Abstract

Background

   The marine diatom Phaeodactylum tricornutum plays a crucial role in
   global carbon and nitrogen cycling. Previous work revealed that lactate
   regulates carbon and nitrogen metabolism of P. tricornutum through
   protein lactylation, significantly affecting growth characteristics,
   photosynthetic efficiency, biochemical composition, and expression of
   genes related to carbon and nitrogen metabolism. However, the
   functional roles of lactate metabolism genes and their regulatory
   mechanisms in carbon-nitrogen homeostasis remain unexplored. This study
   aimed to characterize key lactate metabolism and regulatory genes
   (ldhA, Glo1, Glo2, D-LCR, GlxI, and GPCR) and elucidate their influence
   on carbon and nitrogen metabolic modulation in P. tricornutum.

Results

   Overexpression (OE) and RNA silence (AS) of ldhA, Glo1, Glo2, D-LCR,
   GlxI, and GPCR revealed their roles in lactate metabolism and
   regulation. The overexpression of Glo2 and ldhA enhanced exogenous
   lactate utilization and total lipid accumulation under low nitrogen
   (LN) conditions. Additionally, the overexpression of Glo1 and D-LCR
   facilitated the utilization of exogenous lactate to cope with LN
   conditions. In contrast, the growth and L-lactate consumption rates of
   GlxI and GPCR overexpression strains were significantly lower than or
   not significantly different from those of the WT strain. The key enzyme
   involved in lactate metabolism, LDHA, was selected for further
   functional analysis. Western blot analysis suggested that ldhA
   overexpression promoted the lactylation of an approximately 40 kDa
   lactylated protein in P. tricornutum. ^13C-labeling analysis
   demonstrated that ldhA overexpression in P. tricornutum accelerated
   lactate utilization and the processes of glycolysis, TCA cycle, CCM,
   and Calvin cycle. RNA-Seq analysis revealed that ldhA overexpression
   promoted cell division metabolism and lipid metabolism in P.
   tricornutum under LN conditions and glycerophospholipid metabolism
   under exogenous lactate addition conditions.

Conclusion

   Lactate metabolism and lactylation metabolic processes mediated by
   LDHA, GLO1, GLO2 and D-LCR are important mechanisms by which lactate
   affects the growth of P. tricornutum, rapidly regulates carbon and
   nitrogen metabolism processes, and promotes the accumulation of lipids
   under LN conditions.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12934-025-02779-w.

   Keywords: Phaeodactylum tricornutum, Carbon and nitrogen metabolism,
   Lactate, Lactate metabolism, Lactylation modification

Introduction

   Diatoms, accounting for approximately 20% of global primary
   productivity [[44]1], are key oceanic primary producers [[45]2–[46]4]
   that uniquely regulate carbon and nitrogen balance, enabling them to
   swiftly respond to oceanic carbon and nitrogen fluctuations [[47]5].
   The pennate diatom Phaeodactylum tricornutum, with its sequenced and
   functionally annotated genome [[48]6], serves as a model organism for
   diatom physiological and biochemical studies [[49]7]. Moreover, its
   richness in essential unsaturated fatty acids makes it valuable in
   aquaculture as bait microalgae. Hence, research on carbon and nitrogen
   metabolism regulation and high value-added product synthesis in P.
   tricornutum is crucial for basic research and practical applications.
   Previous studies have shown that under high C: N ratios, P. tricornutum
   increases the recycling and redistribution of nitrogen, stores excess
   carbon skeletons as lipids, and promotes the accumulation of lipids
   [[50]8–[51]10]. Our previous work demonstrated that endogenous lactate
   in P. tricornutum significantly increased under high C: N conditions,
   along with the differential expression of carbon and nitrogen
   metabolism-related proteins [[52]11]. Lactylated proteins were
   significantly enriched in carbon and nitrogen metabolism and energy
   metabolism processes, indicating that endogenous lactate regulates
   these processes through lactylated proteins [[53]11]. Additionally,
   exogenous lactate profoundly impacts the growth, metabolism, and
   synthesis of essential biochemical components in P. tricornutum
   [[54]12]. In conclusion, lactate plays a pivotal role in the carbon and
   nitrogen metabolism of P. tricornutum.

   The function of lactate in cellular processes regulation is
   concentration-dependent. Intracellular lactate production and protein
   chemical modification occur via two main pathways. One pathway involves
   pyruvate conversion to L-lactate catalyzed by L-lactate dehydrogenase,
   followed by lactoyl-coenzyme A (CoA) generation and enzymatic acyl
   transfer to protein lysine residues [[55]13]. The other pathway starts
   with methylglyoxal (MGO), a glycolytic metabolite, forming S,
   D-lactoylglutathione (LGSH) via glyoxalase 1 (GLO1), which then
   undergoes nonenzymatic acyl transfer to generate lactylation.
   Glyoxalase 2 (GLO2) subsequently hydrolyzes LGSH to recover GSH and
   produce D-lactate [[56]14]. Genome-wide annotation of P. tricornutum
   has identified multiple D-lactate dehydrogenases (including LDHA and
   D-LCR), membrane GPCRs acting as lactate-specific receptors, and GLO1
   and GLXI functioning as lactoylglutathione lyase, and
   hydroxylacylglutathione hydrolase GLO2 [[57]6].

   While previous studies in diatoms highlighted the pivotal role of
   lactate in carbon and nitrogen metabolism and its association with
   lactylated proteins, the functional roles of lactate metabolism and
   regulatory genes and their regulatory mechanisms in carbon-nitrogen
   balance remained unexplored. To address these gaps, this study focused
   on key lactate metabolism and regulatory genes, including ldhA, Glo1,
   Glo2, D-LCR, GlxI, and GPCR. We constructed overexpression (OE) and RNA
   silence (AS) strains and examined their physiological and biochemical
   characteristics with wild type (WT) under different conditions. These
   results revealed that LDHA is a key enzyme involved in carbon and
   nitrogen metabolism in P. tricornutum, influenced by lactate. Further
   analysis of LDHA function was conducted through Western blot, ^13C
   labeling analysis, and RNA-Seq. This study aims to clarify the specific
   roles of these genes in lactate metabolism and their impact on carbon
   and nitrogen regulation in P. tricornutum, thereby providing a
   theoretical foundation for understanding lactate metabolism in
   photosynthetic microorganisms.

Materials and methods

Algal culture and experimental treatments

   P. tricornutum Bohlin was obtained from the Microalgae Culture Center
   (MACC) at Ocean University of China and conserved in seawater
   supplemented with f/2 medium [[58]15] under cool white fluorescent
   light (~ 100 µmol·m^− 2·s^− 1) at 20 °C with a 12:12 dark: light cycle.

   The transformed strains and the wild type (WT) strain in the
   logarithmic growth phase were inoculated into sterilized artificial
   seawater supplemented with f/2 (NC), f/2 medium with 1/10 NaNO[3] (LN),
   f/2 medium containing 5 mM lactate (NC + L) with the pH adjusted to
   match that of artificial seawater, and f/2 medium with 1/10 NaNO[3] and
   5 mM lactate (LN + L), with the pH adjusted to match that of artificial
   seawater, each in triplicate. Pre experiments revealed that P.
   tricornutum was able to utilize L-lactate better than D-lactate, so
   L-lactate was used in this laboratory.

Cloning, vector construction, and transformation

   The expression primers for the lactate dehydrogenase gene (ldhA,
   [59]XM_002176401.1), D-lactate dehydrogenase gene (D-LCR,
   [60]XM_002179948.1), glyoxalase I gene (Glo1, [61]XM_002185099.1, GlxI,
   [62]XM_002180396.1), glyoxalase II gene (Glo2, [63]XM_002185180.1), G
   protein-coupled receptor gene (GPCR, [64]XM_002176403.1), and pPha-T1
   expression vector ([65]AF219942.1) were designed for overexpression and
   RNA silence (Table S1), respectively, according to the whole-genome
   sequencing results of P. tricornutum CCAP 1055/1 from NCBI. For gene
   overexpression and knockdown, the PCR amplicons were subsequently
   cloned and inserted into the pPha-T1 expression vector via double
   enzyme digestion.

   The recombinant expression vectors were electroporated into P.
   tricornutum via electroporation systems (BTX, 0.5 kV voltage, 25 µF
   capacitance, and 400 Ω shunt resistance) as reported previously
   [[66]16]. The transformants were screened in f/2 solid medium
   supplemented with 75 mg·L^− 1 Zeocin.

