Thallusin Quantification in Marine Bacteria and Algae Cultures

Thallusin, a highly biologically active, phytohormone-like and bacterial compound-inducing morphogenesis of the green tide-forming macroalga Ulva (Chlorophyta), was determined in bacteria and algae cultures. A sensitive and selective method was developed for quantification based on ultra-high-performance liquid chromatography coupled with electrospray ionization and a high-resolution mass spectrometer. Upon C18 solid phase extraction of the water samples, thallusin was derivatized with iodomethane to inhibit the formation of Fe–thallusin complexes interfering with the chromatographic separation. The concentration of thallusin was quantified during the relevant phases of the bacterial growth of Maribacter spp., ranging from 0.16 ± 0.01 amol cell−1 (at the peak of the exponential growth phase) to 0.86 ± 0.13 amol cell−1 (late stationary phase), indicating its accumulation in the growth medium. Finally, we directly determined the concentration of thallusin in algal culture to validate our approach for monitoring applications. Detection and quantification limits of 2.5 and 7.4 pmol L−1, respectively, were reached, which allow for quantifying ecologically relevant thallusin concentrations. Our approach will enable the surveying of thallusin in culture and in nature and will thus contribute to the chemical monitoring of aquaculture.


Introduction
The cosmopolitan green macroalga Ulva spp. (Ulvales, Chlorophyta) requires specifically associated bacteria that release algal growth-and morphogenesis-promoting factors (AGMPFs) [1,2]. Without such symbioses, Ulva develops into an undifferentiated callus under axenic conditions [1]. Because Ulva is commercially valuable as a source of, for example, bioactive compounds, food, and biofuel [3], beneficial algal-bacterial interactions that promote growth are of particular interest [4]. To identify such AGMPFs, a model system of Ulva compressa L. (cultivar Ulva mutabilis Føyn, recently reclassified [5]) was designed with a microbiome consisting of only two essential bacterial strains forming a tripartite community of U. compressa, Roseovarius sp. strain MS2, and Maribacter sp. strain MS6 [1]. Roseovarius sp. induces cell division in Ulva, while Maribacter sp. promotes primary rhizoid and cell wall formation. Like plant hormones, the Roseovarius-factor functionally resembles a cytokinin, while the Maribacter-factor acts auxin-like, contributing to the attachment of the algae [1,6].
Plant hormones (phytohormones) are naturally occurring small organic molecules and can be categorized into several substance classes: auxins, abscisic acid, brassinosteroids, cytokinins, ethylene, gibberellins, jasmonates, and salicylates. They are produced by plants but also by beneficial and pathogenic microorganisms (e.g., bacteria and microalgae) that can modulate plant growth, physiology, and immunity [7]. Phytohormone research is essential for advancing agricultural and marine science because many genes responsible for important agronomic traits, such as plant height, seed development, and yield, are phytohormone related [8].
Thus, our study aimed to develop a method to determine thallusin concentration with a sufficiently low detection limit, to enable quantitative studies of Ulva development. A reliable quantification method for thallusin with high sensitivity and accuracy is crucial for studying the metabolic, transport, and molecular mechanisms of thallusin in marine macroalgae and, potentially, in terrestrial settings. While several modern methods exist for quantitative analysis of phytohormones in crude plant extracts that were reported [16][17][18][19], no protocols are available for thallusin. Certainly, extraction and purification protocols must be adapted for macroalgae samples [20][21][22][23]. A mass-spectrometry-based analysis is frequently influenced by the sample type (various matrices), the extraction procedure, and the experimental setup, including factors such as the chromatographic separation system and mass analyzer [24]. As such, a validated method is required to compare data generated in different laboratories and experimental setups using an appropriate internal standard.
In this contribution, we describe the development of a method for thallusin quantification in bacterial and algal growth media using an orbitrap high-resolution mass spectrometer. We studied laboratory-grown cultures of different Maribacter and Zobellia strains previously tested for thallusin production [6,25] and algal cultures. The analytical process features solid phase extraction, purification, iodomethane derivatization, and liquid chromatography-mass spectrometry (LC-MS) measurements. Notably, derivatization techniques are frequently employed in phytohormonal analysis to increase the stability and detectability of analytes [8].

