Spinach is not just rich in flavonoids—it contains a surprisingly diverse chemical profile that older methods only partially captured. This study measures thirty-nine compounds in eleven and a half minutes, while processing forty-eight samples in about an hour. Spinach produces an array of unique flavonoids not commonly found in other fruits and vegetables.

These molecules likely serve as defense agents against biotic and abiotic stress and may have health beneficial properties for humans. The problem is that current methods to analyze spinach flavonoids are incomplete and only capture a portion of this uncharacterized pathway. A comprehensive analysis method is needed to determine how genetics, environmental conditions, and other factors influence spinach flavonoid biosynthesis.

Spinach produces a remarkable concentration of flavonoids, but reported values vary between eighty-two and two hundred twenty-four milligrams per one hundred grams fresh weight. That range remains unclear because many studies focus on only a few spinach varieties and a few common flavonoids.

Current extraction methods typically homogenize spinach, add polar solvents, centrifuge the samples, and quantify flavonoids from filtered supernatants. But current extraction methods tend to be time-consuming. More than forty flavonoid species have been identified in spinach, yet most published methods consider only five to ten, with the most comprehensive method quantifying up to seventeen.

Common product ions and accurate-mass collision-induced dissociation experiments make it possible to create a comprehensive analysis method that simultaneously quantifies most of the known flavonoids in spinach. The study developed a high-throughput extraction, processing forty-eight samples per hour, and a comprehensive quantification method measuring thirty-nine flavonoids in eleven and a half minutes using UHPLC-MS/MS.

MS/MS fragmentation experiments were performed on each analyte to report common and unique molecular signatures of spinach flavonoids. A panel of thirty genetically diverse spinach accessions was profiled with both a validated and a high-throughput extraction method to determine how estimates and population structures differed by methodology.

To maximize genetic variation, thirty spinach accessions previously phenotyped and genotyped were selected and grown. The selections reflected differences in carotenoid and mineral concentrations and genetic features that separated accessions into unique subgroups.

Plants were harvested when each plant within a pot had six to eight fully developed leaves, approximately thirty-five days after sowing. Tissue was then frozen and homogenized before extraction and analysis. For the validated extraction, five hundred milligrams of spinach, plus or minus ten milligrams, were spiked with naringin and naringenin internal standards.

The samples were extracted with five milliliters of ninety-eight to two methanol and formic acid, vortexed for one minute, sonicated for twenty minutes, and centrifuged for five minutes at three thousand seven hundred revolutions per minute and four degrees Celsius. The extraction process was repeated twice more, with sonication for only ten minutes in those repeats.

The high-throughput extraction used a Tecan Freedom EVO one hundred fifty and was adapted from a method validated for carotenoids and modified for water-soluble flavonoids. Samples of one hundred milligrams, plus or minus five milligrams, received naringin and naringenin internal standards, followed by two stainless steel balls in each sample tube.

After the internal standards equilibrated with the sample matrix, one milliliter of ninety-eight to two methanol and formic acid was added. Samples were extracted for forty-five seconds at one thousand four hundred revolutions per minute and centrifuged for three minutes at thirteen thousand revolutions per minute.

Spinach flavonoids were separated on a Waters C18 column installed in a ThermoFisher Vanquish Horizon UHPLC, interfaced with a ThermoFisher TSQ Altis triple quadrupole mass spectrometer. Each run lasted eleven and a half minutes, and the autosampler was maintained at twelve degrees Celsius to minimize sample oxidation.

Because glycosylated spinach flavonoids are not commercially available, quantification was relative to quercetin-3-glucoside using the linear portion of a nine-point standard curve. Jaceidin, patuletin, and spinacetin were quantified against authentic standards, while peak areas were normalized to taxifolin to correct for within-day and between-day instrument variability.

Previous mass spectrometry approaches revealed over twenty spinach flavonoid species and provided fragmentation information for putative identification when authentic standards were unavailable. Most flavonoids identified by Cho and colleagues generated either a three hundred thirty-three or three hundred forty-seven mass-to-charge ion in positive mode.

Combining findings from twenty years of method development produced a list of thirty-nine flavonoid molecules discovered in spinach, with varying degrees of information available about them. Table one lists the liquid chromatography tandem mass spectrometry multiple-reaction-monitoring settings used to quantify spinach flavonoids and standards.

For each compound, the authors report retention time, precursor and product mass-to-charge values, collision energy, dwell time, and RF lens voltage—for example, quercetin-three-O-glucoside has a retention time of four point five three minutes and precursor mass-to-charge value of four hundred sixty-three point zero nine.

These parameters define how the UHPLC triple-quadrupole instrument separates and detects the compounds, supporting reproducible quantification across the spinach accessions. Figure one maps representative negative-mode fragmentation pathways for three spinach flavonoids: patuletin in panel A, spinacetin in panel B, and jaceidin in panel C.