Verification of transformants via molecular approaches

   The transformed strains were verified at the DNA and mRNA levels to
   determine whether the overexpression (pPha-T1-OE, POE) and RNA silence
   (pPha-T1-AS, PAS) recombinant plasmids were successfully transformed
   into the algae. The coding sequence of the transformants was amplified
   with pPha-T1-F and BleoR-R (Table S1) to detect the existence of the
   BleoR gene and target gene in the recombinant plasmids. TATA box
   binding protein (TBP) was selected as the internal reference gene to
   normalize the transcription of the target genes in each sample
   [[67]17]. RT-qPCR was conducted via the use of ChamQ Universal SYBR
   qPCR Master Mix (Vazyme, Cat# Q711) and RT-qPCR primers for the target
   genes and TBP (Table S2). At least three biological replicates were
   analyzed for the RT-qPCR assay.

Measurement of physiological indices and biochemical components

   The logarithmic phase of the algal strains was inoculated into sterile
   f/2 medium. The cell density was determined by means of hemocytometry
   and used to plot growth curves throughout the growth cycle.

   The L-lactate consumption rates of the transformed strains were
   analyzed by the K-LATE Lactate Content Assay Kit (Megazyme, Cat#
   K-LATE) to detect the amount of L-lactate in the medium of the
   lactate-supplemented groups (NC + L and LN + L) within 8 days of
   culture. Lactate concentration in the medium was corrected by
   subtracting the degradation values of the cell-free control (only
   medium and lactate).

   The photosynthetic efficiency of each triplicate culture was measured
   using Act-Light (set at 100 µmol·m^− 2·s^− 1) with a PAM-CONTROL
   fluorometer (Walz, Germany) [[68]12]. Pigments, including chlorophyll a
   and carotenoids, were extracted as described previously with minor
   modifications [[69]18]. After centrifugation and collection, the cells
   were flash-frozen in liquid nitrogen and freeze-dried, and total lipids
   were extracted as described previously with minor modifications
   [[70]19]. The total protein content was determined as described
   previously [[71]20]. Total soluble and insoluble sugars were determined
   as previously described with minor modifications [[72]21, [73]22]. Full
   details are given in Supplementary Method S1.

LDHA expression levels and Pan anti-lactyllysine levels were determined by WB

   The protein expression of the ldhA transformed strains was detected by
   Western blot (WB) analysis via the customized LDHA-specific rabbit
   polyclonal antibody (GenScript, Cat# Anti-B7S4E4) and HRP-conjugated
   Affinipure Goat Anti-Rabbit IgG (H + L) (Proteintech, Cat# SA00001-2).
   The OE strains ldhA-POE1 and ldhA-POE2 and the antisense transformed
   strains ldhA-PAS5 and ldhA-PAS6 were inoculated into f/2 medium, and
   cultured for 7 days under NC, NC + L, LN, and LN + L conditions. The WT
   and ldhA transformed strains were collected by centrifugation at low
   temperature. The protein was extracted using a Plant Protein Extraction
   Kit (Keygen, Cat# KGP750) and subsequently concentrated using the
   Protein Quantification Kit (Nanjing Jiancheng, Cat# A045-4).

   Lactylated WB detection was conducted on the WT and ldhA OE strains
   (ldhA-POE1 and ldhA-POE2) using pan anti-lactyllysine antibody (PTM
   Bio, Cat# PTM-1401RM, Lot: L011121) and Goat anti-Rabbit IgG (H + L)
   Secondary Antibody, HRP (Thermo, Cat# 31460). WB analysis included
   three biological replicates. The full details are given in
   Supplementary Method S1.

^13C-labeling treatment and GC-MS analysis

   The cells of the WT and ldhA OE transformed strains in the logarithmic
   growth phase were collected by centrifugation, washed with f/2 medium
   devoid of any carbon source, and inoculated into fresh f/2 medium
   containing either 5 mM sodium lactate (containing 30%
   ^13C-lactate + 70% ^12C-lactate) and 0.174 g·L^− 1 sodium bicarbonate
   or 0.174 g·L^− 1 13C-sodium bicarbonate and 5 mM ^12C-lactate medium,
   with the same initial cell number (18.72 × 10^6 cells·mL^− 1) and three
   biological replicates per treatment. The cells were incubated for
   5 min, 6 h, 4 D, and 7 D, respectively, with the addition of a quencher
   (20% methanol solution diluted with artificial seawater) of one-half
   the volume of algal mixture, and the samples were collected via
   centrifugation.

   The aforementioned samples were subjected to sequential extraction with
   100 µL of a 100% methanol solution, a 50% methanol solution, and pure
   water for the purpose of isolating free amino acids. The samples were
   then vortexed and shaken to ensure thorough mixing and subjected to
   centrifugation at 7000×g for one minute, and the resulting supernatants
   from the three centrifugations were combined and dried at 40 °C for a
   period of one night. A solution of 100 µL of pyridine and 50 µL of
   MTBSTFA was added to the sample, which was then mixed thoroughly. The
   derivatization process was conducted in a water bath at 85 °C for one
   hour. The sample was subsequently centrifuged at 13,000×g for five
   minutes, after which the supernatant was transferred to an injection
   vial for GC-MS analysis [[74]23, [75]24].

   The peaks corresponding to each amino acid were identified according to
   their corresponding mass-to-charge ratios using the GC-MS
   chromatographic results. The background values were deducted, and the
   values of the charge-to-mass ratios and ionic intensities in the mass
   spectra were derived. The values of the mass-to-charge ratios and ionic
   intensities of each fragment corresponding to each amino acid were
   labeled.

   Given that stable isotopes of each element exist stably and in specific
   ratios in nature, the natural abundance of each fragment was corrected
   via a program written in MATLAB [[76]23], and the mass isotope
   distribution vectors and labeling degrees of each fragment were
   calculated by the following equations [[77]25]. Labeling degrees of
   metabolites were normalized in terms of biomass.
   graphic file with name d33e665.gif
   graphic file with name d33e670.gif

RNA-Seq analysis

   The ldhA-POE2 and WT cells were collected at the logarithmic growth
   stage after centrifugation and inoculated in f/2 medium as a control
   (NC) and in f/2 medium supplemented with 1/10 NaNO[3] as a low nitrogen
   (LN) group. Both of these groups were designated the experimental group
   of NC-LN. The ldhA-POE2 and WT cells, which were collected in the
   logarithmic growth phase after centrifugation, were inoculated in f/2
   medium as the control group (C) and in f/2 medium supplemented with 5
   mM lactate as the lactate supplemented group (L), respectively. The two
   groups of media were adjusted such that the pH was consistent with that
   of artificial seawater, and antibiotics (50 m g·L^− 1 ampicillin,
   kanamycin, and streptomycin) were added. This resulted in the formation
   of the experimental group of C-L. Following a 7-day incubation period,
   the cells were harvested by centrifugation, frozen in liquid nitrogen,
   and stored at -80 °C for subsequent RNA-Seq analysis, with three
   biological replicates per treatment.

   Total RNA was extracted via TRIzol^® Reagent (Invitrogen, Cat#
   15596026) and qualitatively characterized as a high-quality RNA sample
   to construct eukaryotic RNA-seq transcriptome libraries, which were
   prepared via Illumina^® Stranded mRNA Prep, Ligation from Illumina (San
   Diego, CA).

   The clean reads in the samples were obtained by trimming and quality
   control of the raw paired end reads by fastp [[78]26] with default
   parameters and were separately aligned to the reference genome (P.
   tricornutum strain CCAP 1055/1 (reference genome version:
   GCF_000150955.2)) in orientation mode via HISAT2 [[79]27] software. The
   mapped reads of each sample were assembled via StringTie [[80]28] in a
   reference-based approach.

   Gene expression levels are calculated by the number of clean reads
   (read counts) located in genomic regions and were quantitatively
   analyzed via RSEM [[81]29] to identify DEGs (differentially expressed
   genes) between two different samples. DEGs with|log[2]FC| ≥ 1 and
   FDR ≤ 0.05 (DESeq2 [[82]30]) or FDR ≤ 0.001 (DEGseq [[83]31]) were
   considered significantly differentially expressed genes. Furthermore,
   GO and KEGG functional enrichment analyses were performed via Goatools
   and KOBAS, respectively, to identify DEGs that were significantly
   enriched in GO terms and metabolic pathways with a Bonferroni-corrected
   p ≤ 0.05 compared with the whole transcriptome background.