Results
Thallusin and its 2:1 iron(III) complex can be detected in the bacterial culture medium and determined by LC-MS analysis, showing the characteristic pseudo molecular ions of [M + H] + and [2M + Fe(III) − 2H] + , respectively (Figure 1) [5]. Thallusin eluted first at 5.21 min (m/z 458.2173) on a C 18 RP silica column, directly followed by the Fe-thallusin complex (m/z 968.3388) at 5.75 min with broad tailing. Adding 0.1% (v/v) formic acid at the start of the preparation process resulted in decomplexation and improved the peak shape, but, unfortunately, it led to a 200-fold decrease in peak intensity. Notably, retention times, peak areas, and heights for metal complexes are often pH dependent; the lower the pH, the smaller the peak areas and heights. Scavenging the iron ions by using a ligand exchange reaction with ethylenediaminetetraacetic acid remained equally unsuccessful. Furthermore, thallusin and its Fe complex interfere with each other in chromatographic separations on C 18 reversed-phase columns ( Figure 1A,B). Likely, a major fraction of thallusin occurs metal bound in the environment, but its Fe(III)-complex is not kinetically stable on HPLC, as indicated by its peculiar HPLC elution profile indicating exchange. Therefore, derivatization by methylation of the carboxyl groups was investigated to suppress metal ion complexation. The synthetic thallusin standard (0.1 mmol L −1 ) and the internal standard (0.1 mmol L −1 ) were analyzed by UHPLC-ESI-HRMS using the single ion mode (SIM) (thallusin: m/z 458.2176; internal standard: m/z 458.2541) combined with C18 SPE. At 5.19 min, the internal standard was eluted, while the thallusin standard was eluted with a strong tailing after 5.21 min-the Fe-thallusin complex (m/z 968.3394) decomposed through in-source fragmentation. (C) Chromatograms after derivatization with iodomethane. The fully methylated thallusin trimethyl ester (m/z 500.264) and the synthetic reference standard were eluted at 6.53 min. The internal standard was eluted at 5.88 min due to methylation (m/z 472.2688). (D) Mass spectra (TIC) of the chromatograms are shown in (B). Left: Mass spectrum of the peak at 5.21 min-thallusin (black, m/z 458.2176) and the internal standard (red, m/z 458.2537) were detected at 5.21 min. Middle: 54

Method Development
Methylation of the carboxyl groups of thallusin was achieved by treatment with iodomethane and potassium hydroxide in DMSO at 20 °C to give methyl esters in high yield (>99%) ( Figure 1A). The thallusin trimethyl ester showed increased peak sharpness, improving the detection limit and calibration capability ( Figure 1B,C). The peak was identified by comparing the associated retention time and the mass spectrum with those of a synthetic reference standard of thallusin trimethyl ester ( Figure 1B-E) [9]. The added internal standard became methylated as well, thus providing evidence of successful

Method Development
Methylation of the carboxyl groups of thallusin was achieved by treatment with iodomethane and potassium hydroxide in DMSO at 20 • C to give methyl esters in high yield (>99%) ( Figure 1A). The thallusin trimethyl ester showed increased peak sharpness, improving the detection limit and calibration capability ( Figure 1B,C). The peak was identified by comparing the associated retention time and the mass spectrum with those of a synthetic reference standard of thallusin trimethyl ester ( Figure 1B-E) [9]. The added internal standard became methylated as well, thus providing evidence of successful derivatization in the analytical procedure. To estimate the methylation yield, we searched for the fully unmethylated thallusin in the total ion chromatogram (TIC). However, unmethylated thallusin and mono-(m/z 472.2330 [M + H] + ) were not detected. The ratio of the peak areas of the double-methylated thallusin (m/z 486.2486 [M + H] + ) to the triple (fully)-methylated thallusin (m/z 500.2643 [M + H] + ) was 1.69 ± 0.35%; this implies that approximately 98% of the thallusin was fully methylated, close to the 99% yield obtained by Avilia-Zárraga and Martinez (2001) for derivatization of a variety of carboxylic acids by iodomethane [26].
The calibration was performed over the entire analytical process for the working range from 0.2 to 2.0 and 2.0 to 10.0 µmol L −1 (Figure 2A,B). The limit of detection (LOD) and limit of quantification (LOQ) were 0.0756 and 0.2217 µmol L −1 , respectively (Table 1); these values constitute adequate detection limits to quantify thallusin in the region of the effective concentration (EC 90 = 7.6 pmol L −1 [6]) in water samples. When 0.75 L of the growth medium was concentrated by solid phase extraction and the final sample volume of eluent was 25 µL, LOD and LOQ values of 2.5 and 7.4 pmol L −1 , respectively, were obtained while accounting for a concentration factor of 30,000. The total error of the analytical procedure s r was 0.8%, indicating an accurate measurement. In order to normalize the thallusin concentration per cell, photometric optical density measurements were correlated with the number of cells counted by flow cytometry ( Figure 2C). The calibrations enabled measuring thallusin concentrations in the bacterial cultures at various stages of their growth and in the laboratory-grown Ulva cultures.  [26]. The calibration was performed over the entire analytical process for the working range from 0.2 to 2.0 and 2.0 to 10.0 µmol L −1 (Figure 2A, B). The limit of detection (LOD) and limit of quantification (LOQ) were 0.0756 and 0.2217 µmol L −1 , respectively (Table 1); these values constitute adequate detection limits to quantify thallusin in the region of the effective concentration (EC90 = 7.6 pmol L −1 [6]) in water samples. When 0.75 L of the growth medium was concentrated by solid phase extraction and the final sample volume of eluent was 25 µL, LOD and LOQ values of 2.5 and 7.4 pmol L −1 , respectively, were obtained while accounting for a concentration factor of 30,000. The total error of the analytical procedure sr was 0.8%, indicating an accurate measurement. In order to normalize the thallusin concentration per cell, photometric optical density measurements were correlated with the number of cells counted by flow cytometry ( Figure 2C). The calibrations enabled measuring thallusin concentrations in the bacterial cultures at various stages of their growth and in the laboratory-grown Ulva cultures.
To quantify thallusin, samples were collected at three different growth phases. Overall, the thallusin concentration increased with the cell density and accumulated in the bacterial growth media (Figure 4).

