Thiamine metabolism genes in diatoms are not regulated by thiamine despite the presence of predicted riboswitches

Abstract

Thiamine pyrophosphate (TPP), an essential co-factor for all species, is biosynthesised through a metabolically expensive pathway regulated by TPP riboswitches in bacteria, fungi, plants and green algae. Diatoms are microalgae responsible for c. 20% of global primary production. They have been predicted to contain TPP aptamers in the 3'UTR of some thiamine metabolism-related genes, but little information is known about their function and regulation. We used bioinformatics, antimetabolite growth assays, RT-qPCR, targeted mutagenesis and reporter constructs to test whether the predicted TPP riboswitches respond to thiamine supplementation in diatoms. Gene editing was used to investigate the functions of the genes with associated TPP riboswitches in Phaeodactylum tricornutum. We found that thiamine-related genes with putative TPP aptamers are not responsive to supplementation with thiamine or its precursor 4-amino-5-hydroxymethyl-2-methylpyrimidine (HMP), and targeted mutation of the TPP aptamer in the THIC gene encoding HMP-P synthase does not deregulate thiamine biosynthesis in P. tricornutum. Through genome editing we established that PtTHIC is essential for thiamine biosynthesis and another gene, PtSSSP, is necessary for thiamine uptake. Our results highlight the importance of experimentally testing bioinformatic aptamer predictions and provide new insights into the thiamine metabolism shaping the structure of marine microbial communities with global biogeochemical importance.

Keywords: Phaeodactylum tricornutum; CRISPR/Cas9; TPP riboswitch; aptamer prediction; diatoms; thiamine biosynthesis; thiamine uptake.

Fig. 1

Multiple sequence alignment of 16 predicted diatom thiamine pyrophosphate (TPP) aptamers and structural comparison with previously characterised eukaryotic riboswitches. (a) Multiple sequence alignment of previously identified (first eight) and a sample of newly identified TPP aptamers in diatoms. Stems are indicated with arrows and are colour coded, asterisks indicate conserved residues across all sequences presented. See Supporting Information Table S1(b,c) for the full sequences of all predicted TPP aptamers. (b) Structural comparison of the predicted Phaeodactylum tricornutum HMP‐P synthase (THIC) aptamer (centre) with experimentally described TPP aptamers in Chlamydomonas reinhardtii (left; Croft et al., 2007) and Neurospora crassa (right; Cheah et al., 2007). The pyrimidine‐binding residues (‘CUGAGA’ motif, red stars) and the pyrophosphate‐binding residues (‘GCG’ motif, blue stars) are highlighted. Green algae and plant aptamers contain an alternative 3′ splicing site used in their mechanisms of action in their P2 stem (AG, boxed). The ‘AACAAA’ sequence overlapping with the PtTHIC aptamer P1 stem (boxed) was predicted by Paspa software to be the most likely polyadenylation site (Ji et al., 2015). Cc, Cyclotella cryptica; Fc, Fragilariopsis cylindrus; Fs, Fistulifera solaris; Hal, Halamphora sp. MG8b; La, Licmophora abbreviata; Pm, Pseudonitzschia multiseries; Pmu, Pseudonitzschia multistriata; Pt, Phaeodactylum tricornutum; To, Thalassiosira oceanica; Tp, Thalassiosira pseudonana.

Fig. 2

Proposed routes for thiamine biosynthesis in diatoms. (a) A tBlastn search using selected algal peptide sequences as queries (please refer to Supporting Information Table S3c) was performed against 19 diatom genomes to determine the presence (full circle P‐value > 10⁻²⁰; half‐full circle P‐value > 10⁻³) or absence (empty circle) of different thiamine‐related genes. The presence of an associated predicted riboswitch in the 3′UTR of the gene is indicated with a hairpin symbol at the right of the circle. The genome abbreviations, accession numbers and references can be found in Table S2. (b) Potential thiamine biosynthetic, salvage and uptake routes in diatoms. The pathway steps with strong support across the diatom lineage are shown in green. AIR, 5‐aminoimidazole ribotide; dA, 5′‐deoxyadenosine; DXP, 1‐deoxy‐d‐xylulose 5‐phosphate; FAMP, N‐formyl‐4‐amino‐5‐aminomethyl‐2‐methylpyrimidine; GA3P, glyceraldehyde 3‐phosphate; HET‐P, hydroxyethyl‐thiazole phosphate; HMP‐P, hydroxymethyl‐pyrimidine phosphate; HMP‐PP, hydroxymethyl‐pyrimidine pyrophosphate; l‐Met, l‐methionine; NAD, nicotinamide adenine dinucleotide; PLP, pyridoxal 5′‐phosphate; SAM, S‐adenosyl methionine; TMP, thiamine monophosphate; TPP, thiamine pyrophosphate. ^& THI5/NMT1 candidates contain an NMT1 Pfam domain (PF09084). ^$ THIG and THIS are encoded in the chloroplast in Phaeodactylum tricornutum, so the results can be biased in genomes that do not include chloroplast sequences. ⁺THID, THIE and HMPK functions are performed by a single peptide in diatoms (TH1). *In some bacteria TenI accelerates a thiazole tautomerisation reaction, but it is not necessary to synthesise HET‐P (Hazra et al., 2011).