Statistical analysis

   Statistical analysis was performed via two-tailed Student’s t-tests,
   one-way ANOVA and two-way ANOVA via GraphPad Prism. Each treatment was
   performed in triplicate. The results are expressed as the means ± SD
   and are expressed as error bars. Significant differences (p < 0.05,
   p < 0.01, p < 0.0005 and p < 0.0001) are marked with “*”, “**”, “***”
   and “****”, respectively, “ns” indicates no significant difference.

Results

Successful construction of transformed strains

   The target genes expression cassettes were successfully inserted into
   the pPha-T1 vector polyclonal sites, and POE and PAS recombinant
   plasmids for the target genes were constructed (Figs. [84]S1 and
   [85]S2). The recombinant plasmids were successfully introduced into P.
   tricornutum cells by electroporation, and individual colonies survived
   on resistant plates containing Zeocin, whereas the WT strain could not
   (Fig. S3). Molecular biological validation at the DNA and mRNA levels
   showed that the transformed strain successfully integrated the target
   gene expression cassette into the genome of P. tricornutum at the DNA
   level (Fig. [86]S4) and successfully overexpressed or interfered with
   the target gene at the mRNA level (Fig. [87]S5).

Physiological and biochemical characteristics of the transformed strains

   The overexpression of ldhA, Glo1, Glo2 and D-LCR accelerated L-lactate
   consumption rates in P. tricornutum under LN conditions (Fig. [88]1).
   ldhA overexpression was beneficial for P. tricornutum to cope with LN
   conditions, utilize exogenous lactate, and promote the accumulation of
   lipids (Fig. [89]2). The overexpression and RNA silence of D-LCR
   inhibited and promoted the growth rate, photosynthetic efficiency and
   accumulation of photosynthetic pigments in P. tricornutum, respectively
   (Fig. [90]S6). D-LCR overexpression facilitated the utilization of
   exogenous lactate to cope with LN conditions and promoted lipid
   accumulation in P. tricornutum (Fig. [91]S6). Although Glo1 and GlxI in
   P. tricornutum belong to the same family of glyoxalase I, their
   overexpression promoted and inhibited the lactate utilization capacity
   of P. tricornutum under LN conditions, respectively. The overexpression
   of Glo1 accelerated the growth and lipid accumulation of P. tricornutum
   under LN conditions (Fig. [92]S7), whereas the overexpression of GlxI
   inhibited the growth and lipid accumulation of P. tricornutum (Fig.
   [93]S9). The overexpression of Glo2 favored the use of lactate in P.
   tricornutum under LN conditions, which accelerated growth,
   photosynthetic efficiency, photosynthetic pigments, and the
   accumulation of sugars (Fig. [94]S8). Nevertheless, the growth and
   L-lactate consumption rates of both the GlxI and GPCR OE strains were
   significantly lower than or not significantly different from those of
   the WT strain (Figs. [95]S9 and [96]S10). In summary, under LN
   conditions, P. tricornutum affects growth, rapidly regulates carbon and
   nitrogen metabolic processes and promotes lipid accumulation through
   lactate metabolism and lactylation metabolic pathways mediated by LDHA,
   GLO1, GLO2 and D-LCR, which in turn affects growth, rapidly regulates
   carbon and nitrogen metabolic processes and promotes lipid accumulation
   in P. tricornutum.

Fig. 1.

   [97]Fig. 1
   [98]Open in a new tab

   L-lactate content in the supernatants of the culture media of the ldhA
   (a), Glo1 (b), Glo2 (c), D-LCR (d), GlxI (e), and GPCR (f) transformed
   strains. Data were analyzed by two-way ANOVA; *p < 0.05, **p < 0.01,
   ***p < 0.001, ****p < 0.0001. Full exact p-values are provided in
   Supplementary Table S5

Fig. 2.

   [99]Fig. 2
   [100]Open in a new tab

   Growth curves of ldhA transformed strains during the whole growth
   period under NC (a), LN (b), NC+L (c), and LN+L (d) conditions;
   F[v]/F[m] (e), YII (f), chlorophyll a (g), carotenoid (h), total
   soluble sugar content (i), total insoluble sugar content (j), total
   lipid content (k), and total protein content (l) of ldhA OE strains.
   Data in a-d were analyzed by two-way ANOVA; data in e-h were analyzed
   by one-way ANOVA; data in i-l were analyzed by two-tailed Student’s
   t-tests; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Full
   exact p-values are provided in Supplementary Table S5

Identification of transformed strains at the protein level

   The LDHA protein levels in the ldhA transformed strains were detected
   by WB under NC, NC + L, LN, and LN + L conditions (Fig. [101]3). The
   target bands were detected in the LN, NC + L and LN + L groups, but not
   in the NC group, suggesting that under normal autotrophic conditions
   (no nitrogen deficiency and no addition of lactate), the expression of
   LDHA in the transformed strains was very low or absent, whereas the
   expression of LDHA was up-regulated under conditions induced by low
   nitrogen or the addition of lactate. Consistent with previous proteomic
   findings of higher LDHA levels under LN than NC conditions [[102]11].
   The ldhA OE strains generally exhibited higher LDHA protein expression
   than the WT and AS strains. In particular, the OE strain ldhA-POE2
   presented the highest LDHA protein expression under the NC + L and
   LN + L conditions. However, there was no significant decrease in LDHA
   expression in the AS strain compared with the WT strain. The above
   results indicated that the recombinant overexpression plasmid
   successfully integrated into the genome of P. tricornutum and
   successfully expressed LDHA protein, whereas the protein expression of
   the AS strain did not significantly differ from that of the WT strain,
   although LDHA expression was significantly downregulated at the mRNA
   level.

Fig. 3.

   [103]Fig. 3
   [104]Open in a new tab

   Verification of the ldhA transformed strains via WB

Pan-lactyllysine WB analyses

   Pan-lactyllysine WB analysis revealed (Fig. [105]4) that lactylated
   proteins were present in P. tricornutum. Compared with the WT strain,
   the ldhA-POE1 and ldhA-POE2 strains presented a new specific band at
   approximately 40 kDa, suggesting that ldhA overexpression promotes
   lactylation of this protein in P. tricornutum.

Fig. 4.

   [106]Fig. 4
   [107]Open in a new tab

   Pan-lactyllysine WB analyses

^13C labeling analysis of free amino acids

   To investigate the effect of ldhA overexpression on the carbon source
   utilization of P. tricornutum, the WT and ldhA OE strains were
   mixed-cultured with ^13C-sodium lactate or ^13C-NaHCO[3] as substrates.
   The degree of ^13C labeling of free amino acids (FAAs) was determined
   via GC-MS, and the degree of FAA labeling in the cells of the WT and
   ldhA OE strains was detailed in Table [108]1.

Table 1.

   Flux of FAA in ^13C-sodium lactate and ^13C-NaHCO[3] labeling
   experiments
                ^13C-sodium lactate      ^13C-NaHCO[3]
   FAA strain   5 min 6 h   4 d   7 d    5 min 6 h   4 d   7 d
   Ser WT       0.014 0.092 0.047 0.035  0.036 0.057 0.028 0.094
       ldhA-POE 0.077 0.092 0.086 0.057  0.043 0.079 0.046 0.067
   Gly WT       0.037 0.077 0.046 0.045  0.056 0.080 0.046 0.055
       ldhA-POE 0.056 0.100 0.071 0.050  0.051 0.091 0.036 0.043
   Phe WT       0.102 0.207 0.145 0.139  0.153 0.208 0.122 0.159
       ldhA-POE 0.141 0.247 0.156 0.157  0.112 0.224 0.093 0.143
   Ala WT       0.069 0.248 0.073 0.053  0.260 0.057 0.029 0.019
       ldhA-POE 0.322 0.229 0.073 0.050  0.058 0.069 0.027 0.017
   Val WT       0.014 0.188 0.064 0.077  0.106 0.025 0.023 0.016
       ldhA-POE 0.098 0.168 0.081 0.080  0.016 0.027 0.020 0.022
   Asx WT       0.018 0.135 0.077 0.070  0.047 0.031 0.026 0.042
       ldhA-POE 0.054 0.136 0.079 0.059  0.022 0.036 0.022 0.024
   Thr WT       0.017 0.112 0.073 0.073  0.029 0.072 0.045 0.048
       ldhA-POE 0.045 0.140 0.105 0.111  0.039 0.073 0.077 0.033
   Glx WT       0.036 0.199 0.115 0.093  0.100 0.059 0.062 0.172
       ldhA-POE 0.163 0.194 0.117 0.092  0.059 0.059 0.047 0.077
   Pro WT       0.026 0.091 0.096 0.076  0.015 0.044 0.023 0.018
       ldhA-POE 0.041 0.078 0.101 0.069  0.030 0.055 0.024 0.018
   His WT       0.288 0.426 0.290 0.405  0.406 0.452 0.440 0.310
       ldhA-POE 0.442 0.456 0.313 0.295  0.405 0.449 0.312 0.338
   [109]Open in a new tab