Species-Dependent Production of Thallusin
After normalization to the bacterial cell count, the thallusin concentration in the media followed a similar pattern of changes for all the tested bacteria strains, with some minor differences (compare Table 2 with Figure 5). Exemplarily, the production of thallusin in Maribacter's growth medium follows the subsequent pattern: After 70 h of incubation, the supernatant of Maribacter sp. comprised 1.5 ± 0.1 nmol L −1 (0.16 ± 0.01 amol cell −1 ) thallusin at the peak of the exponential growth phase. After a further 6 h, the thallusin concentration increased significantly to 4.3 ± 0.8 nmol L −1 (0.27 ± 0.05 amol cell −1 , p < 0.005, t-test, n = 3) before peaking at 11.2 ± 1.7 nmol L −1 (0.86 ± 0.13 amol cell −1 , p < 0.005, t-test, n = 3) during the late stationary phase. Overall

Quantification of Thallusin in Algal Cultures
Using our approach, we set out to detect thallusin in the chemosphere of Ulva in culture. A co-injection experiment allowed for the identification of the peak of the thallusin trimethyl ester by comparison with the synthetic standard ( Figure 6). The thallusin concentration in the pooled stationary culture of a closed cultivation system (U. compressa, three months old, 17.2 g L −1 fresh weight biomass) was 0.22 ± 0.04 nmol L −1 , implying that under the selected laboratory conditions, the actual thallusin concentration in the algal culture was approximately 30-times higher than the EC 90 = 7.6 pmol L −1 [9], indicating strong oversaturation. different species. Maribacter ulvicola grew the fastest and reached the stationary phase after 50 h with an OD620 of 0.31, while Maribacter chungangensis grew the slowest, reaching its stationary phase after 82 h with an OD620 of 0.23 (growth rate µ = 0.158 ± 0.004 h −1 ). Maribacter sp. MS6 (max. OD620 = 0.30, µ = 0.160 ± 0.002 h −1 ), Zobellia galactanivorans (max. OD620 = 0.28), and M. ulvicola (max. OD620 = 0.31) yielded the highest optical densities. The growth rates (µ) between the bacteria varied significantly (one-way ANOVA, p < 0.05), with the highest value obtained for Maribacter sedimenticola (µ = 0.204 ± 0.003 h −1 ; Figure 3H).  ANOVA determined significant differences with a Tukey post-hoc test (p < 0.05); the different letters on top of the boxplots indicate statistically significant differences. The error bars represent the mean ± standard deviation.
To quantify thallusin, samples were collected at three different growth phases. Overall, the thallusin concentration increased with the cell density and accumulated in the bacterial growth media (Figure 4).  Table 2). The error bars represent the mean ± standard deviation (n = 3). A one-way ANOVA determined significant differences with a Tukey post-hoc test (p < 0.05); the different letters indicate statistically significant differences.   Table 2). The error bars represent the mean ± standard deviation (n = 3). A one-way ANOVA determined significant differences with a Tukey post-hoc test (p < 0.05); the different letters indicate statistically significant differences. 3) during the late stationary phase. Overall, the supernatants of the strains Maribacter sp. MS6, M. ulvicola, M. sedimenticola, and Zobellia contained the most thallusin normalized per cell 24 h after reaching the stationary phase, with a thallusin concentration range of 9.2-14.8 nmol L −1 ( Table 2), equivalent to approximately 0.8-1.0 amol cell −1 (Figure 5; one-way ANOVA, Tukey post-hoc test, p < 0.05). This indicates that thallusin accumulated in the medium, in agreement with previous measurements of the biological activity of Maribacter sp. MS6 [6,27]. Figure 5. Quantification of thallusin in the growth medium normalized to the cell number for seven bacterial strains. The different letters indicate statistically significant differences in the thallusin concentration between the bacteria during a given growth phase (one-way ANOVA, Tukey post-hoc test, p < 0.05). The error bars represent the mean ± standard deviation (n = 3).