Fig. 3

Impact of vitamin supplementation on expression of THIC and SSSP in Phaeodactylum tricornutum and Thalassiosira pseudonana. P. tricornutum and T. pseudonana were grown in the absence (blue) or presence (red) of 0.6 μM cobalamin (B₁₂), 10 μM thiamine (B₁) or 10 μM 4‐amino‐5‐hydroxymethyl‐2‐methylpyrimidine (HMP) for 7 d. Three or four biological replicates were analysed by RT‐qPCR in technical duplicate. The technical replicate measurements were averaged for each biological replicate, and transcript levels were normalised for the average transcript levels of three housekeeping genes (H4, UBC, UBQ for P. tricornutum; Actin, EF1a, rbcs for T. pseudonana). Each dot represents the relative expression value for an individual biological replicate and a box plot summarises the data for each gene and treatment. Two‐sided t‐tests between supplemented and control conditions were conducted for all genes. *, P‐value < 0.05.

Fig. 4

Effect of the thiamine antimetabolite pyrithiamine on the growth of Chlamydomonas reinhardtii, Phaeodactylum tricornutum and Thalassiosira pseudonana. C. reinhardtii, P. tricornutum and T. pseudonana were grown for 9 d in the absence (left column) or presence (right column) of 10 μM pyrithiamine and the absence (blue) or presence (red) of 10 μM thiamine in 96‐well plates. Growth was measured as optical density (OD)₇₃₀ every 24 h. Error bars represent the standard deviation of three biological replicates.

Fig. 5

Effect of thiamine supplementation on transformants with PtTHIC promoter and 3′UTR driving expression of the Ble zeocin resistance gene. Transformants carrying a Ble‐Venus reporter controlled by the PtEF2 promoter and PtFCPC 3′UTR (pMLP2047) or the PtTHIC promoter and 3′UTR (pMLP2048) were grown in the absence (left column) or presence (right column) of 10 μM and thiamine the absence (blue) or presence (red) of 75 mg l⁻¹ zeocin. Error bars represent the standard deviation of three biological replicates for each of 10 independent transformants for pMLP2047 and pMLP2048 and three biological replicates for wild‐type (WT).

Fig. 6

Intracellular thiamine and thiamine pyrophosphate (TPP) abundance and PtTHIC protein levels determined in transformants carrying a mutated PtTHIC aptamer. (a) A construct coding for an extra copy of PtTHIC with a targeted mutation in its putative aptamer (pMLP2065) and its respective unmutated control (pMLP2064) were transformed into Phaeodactylum tricornutum. (b) Transformants were grown for 5 d before thiamine and TPP were quantified by high performance liquid chromatography (HPLC) and normalised by fresh weight. Each dot represents the measurement of an independent transformant and a box plot summarises the data. Different letters represent significant differences in average vitamin content between strains in a Tukey honestly significant difference (HSD) test with a 0.95 confidence level. (c) An independent transformant for each construct was grown to c. 5 × 10⁶ cells ml⁻¹ in the presence or absence of 10 μM thiamine, and protein was extracted from 150 ml cultures. A western blot analysis with a primary anti‐HA antibody on total crude extracts normalised to culture optical density (OD) is shown.

Fig. 7

Response of Chlamydomonas reinhardtii carrying the PtTHIC‐CrTHIC chimeric riboswitches to thiamine supplementation. (a) The CrTHI4 riboswitch platform previously developed in our laboratory (Mehrshahi et al., 2020) was cloned in the 5′UTR of a Ble‐eGFP zeocin resistance reporter with a CrPSAD promoter and a CrCA1 terminator. A set of modified aptamers combining five structural parts (P1/2, P3, P3a, L2/4 and P4/5; colour coded) from CrTHIC and PtTHIC aptamers was cloned into the platform and the constructs transformed into C. reinhardtii. (b) Five independent transformants for each modified aptamer design were grown in the presence of 10 mg l⁻¹ zeocin with or without 10 μM thiamine for 4 d. The ratio between the optical density (OD)₇₃₀ in the depleted and the thiamine‐supplemented conditions is shown.

Fig. 8

Determination of genotype and phenotype of Phaeodactylum tricornutum SSSP and THIC CRISPR/Cas9 mutants. (a, b) Schematic representation of the CRISPR‐mediated homology recombination strategy to inactivate SSSP and THIC, respectively. (c, d) Transformants were genotyped with two or three primer pairs colour coded in panel (a). The negative control did not include any template DNA. (e) Wild‐type (WT) and two SSSP knock‐out strains were grown in the absence or presence of 10 μM thiamine for 5 d in biological duplicate, and intracellular thiamine levels were measured in technical duplicate. Different letters represent significant differences in average intracellular thiamine content between strains and conditions in a Tukey honestly significant difference (HSD) test with a 0.95 confidence level. (f) WT and two THIC knock‐out strains were grown in the absence (blue) or presence (red) of 1 μM thiamine in 24‐well plates recording growth as OD₇₃₀ every 24 h. Error bars represent the standard deviation of three biological replicates.

Fig. 9

Phenotype analysis of a PtTHIC knock‐out mutant. (a) The ΔTHIC#1 mutant was initially grown in the absence of supplementation, at day 6 (arrow) 1 μM thiamine (red) or 1 μM 4‐amino‐5‐hydroxymethyl‐2‐methylpyrimidine (HMP; green) was supplemented and growth compared with an unsupplemented control (blue). Error bars represent the standard deviation of three biological replicates. (b) Wild‐type (WT) and the ΔTHIC#1 mutant were grown in increasing concentrations of thiamine (0–500 nM) to determine the thiamine concentration required to support growth of the mutant. Error bars represent the standard deviation of six biological replicates.