   Under ^13C-sodium lactate mixed-culture conditions, the ldhA OE strain
   had greater labeling of various FAAs than the WT strain did at 5 min of
   incubation, especially serine (Ser), alanine (Ala), valine (Val),
   glutamate and glutamine (Glx), and histidine (His), suggesting that
   ldhA-POE utilizes lactate more efficiently than the WT strain does,
   which was consistent with the results shown in Fig. [110]1. The
   metabolic pathway diagram of each amino acid involved in ^13C-sodium
   lactate mixture (Fig. [111]5) indicated that the ldhA OE strain
   utilized lactate more rapidly than the WT strain did and accelerated
   glycolysis, TCA cycle, CCM, and Calvin cycle.

Fig. 5.

   [112]Fig. 5
   [113]Open in a new tab

   Metabolic pathways involving free amino acids in WT and ldhA OE
   transformed strains under ^13C-sodium lactate (a) and ^13C-NaHCO[3] (b)
   mixed culture. The first and second rows of the dataset represent the
   labeling flux of this amino acid at 5 min, 6 h, 4 d and 7 d under
   ^13C-sodium lactate mixed culture of WT and ldhA-POE, respectively

   However, under mixed-culture conditions with ^13C-NaHCO[3], the degree
   of labeling of various FAAs, especially phenylalanine (Phe), Ala, Val,
   and Glx, in the ldhA OE strain was lower than that in the WT strain at
   5 min of incubation. In combination with ^13C-NaHCO[3], the metabolic
   pathway diagram involving various amino acids (Fig. [114]5) shows that
   the WT can utilize NaHCO[3] more quickly than can ldhA-POE.

RNA-Seq analysis

   According to the consistency analysis by PCA, there were obvious
   differences between the four samples in the NC-LN and C-L experimental
   groups, which could be completely distinguished. The biological
   replicates of the samples in each group were clustered into a pile,
   indicating good sample consistency and high-quality RNA-Seq data (Figs.
   [115]S11 and [116]S12). On the basis of the gene expression data,
   pairwise comparisons were made between the four groups of samples in
   the NC-LN and C-L experimental groups, and the DEGs whose expression
   significantly changed (p < 0.05) were obtained (Figs. [117]S13 and
   [118]S14).

   In the A_OE_LN_vs_WT_LN group, GO enrichment analysis results
   (Fig. [119]8a) revealed that the DEGs were significantly enriched in
   the processes of cell division (chromosome organization involved in the
   meiotic cell cycle, mitotic chromosome condensation, cell cycle phase
   transition, mitotic cell cycle phase transition, and cyclin-dependent
   protein kinase holoenzyme complex) and TCA cycle (p < 0.05). It was
   speculated that the overexpression of ldhA affected the cell division
   of P. tricornutum under LN conditions, which was consistent with the
   significantly greater growth rate of the ldhA OE strain than the WT
   strain under LN conditions, as shown in Fig. [120]2b. KEGG pathway
   enrichment analysis (Fig. [121]6b) revealed that the DEGs were
   significantly enriched in fatty acid elongation, unsaturated fatty acid
   biosynthesis and arachidonic acid metabolism processes (p < 0.05).
   These findings indicated that the fatty acid metabolism of the ldhA OE
   strain under LN conditions was altered compared with that of the WT
   strain, which was consistent with the finding that the total lipid
   content of the ldhA OE strain under LN conditions was greater than that
   of the WT strain, as shown in Fig. [122]2k.

Fig. 8.

   [123]Fig. 8
   [124]Open in a new tab

   GO (a) and KEGG (b) enrichment analysis of A_OE_L_vs_WT_L (p < 0.05).
   Pathways related to photosynthesis are highlighted in green and other
   key pathways are highlighted in gray

Fig. 6.

   [125]Fig. 6
   [126]Open in a new tab

   GO (a) and KEGG (b) enrichment analysis of A_OE_LN/WT_LN (p < 0.05).
   Pathways related to cell division and fatty acid metabolism were
   highlighted in blue and yellow, and other key pathway was highlighted
   in gray

   PHATRDRAFT_48112 and PHATRDRAFT_44454 were involved in the biosynthesis
   of cofactors, which were up-regulated 43.15-fold and 2.85-fold in the
   A_OE_NC/WT_NC group, respectively (p < 0.05). It was possible that ldhA
   overexpression accelerates the biosynthesis of cofactors in P.
   tricornutum because ldhA was involved in catalyzing the reversible
   conversion of pyruvate and lactate accompanied by the interconversion
   of the cofactors NADH and NAD^+. In the A_OE_LN/WT_LN group,
   PHATRDRAFT_21354 and PHATRDRAFT_9984, which are involved in cofactor
   synthesis, were the first (5781.49-fold) and second (659.71-fold)
   up-regulated genes, respectively (p < 0.05). PHATRDRAFT_48112, dsCYC2
   and PHATRDRAFT_45192 were significantly up-regulated (p < 0.05) in the
   A_OE_NC/WT_NC group. Among these genes, dsCYC2 and PHATRDRAFT_9984 were
   also significantly up-regulated in the A_OE_LN/A_OE_NC group (p < 0.05)
   (Table [127]2; Fig. [128]7), suggesting that ldhA overexpression may
   affect the biosynthesis process of cofactors in P. tricornutum in
   response to LN conditions.

Table 2.

   Differentially expressed genes in the NC-LN experiment
   Pathway Gene name KO_name FC(WT_LN/WT_NC) FC(A_OE_LN/A_OE_NC)
   FC(A_OE_NC/WT_NC) FC(A_OE_LN/WT_LN)
   Glycolysis / Gluconeogenesis GPI_2 GPI, pgi 16.99 3.88 ns 0.49
   PGAM_7 PGAM, gpmA 78.61 704.49 ns 2.68
   Carbon fixation in photosynthetic organisms PEPCase_1 ppc 1600.45
   402.00 ns ns
   PHATRDRAFT_bd475 GPT, ALT 17.10 3.38 2.52 0.50
   TCA cycle PYC1 PC, pyc ns 0.37 ns 0.01
   PHATRDRAFT_30145 CS, gltA 12.27 9.99 ns 0.33
   Nitrogen metabolism PHATRDRAFT_8155 nirB 75.45 85.80 ns 2.18
   PHATRDRAFT_13154 nirB 979.14 10073.62 ns 11.21
   PHATRDRAFT_27757 nirA ns 5.51 ns 4.62
   OPPP PHATRDRAFT_35590 NDUFB10 0.19 0.28 ns 2.09
   Lactate metabolism PHATRDRAFT_55040 LDHD, dld 4.42 32.80 ns ns
   PHATRDRAFT_bd1469 ldhA 42.94 23.80 3.62 2.01
   PHATR_46664 dld 2.28 ns ns ns
   PHATRDRAFT_48863 GLO1, gloA 4.30 6.99 ns ns
   GLO_2 gloB, gloC, HAGH 0.41 0.26 ns ns
   PHATRDRAFT_43667 gloB, gloC, HAGH; PNKD 7.20 3.50 ns ns
   PHATRDRAFT_15217 GRHPR 0.04 ns ns ns
   PHATRDRAFT_51714 yvgN 8.29 3.69 ns ns
   Arachidonic acid metabolism PHATRDRAFT_13602 LTA4H 0.11 2.35 ns 2.20
   PHATRDRAFT_37658 HPGDS ns 0.29 ns 2.02
   Biosynthesis of unsaturated fatty acids PTD9 SCD, desC ns ns ns 4.48
   PHATRDRAFT_16376 ELOVL6 0.24 0.47 ns 2.06
   Glycerolipid metabolism PHATRDRAFT_46570 PNPLA2, ATGL ns 2.28 ns 2.60
   PHATRDRAFT_40261 DPP1, DPPL, PLPP4_5 0.35 ns ns 2.00
   Biosynthesis of cofactors PHATRDRAFT_44454 K06897 0.39 0.21 2.85 ns
   PHATRDRAFT_48112 pyrD ns 0.41 43.15 5.29
   dsCYC2 lipB ns 3.49 ns 5.47
   PHATRDRAFT_21354 COX15, ctaA ns 0.10 ns 5781.49
   PHATRDRAFT_45192 adk, AK 0.28 ns ns 2.15
   PHATRDRAFT_9984 ubiG ns 301.70 ns 659.71
   [129]Open in a new tab

Fig. 7.