Quantification of Thallusin in Algal Cultures
Using our approach, we set out to detect thallusin in the chemosphere of Ulva in culture. A co-injection experiment allowed for the identification of the peak of the thallusin trimethyl ester by comparison with the synthetic standard ( Figure 6). The thallusin concentration in the pooled stationary culture of a closed cultivation system (U. compressa, three months old, 17.2 g L −1 fresh weight biomass) was 0.22 ± 0.04 nmol L −1 , implying that under the selected laboratory conditions, the actual thallusin concentration in the algal culture was approximately 30-times higher than the EC90 = 7.6 pmol L −1 [9], indicating strong oversaturation.

Discussion
Methylation of thallusin was promoted by potassium hydroxide in methyl sulfoxide and iodomethane to produce the corresponding methyl ester for quantifying thallusin in bacterial and algal cultures using an orbitrap mass spectrometer (Figure 1, Table 1). The validated method was shown to be accurate with an sr < 0.8% and a sufficient LOD. As of now, 0.75 L of the medium is required to quantify thallusin levels as low as the EC90 value in Ulva´s culture ( Figure 6). In fact, methylation by iodomethane has frequently been ap-

Discussion
Methylation of thallusin was promoted by potassium hydroxide in methyl sulfoxide and iodomethane to produce the corresponding methyl ester for quantifying thallusin in bacterial and algal cultures using an orbitrap mass spectrometer ( Figure 1, Table 1). The validated method was shown to be accurate with an s r < 0.8% and a sufficient LOD. As of now, 0.75 L of the medium is required to quantify thallusin levels as low as the EC 90 value in Ulva's culture ( Figure 6). In fact, methylation by iodomethane has frequently been applied in analytics with good yields [28], for example, in quantifying steroidal saponins from switchgrass [29]. Other standard esterification methods use diazomethane or methanol [16], but these were not pursued due to high reagent toxicity or required harsh reaction conditions. Previously, Matsuo et al. (2005) used only trimethylsilyl diazomethane for the derivatization of thallusin preparatively [10].
Thallusin induces the development of basal rhizoids and healthy cell wall formation [6,9] in U. compressa, but also possesses distinguishable context-dependent functions as known for hormones in Plantae [6,30]. For plant hormone profiling, our approach should be thus included in the workflow for the qualitative and quantitative analysis of phytohormones in marine bacteria [31,32] and algae [20,33,34]. Notably, already established methods for the purification and analysis of phytohormones from culture supernatants are particularly suitable for this purpose because they are based on solid phase extractions (SPEs) starting with an Oasis ® HLB-sorbent SPE [35][36][37]. We confirmed that thallusin can be extracted by this sorbent as well (data not shown). However, we preferred the C 18 SPE for thallusin analysis because it involved the extraction of fewer substances, resulting in a less structured matrix and less electrospray ion suppression than upon HLB SPE. Similarly, the bacterial growth medium HaHa_100 contributes to less background interference in the solid phase extraction, reducing ionization suppression and improving the analytical procedure and LOD. The maximum optical density of the bacterial cultures differed between the two artificial media; the growth of Maribacter sp. MS6 on the media led to a maximum optical density (OD 620 ) of 1.78 in marine broth medium but only 0.30 in HaHa_100 (Figure 3) comparable to previously published values [38].
Quantification of thallusin in growth media will help to address whether production is dependent on amino acids and less affected by carbon and nitrogen sources, as demonstrated by the indoleacetic acid (auxin) production of epiphytic bacteria on the red alga Pyropia yezoensis [39]. Land-plant-associated bacteria generally produce auxin, promoting cell division in root cells, similar to thallusin [40]. Ahmed [42]; these are still 10 3 -to 10 4 -times higher than the measured concentrations of thallusin and its EC 50 value (4.9 ± 0.1 pmol L −1 ) [9]. Therefore, the strong contrast between the high biological activity of thallusin and its comparatively low production yields is striking. For large-scale purification Maribacter sp., M. ulvicola, and Z. galactanivorans, the bacteria with the highest thallusin yields in this study and efficient growth performance in the HaHa_100 medium, are preferred if synthetic thallusin should be unavailable.
Because of its plant growth-promoting capabilities, the controlled use of algal biomass as biofertilizers may lessen the environmental impact of aquaculture's excessive use of chemical fertilizers [43]. Thallusin has the potential to make a significant contribution here. The monitoring and quantification of thallusin are, thus, crucial for a deeper understanding of algae growth in complex communities and in natura. It should also be instrumental in studying the dynamic exchange of thallusin precursors and the active substance between algae and bacteria. As thallusin contributes to algal growth promotion, it may be used as an effective non-gene disruptive strategy for sustainable algal biorefineries [44] and a modulator of production rates of secondary metabolites [45,46].

