Abstract
Some organisms in nature have developed the ability to enter a state of suspended metabolism called cryptobiosis1 when environmental conditions are unfavorable. This state-transition requires the execution of complex genetic and biochemical programs1,2,3, that enables the organism to survive for prolonged periods. Recently, nematode individuals have been reanimated from Siberian permafrost after remaining in cryptobiosis. Preliminary analysis indicates that these nematodes belong to the genera Panagrolaimus and Plectus4. Here, we present precise radiocarbon dating indicating that the Panagrolaimus individuals have remained in cryptobiosis since the late Pleistocene (∼46,000 years). Phylogenetic inference based on our genome assembly and a detailed morphological analysis demonstrate that they belong to an undescribed species, which we named Panagrolaimus n. sp. Comparative genome analysis revealed that the molecular toolkit for cryptobiosis in Panagrolaimus n. sp. and in C. elegans is partly orthologous. We show that biochemical mechanisms employed by these two species to survive desiccation and freezing under laboratory conditions are similar. Our experimental evidence also reveals that C. elegans dauer larvae can remain viable for longer periods in suspended animation than previously reported. Altogether, our findings demonstrate that nematodes evolved mechanisms potentially allowing them to suspend life over geological time scales.
Introduction
Organisms from diverse taxonomic groups can survive extreme environmental conditions, such as the complete absence of water or oxygen, high temperature, freezing, or extreme salinity. The survival strategies of such organisms include a state known as suspended animation or cryptobiosis, in which they reduce metabolism to an undetectable level6. Spectacular examples of long-term cryptobiosis include a Bacillus spore that was preserved in the abdomen of bees buried in amber for 25 to 40 million years7, and a 1000 to 1500 years-old Lotus seed, found in an ancient lake, that was subsequently able to germinate8. Metazoans such as tardigrades, rotifers and nematodes are also known for remaining in cryptobiosis for prolonged periods9,10. The longest records of cryptobiosis in nematodes are reported for the Antarctic species Plectus murrayi11 (25.5 years in moss frozen at -20°C), and Tylenchus polyhypnus12 (39 years desiccated in a herbarium specimen).
Intensive research during the last decades has demonstrated that permafrosts (perennially frozen sediments) are unique ecosystems preserving life forms at sub-zero temperatures over thousands of years13,14,15,16. Permafrost remains are an exceptional source for discovering a wide variety of unicellular and multicellular living organisms surviving in cryptobiosis for prolonged periods6,17,18. We recently reanimated soil nematodes that were preserved in Siberian permafrost for potentially thousands of years, and initial morphological observations provisionally described them as belonging to the genera Panagrolaimus and Plectus. Previous studies demonstrated several species of Panagrolaimus can undergo cryptobiosis in the form of anhydrobiosis (through desiccation) and cryobiosis (through freezing)19,20,21,22,23. In various nematodes, entry into anhydrobiosis is often accompanied by a preparatory phase of exposure to mild desiccation, known as preconditioning21,24. This induces a specific remodelling of the transcriptome, the proteome, and metabolic pathways that enhances survival ability2,3,25. Some panagrolaimids evolved adaptive mechanisms for rapid desiccation where most of the cellular water is lost, while others evolved freezing tolerance, without loss of water, at sub-zero temperatures by inhibiting the growth and recrystallisation of ice crystals21. However, the genetic and biochemical mechanisms of long term cryptobiosis in these organisms have not yet been investigated in detail.
Here we present a high-quality genome assembly and a detailed morphological, phylogenetic analysis and define a novel species, Panagrolaimus n. sp. Precise radiocarbon dating indicates that Panagrolaimus n. sp remained in cryptobiosis for about 46,000 years, since late Pleistocene. Furthermore, making use of the powerful model organism C. elegans, we demonstrate that Panagrolaimus and C. elegans dauer larvae utilize similar adaptive mechanisms to survive extreme desiccation and freezing, i.e. upregulation of trehalose biosynthesis and gluconeogenesis.
Results
Discovery site and radiocarbon dating
Previously, we showed that nematodes from the Siberian permafrost with morphologies consistent with the genera Panagrolaimus and Plectus could be reanimated thousands of years after they had been frozen in the permafrost. Several viable nematode individuals were found in two out of more than 300 studied samples of permafrost deposits of different ages and genesis, collected by researchers of the Soil Cryology Lab, Pushchino, Russia, in the course of perennial paleo-ecological expeditions carried out in the coastal sector of the northeastern Arctic17. The detailed description of the study site (outcrop Duvanny Yar, Kolyma River, Fig.1A), sampling and revitalizing procedures are provided in Supplementary Information (SI). Like other late Pleistocene permafrost formations in the northeastern Arctic, Duvanny Yar comprises permanently frozen ice-rich silt deposits riddled with large polygonal ice wedges and divided by them into mineral blocks26,27 (Fig.1B). The sediments include sandy alluvial layers, peat lenses, buried paleosols and Pleistocene rodent burrows (Fig.1C). The burrow (P- 1320), in which Panagrolaimus nematodes were found (Fig.1D), has been taken from the frozen outcrop wall at a depth of about 40 m below the surface and about 11 m above river water level in undisturbed and never thawed late Pleistocene permafrost deposits. The fossil burrow left by arctic gophers of the genus Citellus consists of entrance tunnel and a large nesting chamber up to 25 cm in diameter 26.