   [130]Fig. 7
   [131]Open in a new tab

   DEGs involved in metabolic pathways. The number indicates the no. of
   genes listed in Table S3. Square, WT-LN-vs-WT-NC; circle,
   A-OE-LN-vs-A-OE-NC; triangle, A-OE-NC-vs-WT-NC; rhombus,
   A-OE-LN-vs-WT-LN. Red, green and colorless represent up-regulated,
   down-regulated and no significant difference in gene expression,
   respectively

   In the A_OE_L_vs_WT_L group, GO enrichment analysis (Fig. [132]8a)
   revealed that the DEGs were significantly enriched in the processes of
   photosystem II, photosynthesis, light harvesting, light harvesting in
   photosystem I, thylakoid membrane, chloroplast thylakoid membrane,
   plastid thylakoid membrane, photosynthesis membrane, photosystem, and
   chlorophyll binding (p < 0.05). Light reactions were carried out on
   plastid membranes containing photosynthetic pigments and enzymes
   essential for photosynthesis, suggesting that ldhA overexpression
   affects photosynthesis in algal strains under lactate addition
   conditions. The above results were consistent with the greater
   photosynthetic efficiency and photosynthetic pigment content of the
   ldhA OE strain under lactate addition conditions than those of the WT
   strain (Fig. [133]2e-h). KEGG enrichment analysis (Fig. [134]8b) showed
   that the DEGs were enriched mainly in photosynthesis,
   glycolysis/gluconeogenesis, TCA cycle, lysine biosynthesis, nitrogen
   metabolism, and carbon fixation in photosynthetic organisms (p < 0.05).

   The genes encoding the fucoxanthin-chlorophyll a/c protein complex
   during photosynthesis and light harvesting were most abundant in the
   A_OE_L_vs_WT_L group, and the vast majority of the genes encoding
   Lhcf1-11, Lhcf13-17, Lhcr1-4, Lhcr11-14, and Lhcx2 were significantly
   up-regulated (p < 0.05) (Table [135]3). The genes encoding Lhcf8-9,
   Lhcf14-15, Lhcr2, Lhcr10, Lhcr12-13, Lhcx1-2 and Lhcx4 were
   significantly up-regulated (p < 0.05) in the A_OE_C_vs_WT_C group
   (Table [136]3). Although there was no significant difference in the
   genes involved in photosynthesis between the A_OE_C_vs_WT_C and
   A_OE_L_vs_WT_L groups, as shown in Fig. [137]9, the genes associated
   with the fucoxanthin-chlorophyll a/c protein complex involved in the
   process of light harvesting were significantly up-regulated (p < 0.05).
   This finding was consistent with the results shown in Figs. [138]2e-h,
   which revealed that the photosynthetic efficiency and photosynthetic
   pigment content of the ldhA OE strain were greater than those of the WT
   strain under NC and NC + L conditions. It was speculated that the
   overexpression of ldhA accelerated photosynthesis in P. tricornutum,
   possibly as a result of enhanced light harvesting capacity.

Table 3.

   Differentially expressed genes in the C-L experiment
   Pathway Gene name KO_name FC(WT_L/WT_C) FC(A_OE_L/A_OE_C)
   FC(A_OE_C/WT_C) FC(A_OE_L/WT_L)
   Lactate metabolism PHATRDRAFT_55040 LDHD, dld 0.00 ns 0.38 264.44
   PHATRDRAFT_bd1469 ldhA 27.98 13.17 4.27 2.15
   PHATR_46664 dld 0.48 ns ns ns
   GLXI GLO1, gloA 0.35 0.44 ns ns
   PHATR_46728 GLO1, gloA 2.15 ns ns ns
   PHATRDRAFT_43667 gloB, gloC, HAGH; PNKD 4.26 3.48 ns ns
   GLO_2 gloB, gloC, HAGH ns 0.39 ns ns
   PHATRDRAFT_51714 yvgN 0.00 0.00 0.36 ns
   PHATRDRAFT_15217 GRHPR 3.93 0.22 13.32 ns
   Phosphatidylinositol signaling system gene-PHATRDRAFT_47389
   PIP4K2;PIP5K ns 3.83 0.32 ns
   gene-PHATRDRAFT_48773 CALM; CETN1;CML ns ns 0.36 ns
   gene-PHATRDRAFT_44782 IPMK, IPK2 ns 2.78 0.45 ns
   gene-PHATRDRAFT_49634 trxA; CALM 0.02 0.26 0.38 4.46
   gene-PHATRDRAFT_18323 E3.1.3.25, IMPA, suhB 0.26 5.63 ns 9.21
   gene-PHATRDRAFT_43084 IPPK 0.23 2.32 0.20 2.21
   Glycerophospholipid metabolism gene-PHATRDRAFT_7678 CDC2L; E2.7.7.41,
   CDS1, CDS2, cdsA ns ns ns 2.29
   gene-PHATRDRAFT_38067 glpA, glpD 0.22 ns 0.36 3.02
   gene-PHATRDRAFT_47576 pgsA, PGS1 0.29 ns 0.28 2.52
   gene-PHATR_52110 AGXT2L1, ETNPPL ns 68.45 ns 9.19
   Glycerolipid metabolism gene-PHATRDRAFT_2215 DGD 39.10 19.50 ns 0.19
   gene-HATRDRAFT_41254 MGLL 19.77 ns ns 0.07
   gene-PHATRDRAFT_9619 MGD 378.73 ns ns 0.02
   Arachidonic acid metabolism gene-PHATRDRAFT_37658 HPGDS 5.50 ns 3.05
   0.33
   Ether lipid metabolism gene-PHATRDRAFT_33864 EPT1 18.65 10.79 ns 0.49
   photosynthesis, light harvesting Lhcf1 protein fucoxanthin chlorophyll
   a/c protein ns ns 3.66 4.28
   Lhcf2 protein fucoxanthin chlorophyll a/c protein ns ns 3.72 5.16
   Lhcf3 protein fucoxanthin chlorophyl a/c protein ns 57.02 0.23 7.70
   Lhcf4 protein fucoxanthin chlorophyll a/c protein ns 22.91 0.45 11.00
   Lhcf5 protein fucoxanthin chlorophyll a/c protein 0.15 ns ns 2.94
   Lhcf6 protein fucoxanthin chlorophyll a/c protein 0.24 ns ns 8.35
   Lhcf7 protein fucoxanthin chlorophyll a/c protein 0.30 ns ns 6.96
   Lhcf8 protein fucoxanthin chlorophyll a/c protein 0.32 ns 2.65 10.14
   Lhcf9 protein fucoxanthin chlorophyll a/c protein ns ns 2.84 7.30
   Lhcf10 protein fucoxanthin chlorophyll a/c protein 0.21 0.41 ns 2.30
   Lhcf11 protein fucoxanthin chlorophyll a/c protein ns ns ns 2.82
   Lhcf12 protein fucoxanthin chlorophyll protein 3.42 3.66 ns ns
   Lhcf13 protein fucoxanthin chlorophyll a/c protein ns 2.30 ns 4.44
   Lhcf14 fucoxanthin chlorophyll a/c protein, lhcf type ns ns 2.93 5.83
   Lhcf15 protein fucoxanthin chlorophyll a/c protein 0.03 0.03 4.98 5.49
   Lhcf16 protein fucoxanthin chlorophyll a/c protein ns 2.39 ns 2.03
   Lhcf17 protein fucoxanthin chlorophyll a/c protein 11.30 39.29 ns 3.34
   Lhcr1 protein fucoxanthin chlorophyll a/c protein 0.22 ns ns 6.96
   Lhcr2 protein fucoxanthin chlorophyll a/c protein 2.23 3.40 2.19 3.58
   Lhcr3 protein fucoxanthin chlorophyll a/c protein 0.46 ns ns 4.45
   Lhcr4 protein fucoxanthin chlorophyll a/c protein ns ns ns 5.47
   Lhcr5 protein fucoxanthin chlorophyll a/c protein 0.45 ns ns ns
   Lhcr6 protein fucoxanthin chlorophyll a/c protein 3.21 ns ns 0.39
   Lhcr8 protein fucoxanthin chlorophyll a/c protein 6.59 2.71 ns 0.30
   Lhcr9 fucoxanthin chlorophyll a/c protein, lhcr type 0.33 0.38 ns ns
   Lhcr10 protein fucoxanthin chlorophyll a/c protein 4.84 ns 3.08 ns
   Lhcr11 protein fucoxanthin chl a/c protein ns 3.77 ns 9.30
   Lhcr12 protein fucoxanthin chlorophyll a/c protein 0.49 ns 2.36 6.91
   Lhcr13 protein fucoxanthin chlorophyll a/c protein 3.79 4.99 2.59 3.65
   Lhcr14 protein fucoxanthin chlorophyll a/c protein ns 2.56 ns 7.24
   Lhcx1 protein fucoxanthin chlorophyll a/c protein 21.38 10.63 2.00 ns
   Lhcx2 protein fucoxanthin chlorophyll a/c protein ns ns 4.00 3.49
   Lhcx3 protein fucoxanthin chlorophyll a/c protein 0.37 ns ns ns
   Lhcx4 protein fucoxanthin chlorophyll a/c protein ns ns 2.06 ns
   [139]Open in a new tab