Chemicals, Standards, and Materials
Methanol (MeOH, HPLC grade) and potassium hydroxide (KOH, p.a.) used for sample preparation were purchased from VWR International (Darmstadt, Germany). Dimethyl sulfoxide (DMSO) and marine broth medium were purchased from Carl Roth GmbH + Co., KG (Karlsruhe, Germany). Iodomethane was obtained from Sigma-Aldrich (Steinheim, Germany). Water (H 2 O) for the SPE was purified using a MicroPure system (J.W.T. GmbH, Jena, Germany). Thallusin and a derivative, used as internal standard, were chemically synthesized [9] and dissolved in HPLC grade MeOH to form 2.0 mmol L −1 stock solutions. The internal standard is a monomethylated (−)-thallusin derivative with a switched double bond and one reduced carboxylic acid ( Figure 1A). The C 18 SPE cartridges were purchased from Waters Corporation (Milford, MA, USA). All the solvents used for the sample analysis were of UHPLC grade (CHEMSOLUTE ® ) and were purchased from Th. Geyer GmbH & Co., KG (Renningen, Germany). Formic acid (≥99%) was purchased from Thermo Fisher Scientific (Rockford, IL, USA).

Monitoring of Bacterial Growth
The bacteria were incubated in 200 mL of a HaHa_100 medium [50,52] or marine broth [51] on an orbital shaker (230 V Euro Plug, Standard 5000, VWR). Spectrophotometric measurements were taken every 2 h at a wavelength of 620 nm using a UV-Vis spectrophotometer (Genesys 10S UV-Vis Spectrophotometer, Thermo Scientific, Waltham, MA, USA). To compute cell density, a calibration against the optical density was performed. Nine dilution steps were carried out in triplicate for M. ulvicola (the bacterial culture with the highest optical density) at the end of the exponential growth phase and at the peak of its optical density. The diluted bacterial samples were first analyzed with the UV-Vis spectrophotometer and directly after that with a flow cytometer (BD Accuri C6, Heidelberg, Germany) through the measurement of the forward-scattered light (50 µL, 35 µL min −1 , threshold FSC-H 80,000) [53].

Sample Preparation
Samples were collected during three stages of bacterial growth to determine the thallusin concentration in the bacterial cultures. The growth medium was harvested at the peak of exponential growth (OD 620 = 0.15), at the early (OD 620 = 0.25) stationary phase, and a third sample was taken 24 h later during the late stationary phase. Subsequently, 50 mL aliquots of the cultures were centrifuged (9,500 rpm, 15 min, 10 • C) for sampling. The culture supernatants were then filtered sterile (pore size = 0.22 µm) and loaded onto a C 18 reversed-phase SPE cartridge (Waters Corporation, Milford, MA, USA).

