ecology
Introduction
Adult neurogenesis, i.e., the formation of new neurons that are integrated into the brain of an adult animal, occurs in several different invertebrate and vertebrate animal groups. This process has been studied in quite some detail in decapod crustaceans, where it is well known that newly formed neurons are formed in two specific areas, called clusters 9 and 10, in the olfactory and accessory lobes (Sullivan et al. 2007a). In several studies in the freshwater crayfish Procambarus clarkii, it has been shown that the adult-born neurons probably originate from the circulatory system, and thus have their origin in circulating hemocytes (Beltz and Benton 2017).
The neural precursors are found in a specific structure called a neurogenic niche located on the ventral side of the crayfish brain. There, they undergo cell divisions and migrate along fibrous streams formed by the long processes of niche progenitor cells to specific proliferation zones in clusters 9 and 10 (Brenneis and Beltz 2020). A peculiarity of these stem cells in the neurogenic niche is that after a cell division, both daughter cells migrate away and thus the niche would be depleted of stem cells unless new cells did not assemble and arise there (Benton et al. 2011, 2013). Since neurogenesis continues throughout the life of the animals, this means that there must be another source of stem cells found in the neurogenic niche (Zhang et al. 2009). Such a likely source outside the niche was confirmed by an experiment in which crayfish first received a single injection of BrdU. The presence of BrdU-labeled cells in the niche was then monitored for three weeks, and two peaks of labeled cells could then be observed. The first peak occurred immediately after the injection and showed that cells in the niche took up BrdU at that time. After a few days without labeled cells, a second peak appeared, even though BrdU had already disappeared from the circulation. Since no new cells could be labeled after that time, it suggests that the cells in the second peak came from another, external source (Benton et al. 2014). Since it is mainly cells in the hematopoietic system, i.e. cells that are formed in the hematopoietic tissue that give rise to circulating hemocytes and incorporate BrdU in crayfish, it is likely that this system could be the source of the neural precursors in the niche. Consequently, there are several studies that indicate that this is the case (Benton et al. 2013, 2014; Beltz and Benton 2017).
Studies both in vitro (Benton et al. 2013) and in live animals (in vivo) (Benton et al. 2014) have shown that cells from the crayfish’s blood (hemolymph) are attracted to the niche, while other cell types are not. When EdU-labeled blood cells (hemocytes) from one donor crayfish are transferred to another crayfish, labeled cells can be seen in the niche as early as three days after the transfer. Seven weeks later, some of these cells express signaling molecules typical of nerve cells in clusters 9 and 10. Since these blood cells come from the body’s blood-forming (hematopoietic) tissue, it was concluded that the immune system helps to replenish the niche with new cells (Benton et al. 2014).
Hemocytes in crustaceans are usually categorized according to morphological characteristics, taking into account the amount and type of granules in the cytoplasm into three different groups: hyaline (HC, without granules), semi-granular (SGC, with few and less electron-dense granules) and granular hemocytes (GC, filled with large electron-dense granules) (for reviews see: Bauchau 1981; Söderhäll 2016). Later, cells which were recruited to degenerating axons in the nervous system of the crab Ucides cordatus, were also characterized by light, as well as transmission and scanning electron microscopy, and found to be circulating hemocytes (Chaves-da-Silva et al. 2010). Crustacean hemocyte types can be separated using gradient centrifugation (Söderhäll and Smith 1983) and this method to isolate the hemocyte types has advanced our knowledge about the functions of the different morphotypes in the immune system. However, several studies on the functions of the immune system show that this division is not detailed enough and that there are probably many more types of hemocytes with specific tasks (Söderhäll et al. 2022; Kruangkum et al. 2025).
Regarding the hypothesis that hemocytes can be transported to the neurogenic niche in the brain of the crayfish and develop further and differentiate into neurons, studies have shown that the cells from the circulation that are incorporated into the niche are of hyaline or possibly semi-granular morphology (Benton et al. 2022). However, there is no knowledge so far whether this is a random attraction and that the hemocytes are induced to become neuronal precursors once they have reached the neurogenic niche or whether there are specific hemocytes in the circulation that are predetermined as precursors destined for the nervous system in the brain.
The aim of the experiments reported here was to identify whether there are circulating hemocytes that can become neural precursors by examining the expression of several different transcripts simultaneously in the same individual hemocytes. A further intention was to identify variables that affect the presence of any hemocytes with the potential to form cells in the nervous system. Overall, our obtained data in this study support that neural progenitors already exist in the circulation at a low frequency and that their proportion of the total hemocyte population can be influenced by various internal stimuli that are regulated by environmental cues.
Methods
Animals
Adult male freshwater crayfish, Pacifastacus leniusculus from Lake Erken, (located in east-central Sweden near the Baltic coast (59.8 N 18.6 E), were housed in aquaria containing running aerated tap water at the Department of Organismal Biology, Uppsala University. The ambient temperature in the aquarium room was maintained at approximately 10–12 °C, in a 12:12 light: dark cycle. Only intermolt animals with no visible injuries or signs of weakness were used in the experiments. All animals used in the experiments were treated in accordance with Swedish law.
Ink Administration for Vascular Labeling in the Brain and Tissue Collection
The administration of Indian ink in crayfish was conducted following our previously established methodology (Kruangkum et al. 2025). Briefly, the whole dorsal carapace of crayfish was gently removed before inserting a 27G¾ needle with a 1-mL syringe containing ink (Pébéo, France) into the heart chamber. Around 100–200 µL of ink was gently injected to ensure that no more ink leaked out. The cephalothorax was cut out from the abdomen before being immersed in 4% paraformaldehyde fixative solution (PFA), in PBS (phosphate buffer, containing 137 mM NaCl, 2.7 mM KCl, 8 mM Na_2_HPO_4_, and 2 mM KH_2_PO_4_, pH 7.4). The tissues were kept at 4 °C overnight with gentle shaking. The fixative was replaced with a new solution after ca. 4–6 h. The tissues were then washed with 1 × phosphate-buffered saline (PBS) five times before dissection of the organs under a stereomicroscope (Nikon SMZ1500). For a proper removal of the brain, the nerves that attached the brain to its socket were gently cut before removing it with a micro-forceps with a curved tip. The sheath enveloping the brain was carefully removed from the brain tissue for further observation. The brains were cleaned to avoid any damage during handling. The ventral and lateral sides of the brain were photographed using a digital camera (Nikon DS-Vi1). The hematopoietic tissue (HPT), which remained lined with the dorsal surface of the stomach, was kept intact with the stomach. Only the upper plate of stomach tissues with the HPT was used for further tissue processing. The anterior proliferation center (APC), which was still attached to the brain tissue, was kept in 1x PBS for further use.
Tissue preparation, H&E staining, and Observation
The dissected brains, without any cleaning since they were intact and remained associated with the surrounding connective tissue, and the HPT, which remained on top of the dorsal side of the stomach, were placed in tissue cassettes for tissue processing. The procedure for tissue preparation and staining was as described previously (Kruangkum et al. 2025). Briefly, the tissues were dehydrated by being immersed in serial chambers of increasing ethanol concentration, from 70% to 100%, for 30 min each. They were then moved into xylene for 40 min and then further for 1 h. All steps were performed under gentle agitation. Finally, the tissues were immersed in melted Paraplast (HistoLab, Sweden) for paraffin infiltration and then placed overnight in an oven at 60 °C. The infiltrated tissues were then embedded in the block for further use. The brain tissues were sectioned using a rotary microtome (Leica RM2155) with a thickness of approximately 5–6 μm in both horizontal and sagittal planes.
The paraffin-sectioned tissues were deparaffinized by three steps of xylene for 5 min each, then rehydrated in a series of decreasing ethanol concentrations (100% to 70%, 2 min each). The tissues were immersed in distilled water for 5 min before being briefly dipped for 30 s in a jar containing Mayer’s hematoxylin. The colorization of hematoxylin was activated by rinsing in tap water for 5 min. The slides were then subjected to a 1-minute eosin staining step and subsequently dehydrated. Well-dried slides were placed in a jar of xylene for three steps of 5 min each for clearing. Then they were mounted and covered by VectaMount mounting medium (Vector Laboratories) and a cover slip.
The stained brain tissues were observed by a light microscope (Leica, DM5500B), visualized, and photographed by LASX software (Leica Application Suite X). All photographs were exported from the microscope software and then adjusted for brightness and contrast using Adobe Photoshop 2024 to achieve a higher-quality image.
