2.1 Culture of Chara braunii, growth conditions and dissection of central and nodal cells

Cultures of the non-axenic freshwater strain Chara braunii S276 (KU-MACC) were vegetatively propagated. Chara braunii S276 was originally isolated from Lake Kasumigaura (Ibaraki, Japan), then maintained at Kobe University (Kawai et al., 2022; Sato et al., 2014) and consecutively propagated at the Universities of Marburg and Freiburg. Plantlets were grown using the protocol of Holzhausen et al. (2022) in double autoclaved culture vessels containing sieved compost (Gardol Pure Nature, Bauhaus), lime (Gardol Garten- & Rasenkalk, Bauhaus), quartz sand (0.4–0.8 mm in diameter, Carl Roth GmbH) and distilled water, sealed with surgical tape (Micropore, 3 M). Algae were kept under long-day conditions at 22°C in a 16 h light: 8 h dark cycle with white light lateral illumination (30–70 μmol photons m⁻² s⁻¹, Lumilux L36W/840, OSRAM).

Algae were cut along their central thallus below and above each node, removing internodal and lateral cells (including branchlets), at the University of Marburg using razor blades and a stereomicroscope (Leica DM6000 CS). Harvested nodes from multiple algae were pooled, collected in 1.5 ml Eppendorf tubes, frozen in liquid nitrogen and stored at −80°C until further analysis.

2.2 Preparation of total RNA and northern hybridizations

Total RNA was extracted from complete thalli utilizing a modified acid guanidinium thiocyanate-phenol-chloroform protocol (Chomczynski & Sacchi, 1987, 2006) using PGTX (Pinto et al., 2009), but omitting Triton X-100.

At least 100 mg FW total Chara braunii samples were ground into a homogenate using mortar and pestle under liquid nitrogen. Samples were suspended in 1 ml Z6 buffer (8 M guanidiniumhydrochloride, 20 mM MES, 20 mM EDTA, 50 mM 2-mercaptoethanol, pH 7), mortar rinsed with 500 μl Z6 buffer, and transferred to a screw-cap tube. 3 ml PGTX was added and samples were incubated for 30 min at room temperature while vortexing every 5 min. Two chloroform extractions were performed consecutively: each adding 3 ml chloroform/isoamylalcohol (24:1), incubating samples for 10 min at room temperature under occasional vortexing, centrifugation for 3 min (3.273 g, at room temperature) and transferring the aqueous layer to a fresh tube. After chloroform extractions, RNA was precipitated using one volume 2-propanol overnight at −20°C. RNA was collected by centrifugation for 30 min (13,000 g, 4°C), and washed two times using 70% ethanol and air-dried for 10 min. RNA was resuspended in 100 μl RNase-free H₂O and stored at −80°C until further analysis.

Nodal RNA was extracted similarly. Pooled nodal cells were suspended in 250 μl Z6 buffer in ice-cooled 2 ml reaction tubes. Cell disruption was performed using a mixer mill (MM 400, Retsch; 5 cycles of 90 s with varying frequencies: 30 s 8/s, 30 s 16/s, 30 s 25/s, and 2 min cooled on ice) using one glass bead (2.85–3.45 mm diameter, Carl Roth GmbH). The supernatant was transferred into a new reaction tube, together with 125 μl of Z6 buffer used for rinsing the glass beads. 750 μl PGTX was added and samples incubated for 30 min at room temperature while vortexing every 5 min. Addition of 750 μl chloroform/isoamylalcohol (24:1) was followed by incubation for 10 min at room temperature under occasional vortexing. Afterwards, sample tubes were centrifuged for 3 min (room temperature, 3273 g). Following another chloroform extraction, the upper aqueous phase was precipitated with one volume 2-propanol overnight at −20°C. RNA was collected by centrifugation for 30 min (13,000 g, 4°C) and air-dried for 10 min. RNA was resuspended in 20 μl RNase-free H₂O and stored at −80°C until further analysis.