The sterility of permafrost sampling and age of cultivated biota were discussed in detail in several reviews14,28,29. Based on previous reports, the age of the organisms found in a burrow is equal to the freezing time and corresponds to the age of organic matter conserved in the syncryogenic sediments. This makes it possible to use radiocarbon dating of organic matter to establish the age of organisms. We performed Accelerator Mass Spectrometry (AMS) radiocarbon analysis of plant material obtained from studied borrow P-1320 and determined a direct 14C age of 44,315±405 BP (Institute of Geography, RAS; sample IGANAMS 9137). Calibrated age range is 45,839 – 47,769 cal BP (95.4% probability) (Fig.S1).
Like other parthenogenetic Panagrolaimus the newly discovered species is triploid
The revived animal was cultivated in the laboratory for over 100 generations and initially described as Panagrolaimus aff. detritophagus4 based on morphology. We conducted a detailed morphological analysis of the revived animal (Fig.2, S2, Table S1;BOX1), which confirmed unambiguously that the animal belongs to the genus of Panagrolaimus, in agreement with a previous phylogenetic analysis of the 18S ribosomal RNA sequence4. However, due to the morphological uniformity of Panagrolaimus, unusual even for nematodes, morphology and molecular analysis of a single ribosomal RNA sequence is insufficient to describe a species. We found the species to be parthenogenetic, which further complicates description under most species concepts.
To obtain comprehensive molecular data for species determination using phylogenomics, we generated a genome assembly. Using PacBio HiFi sequencing, we generated 84X coverage in long reads (average length 14,425 bp). Our analysis of repeat and gene content is described in Supplementary Table 3. K-mer analysis of the reads clearly indicated that this animal has a triploid genome (Fig.3A), similar to other parthenogenetic Panagrolaimus species30. Despite the challenges that a triploid genome poses for assembly, we obtained a highly-contiguous contig assembly of the three pseudohaplotypes that comprise about 266 Mb and thus have a similar genome size as other parthenogenetic Panagrolaimus species30. The contig N50 value of all three pseudohaplotypes is 3.8 Mb. Since the three pseudohaplotypes exhibited a noticeable degree of divergence, we further investigated their relationship by using the apparent homeologs in our gene predictions to align long continuous contigs based on micro synteny (Fig. 3B). Links between the contigs clearly shows the triploid state of the genome.
Description of Panagrolaimus n. sp sp. nov.
Description
Body spindle-shaped and usually almost straight after fixation (Fig.2a, b). Cuticle thin and faintly annulated, 10–11 annules per 10 µm in cervical region, 13–14 in midbody, preanally again 10 annules withi 10 µm. Conspicuous convex lateral fields with three incisures 1.5-2 μm wide extended along the body from about ¼–1/3 procorpus length to 2/3 of tail length. In SEM, the lateral field looks as a bolster with a narrow median split. Labial region set off. Mouth opening surrounded with six lips (Fig.2c, d). Anterior sensilla as papillae arranged in two close but separate subsequent circles. Somatic sensilla (i.e., deirids and phasmids) not evident.Buccal cavity cylindro-conoid, and unarmed; its total length 9–13 µm, maximum stoma width 1.7–2.8 µm (Fig.2e). Dorsal stoma wall (dorsal rhabdion) more clearly sclerotized. Anterior part of the buccal cavity comprising cheilostom and gymnostom nearly cylindroids while stegostom conically narrowed and ended with a distinct tight flexion. Pharynx consists of three distinct parts: straight anterior procorpus, narrow medial isthmus and rounded terminal bulb. Procorpus gradually widening to it posterior end, always straight in all specimens, with transversal muscular striation, more prominent in posterior three fourths. Isthmus narrow, cylindroid, bent starkly in nearly all studied specimens. Terminal bulb strongly muscular, with a valvular apparatus at about 40% of the bulb length. Cardia in shape of truncate cone. Intestine (midgut) tissue filled with vacuoles and granules; in the anteriormost region, the granules smaller and look more pallid. Cell borders not visible in the intestine, but internal lumen distinct, sinusoid. No internal content visible in the internal lumen.