Hemocyte Collection
Crayfish [carapace length (CL) ≈ 3–4 cm, 25–35 g bodyweight] were used for bleeding. Hemolymph of each crayfish was drawn into Eppendorf tubes containing anticoagulant solution (Söderhäll and Smith 1983) in 1:1 ratio of volume using an 18-G needle (BD microlane) at the ventral area of the first abdominal segment of crayfish. The hemolymph was gently mixed by flipping the tube before centrifugation at 800 × g for 5–7 min at 4 °C. The supernatant plasma was discarded, while the hemocyte pellet was gently resuspended in 0.15 M NaCl solution, and added onto Superfrost glass slides, into a small drop of 1 M CaCl_2_ solution in a 9:1 ratio. The slides were left for hemocyte attachment for approximately 20 min before being fixed. Formaldehyde (3.7%) in crayfish saline (CFS, 0.2 M NaCl, 5.4 mM KCl, 10 mM CaCl_2_·2H_2_O, 2.6 mM MgCl_2_·6H_2_O, 2 mM NaHCO_3_, pH 6.8) was applied as an overlay over the attached cells for 10–15 min at room temperature. Finally, the fixed cells were washed three times with 1 × PBS, 1 min each, before further use for RNA-FISH.
Note that the prepared hemocytes on glass slides were stored in 1× PBS (prepared with nuclease-free water) for no longer than a week at 4 °C, as recommended.
RNA-fluorescence in Situ Hybridization of Hemocytes
Fixed hemocytes on glass slides were treated as described in detail in our previous reports (Kruangkum et al. 2025; Wang et al. 2025) and in the manual for the QuantiGene ViewRNA ISH Cell and Tissue Assay Kit (Invitrogen). Briefly, the hemocytes were treated with a working protease solution (1:2000 dilution in 1x PBS) at 40 °C for 15 min. The samples were washed three times with 1× PBS before being incubated with the specific probes, as outlined in Supplementary Table 1. The slides covered with a 1:100 dilution of the specific target probes in pre-warmed probe set diluent (QF) were incubated and hybridized at 40 °C in a moisture chamber for 3 h. The samples were rinsed with the washing buffer three times for 1 min each. Preamplification and amplification with a dilution of 1:25 in pre-warmed amplifier diluent QF were sequentially performed for 30 min each step at 40 °C. Three washes with washing buffer were performed at the end of each step of pre-amplification and amplification. The procedure for probe labeling was performed according to the manufacturer’s instructions in the user manual, using a 1:25 dilution of each fluorescent probe. They were incubated at 40 °C for 30 min before performing the final washing. The interference of auto-fluorescence from the cytoplasmic granules was eliminated using the Autofluorescence Quenching Kit (Invitrogen, Thermo Fisher Scientific) for 5 min at room temperature. In the last step, the nuclear stain molecular probes, DAPI (1:100 in 1 × PBS) for 10 min, or Hoechst 33258 in 1:1000 dilution in 1 × PBS for 20–30 min at room temperature, were used for counterstaining. The stained hemocyte samples were mounted with ProLong Gold Antifade mounting medium (Invitrogen, Thermo Fisher Scientific) and overlaid with a coverslip. The tissue slides were kept in a slide box placed in a fridge until they were used for observation.
Probe Synthesis
Custom-branched probes for use with ViewRNA in situ hybridization assays (Invitrogen) were designed according to the sequences listed in Supplementary Tables 1 and were provided by Thermo Fisher Scientific. The specific probes targeting neurogenic transcripts, such as prospero (pros), brain tumor suppressor (brat), Notch signaling repressor (numb), doublecortin (dcx), and soxN, which are supposed to be expressed in neural precursor cells, were used as the main probes for this study. A serotonin receptor 1 (5htr1) is another gene that we expected to play a role in the differentiation and maturation of cells, which could be a primary focus in this study was also included. Astakine 1 (ast1) and hemolectin (hml), were used for co-staining and identification of hemocytes with a role in immunity. Glia cell missing (gcm) is a glial-specific protein that may be present in glial cells and/or neurosecretory cells (Junkunlo et al. 2020). It was used to probe and evaluate some of the hemocytes. A transcript of the proliferation marker, PCNA, was used for co-staining to evaluate the dividing status of the labeled hemocytes.
Microscopic Visualization and Photography
Confocal laser scanning microscopy, CLSM (Leica Stellaris 5) with LASX software (Leica Application Suite X), was used for visualization and photography. Line sequential operative mode [with WILL (Alexa 488, 594, 647, and 750) and diode 405 laser sources] was used to eliminate the interference effect or crosstalk of each fluorescent channel and the background. In this setting mode, a picture is captured for different color channels in an alternating, line-by-line order to prevent spectral crosstalk, and bleed-through and then averaged at each point along the line to reduce noise.
Two or three different areas of each sample were randomly selected for photography at a 20x optical lens magnification (with a 1.5-2x digital zoom, depending on the number of hemocyte populations in the field of view). The photos were taken throughout the entire depth in the Z-axis using Z-stack mode with an optical section at a fixed thickness of 0.68 μm each. Each Z-stack of photos was processed to create a 3D generative image, enhancing image quality through a 3D viewer. The fluorescent intensity and gain adjustments were considered based on the background signal or the signal appearance, which was compared to that of the negative control (without probes). Co-localization of each signal was evaluated in real time, taking into account different focal plane levels. Images generated through optical sectioning and Z-stack reconstruction were used for presentation. After rendering the 3D image in the operative working area, the photos were adjusted in quality to improve visualization using a “volume with shaded” tool. To verify the localization of positive signals presented in the nucleus (such as in Fig. 3C), the optical sectioning with the “sections” tool in a fusing area was used for observation. The separated window channels of each fluorophore and the merged image were saved in a TIFF file for further analysis.
Cell Counting and Statistical Analysis
The “brightness and contrast” were adjusted throughout the entire image to enhance the image quality for better visualization using Photoshop software or “the corrections tool with brightness/contrast adjustment” in PowerPoint software. Moreover, hemocyte counting from each picture, from each individual crayfish, was performed by using the “count” tool of the Photoshop software. The total number of labeled nuclei in an observation area at 20x, was used to represent the total hemocyte number. The co-expression of the putative neurogenic marker transcripts in the same cell was counted, and the total hemocytes and positive cells were recorded in an Excel sheet for further calculation. Two or three values of data (number of hemocytes from two to three areas) from each crayfish were summed up to represent one value as the “representative data from that crayfish”. The raw data were converted into terms of “percent of the neurogenic positive cells” per total cell from the individual. In the experiment in which hemocytes were collected both “before” and “after” different treatments as described below for comparison, data from the same crayfish individual was compared.
RNA-fluorescence in Situ Hybridization in Tissue Sections
The sectioned tissues (brain, APC and HPT) were deparaffinized and rehydrated with xylene and ethanol as mentioned above. They were placed in 1 × PBS for 5 min before being treated with the pretreatment working solution at 90 °C for 10 min. The tissue slides were immersed in a jar containing DEPC-treated water for 5 min, and then washed three times with 1× PBS. The working protease solution from the kit was prepared by diluting 1:100 in a prewarmed 1 × PBS solution and overlaid on top of the tissues for 15 min at 40 °C in an oven. The tissues were washed with 1× PBS before being post-fixed with 4% PFA for 5 min at room temperature. After washing three times 1× PBS, the samples were subjected to specific probes for hybridization by a dilution of 1:40 in pre-warmed probe set diluent QF (please see the details of probe combination in Supplementary Table 2. Then they were incubated for 2 h in a moisture chamber, which was placed in a 40 °C oven. After probe hybridization, the samples were washed with washing buffer three times for 1 min each. The following steps were then followed in a similar manner as the procedures described in the section on hemocytes.
Serotonin (5-HT), Recombinant Astakine 1 (rAst1), Lipopolysaccharide (LPS), and Crayfish Saline (CFS) Injections
The in vivo experiments were designed to compare the hemocyte population “before” and “after” the injections with these compounds as detailed above. The first bleeding was used as a baseline for evaluations of hemocyte numbers (“Before treatment”). The crayfish were then placed in the experimental aquaria to rest for 24 h before being injected with the above chemicals. Twenty-four hours after administration of the substances, a second hemolymph collection was performed and labeled as “After treatment”. The pelleted hemocytes of both “Before and After treatment” from the same individual crayfish were prepared as described above. Note that the hemocyte preparation on glass slides was performed immediately after bleeding. However, the fixed spread hemocytes on glass slides could be stored by immersion in sterile, nuclease-free 1x PBS for no longer than a week for RNA-FISH.