RNA concentration and purity were measured using a NanoDrop ND-1000 spectrophotometer according to manufacturer instructions (PEQLAB Biotechnologie GmbH). Residual DNA was removed using the Turbo DNase-free™ Kit (Thermo Fisher Scientific) following manufacturer instructions. RNA was recovered using RNA Clean & Concentrator kits (Zymo Research). RNA integrity was controlled on a 5200 Fragment Analyzer System, according to manufacturer instructions (Agilent).

Eight μg total RNA was separated by 1.4% denaturing agarose gel electrophoresis and transferred by capillary transfer on Hybond N+ nylon membranes (Cytiva Europe GmbH) overnight. Northern hybridization was performed with radioactively labeled probe for chlorophyll a/b binding protein (cab) mRNA (primers for template; see Table 1 for sequences), generated using [α-32P]-UTP and the Maxiscript T7 in vitro transcription kit (Thermo Fisher Scientific). Blotted RNA was crosslinked to the membrane via 240 mJ using a UV-Stratalinker (Stratagene). Hybridizations were performed in Northern buffer (50% deionized formamide, 7% SDS, 250 mM NaCl and 120 mM Na₂HPO₄/NaH₂PO₄ pH 7.2) overnight at 62°C. The membranes were washed at 57°C in buffer 1 (2× SSC (3 M NaCl, 0.3 M sodium citrate, pH 7.0), 1% SDS), buffer 2 (1× SSC, 0.5% SDS) and buffer 3 (0.1× SSC, 0.1% SDS) for 10, 5, and 2 min. Hybridization signals were detected via Typhoon FLA 9500 imaging system (GE Healthcare) using phosphorimaging and quantified using Quantity One software (Bio-Read Laboratories Inc.).

Templates were amplified from Chara genomic DNA using OneTaq Quick-Load 2× Master Mix with Standard Buffer. To facilitate the use of PCR products as templates for in vitro transcription, the T7 RNA polymerase promoter (TAATACGACTCACTATAGGG) was included in the 5′ sections of the corresponding reverse primers (Table 1).

2.3 Preparation of total DNA from Chara braunii

Genomic DNA was extracted from at least 100 mg FW total Chara braunii samples ground into a homogenate using mortar and pestle under liquid nitrogen. Samples were suspended in 1 ml SET buffer (1 mM EDTA, 25% sucrose, 50 mM Tris) and lysed overnight at 50°C in a screw-cap tube after adding 1/4 volume 0.5 M EDTA, 1/10 volume 20% SDS and 100 μg/ml proteinase K (Scholz et al., 2019).

After the addition of 2 volumes of ROTI Aqua-Phenol (DNA; Carl Roth GmbH) and 2 volumes of chloroform/isoamylalcohol (24:1), samples were incubated for 5 min at room temperature while vortexing intermediately. Samples were centrifuged for 5 min (3273 g, room temperature), the aqueous layer was transferred to a fresh tube and the chloroform extraction step was repeated. Afterwards, DNA was precipitated using one volume 2-propanol overnight at −20°C, collected by centrifugation for 30 min (13,000 g, 4°C), and washed two times using 70% ethanol and air dried for 10 min. DNA was resuspended in 100 μl RNase-free H₂O and stored at −20°C until further analysis.

2.4 cDNA sequencing

cDNA libraries were constructed and sequenced as a service provided by Vertis Biotechnologie AG (Germany). Poly(A)+ RNAs were fragmented using ultrasound (one pulse of 30 s at 4°C), after which an oligonucleotide adapter was ligated to the 3′ termini. First-strand cDNA synthesis was performed using M-MLV reverse transcriptase and the 3′ adapter as primer. First-strand cDNA was purified and the 5’ Illumina TruSeq sequencing adapter ligated to the 3′ end of the antisense cDNA. The resulting cDNA was PCR-amplified using a high fidelity DNA polymerase. The cDNA was purified using an Agencourt AMPure XP kit (Beckman Coulter Genomics).