Ventral excretory-secretory pore and its cuticularized duct situated at the level of anterior part of the bulb. No other details of the excretory-secretory system visible. Vulval lips protruding. Genital branch monodelphic prodelphic and situated dorsally and to the right of the midgut (Fig.2f). Vagina distinctly cuticularized and opens to the elongate uterus. A long oviduct extended anteriad from the anterior uterus; the oviduct folded up and then posteriorward and transforms into an elongate ovary. There are one or two ripe eggs in the uterus in most specimens. Tail short conical, with short acute spike-like mucro (Fig.2g).
Etymology
Species name kolymaensis (Latin) is derived from the Kolyma River area.
Holotype
Senckenberg Natural History Museum, Frankfurt am Main, Germany (collection number SMF 17067).
Paratypes
Senckenberg Natural History Museum, Frankfurt am Main, Germany (collection numbers SMF 17068, SMF 17069) (eighteen paratypes).
Type locality
Frozen fossil rodent burrow buried in permafrost 45,839 – 47,769 cal BP , 40 meters from the surface, outcrop Duvanny yar, Kolyma River, North- East of Siberia, Russia (68° 37ʹ 739ʺ N, 159° 11ʹ 678ʺ E). Frozen material from burrow was collected by Dr.S.Gubin (Soil Cryology Lab, Pushchino, Russia) in august 2002.
Phylogenomics to define Panagrolaimus n. sp sp. nov
To place the species in the genus Panagrolaimus, we conducted a broader multi-gene phylogenomic analysis using Maximum likelihood methods. Our analysis of a concatenated, partitioned alignment of 60 genes, and a coalescence-based approach using a broader set of 12,295 gene trees, retrieves the novel species as sister to all other sequenced Panagrolaimus species, but as an ingroup to Propanagrolaimus31 (Fig.3C; Fig. S3B-C). Thus, the phylogenetic placement provides strong evidence that this animal belongs to a novel species. Further supporting this, there is substantial sequence divergence between this novel species, and Panagrolaimus sp. PS1159 and Panagrolaimus sp. ES5, estimated to be on average 2.06 and 2.11 amino acid substitutions per site in our concatenated alignment, respectively. The substantial divergence is in line with previous data on ages of Panagrolaimus nematodes30, and more broadly seen in nematodes, which can be hyper-diverse32,33. Our data also contradicts the assumption that parthenogenesis is monophyletic trait30 in the Panagrolaimus genus (Fig.3C). Based on the Kolyma River location where the animal was unearthed, we propose the following taxonomic classification and species name:
Phylum Nematoda Potts, 1932
Class Chromadorea Inglis 1983
Suborder Tylenchina Thorne, 1949
Family Panagrolaimidae Thorne, 1937
Panagrolaimus n. sp sp. nov.
C. elegans dauer larvae and Panagrolaimus n. sp. utilize similar mechanisms to enter and remain in cryptobiotic state for prolonged periods of time
In the absence of established genetic methods in Panagrolaimus n. sp. we referred to C. elegans as a powerful model system to gain insights into possible pathways for long term survival3,4,24,25. The high-quality genome of Panagrolaimus n. sp. allowed us to compare its molecular toolkit for cryptobiosis with that of C. elegans. We used orthology clustering and phylogenetics to investigate whether the genome of Panagrolaimus n. sp. contains genes previously implicated in cryptobiosis in the C. elegans dauer larva. Our analysis showed that the Panagrolaimus n. sp. genome encodes orthologs to a C. elegans trehalose phosphate synthase gene (tps-2) and to a trehalose phosphatase gene (gob-1) (Fig. 3D, supplementary file Orthology analysis). Furthermore, we found orthologs to all C. elegans enzymes required for polyamine biosynthesis, the TCA cycle, glycolysis, gluconeogenesis, and glyoxylate shunt (Fig.3D, supplementary file Orthology analysis) suggesting that Panagrolaimus n. sp. utilizes a similar molecular tool kit as C. elegans to facilitate survival of unfavorable conditions.