The solutions were prepared freshly before injection. The different treatments were as follows: (1) 100 µL of 10^− 9^ M 5-HT (ThermoFisher) in CFS (0.2 M NaCl, 5.4 mM KCl, 10 mM CaCl_2_.2H_2_O, 2.6 mM MgCl_2_.6H_2_O, 2 mM NaHCO_3_, pH 6.8); (2) 0.01 µg/g body weight lipopolysaccharide (LPS, Sigma) in CFS; (3) 0.05 µg/g body weight of rAst1 in CFS; and (4) 100 µL CFS as control. All chemicals (in 100 µL) were injected into the ventral hemal sinus of the crayfish. Crayfish were then allowed to rest in their experimental aquarium for 24 h before hemolymph was collected, and RNA-FISH was performed as described in Supplementary Table 2.
Brain Injury
Crayfish without signs of injury or trauma were selected from the aquaria and reared in a separate experimental aquarium. Twice-bled hemolymph samples (before and after injury induction), as outlined in the experimental plan, were used to evaluate the effect of injury induction on the numbers of neurogenic gene-expressing hemocytes in both pre- and post-injury conditions. In the first bleeding, the collected hemocytes were labelled as “Before-induction.” The crayfish were then placed in the experimental aquarium for 24 h before performing “brain injury induction”. The brain injury induction was performed by using a tiny-sized needle (27G ¾”) for wounding in the area of the thinnest cuticle at the center between the eyestalks, below the rostrum, where the brain is localized. Three to four needle sticks were done to ensure injury. The crayfish were closely observed and monitored afterward to ensure that they did not exhibit any severe signs, such as loss of mobility after the injury. They were transferred to the aquaria for 48 h before performing the “second bleeding”. The second hemolymph collection was labeled as “After-induction”. The slides containing hemocytes from both “before” and “after” inductions were used for RNA-FISH as described above and in Supplementary Table 2.
To investigate the effect of brain injury on the number of circulating hemocytes, 50 µL hemolymph samples were taken before and 48 h after brain injury as above. and mixed with 50 µL 4% PFA in PBS. The total hemocyte number was calculated using a Fuchs-Rosenthal counting chamber and compared to un-injured controls (n = 6).
Quantification and Statistical Analysis
For the descriptive analysis in Fig. 3A, the number of neural lineage marker-positive cells and total hemocytes, collected before the injections or before brain injury induction, was used as a baseline or normal condition for further analysis. The data are presented as “mean \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\pm\:\:$$\end{document} SD” without performing statistical comparisons. The effect of chemical treatments and needle injury in the brain was compared between “before” and “after” treatments in each individual, and analyzed with “Paired T-test” (two-tailed), and normal (Gaussian) distribution was tested with the “Shapiro-Wilk normality test” and “Q-Q plot” (Figs. 3A and 6). All data were represented as “mean \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\pm\:\:$$\end{document} SD” by GraphPad Prism software. The number of individual animals in each experiment was n = 4–6 for injection experiments and n = 9 for brain injury, and n = 6 for total hemocyte count after brain injury.
Results
Neural Lineage Marker Genes Are Expressed in Circulating Hemocytes
Previous experiments have indicated that neural progenitors are present in the circulation of freshwater crayfish, and that these are recruited to the neurogenic niche to form neurons in the brain area containing neurons innervating the olfactory lobe of the crayfish’s brain (Benton et al. 2022). Figure 1 shows a schematic illustration of the proposed mechanism of adult neurogenesis, in which hemocytes released from the hematopoietic tissue (HPT) and the anterior proliferation center (APC) are the sources of neural precursors. These precursor cells form mature neurons in the brain through a developmental process in the neurogenic niche, as shown in several earlier reports (Fig. 1) (Benton et al. 2011, 2013, 2014, 2022; Chaves-da-Silva et al. 2013). In order to find crustacean neurogenic specific marker transcripts in vivo, we searched in the P. leniusculus transcriptome from hemocytes and the HPT (Accession: PRJNA259594) for some candidate genes related to the formation of new neurons, including prospero (pros),* brain tumor suppressor (brat), Notch signaling repressor (numb), doublecortin (dcx), and sox-neuro (soxN).* Accordingly, the circulating hemocytes were examined for these transcripts using multiplex RNA-FISH.
Fig. 1. How adult-born neurons can originate from the immune system. Schematic illustration presenting the proposed mechanism of new neuron formation and integration in the brain of adult decapod crustaceans, which is modified from Benton et al. 2022. Previous experiments indicate that the immune and nervous systems interact to generate neurons in the adult brain of crayfish, and that the neural precursors responsible for producing these adult-born neurons originate from the hematopoietic tissue (HPT) and that they travel to the neurogenic niche via the circulatory system
As shown in Fig. 2A, we first used specific RNA probes to visualize the expression of the neural lineage-specific markers numb,* brat*, and pros in combination with the hemocyte-specific marker transcript Hml. Surprisingly, we found that these three putative neural lineage-specific markers were co-expressed in very few, but in the same hemocytes, and most importantly, not together with Hml expression (Fig. 2A–A″ and Supplementary Fig. 1A). Next, Fig. 2B shows another combination of probes, and the expression of brat,* dcx*,* and pros* was detected in the cytoplasm of the cell, but also here all transcripts were present in the same cell. This suggests that these are mature mRNAs localized in some specific circulating hemocytes (2B arrows, and Supplementary Fig. 1A, B).
Fig. 2A low number of hemocytes expresses markers for neuronal precursor cells. Circulating cells are stained using different combinations of RNA probes specific for neurogenic transcripts by the use of multiple RNA-FISH. A A low-magnification micrograph showing that the putative neurogenic markers (numb in red, brat in yellow, and pros in green) co-localize in the same hemocyte, scale bar = 50 µm (magnified in inserts A’ and A”). However, they do not co-express with hml^+^ (in pink) in those expressing cells (inserts A’ and A”, scale bar = 10 μm). B In this micrograph one pros/brat+ hemocyte was also positive for the dcx transcript (arrow) scale bar = 20 μm. C A hemocyte expressing 5htr1, dcx and numb (arrow), whereas ast1, (in red) was expressed in several other hemocytes, scale bar = 20 μm. D Hemocytes (arrows) positive for soxN, numb and pros transcripts, scale bar = 20 μm. E and F The observed hemocytes positive for putative neurogenic markers can be classified into two major morphotypes: the hyaline-like (in E, scale bar = 10 μm) and semi-granular-like (in F, scale bar = 5 μm) cells. G Some of these putative neurogenic hemocytes were positive for “glia cell missing (gcm)” mRNA and the proliferation cell marker (pcna) (arrow). Hoe = Hoechst 33342 nuclear staining, scale bar = 20 μm
Since astakine1 (ast1) as well as serotonin (5-HT) have been shown to be associated with adult neurogenesis in Procambarus clarkii (Benton et al. 2022), we accordingly used a probe combination for co-localization of ast1 and a serotonin receptor type 1 (5htr1) expression together with putative neural lineage markers such as numb,* pros*,* soxN* and dcx in different combinations (Fig. 2C-D).
Figure 2C shows that numb was expressed in the same cells as dcx and 5htr1, and with a different probe combination as shown in Fig. 2D, numb mRNA was detected in the same cells as pros, and soxN. Figure 2C, and 2E-F show that ast1 and 5htr1 were expressed together with neural lineage markers in some cells, but the ast1 transcript was detected in several different hemocytes, indicating that this gene has a more general function (Fig. 2C, D and Supplementary Fig. 1C, D). When considering the transcript of “numb”, which was used for staining together with different probe combinations we conclude that also pros+ hemocytes could also express the 5htr1 in some cases (Fig. 2C and D).
When the morphology of neural lineage marker positive hemocytes was examined, hyaline (Fig. 2E and Supplementary Fig. 1E) and semi-granular (Fig. 2F and Supplementary Fig. 1F) hemocytes were the dominant cell types, which is consistent with earlier findings and descriptions in the crayfish P. clarkii. As shown in Fig. 2G, we observed expression of the presumed glia-specific gene, glial cell missing (gcm), and the cell proliferation marker gene (pcna) in the hemocytes, and they were co-localized with some of the neural lineage markers (pros and numb).
