CTP synthase forms cytoophidia in archaea

CTP synthase (CTPS) is an important metabolic enzyme that catalyzes the rate-limiting reaction of de novo synthesis of the nucleotide CTP. Since 2010, a series studies have demonstrated that CTPS can form filamentous structures termed cytoophidia in bacteria and eukaryotes. However, it remains unknown whether cytoophidia exist in archaea, the third domain of life. Using Haloarcula hispanica as a model system, here we demonstrate that CTPS forms distinct intracellular compartments in archaeal cells. Under stimulated emission depletion (STED) microscopy, we find that those HhCTPS compartments are elongated filamentous structure, resembling cytoophidia in bacteria and eukaryotes. When Haloarcula cells cultured in low-salt medium, the occurrence of cytoophidia increases dramatically. Moreover, overexpression CTPS or glutamine analog treatment promote cytoophidium assembly in H. hispanica. Our study reveals that CTPS forms cytoophidia in all three domains of life, suggesting that forming cytoophidia is an ancient property of CTPS. Summary Author CTP synthase (CTPS), as a textbook molecule, has been studied biochemically for almost 70 years. It catalyses the last step of making up the nucleotide CTP. Since biochemical properties and the regulatory mechanisms of CTPS were thoroughly studied in the past decades, it came as a surprise when a new feature of CTPS was revealed in 2010. Multiple studies reveal that CTPS can form snakeshaped structures termed cytoophidia in bacteria and eukaryotes. However, it is not clear whether cytoophidia exist in archaea, the third domain of life. Here we use halophilic archaeon Haloarcula hispanica as a model and show that CTPS can form defined structures in archaea. Using super resolution microscopy, we confirm that those CTPS-containing structures are elongated filaments, similar to cytoophidia described in bacteria and eukaryotic cells. Therefore, this study demonstrates that CTPS forms cytoophidia not only in bacteria and eukaryotes, but also in archaea.


Introduction
CTP (cytidine triphosphate) is the basic building block of RNA and DNA and the key precursor in the biosynthesis of membrane phospholipids (Carman and Henry, 1989). The synthesis of CTP is the last committed step in pyrimidine nucleotide de novo synthesis catalyzed by CTP synthase (CTPS) (Koshland Jr and Levitzki, 1974). This catalytic reaction includes the ATP-dependent phosphorylation at the C-4 position of UTP to form intermediate 4phosphoryl UTP, which is reacted with ammonia from glutamine hydrolysis to generate CTP (Lieberman, 1956;Long and Koshland, 1978). GTP activates glutamine hydrolysis by allosterically binding to CTPS to form a covalent glutaminyl enzyme intermediate (Levitzki and Koshland, 1972;Long and Pardee, 1967). The product CTP is a competitive inhibitor of the enzyme, which is critical to maintain the intracellular CTP level (Endrizzi et al., 2005;Ostrander et al., 1998).
Compartmentalization is the process of formation of cellular compartments which play a key role in homeostasis (Liu, 2016).
Compartments are often defined as membrane-enclosed regions such as mitochondria and chloroplasts (Diekmann and Pereira-Leal, . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 5 / 37 2013), while also including protein-based membraneless organelles such as purinosomes  and Cajal bodies (Cajal, 1903). In 2010, CTPS was found to compartmentalize into filamentous structures in Drosophila. These structures are termed cytoophidia (cellular snakes in Greek) (Liu, 2010). Since then, independent studies have discovered that CTPS can form cytoophidia in bacteria (E. coli and Caulobacter crescentus) (Ingerson-Mahar et al., 2010), and eukaryotes such as budding yeast, fission yeast, human cells and plant cells (Carcamo et al., 2011;Chen et al., 2011;Daumann et al., 2018;Ingerson-Mahar et al., 2010;Noree et al., 2010;Zhang et al., 2014). The phenomenon that CTPS forms cytoophidia is conserved across bacteria and eukaryotes.
Life can be divided into three domains: archaea, bacteria and eukarya (Woese et al., 1990). Several studies have found evidence of CTPS forming cytoophidia in bacteria and eukaryotes, which raises a very interesting question: we would like to know if this process occurs in archaea. The reason for this interest is that many archaea survive in extreme environments in which most microorganisms are not able to survive. So we want to ask the following three questions: 1) Can CTPS form intracellular structures . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020.  (Madern et al., 2000;Thombre et al., 2016). H. hispanica was isolated from a solar saltern in Spain, and is a classic model in archaeal research (Juez et al., 1986). H. hispanica can survive well in a medium containing optimal NaCl concentration at 3.5-4.2 M. It grows slowly in the case of low salt. Other conditions are also important for H. hispanica growth; for example, it requires at least 0.005 M Mg 2+ concentration. In addition, the optimum growth temperature is 35-40°C and the optimal pH is 7.0 (Garrity, 2012). We found that there were few cytoophidia (about 1%) under normal growing conditions (AS168 medium, 37 ℃ ).
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 7 / 37