Fig. 9.

   [140]Fig. 9
   [141]Open in a new tab

   Differentially expressed genes involved in metabolic pathways. The
   number indicates the no. of genes listed in Table S4. Square,
   WT-L-vs-WT-C; circle, A-OE-L-vs-A-OE-C; triangle, A-OE-C-vs-WT-C;
   rhombus, A-OE-L-vs-WT-L. Red, green and colorless represent
   upregulated, downregulated and not significantly different in gene
   expression, respectively

   Most of the genes involved in the oxidative phosphorylation process
   presented similar expression trends in the WT_L/WT_C and A_OE_L/A_OE_C
   groups. However, there were more DEGs involved in oxidative
   phosphorylation in the A_OE_L/A_OE_C group (Table [142]3; Fig. [143]9),
   suggesting that exogenous lactate had a greater effect on the ldhA OE
   strain than on the WT strain in terms of energy balance. Most of the
   genes involved in the Calvin cycle were significantly up-regulated
   (p < 0.05) in the A_OE_L/A_OE_C and A_OE_L_vs_WT_L groups but were
   down-regulated or not significantly different in the WT_L/WT_C group
   (Table [144]3; Fig. [145]9), indicating that ldhA overexpression
   accelerated the Calvin cycle of P. tricornutum under lactate addition
   conditions. In addition, most of the genes involved in photosynthesis
   were significantly down-regulated (p < 0.05) in the WT_L/WT_C group but
   significantly up-regulated (p < 0.05) in the A_OE_L_vs_WT_L group
   (Table [146]3; Fig. [147]9), suggesting that ldhA overexpression
   accelerated photosynthesis in P. tricornutum under lactate addition
   conditions.

   The expression of PHATRDRAFT_bd1469 (encoding LDHA), a differential
   gene involved in lactate metabolism, was significantly up-regulated
   (p < 0.05) in the 4 different groups of the C-L experimental group.
   PHATRDRAFT_55040 and PHATRDRAFT_51714 were significantly down-regulated
   (p < 0.05) in the A_OE_C_vs_WT_C group. These genes are involved in the
   reversible reactions from D-lactate to pyruvate and from MGO to
   L-lactate, respectively. In the A_OE_L_vs_WT_L group, PHATRDRAFT_55040
   involved in the reversible reaction of D-lactate to pyruvate was
   significantly up-regulated by 264.441-fold (p < 0.05) (Table [148]3),
   indicating that the overexpression of ldhA accelerated the reversible
   reaction of D-lactate to pyruvate in P. tricornutum under lactate
   supplementation.

   According to Table [149]3, in the A_OE_C/WT_C group, the genes involved
   in the phosphatidylinositol signaling system PHATRDRAFT_47389,
   PHATRDRAFT_48773, PHATRDRAFT_44782, and PHATRDRAFT_43084 were
   significantly down-regulated (p < 0.05). Whereas, in the A_OE_L_vs_WT_L
   group, PHATRDRAFT_49634, PHATRDRAFT_18323, PHATRDRAFT_43084 and
   PHATRDRAFT_7678, which participate in the phosphatidylinositol
   signaling system, were significantly up-regulated (p < 0.05),
   demonstrating that ldhA overexpression affects the phosphatidylinositol
   signaling system under lactate addition conditions, which may affect
   the energy balance and carbon and nitrogen balance of P. tricornutum.

   In addition to the fatty acid synthesis process shown in Fig. [150]9,
   other genes related to lipid metabolism were differentially expressed
   (Table [151]3). PHATRDRAFT_38067, PHATRDRAFT_47576 and PHATR_52110,
   which are involved in glycerophospholipid metabolism, were
   down-regulated or not significantly different between the WT_L/WT_C and
   A_OE_L/A_OE_C groups, but were significantly up-regulated in the
   A_OE_L_vs_WT_L group (p < 0.05), indicating that the overexpression of
   ldhA promoted glycerophospholipid metabolism in P. tricornutum under
   lactate addition conditions. This finding was consistent with the
   results shown in Fig. [152]2k, which revealed that the total lipid
   content of the ldhA OE strain was greater than that of the WT strain
   under lactate-supplemented conditions. The genes involved in
   glycerolipid metabolism, arachidonic acid metabolism and other lipid
   metabolism, PHATRDRAFT_2215, PHATRDRAFT_41254, PHATRDRAFT_9619,
   PHATRDRAFT_37658, and PHATRDRAFT_33864, were significantly up-regulated
   in the WT_L/WT_C group (p < 0.05) but significantly down-regulated
   (p < 0.05) in the A_OE_L_vs_WT_L group, suggesting that lactate
   promoted lipid accumulation in P. tricornutum and that ldhA
   overexpression affected lipid metabolism in P. tricornutum under
   lactate addition conditions.

Discussion

   Lactate is a metabolic product of cellular anaerobic glycolysis and
   acts as a critical role in the regulation of cellular metabolism.
   First, lactate can act as a signaling molecule and with GPCR to inhibit
   the activity of adenylate cyclase, reduce the synthesis of the second
   messenger cAMP, and regulate gene transcription through the cAMP
   signaling pathway, thus affecting cellular metabolism [[153]32].
   Second, lactate can modify lysine residues of proteins by lactylation
   to regulate gene transcription (lactylation of histones) or modulate
   enzyme activity (lactylation of nonhistone proteins) [[154]33–[155]35].
   Protein lactylation is considered to be the main way in which lactate
   performs its metabolic regulatory function and is very well conserved
   during the evolution of species [[156]33].

   In contrast to microalgal lactylation, in plants, including rice (Oryza
   sativa) [[157]34], wheat (Triticum aestivum L.) [[158]36], maize (Zea
   mays L.) [[159]37], and sugarcane (Saccharum spp.) [[160]38],
   lactylation modification was mainly related to photosynthesis and
   carbon metabolism, especially in the regulation of photosynthesis and
   photoenergy conversion. Previous studies have shown that in
   Nannochloropsis oceanica, lactylation was significantly enriched in
   lipid metabolism, carbon fixation, glycolysis/gluconeogenesis, TCA
   cycle, and photosynthesis, primarily responding to nitrogen stress by
   inhibiting carbon fixation and activating lipid synthesis pathways
   [[161]39, [162]40]. In contrast, in P. tricornutum, lactylation
   modification were more focused on coordinating carbon and nitrogen
   metabolism with environmental adaptation, reflecting the unique
   ecological strategy of diatoms [[163]11]. Interestingly, under nitrogen
   stress, the phenomenon of enhanced lactylation occurs in both P.
   tricornutum (LN condition) and N. oceanica (nitrogen deprivation
   condition). In summary, lactylation modifications were closely related
   to energy metabolism in both above mentioned plants and microalgae,
   indicating the functional diversity and evolutionary conservation of
   lactylation in different organisms.