For Illumina NextSeq sequencing, samples were pooled in approximately equimolar amounts, purified using the Agencourt AMPure XP kit (Beckman Coulter Genomics) and analyzed by capillary electrophoresis. Primers used for PCR amplification were designed for TruSeq sequencing according to manufacturer instructions. The cDNA pool was single-read sequenced on an Illumina NextSeq 500 system using 75 bp read length. Single replicates were sequenced for the pooled nodal cells and the control cell samples, respectively.

2.5 Bioinformatic analyses

Data analysis was performed on the European instance galaxy server (The Galaxy Community, 2022) following the available guidelines for reference-based RNA-seq analysis (Batut et al., 2018, 2022). The quality of raw reads was assessed using FastQC (Andrews, 2010) and MultiQC (Ewels et al., 2016). Trimming of adapter and barcode contamination was performed using cutadapt (Martin, 2011). Mapping and counting of reads against the published Chara genome (Nishiyama et al., 2018) was performed via RNA STAR (Dobin et al., 2013). Differential gene expression analysis was performed via log₂FC calculation following the given guidelines. Differentially expressed genes were determined with a cutoff as log₂FC ≥ |1|. Gene ontology (GO) analysis of RNA-seq data was performed using goseq (Galaxy version 1.44.0; Young et al., 2010).

2.6 Phylogenetic analysis

A set of amino acid sequences (Table S1) was selected based on position within the green lineage and sequence similarity to Chara braunii putative amylase homologous genes of interest, g31563 and g41182, and putative MurE-like PAP11, g32224. The sequences were aligned with M-coffee (Di Tommaso et al., 2011; Notredame et al., 2000) and further analyzed using the BEAST 2 software package version 2.7.3 (Bouckaert et al., 2019). Calculations were performed using standard BEAUTi settings utilizing the tree prior yule model, substitution model Blosum62 and MCMC chain length of 1e6, with logged parameters every 1e4 steps. Tracer v1.7.2 software (Rambaut et al., 2018) was used to validate the effective sample sizes (ESS > 200). TreeAnnotator was used to build maximum clade credibility trees using burnIn of 50% and a posterior probability limit of 0.5 for median node heights. FigTree v1.4.4 software (Rambaut, 2018) was used to visualize the generated dendrograms.

3.1 The most abundantly transcribed genes in Chara nodal cells

The nodal cells of Chara braunii are embedded in the zone between two adjacent large internodal cells from where the branchlets are emanating (Figure 1). By cutting the algae below and above each node, only the nodal cells from the central thallus were collected. Separated nodes were kept in liquid nitrogen until RNA extraction. RNA was extracted from 28 pooled nodes, yielding 0.68 μg of total RNA after DNase treatment. In parallel, RNA was extracted from fresh total algal material. We decided to use complete above-ground thalli, including nodal cells, as control. The inclusion of nodal cells in the control allowed, after cDNA sequencing, the direct calculation of log₂ fold change (log₂FC) factors for the differentially expressed genes. Accordingly, positive log₂FC factors indicated mRNAs enriched in nodal cells. Judged by the visible sharp rRNA bands and the test hybridization for cab mRNA, the prepared RNA samples were of high quality (Figure S1). After polyA-mRNA enrichment, cDNA libraries were generated for the two different Chara braunii RNA sample types and sequenced, yielding a total of 54.042.636 raw single read counts for enriched nodal cells and 53.423.073 for total cells. From these, 10.991.291 (20.6%) reads for total and 13.763.747 (25.5%) for the enriched sample remained unmapped. After trimming and mapping steps, 4.994.360 nodal (9.4%) and 2.628.740 total (4.9%) reads were uniquely mapped to the Chara genome (Figure 2). Mapped reads matched 11.319 putative genes for the nodal sample and 15.627 genes for total sample. The 15 most highly expressed genes for each tissue sample are given in Tables 2 and 3, while the full list of genes is given in Table S2.

In the control sample, the 15 most expressed genes were mostly associated with photosynthesis and protein synthesis, such as chlorophyll a-b binding proteins (5/15 instances), the photosystem (PS) II 5 kDa protein PsbT, or the ribosomal proteins RpL6-like, RpL35a-3-like, RpL10, RpS19-3, and RpS8 (Table 2). Moreover, mRNA for g46838, a gene encoding a GC-rich sequence DNA-binding factor-like protein with similarity to the PAX3- and PAX7-binding proteins, was found enriched in both samples (Tables 2 and 3). Last, three of these highly expressed mRNAs encode unknown proteins.