Solid Phase Extraction
The C 18 cartridge was preconditioned with 5 mL MeOH and equilibrated with 10 mL MicroPure water. The culture supernatant was then loaded onto the cartridge. The C 18 cartridge matrix was washed with 10 mL MicroPure water and then with 4 mL of 25% (v/v) MeOH. Subsequently, 4 mL of 75% MeOH was used allowing thallusin to elute from the matrix. The solvent of the eluates was evaporated with nitrogen (Biotage-TurboVap ® LV, Uppsala, Sweden), and the eluates were either directly derivatized with iodomethane or dissolved in 100 µL of 75% MeOH. The analyte solution was filtered (PVDF, 0.22 m, 4 mm, Millex ® -GV), transferred to UHPLC-MS vials, and stored at −20 • C until measurement.

Derivatization of Thallusin
Thallusin was derivatized by permethylation by using iodomethane, according to the protocol of Avila-Zárraga and Martínez (2001) [26], with modifications. In brief, the dried residue of the eluates was dissolved in 150 µL of a DMSO/KOH solution (0.89 mol L −1 ), prepared by dissolving 1 g of powdered KOH (17.8 mmol) in 20 mL of DMSO with constant stirring for 30 min at 20 ± 1 • C. Iodomethane (30 µL) was then added to these sample solutions, and the mixtures were stirred for 2 h at 20 ± 1 • C. To inactivate the iodomethane, 50 µL formic acid was added to the derivatized sample, and stirring was continued for 1 h at 20 ± 1 • C. The sample mixtures were then diluted in 50 mL water and again extracted for clean-up using the C 18 SPE protocol (see Section 4.5).

Ultra-High-Performance Liquid Chromatography (UHPLC) Coupled with Electrospray Ionization (ESI) High-Resolution Mass Spectrometry (HRMS) Measurements
An UltiMate HPG-3400 RS binary pump was used to perform the measurements involving UHPLC coupled with high-resolution Orbitrap MS (Thermo Fisher Scientific, Germany). The Kinetex ® C−18 RP chromatography column (50 × 2.1 mm; 1.7 µm; 100 Å; Phenomenex, Aschaffenburg, Germany) was kept at 25 • C within the TCC-3200 column compartment. Eluent A contained 2% acetonitrile and 0.1% (v/v) formic acid in H 2 O. Neat Acetonitrile with 0.1% (v/v) formic acid was used as eluent B. The initial conditions were as follows: 0. Mass spectra were collected using a Thermo Fisher Scientific Q Exactive TM hybrid quadrupole-Orbitrap mass spectrometer coupled to a heated ESI source. For the analysis of thallusin, the internal standard, and the methyl ester of thallusin, targeted selected ion monitoring (tSIM) was used in the positive ionization mode with the following instrument The parameters of a simultaneous full scan were set to m/z = 100-1000; resolution 70,000; AGC target, 3 × 10 6 ; and maximum IT, 200 ms. Further, the sheath gas flow rate was set to 60; auxiliary gas flow rate, 20; sweep gas flow rate, 5; spray voltage, 3.0 kV; capillary temperature, 360 • C; S-lens RF level, 50; and the auxiliary gas heater temperature, 400 • C.

Calibration and Statistical Analysis
An external calibration curve with 5-7 calibration standards was generated in triplicate to measure the concentration of thallusin [54]. The linear regression model was calculated, plotted, and validated with SigmaPlot v.14.5 (Systat, Germany). The parameters for the calibration procedure (s x,o , s yx , s r , LOD, and LOQ) were determined according to DIN 32645. The lack-of-fit test proved linearity from the analysis of variance [55]. The test value F = 1.033 was smaller than the critical value F one-side = 2.958, with P = 95%, df LOF (lack of fit) = 5, and df PE (pure experimental error) = 14.

Conclusions
We developed a simple, quantitative LC-HRMS method for determining the algal growth-and morphogenesis-promoting factor thallusin. Due to its high affinity to iron, the derivatization of the carboxyl groups of thallusin is essential to obtain symmetrical peaks for the analyte and efficient chromatographic separation of the analyte. The use of iodomethane for derivatization is highly effective and operationally simple. To handle matrix effects, calibration was performed throughout the entire analytical procedure. Our advanced method for the quantitative analysis of the morphogen thallusin can, thus, be used to broadly monitor thallusin in the mutualistic relationship between Ulva and its associated bacteria in culture as well as in natura. It may thereby contribute to a better understanding of the maintenance of sustainable aquacultures. The method is currently being expanded toward detecting thallusin in relevant bacterial and algal culture settings and should also be useful for green tide monitoring. Chemically, our study suggests that traces of other carboxylic-acid-containing metabolites released into the water may be equally quantified by implementing a similar workflow.