Our earlier findings established that among several developmental stages of C. elegans, only the dauer larva, formed during unfavorable conditions (such as low nutrients and high population density) could survive anhydrobiosis and freezing4,24. To survive extreme desiccation, C. elegans dauer larvae need to be first preconditioned at high relative humidity (98% RH) for 4 days24. During preconditioning, dauer larvae upregulate trehalose biosynthesis that ensures their survival to harsh desiccation24,25. We tested whether preconditioning also facilitates survival of P. kolymaensis. Although a small proportion of Panagrolaimus n. sp. individuals survive harsh desiccation and freezing without preconditioning (Fig.4A), mixed populations of Panagrolaimus n. sp. survive significantly (p value< 0.0001) higher to harsh desiccation upon preconditioning (Fig.4A). Similarly, preconditioning and desiccation further enhanced survival rate of Panagrolaimus n. sp. to freezing (-80°C). Like C. elegans, Panagrolaimus n. sp. upregulate trehalose levels up to 20-fold upon preconditioning (Fig.4B). We previously reported that, to upregulate trehalose levels upon preconditioning, C. elegans dauer larva dissipate their fat reserves (Triacylglycerols) by activating the glyoxylate shunt and gluconeogenic pathway25. Upon preconditioning, we found that triacylglyceride (TAG) levels are significantly decreased in Panagrolaimus n. sp. (Fig.S5A&B). To further investigate whether the acetyl-CoA derived from degradation of TAGs culminate in trehalose, we applied the previously developed method of metabolic labelling with 14C-acetate in combination with 2D-TLC 4,25. As shown in Fig. 4C, preconditioning led to huge increase of radioactivity in trehalose and to smaller extent in some amino acids (glycine/serine, phenylalanine; panels c and d). Interestingly, Panagrolaimus n. sp. displayed an additional spot (Fig. 4D, enumerated as 7), that was not found in C. elegans, which we identified as trehalose-6-phosphate (Fig.S5C-H), an immediate precursor of trehalose, using mass spectrometry. Thus, to resist harsh desiccation, like C. elegans, Panagrolaimus n. sp. utilizes glyoxylate shunt and consequently acetate derived from TAGs to synthesize trehalose. Detection of the immediate precursor (trehalose-6-phosphate) suggests the flux of metabolites is intense in the latter.
Finally, we investigated whether C. elegans can also survive in prolonged cryptobiotic state. Despite of preconditioning, the survival ability of desiccated dauer larvae at room temperature declines very rapidly, with most larvae dead after around 10 days (Fig. 4E) (Fig. S4A&B). Direct freezing at -80 C, leads to almost instant death of the animals. To test whether combining these conditions could extend the viability of dauer larvae (Fig. S4A&B), we transferred the desiccated larvae to -80 C. Remarkably, under these conditions, there was no significant decline in viability even after 480 days (Fig. 4E). Moreover, after thawing, the animals resumed reproductive growth and produced progeny in numbers like those of animals kept under control conditions (Fig. 4F). Since no reduction in survival was observed at any time point in extended experiments, these results suggest that the combination of anhydrobiosis and freezing can prolong the survival ability of dauer larvae. Thus, C. elegans, when exposed to combination of cryptobiotic states can survive for extremely long periods of time.
Discussion
Our findings here demonstrate the ability of the nematodes to survive in suspended animation for geological time scales. Along with the adaptive mechanisms of an organism, a naturally preserving ecosystem would synergistically aid the survival for prolonged periods. The Siberian permafrost is a unique repository for preserving organisms in sub-zero temperatures for millions of years. Expeditions in the past decade revived several organisms across various taxa from the Siberian permafrost5,34,35,35,36. The possibility to exploit permafrost as a source for reanimating multicellular animals was recognized already in 1936. A viable Cladocera crustacean Chydorus sphaericus preserved in the Transbaikalian permafrost for several thousand years37,38 was discovered by P.N. Kapterev, who worked at the scientific station Skovorodino as a GULAG prisoner. Unfortunately, this extraordinary observation remained unnoticed for many decades.
The new species can now be placed into the genus Panagrolaimus39, which contains several described and undescribed parthenogenetic and gonochoristic species and strains30,40. Many Panagrolaimus display adaptation to survival in harsh environments21 and the genus includes the Antarctic species P. davidii22. The genus Panagrolaimus is exceptional in its morphological uniformity even among nematode species that are hard to classify based on morphology in general. Thus, species designation via microscopic (including SEM) analysis is unreliable, which is further complicated by the absence of males in parthenogenetic species makes. Males have important diagnostic features as spicules and pericloacal papillae, females differ from one species to another mainly by morphometrics, whereby the interspecies differences (absolute measures and ratios) may be very fine. Our specimens are close by absolute sizes and ratios to females of the bisexual species Panagrolaimus detritophagus41. The only non- overlapping morphometric character is index “b” (body length: pharynx length): 5.6–6.8 in Panagrolaimus n. sp. versus 4.4-5.1 in P. detritophagus.
Consequently, we turned to phylogenomics methods to place the species on the tree. This showed that species is an outgroup to other known Panagrolaimus species, raising for the possibility of a second independent evolution of parthenogenesis in the genus, in contrast to previous findings30,31,40. This will need to be confirmed through further sampling and genome sequencing of Panagrolaimus species. We found Panagrolaimus n. sp. to be triploid and thus a hybrid origin is possible, as seen in other parthenogenetic Pangrolaimus30. The highly contiguous genome of Panagrolaimus n. sp. will allow for analyses of this trait in comparison to other Panagrolaimus species currently being genome sequenced. The successful phylogenomic species identification we conducted here demonstrates the utility of this method for morphologically hard to identify species, in particular when the additional data yielded by a fully sequenced genome are considered.