In the sample enriched for nodal cells, the 15 most strongly expressed genes differed from the control by lacking all genes related to photosynthesis. However, even 9/15 mRNAs were encoding ribosomal proteins, indicating that the drive for protein synthesis was high in this tissue (Table 3). g26356, encoding a thioredoxin-encoding mRNA, and g46838, a protein with similarity to the PAX3- and PAX7-binding proteins, were also found enriched. Another four enriched mRNAs encode unknown proteins.

When looking at the full list of genes expressed, 117 different mRNAs for proteins with domains of unknown function (DUF) were identified in both samples. Among them, we found 21 instances of genes encoding DUF4360-containing cysteine-rich proteins. While DUF4360-containing proteins exist in bacteria and eukaryotes (Paysan-Lafosse et al., 2023), it is striking that these Chara braunii DUF4360-containing proteins are more similar to homologs from fungi (mainly Ascomycota) than to any other. Although potential instances of horizontal gene transfer events between algae and fungi or soil bacteria were reported (Donoghue & Paps, 2020), we cannot unambigiously exclude the possibility of fungal contamination given the non-axenic nature of the Chara cultures. However, the genes encoding DUF4360-containing proteins are part of the Chara braunii genome annotation. Therefore, if these reads resulted from a contaminating fungus, the same contamination had to be present already in the cultures that were taken for the initial genome analysis. Most of the mRNAs for DUF4360-containing proteins were more abundant in the Chara total control sample (e.g., genes g9217, g68682, g2717, g2719, g58375, g11068, g36724, g68700, g9208, g68689, g11072, and g76151; for details see Table S1).

3.2 Functional assignments to differentially expressed genes

To obtain an overview of genes differentially expressed between the two samples, we set cutoffs of log₂FC ≥ |1| and reads per kb (RPK) ≥10, leaving a set of 3145 differentially regulated genes (2762 up, 383 down in the sample enriched for nodal cells compared to control). The full list of genes can be found in Table S2.

Gene ontology (GO) assignments were used to classify the putative functions of the differently expressed Chara braunii genes (Figure 3). We sorted the GO assignments according to biological process, cellular components and molecular function, and into the up- or down-regulated categories. Most prevalent among the biological processes upregulated in the nodal sample were the GO terms “DNA integration” (225 genes), “protein phosphorylation” (121 genes), “proteolysis” (119 genes), “metabolic process” (63 genes) and “regulation of transcription, DNA-templated” (50 genes). Prominent cellular component GO terms were “integral component of membrane” (103 genes), “nucleus” (91 genes), “membrane” (78 genes), “intracellular anatomical structure” (59 genes) and “cytoplasm” (39 genes). Among the upregulated GO terms for molecular function, “protein binding” (459 genes), “nucleic acid binding” (517 genes), “ATP binding” (287 genes), “zinc ion binding” (138 genes), and “DNA binding” (138 genes) were highly enriched.

In contrast, the biological process GO terms “protein phosphorylation” (35 genes), “metabolic process” (30 genes), “proteolysis” (27 genes), “transmembrane transport” (27 genes) and “DNA integration” (24 genes) were downregulated. Among downregulated cellular component GO terms, “integral component of membrane” (57 genes), “membrane” (52 genes), “cytoplasm” (24 genes), “nucleus” (14 genes), and “intracellular anatomical structure” (12 genes) were most prevalent. Prominent GO terms for downregulated molecular functions were “protein binding” (101 genes), “ATP binding” (84 genes), “nucleic acid binding” (69 genes), “catalytic activity” (63 genes), and “protein kinase activity” (35 genes). Interestingly, 30 genes associated with the GO term “ribosome” were upregulated and 10 genes downregulated. Within this subset, two paralogous genes encoding 40S ribosomal S21-2 proteins were inversely regulated (g8281, log₂FC 2.06 and g28482, log₂FC −1.59), similarly two genes for the 60S ribosomal L22-2 protein were inversely regulated (g50176, log₂FC 1.40 and g41227, log₂FC −1.20). Besides their crucial function in translation, ribosomal proteins can have extraribosomal functions. One intriguing example of this is provided by RPL18aB, which is involved in sexual reproduction and male gametophyte development in Arabidopsis (Xiong et al., 2021; Yan et al., 2016). A putative 60S ribosomal protein L18-2 was found downregulated within the ribosomal subset (g50118, log₂FC −3.06). A full list of GO terms can be found in Table S3.