CTPS forms cytoophidia in H. hispanica
CTPS is evolutionarily conserved among all three domains of life: archaea, bacteria and eukarya (Figure 1). The H. hispanica CTP synthase (HhCTPS) gene is located in Chromosome I (Liu et al., 2011b), and HhCTPS contains two conserved regions: N terminal (residues 16 to 277, ligase domain) and C terminal (residue 303 to 535, glutaminase, type 1 glutamine amidotransferase). We tried to align HhCTPS with some classical cytoophidium-forming CTPS types and well-studied crystal structures of CTPS. There is a significant similarity between CTP synthase from H. hispanica and proteins from Thermus thermophiles, humans, Saccharolobus solfataricus, Drosophila, E. coli, Mycobacterium tuberculosis and Saccharomyces cerevisiae, which share 46% -53% sequence identity ( Figure S1; Table S1).
In order to visualize the H. hispanica CTPS in H. haloarcula, we fused modified pSMRSGFP to the C-terminus of CTPS and visualized the distribution of HhCTPS-GFP (pSMRSGFP knock-in strain) under laser scanning confocal microscopy and superresolution microscopy (Figure 2a-c). We found that CTPS is mostly . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. Therefore, we concluded that cytoophidia do exist in this species. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 9 / 37

H. hispanica CTPS forms cytoophidia in E. coli
To determine whether HhCTPS can form cytoophidia in other species independent of its own intracellular environment, we expressed mCherry-labeled HhCTPS in E. coli. We found that HhCTPS can form cytoophidia in E. coli (Figure 2d), but they are somewhat different to those formed by the E. coli CTPS (EcCTPS).
In E. coli cells, homogenous EcCTPS forms longer cytoophidia in more cells than heterogenous HhCTPS. The results demonstrate that Haloarcula CTPS has the ability to form cytoophidia in E. coli.

A glutamine analog promotes cytoophidium assembly in H.
hispanica Eukaryotic CTPS uses glutamine as a substrate. It was found that 6-diazo-5-oxo-l-norleucine (DON), an analog of glutamine, inhibits CTPS activity by irreversibly binding to its glutamine amidotransferase domain. Meanwhile, DON promotes cytoophidium assembly in Drosophila and human cells (Boeke et al., 1984;Levitzki and Koshland Jr, 1971), but inhibits the formation of cytoophidia in bacteria (Ingerson-Mahar et al., 2010). To understand how DON affects HhCTPS in H. hispanica, cells were cultured in medium supplemented with DON. We observed that DON treatment . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint induces cytoophidium assembly in H. hispanica cells ( Figure S3).
Meanwhile, DON treatment inhibits cell growth in a concentrationdependent manner ( Figure S3b). Our data indicate that the effect of DON on HhCTPS-forming cytoophidia is similar to that on Drosophila and human CTPS, but different to that on bacterial CTPS.

Low salinity induces cytoophidium formation in H. hispanica
Next, we wanted to know how formation of cytoophidia is related to the physiological characteristics of H. hispanica. We cultured H.
hispanica under different conditions. We found that the H. hispanica  (Han et al., 2009) or resuscitation medium to arrest or promote cell growth. We found that cells were also deformed into a bigger or longer shape, but the . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint cytoophidia had no effects ( Figure S5).
In addition, to our surprise, HhCTPS can form a large number of intracellular structures under low-salt conditions. In 1.5M NaCl AS168 medium (Figure 4a, b), the H. hispanica cells grew very slowly, with growth almost stopping (Figure 4c . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint

Overexpressing Haloarcula CTPS promotes cytoophidium formation
We considered whether the level of CTPS has an effect on cytoophidium formation. Cytoophidia were induced in many H. hispanica cells when CTPS was overexpressed (Figure 5a). STED images showed that cytoophidia induced by HhCTPS overexpression were indeed elongated structures (Figure 5b).
Moreover, the formation of cytoophidia changes with the growth phases of archaea (Figure 5c,d). There was a dramatic decrease of abundance of foci on the first day. Narrowing the time range to 0-60h then revealed an obvious drop followed by a rise in abundance ( Figure S7). The number of cytoophidia in the early stage of growth was reduced, but the GFP signal became stronger. The abundance of cytoophidia in the logarithmic phase and subsequent growth phases was almost unchanged, but the size of these structures decreased initially and then increased (Figure 5e,f).