Our results provide a deeper insight into the homology of molecular and biochemical mechanisms between C. elegans and P. kolymaensis, which are not only taxonomically but also ecologically distinct. C. elegans can mostly be found in rotting fruits and plants in temperate regions42,43, while Panagrolaimus species are globally distributed and prevalent in leaf litter and soil40, including in harsh environments21. We show through orthology analysis that the well-studied molecular pathways used by C. elegans larvae to enter the dauer state, such as insulin44,45 (DAF-11, DAF-2 & DAF-16), TGF-ϕ346 (DAF-7), steroid47 (DAF-9, DAF- 12) are present in the genome of the Panagrolaimus n. sp. (Fig.S4C). While further functional analyses are needed to study molecular pathways detail, our results hint at convergence or parallelism in the molecular mechanisms organizing dauer formation and cryptobiosis.
As mentioned earlier preconditioning enhances the survival of Panagrolaimus n. sp. to desiccation and freezing. We previously reported that preconditioning induces trehalose upregulation in C. elegans dauer larvae and the elevated trehalose renders desiccation tolerance by protecting the cellular membranes24. It is not surprising that Panagrolaimus n. sp. upregulates trehalose, however the magnitude of trehalose elevation is higher than C. elegans dauer larvae. This indicates that central regulators (DAF-16, DAF-12) of trehalose upregulation may differentially regulate tps-2 in Panagrolaimus n. sp 48,49. Although Panagrolaimus n. sp. utilizes glyoxylate shunt and gluconeogenesis to upregulate trehalose levels, it is intriguing to observe that they accumulate substantial levels of trehalose-6- phosphate. Further investigation of this observation using perturbation experiments will provide insights into metabolic regulation in Panagrolaimus n. sp. upon preconditioning. Our findings for the first time demonstrate that C. elegans dauer larvae possess an inherent ability to survive freezing for prolonged periods if they undergo anhydrobiosis. It is tempting to speculate that undergoing anhydrobiosis might be a survival strategy of C. elegans to survives seasonal changes in nature.
In summary our findings indicate that by adapting to survive cryptobiotic state for short time frames in environments like permafrost, some nematode species gained the potential for individual worms to remain in the state for geological timeframes. This raises the question whether there is an upper limit to the length of time an individual can remain in the cryptobiotic state, potentially only limited by drastic changes to the environment such as strong fluctuations in ambient temperature, natural radioactivity, or other abiotic factors. These findings have implications for our understanding of evolutionary processes, since generation times could be stretched from days to millennia, species ages might be much longer than anticipated, and long- term survival of individuals of species can lead to the refoundation of otherwise extinct lineages. Finally, understanding the precise mechanisms of long-term cryptobiosis and cues that lead to successful revivals can inform new methods for long term storage of cells and tissues.
Methods
Materials and C. elegans strains
[1-14C] -acetate (sodium salt) from Hartmann Analytic (Braunschweig, Germany). All other chemicals were purchased from Sigma-Aldrich (Taufkirchen, Germany). The Caenorhabditis Genetic Centre (CGC) provided daf-2(e1370) and E. coli NA22 strains.
Genomic DNA isolation from Panagrolaimus n. sp nematodes
Panagrolaimus n. sp. nematodes were grown on several plates of NGM agar plated with E. Coli NA22 bacteria at 20°C. Worms were collected from the plates, washed with water at least three to five times by centrifugation at 1000 g to remove any residual bacteria and any debris. The worm pellet was dissolved in 5 volumes of worm lysis buffer (0.1M Tris-HCl pH=8.5, 0.1M NaCl, 50mM EDTA pH=8.0) and distributed in 1.5 ml of microcentrifuge tubes. These tubes are incubated at -80°C for 20 minutes. 100 µl of Proteinase ‘K’ (20 mg/ml) was added to each tube and they are incubated at 60°C overnight. 625 µl of cold GTC buffer (4M Guanidinium Thiocynate, 25mM Sodium citrate, 0.5% (v\v) N-lauroylsarcosine, 7%(v/v) Beta Mercaptoethanol) was added to the tube, incubated on ice 30 min, and mixed by inverting every 10 min. 1 volume of phenol–chloroform-isoamyl alcohol (pH=8) was added to the lysate and mixed by inverting the tube 10-15 times. Tubes were centrifuged for 5 min at 10,000 g at 4 °C to separate the phases. The upper aqueous phase was carefully collected into a fresh tube. One volume of fresh chloroform was added and mixed by inverting the tubes for 10-15 times and centrifuged for 5 min at 10,000 g at 4°C to separate the phases. One volume of cold 5 M NaCl was added, mixed by inverting the tubes and incubated on ice for 15 min. After incubation these tubes were centrifuged for 15 min at 12,000–16,000 g at 4 °C. The supernatant containing the nucleic acids were slowly transferred into a fresh tube. One volume of isopropanol was added to the tube, inverted few times, and incubated on ice for 30 minutes. After incubation, the tubes were centrifuged at 3000 g for 30–45 min at 25°C and the supernatant was discarded without disturbing the pellet. The pellet was washed twice with 1 ml of 70% ethanol, tubes were centrifuged at 3000 g for 5 min and supernatant was discarded and incubated at 37°C for 10-15 min to dry the pellet. The pellet was resuspended carefully in TE buffer. The quality of the genomic DNA was analyzed with pulse field gel electrophoresis.