   Studies related to the involvement of lactate as a signaling molecule
   in signaling pathways are currently limited to mammals and have not
   been reported in algae. Previous studies in our laboratory revealed
   that endogenous lactate increased significantly in P. tricornutum under
   high C: N ratio conditions, accompanied by differential expression of
   proteins related to carbon and nitrogen metabolism. Lactylated proteins
   were significantly enriched in carbon and nitrogen metabolism and
   energy metabolism, suggesting that endogenous lactate regulates carbon
   and nitrogen metabolism and energy metabolism in P. tricornutum through
   the lactylation of proteins [[164]11]. In addition, exogenous lactate
   significantly affected the growth, metabolism and synthesis of
   important biochemical components in P. tricornutum [[165]12]. In
   conclusion, lactate plays an essential role in the carbon and nitrogen
   metabolism of P. tricornutum. In view of these findings, the present
   study was conducted to investigate the lactate metabolic pathway and
   its regulatory genes. The aim of this study was to provide a
   theoretical basis for studying the regulatory mechanism of carbon and
   nitrogen metabolism in P. tricornutum based on lactate and to provide
   theoretical guidance for developing technology for producing P.
   tricornutum with a high oil yield.

   In this study, we successfully constructed strains that overexpressed
   key endogenous genes for lactate metabolism and regulation (ldhA, Glo1,
   Glo2, D-LCR, GlxI, and GPCR). The overexpression of ldhA, Glo1, Glo2,
   and D-LCR accelerated L-lactate consumption rates of the WT strain
   (Fig. [166]1), which is beneficial for P. tricornutum to cope with low
   nitrogen conditions (Figs. [167]2 and [168]S6-S8). In contrast, the
   growth and L-lactate consumption rates of the GlxI and GPCR OE strains
   were significantly lower than or not significantly different from those
   of the WT strain (Figs. [169]S9-S10). The above results indicated that
   LDHA, GLO1, GLO2 and D-LCR mediate lactate metabolism and lactylation
   metabolic processes in P. tricornutum under low nitrogen conditions,
   which constitutes the molecular basis by which lactate influences
   growth, rapidly regulates carbon and nitrogen metabolism processes, and
   promotes lipid accumulation.

   Lactate dehydrogenase is a tetramer that catalyzes the reversible
   conversion of pyruvate to lactate (the last step in glycolysis)
   [[170]41]. Previous research on lactate dehydrogenase in microalgae has
   focused mainly on the use of microalgae to produce lactate through
   lactate fermentation, which is widely used as a commercial chemical in
   various fields. For example, C. merolae produced L-lactate by L-lactate
   dehydrogenase-catalyzed lactate fermentation under dark anaerobic
   conditions [[171]42], and nitrogen deprivation prior to dark anaerobic
   incubation enhanced L-lactate production [[172]43]. Synechocystis sp.
   PCC 6803, an engineered alga carrying lactate dehydrogenase, produced
   D-lactate from anaerobic digestion wastewater [[173]44]. In this study,
   on the basis of the prediction of protein structural domains (Fig.
   [174]S15), LDHA in P. tricornutum exhibited a function similar to that
   of D-lactate dehydrogenase, with NAD as a cofactor [[175]45]. While,
   D-LCR was a D-lactate dehydrogenase that catalyzes the oxidoreductase
   of a specific substrate through the binding of flavin adenine
   dinucleotide (FAD) as a cofactor. It is an oxidoreductase that
   catalyzes a specific substrate by binding with FAD, and the histidine
   domain bound with FAD group is conserved. LDHA and D-LCR, which also as
   lactate dehydrogenases, performed their catalytic functions with
   different cofactors, so the functions of LDHA and D-LCR may differ in
   P. tricornutum. Under NC conditions, the effects of ldhA overexpression
   and RNA silence on the growth of P. tricornutum did not differ
   significantly (Fig. [176]2a), but the overexpression and RNA silence of
   D-LCR inhibited and promoted the growth rate, photosynthetic
   efficiency, and accumulation of photosynthetic pigments in P.
   tricornutum, respectively (Fig. S6). In contrast, the growth rate of
   ldhA-POE2 was significantly greater (p < 0.05) than that of WT under LN
   conditions (Fig. [177]2b), and the overexpression and RNA silence
   strains of D-LCR presented the same growth trend as that under NC
   conditions (Fig. [178]S6b). These results suggest that the effects of
   ldhA and D-LCR on P. tricornutum differ greatly, and that D-LCR may
   preferentially participate in the reversible transformation process of
   pyruvate and D-lactate. Hence, the overexpression and RNA silence of
   D-LCR directly affected the growth and photosynthesis of P. tricornutum
   under NC conditions and thus had opposite effects. Nevertheless, the
   effect of ldhA overexpression on P. tricornutum was more obvious under
   LN conditions, indicating that ldhA may preferentially perform its
   catalytic function under low nitrogen conditions. Under low nitrogen
   conditions, the overexpression of ldhA could increase the ability of P.
   tricornutum to utilize lactate and accelerate lactate metabolism, which
   in turn could promote central carbon metabolic processes, such as lipid
   metabolism, the Calvin cycle, glycolysis, and the pentose phosphate
   pathway, and facilitate its growth and lipid accumulation.

   Glyoxalase is a key enzyme in lactylation [[179]33] and can degrade
   MGO, a byproduct of glycolysis, among which glyoxalase I (GLO1 and
   GLXI) is a lactoylglutathione lyase that prefers divalent cationic zinc
   as a cofactor [[180]46]. Glyoxalase I catalyzes the intramolecular
   redox reaction of hemithioacetal (formed by MGO and GSH) to form
   S-D-LGSH. Eukaryotic glyoxalase I tends to use divalent cationic zinc
   as a cofactor, whereas Escherichia coli and other prokaryotes use
   nickel [[181]47]. Under LN + L conditions, the growth rates of
   Glo1-POE8 and GlxI-PAS3 were significantly greater than those of WT
   (p < 0.05), but the growth rates of Glo1-PAS3 and GlxI-POE6 were
   significantly lower than those of WT (p < 0.05) (Figs. [182]S7 and
   [183]S9). These results indicated that although both Glo1 and GlxI in
   P. tricornutum were associated with glyoxalase I, their effects on the
   growth of P. tricornutum under low nitrogen conditions were completely
   opposite, probably because the overexpression of Glo1 increased the
   ability of P. tricornutum to utilize exogenous lactate under low
   nitrogen conditions, whereas the overexpression of GlxI inhibited its
   ability to utilize lactate. In summary, the overexpression of Glo1
   accelerated the utilization of lactate by P. tricornutum under low
   nitrogen conditions, thereby accelerating lactate metabolism and
   facilitating the growth and lipid accumulation of P. tricornutum,
   whereas the overexpression of GlxI inhibited the ability of P.
   tricornutum to utilize lactate and inhibited its growth and lipid
   accumulation. GLO2 belongs to glyoxalase II in the glyoxalase system
   and acts as a hydroxyacylglutathione hydrolase that catalyzes the
   hydrolysis of S-D-LGSH to form glutathione and D-lactate [[184]48].
   Under NC conditions, the growth rate and photosynthetic pigment content
   of Glo2-PAS6 were significantly lower (p < 0.05) than those of the WT
   (Fig. [185]S8a), suggesting that RNA silence of Glo2 inhibited the
   accumulation of photosynthetic pigments in P. tricornutum. The growth
   rates of Glo2-POE13 and Glo2-PAS6 were significantly greater than those
   of the WT (p < 0.05) under LN + L conditions (Fig. [186]S8d),
   indicating that the overexpression of Glo2 facilitates the utilization
   of lactate by P. tricornutum under low nitrogen conditions, which in
   turn promotes its growth.