Yeast extract removal affects cytoophidium formation in H. hispanica
Low salinity (1.5 M NaCl AS168 medium) induced cytoophidium formation. We next investigated whether altering other components . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. have influence on either cell shape or cytoophidia formation among these components. We found that Mg 2+ deprivation made the cells rounder, similarly to low salt influence, but did not affect HhCTPS cytoophidium formation ( Figure S8). Interestingly, we found that HhCTPS also formed a number of cytoophidia (Figure 6a,b), noted as elongated structures under STED microscopy (Figure 6c), in the cells cultured in yeast extract-subtracted AS168 medium. Moreover, the growth curve of HhCTPS-GFP indicated that the growth of H. hispanica is inhibited in yeast extract-subtracted AS168 medium (Figure 6d).
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 14 / 37

Discussion
In summary, we found that CTPS can form cytoophidia in H.
hispanica. Our study demonstrates that CTPS forms cytoophidia not only in bacteria and eukaryotes but also in archaea. Cytoophidia exist in all three domains of life, suggesting that this is an ancient and fundamental property of CTPS. Furthermore, we found that environmental stress, such as low salinity or yeast extract deprivation, can promote cytoophidium assembly in H. hispanica.

CTPS may be mostly diffused under normal circumstances.
However, in the case of low salt, CTPS can aggregate to form cytoophidia. Meanwhile, the cell shape becomes rounder, which is similar to Mg 2+ deprivation in H. hispanica. Research has shown that the Na + /Mg 2+ ratio has an influence on amylase activity (ENACHE et al., 2009). When we decrease the concentration of Na+ or Mg2+, is the HhCTPS activity affected and are cytoophidia formed in response? This deserves further study. Some previous in-vitro studies indicated that the product nucleotide-bound filament is an inactive form of CTPS in prokaryotes, but an active form of CTPS in eukaryotes. (Barry et al., 2014;Lynch et al., 2017). However, two recent studies showed that Drosophila CTPS and human CTPS2 can form two kinds of filament: product-induced and substrate-. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 15 / 37 induced (Lynch and Kollman, 2019a, b;Zhou et al., 2019). It is therefore necessary to further define the CTPS activity state with regard to the different types of filament. We worked to purify the HhCTPS expressed by E. coli, and found that it has no enzyme activity. However, the HhCTPS protein purified from halophiles has obvious enzyme activity, which indicates that halophilic proteins need the salt environment to support normal translation or protein folding (unpublished data). We wonder which kind of filament HhCTPS can form in vitro: product-bound, substrate-bound, or both?
In addition, do different Na + /Mg 2+ ratios, or Na + or K + concentrations affect filament formation or enzyme activity?
Low salt induced cell shape change might be related to osmotic pressure. The amino acid composition of halophilic archaea proteins is enriched in acidic residues on the surface of the protein compared with those from non-halophilic archaea. An increase in the number of acidic residues is thought to promote the correct folding of proteins under high salt conditions (Fukuchi et al., 2003). Meanwhile, the walls are assembled by surface layer proteins that form the S layer. In most archaea (Kandler and König, 1998). The S-layer glycoproteins are also enriched in acidic residues (Eichler, 2003).
Under low salt conditions, S-layer glycoproteins might lose function.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint

/ 37
It is hard to maintain normal cell shape. Halophilic archaea can grow in environments where water activity (a w ) is close to 0.75, but a w below 0.90 has an inhibitory effect on most microorganisms (Stevenson et al., 2015). The amount of solute is related to the osmotic pressure of the microorganisms. When there is a significant imbalance in the osmotic pressure of the cells, it may lead to cell lysis, protein unfolding, and enzyme inactivation (Cheftel, 1995).
Halophilic archaea overcome this problem by keeping compatible solutes such as potassium chloride (KCl) in the cell to balance osmotic pressure (Antranikian, 1998). The way halophilic archaea retains these salts is called "salt-in". The largest component in the AS168 medium is NaCl. When we lowered the NaCl concentration in AS168, HhCTPS tended to compartmentalize. We deduced that there is no need for more acidic residues to resist extracellular salt during low salt conditions. As a result, these proteins tend to cluster together in order to hide the acidic residues. DON treatment inhibited cell growth but promoted HhCTPS assembly. What's more, more than half of the cytoophidia were located in the middle of the H. hispanica cell. It is known that the glutamine analog DON promotes CTPS cytoophidium formation in the human cell and Drosophila cell, but diminishes CTPS . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 17 / 37 compartmentalization in bacteria (Chen et al., 2011;Ingerson-Mahar et al., 2010). CTPS utilizes either L-glutamine or ammonia as the source of nitrogen (Endrizzi et al., 2004;Willemoes, 2004). Lglutamine is the primary amino donor in E. coli as in animals, but the utilization of ammonia was stimulated by DON (Chakraborty and Hurlbert, 1961). This difference might be due to the bias of nitrogen donors while treating with DON. A study in S. pombe showed the asymmetric inheritance of both cytoplasmic and nuclear cytoophidia during cell division (Zhang et al., 2014). The cell growth inhibition and cell enlargement induced by DON treatment might be due to cell cycle arrest in H. hispanica. The H. hispanica cells can't divide normally into daughter cells, which leads to larger or longer cells.
Also of interest is that the CetZ proteins related to tubulin and FtsZ play a key role in controlling the archaeal cell shape. And some CetZ proteins are located at the middle of the cell during the log phase growth stage while being localized at one or both poles in the motile halo cells (Duggin et al., 2015). Many HhCTPS cytoophidia are located near the cell envelope during culture in the normal AS168 medium. DON treatment facilitates the cytoophida to gather at the middle of the cells, which indicates that cell cycle arrest stops cytoophidia from being inherited by daughter cells.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 18 / 37 H. hispanica is one example of archaea. Our work shows that CTPS can form cytoophidia not only in eukaryotes and bacteria, but also in archaea. This work and studies in several laboratories (including ours) in the first decade of the cytoophidium field indicate that filament forming is an ancient and evolutionarily conservative phenomenon. In order to understand more about the role of the cytoophidia formed by CTPS in archaea, further exploration is necessary for other archaea species, such as thermophiles and methanogens.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint

Strains and growth conditions
The strains used in this study are listed in Supplementary Table   S2. Escherichia coli was grown in Luria-Bertani medium at 37°C (CHONG, 2001), with ampicillin at a final concentration of 100 μg/ml.

Plasmid construction, transformation, and gene mutants
The plasmids used in this study are listed in Supplementary Table   S2 and primers are listed in Supplementary Table S3. For knock-. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020.  Table   S3) from plasmid pWL502-GFP and genome. These three DNA fragments were linked and inserted into the suicide plasmid pHAR (predigested with BamHI and KpnI), and transferred into DF60 cells to knock in PSMRSGFP by the pop-in/pop-out method described previously (Liu et al., 2011a). The expression plasmid pWL502 (Cai et al., 2012) was used to overexpress HhCTPS-GFP. A DNA fragment of HhCTPS with PSMRSGFP was amplified from HhCTPS-GFP knock-in strain containing a BamH I site at 5′ end and a Kpn I site at 3′ end. This DNA fragment was inserted into plasmid pWL502 (predigested with BamHI and KpnI) and transferred into DF60 cells. pWL502-GFP (pWL502-pSMRSGFP) was used as control. All transformants were screened by PCR verification and confirmed by sequence analysis.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint

Heterogeneous expression
For heterogeneous expression of HhCTPS in E. coli, DNA fragments of HhCTPS or EcCTPS operon with linker (GGGS) and mCherry gene were amplified by primer pairs ECCTPS-F / ECCTPS-R, mCherry-F / mCherry-R. Two DNA fragments were linked and inserted into plasmid pET28a, and transferred into Transetta (TRANSGEN BIOTECH). IPTG with a final concentration of 1 μM was added when needed.

Preparation of fixed H. hispanica
Paraformaldehyde containing 20% NaCl was added to H. hispanica cell culture to give a final concentration of PFA of 4%. After fixing for 10 min at room temperature, cells were spun down for 3 min at 6000 rpm/min, then washed once in basal salt solution (BSS; per liter, 200 g NaCl, 20 g MgSO 4 · 7H 2 O, 2 g KCl, and 3 g trisodium citrate).
Samples were stored in BSS at 4°C.
Images were acquired under 63x objectives with differential . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020.

Growth analysis
H. hispanica cells were diluted in growth medium to a start optical density of approximately 0.1 at 600 nm (OD 600nm ) (with growth medium used for blank). 5 ml cell cultures were incubated in 12 ml flasks or 20 ml cell cultures were incubated in 100 ml flasks at 37°C with shaking (220 rpm). The OD 600nm was measured for three to ten . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint 23 / 37 days in each flask. The growth curves represent the average of three independent samples. The cells grown in 1.5 M NaCl AS168 were centrifuged for 3 min at 6000 rpm/min and the supernatant was removed. Then the cells were washed once with AS168 or 1.5 M NaCl AS168, and finally transferred to AS168 or 1.5 M NaCl AS168 for continuous cultivation. The OD 600nm was measured every day in each flask. For cell growth on the agar plate, 2 μl of culture in the growth medium was inoculated onto the subsurface of another agar plate, with up to four inoculations per plate from the same liquid culture. Plates were incubated in a closed plastic box at 42°C for 10 days, and then images of the plates were acquired using a camera. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted January 23, 2020. ; https://doi.org/10.1101/830091 doi: bioRxiv preprint