Because there were many genes with unclear annotation or similarity to retrotransposons, we focused on genes with an available functional annotation (Table 4). Among these genes, g31563 (log₂FC + 5.19) and g41182 (log₂FC + 4.38), which encode proteins with similarity to 1,4-alpha-glucan-branching enzymes/alpha/maltogenic amylase and beta amylase, were enriched in nodal cells. While it is known that streptophyte algae use starch as a storage product (García, 1994), the upregulation of both genes in the nodal cell sample indicates active carbohydrate storage molecule metabolism in the respective cells of Chara braunii.

With their publication of the Chara draft genome, Nishiyama et al. (2018) predicted the presence of certain factors involved in phytohormone biosynthesis and signaling pathways. Within our data set, we detected several upregulated genes belonging to these pathways. Related to ethylene signaling, we detected three upregulated genes in the nodal samples. The genes g30098 (log₂FC +2.1) and g17647 (log₂FC +2.4) encode putative homologs of the ETHYLENE INSENSITIVE 2 and 3 proteins (EIN2 and EIN3) (Nishiyama et al., 2018), proteins that are central in the regulation of ethylene-mediated responses (Chao et al., 1997; Solano et al., 1998). Furthermore, we detected putative serine/threonine-protein kinases EDR1, g19355 (log₂FC +2.6) and g55470 (log₂FC +2.59), homologs of two proteins involved in crosstalk between ethylene-, ABA- and salicylic acid signaling in A. thaliana (Tang et al., 2005; Wawrzynska et al., 2008). Additionally, we detected with a log₂FC of +4.1 enrichment of g23946, encoding a homolog of the trigalactosyldiacylglycerol 3 (TGD3) protein, involved in chloroplast lipid import (Lu et al., 2007), jasmonic acid signaling and plant defense responses via phosphatidic acid signaling (Tagami et al., 2021).

Regarding further upregulated, signaling-related genes, g197 (log₂FC +2.3), an mRNA encoding a putative ADP-ribosylation factor GTPase-activating protein (ARF), was detected. Another upregulated gene, g48048 (log₂FC +4.6), encoding a putative mediator of RNA polymerase II transcription subunit 19a-like, points towards an increase in transcriptional activity for the actively dividing nodal cells. Another potential regulatory factor enriched in the nodal sample (log₂FC +4.6) is a putative spastin-like protein encoded by g19216.

Within our data set, we detected several genes encoding proteins involved in photosynthetic processes expressed at lower levels in nodal cells. Among them, the genes g19126, g20155, g19141, g20151, g20157 and g19133 (log₂FC −6.8, −5.4, −5.2, −3.6, −3.3, −3.2) encode chlorophyll a/b binding proteins, apoproteins of the light-harvesting complex of photosystem II (Liu et al., 2013). Additionally, g31512 (log₂FC −2.9; encoding the PSI reaction center subunit VI), g50764 (log₂FC −3.3; encoding the putative PSII repair protein PSB27) and g602 (log₂FC −3.9; encoding a putative LHCP translocation defect protein) point at less active photosynthetic processes in central and nodal cells. Interestingly, g61495 (log₂FC −3.4), a putative peroxiredoxin involved in cellular protection from reactive oxygen species (ROS), was downregulated as well. Besides their crucial scavenging capabilities, however, peroxiredoxins also function as modulators of stress response pathways with their interaction with transcription factors (Bréhélin et al., 2003; Hopkins & Neumann, 2019).