   GPCRs with seven transmembrane structural domains belong to class C
   GPCRs that transmit physiological signals from the extracellular space
   to the intracellular space, induce conformational changes, activate
   heterotrimeric G proteins, and activate many downstream effector
   proteins [[187]32]. From the genome of P. tricornutum, it is known that
   GPCR, pkA and ldhA are arranged in tandem [[188]49] (Fig. [189]10), and
   it is hypothesized that lactate may activate the cAMP signaling pathway
   of P. tricornutum through the signaling molecule-GPCR-protein
   kinase-transcription factor gene expression-regulated delivery chain
   consisting of LDHA, PKA and GPCR. The cAMP signaling pathway in P.
   tricornutum is involved in the perception of carbon concentration and
   conductance and the regulation of carbon assimilation
   [[190]50–[191]53]. The overexpression of GPCR inhibited the
   photosynthetic efficiency and accumulation of photosynthetic pigments
   in P. tricornutum, whereas the RNA silence of GPCR increased the
   photosynthetic efficiency of P. tricornutum (Figs. [192]S10e-h). The
   overexpression of GPCR was detrimental to the response of P.
   tricornutum to low nitrogen conditions, but was conducive to protein
   accumulation under low nitrogen conditions (Fig. [193]S10l).

Fig. 10.

   [194]Fig. 10
   [195]Open in a new tab

   Metabolic pathways of lactate production and lactoacylation

   Based on previous studies in our laboratory and the present study, the
   process of lactate generation and lactylation in P. tricornutum under
   LN conditions (Fig. [196]10). LN conditions lead to the accumulation of
   central carbon metabolites such as pyruvate. On the one hand, a large
   amount of pyruvate was catalyzed by L-lactate dehydrogenase to generate
   L-lactate and lactylated proteins in the cytoplasm. On the other hand,
   a large amount of pyruvate was generated as GMO, which was catalyzed by
   GLO1 to generate S-D-LGSH, and modified by lactylation in a
   non-enzyme-dependent acyl transfer reaction, which further accelerated
   the hydrolysis of LGSH by GLO2, recovered GSH and generated D-lactate,
   which could be reversibly converted to pyruvate by the catalytic action
   of LDHA or D-LCR. The intracellularly generated lactate may initiate
   the P. tricornutum cAMP signaling pathway under LN conditions through a
   signaling molecule-GPCR-protein kinase-transcription factor gene
   expression-regulated delivery chain consisting of LDHA, PKA, and GPCR,
   which may influence photosynthesis and protein synthesis in P.
   tricornutum by participating in the regulation of carbon concentration
   sensing, conductance, and carbon assimilation. In addition, the local
   concentration of lactate varies in different cellular compartments, and
   L-lactate transporter enzyme (LCTP) directly affects the distribution
   of lactate in different cellular compartments. The only enzyme
   responsible for the transmembrane transport of L-lactate, LCTP, which
   is present in the genome of P. tricornutum, was detected under high C:
   N ratios [[197]11]. Members of our laboratory have resolved the
   function of LCTP and confirmed that LCTP is involved in the transport
   and metabolism of exogenous lactate.

   Proteins in multimeric form typically have a larger molecular weight
   than monomers do. Larger than expected bands were observed by WB,
   probably due to the multimeric structure of LDHA. While
   pan-lactyllysine blots confirmed lactylation in ldhA OE strains
   (Fig. [198]4), the direct functional impact on enzyme activity remains
   to be elucidated. In the future, we can identify lactylation targets
   through proteomics and validate the regulatory effect of lactylation
   sites on enzyme activity by site-directed mutagenesis. In mammals, a
   PDK1 inhibitor (dichloroacetate sodium) reduced intracellular lactate
   levels by inhibiting the PDH-catalyzed production of acetyl coenzyme A
   from pyruvate, whereas an LDH inhibitor (oxalate) decreases histone
   lactylation levels by inhibiting LDH activity. In contrast, rotenone
   (an inhibitor of mitochondrial respiratory chain complex I, which
   drives cells toward glycolysis) increased intracellular lactate and
   histone lactylation levels [[199]35]. It was speculated that LDHA in P.
   tricornutum functions similarly to that in mammals, regulated
   intracellular lactate content through LDHA and thus affected the level
   of lactylation modifications. ^13C labeling analysis revealed that ldhA
   overexpression promoted lactate utilization efficiency in P.
   tricornutum and accelerated glycolysis, TCA cycle, CCM, and Calvin
   cycle (Fig. [200]5).

   GO enrichment analysis of the A_OE_LN_vs_WT_LN group indicated that the
   overexpression of ldhA affected cell division metabolism under low
   nitrogen conditions (Fig. [201]6), which was consistent with the
   results shown in Fig. [202]2b, which revealed that the growth rate of
   the overexpression strain of P. tricornutum was significantly greater
   than that of the WT strain under low nitrogen conditions. The genes
   involved in lipid metabolism (arachidonic acid metabolism, unsaturated
   fatty acid biosynthesis, and glycerolipid metabolism) and
   nitrite-to-ammonium conversion in nitrogen metabolism were
   significantly up-regulated (p < 0.05) (Table [203]2), which was
   consistent with Fig. [204]2k and demonstrated that the total lipid
   content of the ldhA OE strain was greater than that of the WT strain
   under LN conditions, suggesting that the overexpression of ldhA
   affected the energy homeostasis and lipid metabolism of P. tricornutum
   under low nitrogen conditions.

   GO and KEGG enrichment analysis of the A_OE_L_vs_WT_L group indicated
   that ldhA overexpression affected photosynthesis in P. tricornutum
   under lactate addition conditions. The genes encoding the
   fucoxanthin-chlorophyll a/c protein complex were significantly
   up-regulated (p < 0.05) in the A_OE_C_vs_WT_C and A_OE_L_vs_WT_L groups
   (Table [205]3), which was consistent with the results showing that
   F[v]/F[m], YII, chlorophyll a, and carotenoid contents of the ldhA OE
   strain were significantly greater than those of the WT strain under NC
   and NC + L conditions (Fig. [206]2e-h), suggesting that the
   overexpression of ldhA accelerated photosynthesis in P. tricornutum,
   possibly due to the increased light harvesting capacity. The
   fucoxanthin-chlorophyll a/c protein complex is a light harvesting
   antenna protein that plays an important role in harvesting blue-green
   light protection and is the energy basis for the rapid growth and
   reproduction of diatom cells [[207]54].

   The phosphatidylinositol signaling system may play a crucial role in
   the regulation of energy, carbon, and nitrogen balance and homeostasis
   maintenance in P. tricornutum [[208]55]. The expression of genes
   involved in the phosphatidylinositol signaling system was up-regulated
   (p < 0.05) in the A_OE_L_vs_WT_L group (Table [209]3), indicating that
   the overexpression of ldhA might affect the energy homeostasis of P.
   tricornutum under lactate addition conditions. On the other hand, genes
   involved in glycerophospholipid metabolism were significantly
   up-regulated (p < 0.05) in the A_OE_L_vs_WT_L group (Table [210]3),
   which was consistent with the results that the total lipid content of
   the ldhA OE strains was greater than that of the WT strain under
   lactate addition conditions (Fig. [211]2k), indicating that the
   overexpression of P. tricornutum promoted glycerophospholipid
   metabolism under lactate addition conditions.

Conclusions

   LDHA, GLO1, GLO2 and D-LCR mediate lactate metabolism and protein
   lactylation in P. tricornutum under low nitrogen conditions, which are
   the molecular basis for the ability of lactate to influence its growth,
   rapidly regulate carbon and nitrogen metabolism processes, and promote
   lipid accumulation. Among them, ldhA overexpression in P. tricornutum
   promoted lactylation of an approximately 40 kDa protein; accelerated
   the ability to utilize lactate, accelerated glycolysis, TCA cycle, CCM
   and Calvin cycle; accelerated lipid accumulation in P. tricornutum
   under low nitrogen and lactate addition condition. This
   lactate-mediated metabolic optimization establishes a molecular
   foundation for engineering high-lipid P. tricornutum strains.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [212]12934_2025_2779_MOESM1_ESM.docx^ (5.9MB, docx)

   Supplementary Material 1: Supplementary Figures
   [213]12934_2025_2779_MOESM2_ESM.docx^ (27.9KB, docx)

   Supplementary Material 2: Supplementary Method S1
   [214]Supplementary Material 3^ (77.1KB, xlsx)
   [215]Supplementary Material 4^ (9.4KB, xlsx)
   [216]Supplementary Material 5^ (9.2KB, xlsx)
   [217]Supplementary Material 6^ (18.7KB, xlsx)
   [218]Supplementary Material 7^ (23.6KB, xlsx)

Acknowledgements