Genome sequencing and assembly
Long insert library has been prepared as recommended by Pacific Bioscienes according to theprotocol ‘Procedure & Checklist-Preparing gDNA Libraries Using the SMRTbell® Express Template Preparation Kit 2.0’. In summary, RNAse treated HMW gDNA has been sheared to 20Kb fragments on the MegaRuptorTM device (Diagenode) and 10 µg sheared gDNA have been used for library preparation. The PacBio SMRTbellTM library was size selected in two fractions (9-13kb, > 13kb) using the BluePippinTM device with cassette definition of 0.75% DF MarkerS1 3-10 kb Improved Recovery. The second fraction of the size selected library has been loaded with 95 pM on plate on a Sequel SMRT cell (8M). Sequel polymerase 2.0 has been used in combination with the v2 PacBio sequencing primer and the Sequel sequencing kit 2.0EA, runtime was 30 hours. We created PacBio CCS reads from the subreads.bam file using PacBio’s ccs command linetool (version4.2.0). We obtained 8.5Gb hiqh quality CCS reads (HiFi reads) with a N50 of 14.4 Kb. We ran HiCanu (version 2.2)39 to create the contig assembly. Blobtools (version 1.1.1,CITE:) was used to identify and remove bacterial contigs. The final triploid contig assembly consists of 856 contigs has a N50 of 3.82 Mb and a size of 266Mb. The mitochondrial genome was created with the mitoHifi pipeline (version 2, https://github.com/marcelauliano/MitoHiFi) based on the assembled contigs and the closely related reference mitochondrial genome of Panagrellus redivivus (strain: PS2298/MT8872, ENAaccession: AP017464). The mitoHifi pipeline identified 49 mitochondrial contigs ranging from 13-32Kb. The final annotated circular mitochondrial genome has length of 17467 bp. To identify pseudohaplotypes in the P. kolymanensis sp. nov genome assembly, we selected the longest isoform of each predicted protein-coding gene in our assembly and in the C. elegans genome (downloaded from WormBase Parasite, release WBPS15) using AGAT (version 0.4.0) and clustered them into orthologous groups (OGs) using OrthoFinder (version 2.5.2). We identified OGs that contained three Panagrolaimus sequences (i.e. groups that were present as single-copy in all three pseudohaplotypes) and used these to identify trios of multi-megabase size contigs derived from the three pseudohaplotypes. We visualized synteny between the three pseudohaplotypes using Circos to plot the positions of each homeolog (version 0.69-8).
Genome annotation
We used RepeatModeler 1.0.8 (http://www.repeatmasker.org/ ) with parameter ‘-engine ncbi’ to create a library of repeat families. This library was used with RepeatMasker 4.0.9 to soft- mask the Panagrolaimus genome. To annotate genes, we cross mapped protein models from an existing Panagrolaimus as external evidence in the Augustus51 based pipeline. We evaluated the completeness of our predictions using BUSCO on the gVolante web interface.