The primary cell walls of streptophyte algae contain pectin, hemicellulose and cellulose, and resemble a basal version of those in land plants (Domozych & Bagdan, 2022). Within our data set, we detected mRNAs of several cell wall-associated genes. We found putative cellulose synthase catalytic subunits (g8592 and g8591) upregulated in the nodal sample, as well as an expansin-like protein (g49779) and an mRNA encoding a putative polygalacturonate 4-alpha-galacturonosyltransferase-like protein (g11196). Additionally, a putative cellulose synthase catalytic subunit (g31308) was downregulated, as well as an expansin like protein (g3302), and a putative polygalacturonate 4-alpha-galacturonosyltransferase-like protein (g3195). Furthermore, we detected six putative pectine esterases; Four of which were downregulated (g9040, g11990, g12130, and g17556) and one upregulated (g20314), while one putative pectinesterase/pectinesterase inhibitor 24-like protein was upregulated (g32217). Finally, we detected g29739 (log₂FC +4.9), a putative caffeoylshikimate esterase (CSE) involved in the biosynthesis of lignin as enriched in the nodal material (Table 4).

3.3 Divergent phylogenies for paralogous genes highlight evolutionary innovations

Another set of proteins with central functions in the chloroplast are the PEP-associated proteins (PAPs), which control the activity of the plastid-encoded RNA polymerase (PEP) (Pfalz & Pfannschmidt, 2013). Here, we detected a lower expression for several of these PAPs in the nodal sample compared to the control, with log₂FC of −3.63, −2.57, and −1.45 for PAP5, PAP9 and PAP10 (genes g48045, g17032, and g51450), which hence is consistent with their role in photosynthetically active cells. However, we detected for one of these proteins, the MurE-like PAP11 (gene g32224), enrichment in the nodal material (log₂FC +1.17). Phylogenetic analysis of the putative MurE-like PAP11 gene (g32224) indicated that it is most closely related to homologs from Closterium sp. NIES-68 (67/83% identical/similar amino acids in a 531-residue overlap between g32224 and protein GJP33877.1) and K. nitens (60/75% identical/similar amino acids in a 586-residue overlap with protein GAQ90408.1), which form a shallow sister group to mosses, ferns and land plant PAPs (Figure 4) and, therefore, do not contradict a divergent functionality of the PAP11 homolog in Chara braunii. Surprisingly, we detected a deep divergence among the homologs in some green algae. These homologs, found in Chloropicon primus, Trebouxia, Chlorella variabilis, Tetrabaena socialis, Chlamydomonas reinhardtii, and Ostreococcus tauri form a distinct group against homologs from all other organisms (Figure 4, 43/56% identical/similar amino acids in a 589-residue overlap between g32224 and protein QDZ19900.1 of Chloropicon primus).

The putative 1,4-alpha-glucan-branching enzyme/alpha amylase g31563 and the beta amylase g41182 highly activated in nodal cells also showed an interesting phylogenetic divergence (Figures 5 and 6). g31563 encodes a protein with high sequence conservation to homologs from other algae (51% identical and 68% similar amino acids in a 572-residues overlap with the homolog in the Klebsormidiophyceae alga K. nitens (accession no. GAQ89829) and 47% identical and 65% similar amino acids in a 585 residues-long overlap with the homolog in the Zygnematophyceae algae Closterium sp. NIES-68 (accession no. GJP44716)). The next-closely related proteins are mainly from Apicomplexa, such as Toxoplasma gondii, and other Alveolata, the Chlamydomonadales and other green algae.

The most closely related protein to the protein encoded by g41182 is, again, a homolog from K. nitens (52/69% identical/similar amino acids in a 485-residues overlap with protein GAQ79925) and homologs from Chlorella vulgaris, Chlamydomonas reinhardtii, and Closterium sp. NIES-68 (Figure 6). However, these form a shallow sister group to proteins from land plants, indicating a closer relationship to homologs in land plants (e.g., 49/61% identical/similar amino acids in a 505 residues-long overlap between the Chara braunii g41182 gene product and protein KAJ0630873 from sunflower (Helianthus annuus)).