To assess whether hemocytes that were positive for other neural lineage markers were in a proliferative state, we examined their co-expression with pcna. As shown in Supplementary Fig. 2A–C, we divided the results for the hemocyte population into two sets with different patterns of mRNA expression; (1) pcna+ cells without any neural lineage marker transcripts (white arrow in Supplementary Fig. 2A–C), and (2) co-expression of the pcna+ cells with neural lineage marker transcripts (yellow arrow in Supplementary Fig. 2A–C). However, in the latter group, we were able to find variations in the level of gcm transcript. Some cells had a moderate expression of pcna, high expression of pros/numb, and high expression of gcm (yellow arrow in Supplementary Fig. 2A), while other cells had very low pcna expression, high expression of pros, moderate expression of numb and low expression of gcm (yellow arrow in Supplementary Fig. 2B). There were also some cells characterized by low pcna expression, moderate to low expression of pros, and absence or very low expression of numb/gcm (yellow arrow in Supplementary Fig. 2C). The negative controls, without gene specific probes, had no positive signals in any fluorescent channel (Supplementary Fig. 1H). These results suggests that there is a variation of developmental stages in neural lineage marker positive hemocytes. Therefore, we conclude that there is evidence for the presence of a small population of circulating hemocytes that express neural lineage marker transcripts.
Circulating Neural Progenitor Cells in Hemolymph
Next, we investigated the abundance of circulating hemocytes expressing neural lineage marker transcripts in naïve animals. As shown in Fig. 3A, we used two different probe combinations. First co-expression of 5htr1 together with soxN/numb/brat or soxN/numb/gcm or soxN/numb/dcx, and secondly, co-expression of pros together with numb and either brat,* gcm or dcx* (Fig. 3A). The reason for these two different probe combinations was technical limitations, since only four different fluorophores could be used at the same time.
Fig. 3. Circulating hemocytes positive for neural lineage marker transcripts account for approximately 1% of the entire hemocyte population. Violin plots show the percent of neural lineage positive cells per total hemocyte population and their spatial expression patterns of the neural lineage-specific transcripts in the hemocytes and analyzed by multiple RNA-FISH. A A violin plot showing the percentage of cells positive for co-localization of different neural lineage marker transcripts. Upper panel: proportion of 5htr1^+^ hemocytes co-localized with other neural lineage markers: either soxN/numb/brat, or soxN/numb/gcm, or soxN/numb/dcx. Middle panel: proportion of pros^+^ hemocytes co-localized with other neural lineage markers: either numb/brat or numb/gcm or numb/dcx; Lower panel: total proportion of hemocytes expressing neural lineage marker transcripts. The percentage of 5htr1 positive cells co-localized with other neurogenic markers (pink color) was 0.5 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\pm\:$$\end{document} 1.03% of the total population, and the percentage of pros positive co-expressed with the other neurogenic markers (purple color), was 1.06 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\pm\:$$\end{document} 1.53% of the total population of hemocytes. The black dots represent the individual values (percent of positive cells per total hemocytes) from each animal. B–B′ Positive signals of neural lineage marker transcripts were observed in the cytoplasm of some hemocytes without (B) and with phase contrast overlay (B′), scale bars = 10 μm. This example shows dcx, 5htr1, numb and Ast1. C–C′ Positive signals of neural lineage marker transcripts were observed within the nucleus of some hemocytes, scale bar C = 10 μm and C’=5 μm. The right panels of Ci–Civ show the different axes of the optical sectioning, as shown in a 3D-generated image of a positive hemocyte obtained by confocal microscopy, which indicates the localization of positive signals within the nucleus (white arrowheads). This example shows SoxN, 5htr1, brat and numb. D Positive signals of gcm transcript were observed, with (white arrow) or without (yellow arrow) co-localization with other neural lineage marker transcripts, scale bar = 10 μm
Figure 3A shows a quantitative analysis as violin plots and the percentage of cells positive for the co-localization of these different neural lineage marker transcripts. The proportion of 5htr1^+^ hemocytes co-localized with other markers (soxN+, numb+, and brat + or gcm + or dcx+), was approximately 0.5% (Fig. 3A, upper panel). The proportion of pros+ hemocytes co-localized with other neural lineage markers (numb + and brat + or gcm + or dcx+) was approximately 1% (Fig. 3A, middle panel). The proportion of total hemocytes expressing neural lineage marker transcripts was in total around 1%, which indicates that 5htr1 is not expressed in all potential neural precursor cells (Fig. 3A, lower panel, with a yellow highlight).
As shown in Fig. 3B–B′ a cytoplasmic localization of the neurogenic marker transcripts was found in some cells (Fig. 3B–B′, yellow arrowhead), as well as within the nucleus in some (Fig. 3C–C′, white arrowhead). Nuclear localization was confirmed by Z-stacking fluorescent images of the positive hemocyte using confocal microscopy in three-dimensional axes (Fig. 3C_i_–C_iv_).
Figure 3D shows that a few cells were found to express gcm, but without the presence of other neural marker transcripts (Fig. 3D, yellow arrow), and these cells had a more granular morphology as was previously observed in isolated brain cells in the crayfish P. leniusculus by Junkunlo et al. (2020). In summary these results so far indicate that there are about 1% of neural lineage progenitor cells in crayfish hemolymph that express several neural marker transcripts, and in addition a few cells only express gcm which imply that these may be precursors of glial cells.
The Hematopoietic tissue, a Source of Neural Precursor Cells
To further address the origin of the circulating neural progenitors, we examined horizontal and sagittal HPT tissue sections using different sets of neural lineage-specific RNA-probes. We found a few cells expressing these marker genes within the hematopoietic tissue (HPT), and some examples are shown in Fig. 4A (pros^+^/numb^+^), and Fig. 4D (5htr1^+^/dcx^+^/numb^+^/brat^+^). A common characteristic of these cells was their condensed nuclei. Moreover, we found some pros^+^/numb^+^/soxN^+^ HPT cells that appear to have been released from the HPT-lobule into a hemal sinus. One such cell is shown in Fig. 4B (white arrow). A sagittal section of the HPT illustrates two positive cells with different patterns of their mRNA marker expression (Fig. 4C). One cell positive for 5htr1,* numb*,* brat*,* and soxN* was located close to the apical surface of the HPT, (Fig. 4C, yellow arrow), and another cell expressing soxN and 5htr1 but not brat or numb was found close to the basal side (Fig. 4C, white arrow). Most cells positive for any neural lineage marker transcripts shared some common characteristics, i.e., round shape, low amount of cytoplasm, intense nuclear staining and located at the periphery of a lobule and that some were recently released from a lobule. In contrast to the HPT, no positive signals of the specific neural progenitor transcripts were detected in cells of the anterior proliferation center (APC). We could only detect some cells expressing 5htr1, and they were also co-expressing hml This observation points to a role in immunity and this assumption is consistent with some previous reports (Noonin 2018; Tong et al. 2020) (Supplementary Fig. 3).
Fig. 4. Localization of cells expressing neural lineage marker transcripts in the hematopoietic tissues (HPT) analyzed by multiple RNA-FISH. A A horizontal section of the HPT showing co-localization of the neural lineage marker transcripts pros and numb with the proliferation marker pcna and ast1 in a cell that remains inside a HPT lobule (white arrow), scale bars = 20 μm (insert scale bars = 10 μm. B High magnification of another area in the HPT, where a released cell positive for (pros^+^/numb^+^/soxN^+^) was found (white arrow), scale bars = 10 μm. C A sagittal section of HPT tissue showed different cells positive for neural lineage markers. One 5htr1^+^/soxN^+^ cell (white arrow) was observed close to the basal margin of the HPT (red arrow heads indicate the basal lamina), and one 5htr1^+^/soxN^+^/numb^+^/brat^+^ cell (yellow arrow) was observed in the loose organized HPT lobule, where it is located close to the apical surface (indicated by a dashed line) of HPT, scale bars = 20 μm. D A similar pattern could be observed in other areas of the HPT tissue, where a few cells positive for 5htr1^+^/numb+/ brat+/dcx^+^ were observed in the apical (dashed line) and basal areas (close to the thin lining of the basal membrane, indicated by red arrowheads) of the HPT section (white arrows). Abbreviation: HS; hemal sinus, scale bars = 20 μm
Neural Progenitor Hemocytes Increase in the Circulation Following Stimuli Treatment
Given that earlier studies have shown an increase of cells in the neurogenic niche after injection of recombinant astakine1 (rAst1) (Benton et al. 2014, 2022), or treatment with 5-HT (Benton et al. 2022), and that hemocytes with transcripts related to the nervous system were increased after LPS injection in crayfish (Xin and Zhang 2023), we investigated the proportion of circulating neural progenitors after such stimuli. Figure 5A describes the experimental design used to determine the number of neural lineage marker transcript-positive cells among circulating hemocytes, before and after different treatments. As a negative control, injection with crayfish saline (CFS) was used (Fig. 5A). Apart from neural lineage marker transcripts, the expression of 5htr1 was also included in this study since 5-HT has been shown to influence adult neurogenesis in crayfish (Zhang et al. 2011; Benton et al. 2022). We used two different probe combinations with numb as an overlapping probe in order to get as much information as possible for neural lineage marker labelling and quantification (Supplementary Table 2).