Orthology analysis
We conducted a gene orthology analysis using genomic data from Panagrolaimus n. sp, the plectid nematode species from the permafrost, as well as genomic data from WormBase Parasite (https://parasite.wormbase.org; accessed 17/12/2020): Caenorhabditis elegans, Diploscapter coronatus, Diploscapter pachys, Halicephalobus mephisto, Panagrellus redivivus, Panagrolaimus davidi, Panagrolaimus sp. ES5, Panagrolaimus sp. PS1159, Panagrolaimus superbus, Plectus sambesii, and Propanagrolaimus sp. JU765. For plectids, genomic resources are scarce. We therefore added transcriptome data of Plectus murrayi, Anaplectus granulosus, Neocamacolaimus parasiticus, and Stephanolaimus elegans. The latter three transcriptomes were kindly provided by Dr. Oleksandr Holovachov (Swedish museum of natural history). The Anaplectus granulosus, and Neocamacolaimus parasiticus transcriptomes have been published 52,53. All three transcriptomes were assembled de novo with Trinity54. The exact procedures are described in the respective publications52,53 . The Stephanolaimus elegans transcriptome was assembled in the same way as described for Neocamacolaimus parasiticus. The Plectus murrayi transcriptome was built from raw reads deposited at NCBI (https://sra-downloadb.be-md.ncbi.nlm.nih.gov/sos2/sra-pub-run-13/SRR6827978/SRR6827978.1; accessed 22.12.2020). The transcriptome was assembled using Galaxy Trinity version 2.9.154,55 using all default options and including in silico normalization of reads before assembly. Transdecoder (conda version 5.5.0)56 was used to translate to amino acid sequence. Identical reads were removed with cd-hit version 4.8.157,58, and the Trinity get_longest_isoform_seq_per_trinity_gene.pl command56 (Trinity conda version 2.8.5) {Anaconda Software Distribution, Conda, Version 4.9.2, Anaconda, Nov. 2020; was used, to remove shorter isoforms. Amino acid translated longest isoforms from genomic data were extracted with AGAT {Dainat, https://www.doi.org/10.5281/zenodo.3552717} from genome assembly FASTA files and genome annotation GFF3 files, using the agat_convert_sp_gxf2gxf.pl command to convert all GFF3 files to the required GFF3 format, and the agat_sp_keep_longest_isoform.pl as well as the agat_sp_extract_sequences.pl commands to extract amino acid translated FASTA files. The headers and names of all FASTA files were modified to allow for simple species assignment of each sequence in subsequent analysis. The orthology analysis was conducted with OrthoFinder v. 2.5.159,60 using default settings. For genes of interest, we constructed alignments with MAFFT v. 7.47561 using the localpair and maxiterate (1000) functions. We removed spurious sequences and areas that were not well aligned with Trimal v. 1.4.rev2262 (procedure stated in supplementary file Orthology analysis below each phylogeny). We then ran phylogenetic analysis with Iqtree2 v. 2.0.663, with -bb 1000 option, testing the model for each analysis (models eventually used stated in supplementary file Orthology analysis). We also checked for PFAM domains using Interproscan v. 5.50-84.064. Part of the analysis was performed on the HPC RRZK CHEOPS of the Regional Computing Centre (RRZK) of the University of Cologne. The phylogenies were visualized with Dendroscope 3.7.665 and figures were created with Inkscape {https://inkscape.org}.
Phylogenomics
Sequences of 18S and 28S genes from 44 taxa across the Propanagrolaimus, Panagrolaimus, Panagrellus and Halicephalobus genera were aligned (MAFFT L-INS-I v7.475)61, concatenated 66and used to infer a species tree using maximum likelihood via (IQTREE)67 and partitioned by best-fit models of sequence evolution for both68. Nodal support was determined using 1000 bootstrap pseudoreplicates. A further 60 genes from 101 taxa were used to confirm the taxonomic position using the supermatrix concatenation methods outlined above. Given the limitations of differential gene sampling, we expanded our phylogenomic analyses to include a coalescence approach using 12,295 ML gene trees inferred for orthogroups containing the target animal. Instances of multiple genes per species per group were treated as paralogs/orthologs and analysed using ASTRAL-Pro69.
Desiccation survival assay
C. elegans dauer larvae desiccation assays were performed as described in. Panagrolaimus n. sp desiccation assays were performed with mixed populations of the worms.
Exposure of nematodes to extreme environments
Dauer larvae or Panagrolaimus n. sp nematodes were preconditioned and desiccated as described24 and then transferred to elevated temperature of 34°C, freezing (-80°C) and anoxia. Anoxic environment was generated in a desiccation chamber at 60%RH by flushing the Nitrogen gas into the chamber. The concentration of oxygen inside the chamber was monitored. After each timepoint they were rehydrated with 500 μl of water for 2-3 hours. Rehydrated worms were transferred to NGM agar plates with E. Coli NA22 as food. Survivors were counted after overnight recovery. Each experiment was performed on two different days with at least two technical replicates.
Trehalose quantification from nematode lysates
Trehalose measurements were performed as described in previous reports25.
Radiolabeling, metabolite extraction and 2D-TLC
The above-mentioned procedures were performed according to previous reports3,25.
Identification of trehalose-6-phosphate from TLC plates
Normalised aqueous fractions from the non-preconditioned and preconditioned samples were separated by high performance thin layer chromatography (HPTLC), using 1-propanol- methanol-ammonia (32%)-water (28:8:7:7 v/v/v/v) as first, dried for 15 min and 1-butanol- acetone-glacial acetic acid–water (35:35:7:23 v/v/v/v) second dimension respectively. Using the trehalose as a standard on both dimensions of the TLC, the regions of interest were scrapped out from the TLCs. The scraped-out silica was extracted with 10 ml of 50% methanol twice. The fractions were combined, dried under vacuum and dissolved in 100 µl of Ms mix solution containing 4:2:1 (Isopropanol:Methanol:Chloroform) with 7.5 mM ammonium formate. Mass spectrometric analysis was performed on a Q Exactive instrument (Thermo Fischer Scientific, Bremen, DE) equipped with a robotic nanoflow ion source TriVersa NanoMate (Advion BioSciences, Ithaca, USA) using nanoelectrospray chips with a diameter of 4.1 µm. The ion source was controlled by the Chipsoſt 8.3.1 soſtware (Advion BioSciences). Ionization voltage was + 0.96 kV in negative mode; backpressure was set at 1.25 psi. The temperature of the ion transfer capillary was 200°C; S-lens RF level was set to 50%. FT MS spectra were acquired within the range of m/z 50–750 at the mass resolution of R m/z 200 = 140000; automated gain control (AGC) of 3×106 and with the maximal injection time of 3000 ms. FT MS/MS spectra were acquired within the range of m/z 50–750 at the mass resolution of R m/z 200 = 140000; automated gain control (AGC) of 3×104 and with the maximal injection time of 30 s.