The arrows mark diagnostic changes, including proton loss, retro Diels–Alder cleavage, demethylation, carbon monoxide loss, and carbon dioxide loss, alongside the resulting formulas and mass-to-charge values. These patterns mattered because the authors used them to screen extracts for potential precursor molecules sharing the same structural motifs and to support confident identification.

The method sought five product ions per analyte, then retained the top two multiple-reaction-monitoring transitions with the highest signal-to-noise ratio for quantification. Fifteen-second windows were added on each side of a retention time to maximize the instrument’s duty cycle and allow twelve to fifteen points across a peak.

A spike addition experiment generated quantitative recovery estimates for the high-throughput method. Spinacetin and patuletin were excluded because these aglycones are rarely present at detectable levels in spinach. Naringin recovery was one hundred two point one plus or minus two point one percent, naringenin recovery was one hundred point five plus or minus four point seven percent, and jaceidin recovery was one hundred seven point eight plus or minus six point three percent.

These findings support that the high-throughput extraction works efficiently with analytes varying in polarity. Interday and intraday variability experiments quantified the combined variation of the high-throughput extraction and the analysis. One individual extracted six randomly selected samples and analyzed them with the UHPLC-MS/MS method, then repeated the experiment twice more to determine interday variation.

The findings indicate that the high-throughput extraction can reliably estimate analyte concentrations within and between days, with taxifolin helping compensate for day-to-day mass-spectrometer variation. With the high-throughput method, total flavonoid concentrations ranged from seventy-five point one to one hundred seventy point one milligrams per one hundred grams fresh weight.

Approximately seventy percent of total flavonoids was represented by six measured mass-to-charge signals. Using the traditional method, concentrations ranged from ninety-three point one to one hundred eighty-seven point two six milligrams per one hundred grams fresh weight.

The proportion of each flavonoid relative to total flavonoids was nearly identical between methods, indicating that differences were largely due to extraction efficiency. Given the limited environmental variability in the study, the concentration range primarily reflected genetic differences, although light and temperature could drastically alter spinach flavonoid concentrations.

Figure two uses principal components analysis to project the samples into two dimensions, with colors identifying varieties. Panel A shows traditional extraction, where principal component one explains twenty-nine point sixty-nine percent of variation and principal component two explains nineteen point ninety-four percent; panel B shows high-throughput extraction, with twenty-eight point nineteen and twenty point sixty-three percent, respectively.

The accession groupings are broadly visible in both panels, while replicate spacing and cluster separation illustrate how extraction method can affect the apparent population structure. Random-effects modeling partitioned variance associated with genetic background, extraction method, replication, and within-method variability.

In general, genetic background explained most variation, from forty point zero to ninety-seven point five percent, compared with extraction method, from zero to forty-nine point zero percent. Spinacetin was a notable exception: it was detectable only in the traditional extraction method, possibly because low pH and long sonication caused hydrolysis of glycosylated flavonoids.

The contribution of extraction method tended to increase inversely to analyte polarity, and analyte estimates were universally lower in high-throughput extracts for these later-eluting molecules. Figure three breaks each analyte’s total variance into genetics, replication, extraction method, replication within extraction, and residual error.

The stacked bars show that genetic background explains most variation for many analytes, ranging from forty point zero to ninety-seven point five percent, while extraction method accounts for zero to forty-nine point zero percent overall. This matters because it distinguishes biological differences among spinach samples from variation introduced by the extraction workflow, including the notable spinacetin signal detectable only with traditional extraction.

A key advantage of the high-throughput method is that forty-eight samples can be simultaneously extracted in roughly sixty minutes, using three milliliters of solvent per sample compared with fifteen milliliters per sample. That approach could profile large spinach populations, while the traditional approach could later be used on a subset requiring additional accuracy and precision.

The study presents a comprehensive UHPLC-MS/MS method for spinach flavonoids and an accompanying high-throughput extraction method. The method can separate and quantify thirty-nine flavonoids uniquely produced by spinach in eleven and a half minutes. A limitation is that the absence of authentic standards for spinach flavonoids makes it impossible to know the true ionization efficiency of each molecule.

Future studies need to establish ionization-efficiency ratios between quercetin-3-glucoside and authentic spinach standards to estimate concentrations more accurately. The high-throughput approach is suited to profiling a large spinach population, while the traditional approach can be used on a subset requiring additional accuracy and precision for more nonpolar flavonoids.

Given sporadic reports in the literature, the data represent benchmark values and ranges for many spinach flavonoids. These tools enable understudied analytes to be examined in plant biology, food science, and human health research. The study delivers a fast, broad spinach-flavonoid workflow that exposes substantial genetic variation, while showing that extraction efficiency still matters—especially for less polar compounds and for estimates based on limited authentic standards.