Fig. 5. Illustration showing the procedures of the in vivo assays. A Method for collecting circulating hemocytes for RNA-FISH before and after stimulation with the putative stimulating agents, injection of 5-HT (10^− 9^M 5-HT in CFS), recombinant Ast1 (0.05 mg/g body weight), or LPS (0.01 mg/g body weight), as well as a control using CFS injection. B Method for collecting circulating hemocytes for RNA-FISH before and after brain injury to study the effect of injury on the number of hemocytes positive for neurogenic markers
Figure 6 shows the percentage of positive hemocytes in each individual before and after the different injection treatment. The effect of injection of 100 µL, 5-HT at 10^− 9^M, is shown in Fig. 6A. The percentage of neural lineage marker transcript-positive cells was below 1% before treatment. One day after the treatment, the percentage of positive hemocytes significantly increased as determined with two different probe combinations. The percentage of hemocytes with neural lineage pros/gcm/numb positive transcripts was 3.35 ± 1.22 (P = 0.0011), and of neural lineage soxN/numb/brat positive transcripts in the hemocytes was 3.13 ± 1.10 (P = 0.0009). In a combination of data with both sets above, the percentage of positive transcripts in the hemocyte was 3.15 ± 0.67 (P = 0.0001) (Fig. 6A in left, middle, and right (framed) panel, respectively).
In Fig. 6B, the effect of LPS injection is shown, and the percentage of pros/numb/brat-positive cells before treatment was variable around 1%, and if dcx was included it was lower. No statistic significant difference was detected one day after injection when experiments probed with dcx/numb and pros/numb/brat neural lineage marker transcripts were analyzed separately. In contrast, a combination of data from the (dcx/numb and pros/numb/brat) positive signals resulted in a significant increase when comparing before and after treatment [1.68 ± 0.82 (P = 0.0270)] (Fig. 6B right framed panel).
Fig. 6. The percentage of hemocytes positive for neural lineage marker transcripts in the circulation “Before” and “After” treatments in the in vivo assays. The percent of hemocytes positive for neural lineage marker transcripts in different probe combinations; A before and after injection of 5-HT (100 µl 10^− 9^M), n = 6. B before and after injection of LPS (0.01 mg/g BW), n = 4. C before and after injection with CFS (100 ml) as control, n = 4. D before and after injection of Ast1 (0.05 mg/g BW), n = 6. E Co-localization of the 5htr1 + signal with the other neural lineage marker transcripts in 5-HT (right, n = 6) and LPS (left, n = 4) injection assays. Comparison was done between “before” and “after” treatment samples for each individual, and analyzed with “Paired T-test” (two-tailed), and normal (Gaussian) distribution were tested with the “Shapiro-Wilk normality test” and “Q-Q plot. The asterisks (), (), (), (****) refer to statistic significant differences: P \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\le\:$$\end{document} 0.05, P \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\le\:$$\end{document} 0.01, P \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\le\:$$\end{document} 0.001, P \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\le\:$$\end{document} 0.0001, respectively, “ns” indicates “no statistic significant difference”
As shown in Fig. 6C, the percentage of positive neural lineage marker transcripts was not significantly different for pros+/gcm+/numb + or soxN+/numb+/brat+, or their combination in the circulating hemocytes before and after CFS control injection (Fig. 6C; P = 0.67, P = 0.18, and P = 0.28, respectively).
In Fig. 6D the effect of injecting 0.05 µg/g body weight recombinant Ast1, it was found that the percentage of positive neural lineage marker transcripts, pros+/numb+/gcm+, was significantly increased (3.26 ± 2.77, P = 0.0344) (Fig. 6D).
As we had noticed that some hemocytes which expressed neural lineage markers at the same time expressed the serotonin receptor 5htr1, we also tested whether the expression of this receptor was affected by 5-HT injection. Thus, we counted positive signals of 5htr1 transcript, co-localized with soxN+/numb+/brat + in hemocytes, before and after 5HT-injection assay, and as shown in Fig. 6E, a large increase in positive hemocytes was found. (Fig. 6E, left panel). This effect was specific for 5-HT injection since LPS injection did not give similar result (Fig. 6E, right panel).
Brain Injury Induced an Increase in Neural Lineage Marker Positive Hemocytes
Next, we punctured the brain with a thin needle to induce a penetrating traumatic brain injury, in order to evaluate the proportion of neurogenic hemocytes as a result from such injury of the brain. Hemocytes were collected from the same individual before and then 2 days after the injury to the crayfish brain (Fig. 5B). This was followed by experiments using RNA-FISH in circulating hemocytes using two probe combinations.
As shown in Fig. 7 the percentage of neural lineage marker transcript-positive cells before induction was below 1%, and 48 h after brain injury the percentage of neural lineage marker transcript-positive hemocytes increased significantly, both for the probe combination dcx+ (or soxN+),* numb+, gcm+* cells (Fig. 7, left panel, P = 0.0143) and for the probe combination pros+,* brat+, numb+* cells (Fig. 7, middle panel, P = 0.0205). A combination of the two probe sets revealed that the percentage of neural lineage marker transcript-positive cells after brain injury was significantly higher and constituted 1.89% of the total cells compared to before induction (Fig. 7, right framed panel, P = 0.0005). As shown in Supplementary Fig. 4. the number of circulating hemocytes after brain injury compared to a bleeding control did not change significantly although for some animals there was an increase.
Fig. 7. The percent of hemocytes positive for neural lineage marker transcripts in the circulation “Before” and “After” performing “brain injury induction”. Left panel: percentage of positive hemocyte for dcx^+^ (or soxN+), numb+, and gcm + in the circulation. Middle panel: percentage of positive hemocyte for pros^+^, brat+, and numb + in the circulation. Right panel: A summarized inclusion of two probe combination. Comparison was done between “before” and “after” treatment samples for each individual, and analyzed with “Paired T-test” (two-tailed), and normal (Gaussian) distribution were tested with the “Shapiro-Wilk normality test” and “Q-Q plot. The asterisks () and (**) refer to statistic significant differences: P \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\le\:$$\end{document} 0.05 and P \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\le\:$$\end{document} 0.001, respectively, n=9
Crayfish brain, its organization, and Neurogenic Cell Identification
The brain histology of P. leniusculus was analyzed in two different anatomical planes as shown in Fig. 8, including horizontal and sagittal views, and then stained with a routine hematoxylin and eosin (H&E). Brains, collected from crayfish that had been injected with Indian ink, were used for histological studies to localize the vascular area associated with the brain tissue.
Fig. 8. Histology of brain tissue sections stained with H&E in horizontal and sagittal planes. A Low magnification of the right hemisphere of the intact brain tissue shows the brain sheath (or perineural sheath) surrounding the brain mass (including the neuropils and cell bodies), scale bar = 250 μm. B High magnification of the part labelled B in panel A, of the brain tissue near the neuronal cluster 10, which shows a loose connective tissue surrounding the brain mass, known as “intracerebral connective tissue,”. This tissue is enwrapped in a thick brain sheath, separating the external and internal environments. The brain sheath is lined internally by some unknown cell types shown by blue arrowheads in the insert, but their organization appears to be similar to that of the “perineural glia” of Drosophila (Hartenstein 2011), scale bar = 75 μm. C, D Parasagittal sections passing through: C at the area of the antennular nerve (An_1_Nv) entering the brain and terminating in the olfactory lobe (OL), and: D lateral to the area in C as indicated in the small brain pictures (Br) (please see Fig. S4D), scale bar = 250 μm. E A high-magnification area from part of D, a bundle of An_1_Nv was found at the location of the area between the olfactory lobe (OL) and accessory lobe (AcL), a particular area of interest for observing the neurogenic niche or its migratory stream (yellow area), (blue arrow indicated ink-signal). Abbreviations: PT; protocerebral tract, AMPN and PMPN; anterior and posterior median proto-cerebral neuropils respectively, CB; central body, OGTN; olfactory globular tract neuropil, MAN and LAN; medial and lateral antennular neuropils, AnN; antennular neuropil, ExCCT; external cerebral connective tissue, AV; anterior-ventral, PI; posterior-interior, scale bar = 75 μm
In the following text we used the term glia in the same way as for Drosophila species (Hartenstein 2011), and for brain the terminology of Sandeman et al. (1992). The tissue organization consists of intact structures, aligning from the periphery to the central brain mass. The brain contains neuronal clusters, associated glia, as well as puncta of neuronal fibers and neuropils (Fig. 8A–E). In the periphery, the brain is covered with a thick layer of “brain sheath” (Fig. 8A, B), which separates the connective tissues wrapping the brain into the “external cerebral connective tissue (ExCCT)” and “internal cerebral connective tissue (InCCT)” (Fig. 8B). The adjacent cells lining beneath the brain sheath are called the “perineurial glial-like cells,” which are the most peripheral cells of the brain that have been found (Fig. 8B, arrowheads). Inwardly, a loose connective tissue (InCCT) surrounds the brain mass, which is occupied by “sub-neural glia” (Fig. 8B). The neuropil-associated or ensheathing glia and cortex-associated glia (as in Drosophila) and are situated around neural bodies in the cortex) are the innermost layers that are located and form part of the brain mass (Fig. 8B).