Triacylglycerols measurement from Panagrolaimus n. sp lysates
Non-preconditioned and preconditioned pellets were lysed in 200 µl of isopropanol with 0.5 mm Zircornium beads twice for 15 min. The lysates were centrifuged at 1300 g for 5 min. The supernatant was carefully collected without any debris, 20 µl of the lysate was used for protein estimation. Normalization was performed according to soluble protein levels, supernatant volumes corresponding to 50-100 µg of proteins were dried in the desiccator. 700 µl of IS ((10:3 (Methyl tert-butyl ether: ethanol)) mix (warmed to room temperature) was added to dried samples and left on a shaker for 1 hour. The samples were centrifuged at 1400 rpm and 4°C. 140 µl of water was added and left on shaker for 15 min. These samples were centrifuged at 13400 rpm for 15 min. The upper organic fraction was collected and transferred to 1.5 ml glass vial and left for drying in the desiccator. The dried samples were reconstituted in appropriate volume of 300 µl of 4:2:1 (Isopropanol:Methanol:Chloroform) and volume corresponding 1 µg was used for injection.
LC-MS/MS analysis was performed on a high-performance liquid chromatography system (Agilent 1200 HPLC) coupled to a Xevo G2-S QTof (Waters). The samples were resolved on a reverse phase C18 column (Cortecs C18 2.7um from Waters) with 50:50:0.1:1% (Water:methanol:formic acid: 1M Ammonium formate) and 25:85:0.1:1% (Acetonitrile:Isopropanol:Formic acid:1M Ammonium formate) as mobile phase. The following gradient program was used: Eluent B from 0 % to 100 % within 12 min; 100 % from 12 min to 17min; 0 % from 17 min to 25 min. The flow rate was set at 0.3 ml/min. The samples were normalised according to the total protein concentration and the worm numbers. TGD 50:00:00 was used as internal standard. Skyline software was used to analyse the raw data. The TG were extracted from Lipidmaps database.
Author contributions
AS, VG, PS, TK conceived and designed the study. AS, VG, TH, MP, AT, GH, MH, ER, PS and TK contributed to the original draft. AS performed isolation and cultivation of nematodes. VG performed desiccation survival assays, trehalose measurement, 2D-TLC of metabolites, trehalose-6-phosphate detection, combination of cryptobiosis experiments, genomic DNA isolation from Panagrolaimus n. sp, assembled the data from all the authors and prepared the figures. MP performed genome assembly. TH conducted orthology analyses and single gene phylogenies and worked on figures. AT provided morphological description, light, and scanning electron microscopy. LS analyzed the genome assembly and created figures. GMH performed the phylogenetic analyses. ST performed triacylglycerols measurements, trehalose- 6-phosphate detection with suggestions from Andrej Shevchenko. MH annotated supplementary table 3. PS performed genome annotation and supervised TH and LS. EWM supervised MP and had overall responsibility for sequencing and assembly including funding it.
Conflict of interest
The authors declare they have no conflict of interest relating to the content of this article.
Figure legends
Acknowledgements
VG is thankful to Andrej Shevchenko, members of Kurzchalia lab for helpful discussions and core facilities of MPI-CBG for assistance. We are grateful to Dr. S. Gubin for field study and sampling, our colleagues in Soil Cryology Lab, Pushchino and North-East Scientific Station in Chersky, Republic of Sakha (Yakutia) for their help and cooperation. The authors thank Long Read Team of the DRESDEN-concept Genome Center, DFG NGS Competence Center, part of the Center for Molecular and Cellular Bioengineering (CMCB), Technische Universität Dresden and MPI-CBG. GMH is funded by a UCD Ad Astra Fellowship. This work was supported by the Russian Foundation for Basic Research (19-29-05003-mk) to AS and ER. VG and TK acknowledge the financial support from the Volkswagen Foundation (Life? research grant 92847). PS and TH are supported by a DFG ENP grant to PS (DFG project 434028868). The authors are thankful to Richard Roy and Jens Bast for critical reading the manuscript. We thank Iain Pattern for suggestions in writing the manuscript. The funders have no role in the design of the study.