The outer cerebral vascular plexus (OVs) was identified at histological levels (Fig. 8D, E). The sagittal planes show where the antennular nerve (An_1_Nv) passes through the brain, ventrally (Fig. 8C, D), and Supplementary Fig. 5 shows the ventral morphology of the brain with the section planes for Fig. 8C, D). Ventromedially to the brain, the An_1_Nv bundle was found, which penetrates the brain sheath before connecting to the olfactory lobe (OL) (Fig. 8C). In another parasagittal section, laterally, a tissue bundle of the An_1_Nv remains present. At the same time, the neurogenic niche and/or the neurogenic streams were identified underneath that area (Fig. 8D-E).
As shown in Fig. 9 localization of cells that expressed neural lineage marker transcripts was detected in the sagittal tissue sections of the brain and pros^+^/dcx^+^/gcm^−^/numb^−^ cells were located both inside (white arrow in Fig. 9A-B) and outside (yellow arrow in Fig. 9A-B) the brain sheath. The boundary of which is shown in Fig. 8B, and Fig. 9B-C. The positive cells were located outside the brain and situated near the nerve bundle (indicated by the yellow arrow in Fig. 9A). The flattened perineurial-like glial cells inside the brain sheath were pros^+^/dcx^+^ (Fig. 9B, green arrows), in addition to some round cells inside the hemal sinus (Fig. 9B, yellow arrow). Interestingly, the pros^+^/dcx^+^ cell located near the brain sheath co-expressed gcm in the same cell, while any expression of the numb gene was not detected (Fig. 9B, yellow arrow). Control experiments for any autofluorescence and non-specific signal are shown in Supplementary Fig. 6)
Fig. 9. Cells positive for neural lineage marker transcripts are present in brain tissue sections. A Sagittal brain section at the area of the olfactory (OL) and accessory lobes (AcL) showing pros^+^/dcx^+^ positive signals in the perineural glial-like cells (indicated by white arrows), which line inside the brain sheath (red dashed line) and the undefined cells associated with the An_1_Nv bundle (yellow arrows), scale bar = 200 μm, and magnified are scale bars = 20 μm. B A parasagittal section passing through the outer vascular plexus or vessels (OVs), which is associated with the An_1_Nv (in the area of white dashed line), showing pros^+^/dcx^+^ cells in the perineural glia-like cells (green arrow), and in round-shaped cells, which are located inside the ink-labelled vasculature (yellow arrow). A red dashed line indicates the brain sheath, scale bar = 20 μm. C A round-shaped pros^+^/dcx^+^/gcm^+^ cell (yellow arrow) was found close to the area of the brain sheath (red dashed line). Abbreviations: S; superior, V; ventral, scale bar = 20 μm and magnified are scale bars = 10 μm
Discussion
Neural Lineage Marker Genes Are Expressed in Circulating Hemocytes
A significant effort to study adult neurogenesis in crustaceans was initiated around the early 2000s, and comparisons to phenomenon in vertebrates and other invertebrates, have been made (Harzsch et al. 1999; Sullivan et al. 2000, 2007b; Goergen et al. 2002; Beltz and Sandeman 2003; Sandeman et al. 2011). Interestingly, the concept that hemocytes in the circulation can serve as an external source for adult neurogenesis has been reported for a few decades (Zhang et al. 2009; Benton et al. 2011), indicating that crustacean hemocytes have the potential to transdifferentiate into a new neuron in the nervous tissues (Benton et al. 2011, 2014; Beltz et al. 2011; Chaves-da-Silva et al. 2013) or that neuronal precursors are present in the hematopoietic system. Previous reports have made significant contributions to our understanding of crustacean adult neurogenesis through cellular tracing of potential hemocyte precursors to neuronal cells using mitotic cell assays, immunohistochemistry with universal neural-specific marker antibodies, and membrane-labelling assays. A recent study in Ucides cordatus, showed that isolated hemocytes cultured in vitro, could be stimulated by Poly-D-Lysine substrate and bovine pituitary extract, to become immunoreactive for the neurogenic markers beta III tubulin and NeuN (Wajsenzon et al. 2025). Moreover, neurogenic marker transcripts have been localized during embryogenesis in the marble crayfish Procambarus viriginalis (Brenneis et al. 2021). However, so far, no molecular characterization of such putative neural precursors present in the circulation or hematopoietic tissues have been performed, and no neurogenic-specific differentiation markers have been detected in the hemocytes of crayfish or any other crustacean in vivo. Recently, we could detect neural-related transcripts in some specific cell clusters in a single cell transcriptomic study of the crayfish hemocytes and hematopoietic tissue (Söderhäll et al. 2022). This led us to investigate further in what cell types, and in what proportion of the total hemocyte population we could find such transcripts. In the present study we could show that neural lineage marker genes, including prospero (pros),* brain tumor suppressor (brat), Notch signaling repressor (numb), doublecortin (dcx)*, and sox-neuro (soxN), were expressed in the circulating hemocytes by using RNA-FISH labelling techniques. The advantage of multi-probe labelling using RNA-FISH is that it provides valuable insights into the pattern of localization and expression. Our study demonstrates that the neural lineage marker transcripts mentioned above are expressed in the circulating hemocytes which is novel information and gives a clearer insight compared to only use of single-cell RNA sequencing (Söderhäll et al. 2022). Interestingly, the aforementioned transcripts share their co-localization in the same cell, and they are present in only 1% of the whole circulating hemocyte population. Our finding of the proportion of putative neural progenitors to total hemocytes aligns with the study in P. clarkii, where around 1–3% of the circulating hemocytes were positive for an antibody against glutamine synthase, a proposed marker for neural progenitor cells (Benton et al. 2011). This finding also supports that the crustacean hemocytes have the potential to develop into neural tissue by expressing these neural-specific markers (Zhang et al. 2009; Benton et al. 2011, 2014, 2022; Beltz et al. 2011; Chaves-da-Silva et al. 2013), while the exact mechanism behind this process requires further study.
The Existence of Neural Lineage Marker Transcripts Reflects Hemocyte Diversity and Neural Progression
Our findings based on co-localization of those neural-specific transcripts in the hemocytes by RNA-FISH reveal that their pattern of co-localization varied in different cells, as well as the fluorescent signal intensity of each transcript. This suggests that the small proportion of neural-specific hemocytes present in the circulatory system exhibits some heterogeneity or plasticity (Söderhäll et al. 2022, 2025; Kruangkum et al. 2025), rather than persisting in a single developmental stage prior to incorporation into the neural tissue. It was evident that cells positive for neural lineage–specific transcripts do not co-express hemolectin (hml), a specific marker for immunologically active hemocytes. Thus, it is reasonable to suggest that neural-determined hemocytes constitute a distinct lineage separate from other hemocyte lineages (Beltz and Benton 2017; Benton et al. 2022). Moreover, expression of the cytokine Ast1 was observed in some putative neuronal precursors as well as in other hemocyte types. This is in line with evidences reported that the expression of an ast transcript was found not only in hemocytes of the giant freshwater prawn, Macrobrachium rosenbergii, but also in nervous tissues (Wen et al. 2022). Mammalian prokineticin 2, a homolog of crustacean astakine, functions as a chemoattractant for SVZ-derived neuronal progenitors for adult neurogenesis in the olfactory bulb (Ng et al. 2005). Our results suggest that Ast1 could play a role in neural progression within a specific hemocyte populations in crayfish, in addition to its role in immunity and hematopoiesis (Söderhäll 2016).
In P. leniusculus, hyaline and semi-granular hemocyte morphotypes, were the predominant cells showing positive co-localization of neural lineage-specific transcripts. This corresponds to Benton and coworkers finding (Benton et al. 2011), that in the crayfish P. clarkii, the upper layer of extracted hemocytes by Percoll gradients, mostly contains hyaline and semi-granular type of hemocytes and some of these could be highly potent to form a new neuron in the crayfish brain (Benton et al. 2011). The hemocytes expressing neural lineage marker transcripts are related to the phase of neural development corresponding to the reports in vertebrates and Drosophila (Crémazy et al. 2000; Carl and Russell 2015). The localization patterns of those lineage marker transcripts in our study could reflect some putative roles based on their expression. In the early phase of development, SoxNeuro (soxN), the homolog of mammalian group B Sox1, Sox2, and Sox3 genes, play a crucial role as an early regulator of embryonic neural development, orchestrating gene networks that drive neural specification and differentiation from the neural ectoderm in Drosophila (Ferrero et al. 2014; Harding and White 2018). Evidences for the involvement of soxN in adult neurogenesis are scarce (Fernández-Hernández et al. 2013), and our finding of soxN transcript expression in hemocytes reflects an important role in adult neurogenesis.
The genes pros,* numb*, and brat are involved in asymmetric division, fate determination, transition of neuroblasts (NBs) to differentiation into ganglion mother cell (GMC), and amplifying of NB lineage (or self-renewal) in Drosophila (Betschinger et al. 2006; Egger et al. 2008; Bowman et al. 2008; Froldi et al. 2015; Harding and White 2018). A major mechanism that amplifies neural progenitor cells involves symmetric cell division. This process is regulated by a group of cell fate determinants by apical-basal polarity discrimination of the cell, including the co-expression of pros,* numb*, and brat complex in the GMC (Egger et al. 2008; Froldi et al. 2015). The co-expression of these genes within the same hemocytes in our study suggests that they may contribute to neural fate determination, distinguishing these cells from other hemocytes in the crayfish circulatory system as putative precursors of neurons or glial cells. However, the fluorescent intensity of each gene represents the dynamics of gene expression (Kim et al. 2020; Wu et al. 2024), which points to its particular functions at each time of observation.
Glial cells have been described in crustaceans during the mid 1900th (for review see (Radojcic and Pentreath 1979), and morphologically characterized in several species (Allodi et al. 1999; Allodi and Taffarel 1999; Harzsch et al. 1999). Later, immunoreactivity against glutamine synthase, as well as glial fibrillary acidic protein have been used to define glial cells in Panulirus argus (Linser et al. 1997; Schmidt and Derby 2011), in Macrobracchium rosenbergii (Allodi et al. 2006), in Ucides cordatus (Chaves-da-Silva et al. 2010; Hollmann et al. 2021) and in P. clarkii (Sullivan et al. 2007b; Beltz and Benton 2017). Function of the gcm gene in crustaceans is unexplored in comparison to that in Drosophila, where cells with glial properties have been described during embryogenesis (Jones et al. 1995; Schreiber et al. 1997). In crustacean, glia-like cells, immunoreactive for glutamine synthase have been described during adult neurogenesis (Sullivan et al. 2007a; Benton et al. 2011), although it is not known if these express the gcm gene. A glia cell missing gene (gcm) is reported in P. leniusculus by Junkunlo et al. (2020), and was found to mainly be expressed in the nervous tissues, including the brain and nerve. It lacked any expression in the hemocytes and hematopoietic tissue (HPT) as determined by RT-qPCR. However, our findings indicate that there exist gcm-positive cells co-localized with other neural lineage marker transcripts in hemocytes. The lack of detection by RT-qPCR is likely due to the low abundance of the gcm transcripts, which may fall below the detection threshold of conventional PCR assays when a large hemocyte population is used for RNA extraction. Our observations indicate that the gcm transcript does not exhibit a unique localization pattern comparable to that of pros,* brat*, and numb transcripts. It was detected both in cells that co-express and those that lack co-expression with other neural lineage marker transcripts. This suggests that gcm might be an intermediate factor in switching the role of neuronal or glial determinants (Hosoya et al. 1995).
Doublecortin (dcx) is a microtubule-associated protein that is well known for regulating neuronal migration and early post-mitotic maturation in the vertebrate brain (Vargas et al. 2005). It is expressed in the migrating neurons by controlling the organization and stability of microtubules (Gleeson et al. 1999). Drosophila does not have a direct ortholog of vertebrate dcx but it has proteins with doublecortin domains, named echinoderm–microtubule-associated proteins or dcx-emap. It was suggested that these play a role in microtubule dynamics within sensory structures, but they could not be related to neuronal migration or differentiation (Bechstedt et al. 2010). In crustaceans, information about dcx function is lacking, but the existence of the dcx transcript and its co-expression with specific neural lineage markers in the crayfish hemocytes indicate that this gene may play a role in contributing to neural commitment in a neural lineage.
We also analyzed the expression of pcna together with neural lineage markers in order to understand if these cells were in active cell division cycle. We could then detect some cells expressing varying levels of pcna, while others were negative, indicating different developmental stages of the putative neural progenitor cells. Two patterns of spatial localization of pcna mRNA were recognized by RNA-FISH in different hemocytes in this study; cytoplasmic as well as intranuclear, indicating that in some cells the presence of mature pcna mRNA is ready for translation. In contrast, the nuclear foci localization relates to the immature or nascent transcripts being synthesized or undergoing the splicing process in the nucleus (Femino et al. 1998; Vargas et al. 2005; Quinodoz et al. 2021).
The Hematopoietic Tissue Serves as a Source of Cells Capable of Differentiating into a Neural Lineage
It is well known that the hematopoietic tissue (HPT) serves as a stem cell source for hemocyte production in crayfish (Söderhäll 2016; Söderhäll et al. 2022; Söderhäll and Söderhäll 2022; Kruangkum et al. 2025). Moreover, it is proposed to be involved in the process of adult neurogenesis by giving rise to neural precursors (Chaves-da-Silva et al. 2013; Benton et al. 2022). In our present study we utilized a set of specific RNA probes for neural lineage markers, including pros/numb/brat, and soxN, to investigate their localization and expression levels in HPT. We could detect pros/numb/soxN/brat mRNA present in the same HPT cells, which were located peripherally to the lobule at both the apical and basal sites of the tissue. In addition, some positive cells were detached from the lobule, but notably, not all the peripheral cells expressed the neural lineage markers. Moreover, we could not detect any neural lineage transcripts in cells of the anterior proliferation center (APC) in P. leniusculus, which is an interesting finding. However, previous studies using BrdU, a cell proliferation tracer, and immunohistochemistry with a tyrosinated tubulin antibody, or CellTracker molecular probe, have revealed and suggested that adult neurogenesis requires progenitor cells from a source outside of the neural tissues, such as HPT and APC (Zhang et al. 2009; Beltz et al. 2011; Chaves-da-Silva et al. 2013). Based on our current results, we cannot draw a definitive conclusion that only HPT, rather than APC, is the source of neural stem cell production using RNA-FISH localization alone but some evidence points to this assumption.
Serotonin, astakine1, LPS, or Acute Brain Injury Could Induce Enrichment of the Neural Fate Determinant Hemocytes in the Circulation
To further investigate whether it was possible to manipulate the number of putative neuronal precursor cells in the circulation, we treated crayfish by injecting substances which are known to affect adult neurogenesis. We then showed that administering 5-HT into the P. leniusculus circulation can significantly enhance the proportion of neural lineage marker-positive hemocytes in the circulation, as well as in a group receiving Ast1 injection. This finding conforms with previous studies, which reported that these molecules are involved in increasing the total number of hemocytes, thereby increasing the proportion of niche cells in the adult crayfish brain (P. clarkii) (Benton et al. 2014, 2022). However, the mechanism by which this increase occurs still remains unclear. It is known that 5-HT injection specifically can cause a rapid increase in the number of circulating hemocytes already after a few minutes (Noonin 2018; Benton et al. 2022). It is also shown that 5-HT treatment cause release of Ast1 from hemocytes in vitro (Noonin 2018). Thus, it is likely that the injection of 5-HT in to the crayfish resulted in an increased level of plasma Ast1, which then is known to induce both proliferation of cells in the HPT, and their differentiation and release into the circulation (Söderhäll et al. 2005). Since we show that the 5-HT receptor is expressed on the surface of hemocytes, it is reasonable to suggest that this release of Ast1 after 5-HT administration as shown by Noonin (2018) is 5-HTR1-mediated. However, whether the simultaneous increase in circulating hemocytes that express neural lineage marker transcripts is due to specific induction of neural differentiation pathways, or to a general increase and change in the hemocyte population remains to be clarified. A saline-based solution injection was used as the control and had no effect on the quantity of neural lineage marker-positive hemocytes in the circulation.
The serotonin receptor 1 alpha (5-HTR1α) is expressed in several tissues, including the brain, HPT, and hemocytes in P. clarkii (Benton et al. 2011; Beltz and Benton 2017). The interaction of 5-HT has been proposed to play a role in adult crayfish neurogenesis, which is explained by the presence of 5-HT immunoreactivity in the rim of the vascular cavity, while its receptor, especially 5-HT1α receptor, is expressed in the hemocytes. This supports a communication between the putative cells and the niche (Benton et al. 2011; Beltz and Benton 2017). Here, we could also show that injection of 5-HT increased the expression of this receptor in hemocytes and specifically in cells that co-expressed neural lineage marker transcripts. This suggest that such cells may respond to increased levels of 5-HT. A possible explanation for this is proposed by Zhang and colleagues who suggested that 5-HT plays a role in the neural fate determination and development (Zhang et al. 2011). Our study points to that the 5Htr1 receptor is a shared characteristic of the neural stem cell population, which could be involved in regulating neural fate in hemocytes, with the potential to enter into the neurogenic niche from the circulation. In Drosophila, a study on the interactive linkage of nervous and immune systems revealed that 5-HT released from neurons signals to the vascular niche, the organ responsible for hematopoiesis. Serotonin is then bound to the 5-HT1B receptor in the aorta, which controls the process of lymph gland premature rupture prevention and lamellocyte differentiation (Liu et al. 2025). Our study implies that the existence of 5-HT and its receptor in the hemocytes, as well as in the HPT-cells in crayfish, could facilitate these processes, in a similar manner as in the Drosophila model.
There is no direct evidence regarding the role of LPS in neurogenesis. However, there is one piece of evidence by Xin and Zhang from a single-cell transcriptomic study done in the crayfish, P. clarkii. This study reported that LPS injection increased the proportion of hemocyte cluster 8 cells expressing transcripts related to the nervous system, as determined by single-cell RNA sequencing (Xin and Zhang 2023) and in this study LPS induced an increase in cells expressing some neural lineage marker transcripts. The mechanism by which LPS increases the proportion of neural lineage-positive cells as shown in the single-cell transcriptomic study by (Xin and Zhang 2023), as well as in the present study is not clear. It is known that LPS indirectly induces astakine release from the hemocytes, resulting in a boost in HPT cell proliferation and hemocyte differentiation (Söderhäll 2013), which in turn may cause a greater proportion of neural lineage marker positive cells in a similar way as 5-HT or Ast1 injection.
An induction of acute brain injury also led to an enhancement of the proportion of neural lineage marker-expressing cells in the circulation, as demonstrated by the use of a needle for brain injury induction. Several previous studies have revealed that neural injury, such as antennule/brain damage or any neurodegenerative inductions, can cause an increase in hemocytes in the circulation recruited to the damaged area, perhaps released from the HPT (Benton et al. 2014, 2022; Chaves-da-Silva et al. 2020). So far, no direct evidence in crustaceans supports that adult neural tissue damage stimulates the expression of neural fate determinant transcripts, whereas we now in this study showed an increased proportion of the neural lineage marker-positive hemocytes as a response to neural tissue injury in a similar manner as described in Drosophila that adult neurogenesis increases after brain injury (Fernández-Hernández et al. 2013; Li and Hidalgo 2020; Crocker et al. 2021; Simões et al. 2022). Acute brain damage activates the proliferation of adult neuroblasts in the brain (Fernández-Hernández et al. 2013) and upregulates neural progenitor transcripts for neurogenesis (Crocker et al. 2021). Unlike Drosophila, crayfish require the hemocytes as the outsourced neuro-progenitor to generate new neurons (Benton et al. 2011, 2014). We suggest that an increase in the percentage of neural lineage marker-positive cells in an acute brain injury could be a mechanism for cell recruitment response to the neurodegeneration in crustaceans.
Some, but not all, Neural Lineage Marker Transcripts Were Detected in Cells Lining the Brain Sheath and in Hemocytes Associated with the Vascular Plexus
The crayfish brain morphology and histology show the interconnection of the brain mass, its membrane envelop or brain sheath, and the projection of nerve bundles, especially the antennular nerve. Here we show that the antennular nerve comes through the brain at the ventral-medial side of the brain, where it is located close to the neurogenic niche and neurogenic stream on each side. These are enclosed by a thin connective tissue, the brain sheath, which is internally lined with a flattened cell layer of perineurial glia-like cells, as in Drosophila (Hartenstein 2011). The connection between the brain and HPT has been revealed in P. clarkii, mediated by the vasculature (Chaves-da-Silva et al. 2012, 2013), which are supposed to deliver the neural progenitor hemocytes along the ophthalmic artery, intracerebral vessels into the neuronal niche of the adult brain to form newborn neurons (Chaves-da-Silva et al. 2012, 2013; Sintoni et al. 2012; Benton et al. 2014). Histological examination combined with the RNA-FISH technique, specific to the early (pros and numb) and late (dcx and gcm) neural lineage-specific transcripts, showed pros positive cells among the perineurial glia-like cells and hemocytes within the vascular plexus associated with the nerve. Moreover, some of these cells were pros^+^/dcx^+^/gcm^+^ without expression of the numb gene in the intact brain tissue. It can be expected that the positive hemocytes have not only adopted a neural fate (pros⁺), but are also beginning to differentiate along the neuronal lineage (dcx⁺) (Gleeson et al. 1999; Harding and White 2018) or are developing glial characteristics (Hosoya et al. 1995; Cattenoz and Giangrande 2016). Interestingly, the positive signals of pros and dcx could be observed in the perineurial-like glia only at the ventral side, not in the dorsal side of the brain in this study. This glial type may possess greater neurogenic potential within the brain, in addition to that contributed by hemocytes as reported in previous reports (Benton et al. 2011, 2014, 2022; Chaves-da-Silva et al. 2012). The pattern of adult neurogenesis contributed by neuroglia trans-differentiation was reported in Drosophila (Crocker et al. 2021; Simões et al. 2022; Casas-Tinto et al. 2025), especially the ensheathing and astrocyte-like glia (Simões et al. 2022; Casas-Tinto et al. 2025). This observation may constitute further evidence of a conserved evolutionary pattern associated with our findings. The neural lineage marker transcripts that are expressed in the glial cells of the crayfish brain could be another possible mechanism for further study in crayfish neurogenesis.
Conclusion and Limitations
In summary, our present study is the first report showing molecular evidences for the presence of putative neural progenitors among hemocytes in crustacean hemolymph using multiplex RNA-FISH for several neural-lineage specific markers. We could also show that the proportion of circulating hemocytes expressing neural-lineage specific marker transcripts increased after stimuli with 5-HT, Ast1 or by a penetrating traumatic brain injury. The results show that there are cells with expression like ganglion mother cells in the crayfish circulation, and we suggest that these cells may have the capacity to develop into glial cells or neurons after being incorporated into the neurogenic niche in the brain.
Due to the use of an adult non-model organism, there are some technical limitations, resulting in that we are not able to link the presence of hemocytes expressing neural-lineage specific markers, with certainty to developing neurons in the brain. It is also not feasible to perform robust lineage tracing using, for example, CRISPR-based genome scarring or DNA barcoding in these animals. However, our findings are clearly in line with earlier studies, and add molecular evidences that strengthen the hypothesis that there is a specific subset of hemocytes that are neural precursors.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Benton JL, Li E, Weisbach E et al (2022) Adult neurogenesis in crayfish: identity and regulation of neural progenitors produced by the immune system. 10.1016/j.isci.2022.103993. i Science 10399310.1016/j.isci.2022.103993 PMC 894120335340434 · doi ↗ · pubmed ↗
- 2Söderhäll I, Fasterius E, Ekblom C, Söderhäll K (2022) Characterization of hemocytes and hematopoietic cells of a freshwater crayfish based on single-cell transcriptome analysis. 10.1016/j.isci.2022.104850. i Science 10485010.1016/j.isci.2022.104850 PMC 939157435996577 · doi ↗ · pubmed ↗