The team, led by corresponding author Dinesh Upadhya of the Department of Anatomy, used human pluripotent stem cells, or PSCs, as their starting material. These cells were differentiated into neural stem cells and then into neurons, creating a controlled developmental trajectory that mirrors the earliest chapters of human brain formation. At each stage, the researchers isolated both whole-cell lipid extracts and purified mitochondrial fractions, verifying their purity with marker proteins such as ATP5A1 for mitochondria, GRP94 for the endoplasmic reticulum, and CD63 for other membranes. They then employed liquid chromatography coupled with tandem mass spectrometry, a sensitive lipidomics workflow, to quantify hundreds of individual lipid species across the glycerolipid and glycerophospholipid families.

The first broad observation was a gradual increase in relative lipid content as cells progressed from pluripotency toward neuronal identity, a trend seen in parallel at both the cellular and mitochondrial levels. Remarkably, the ratio of unsaturated to total lipids remained relatively stable across these transitions, suggesting that developing cells maintain a consistent degree of membrane fluidity even as they accumulate more lipid material. This constancy of unsaturation may reflect an underlying homeostatic program that protects membrane properties during the intense structural reorganization that accompanies neural differentiation.

Beneath this apparent calm, however, the cellular lipidome was anything but static. Relative levels of diacylglycerols, short-lived signaling lipids best known for activating protein kinase C pathways, rose steadily and significantly as PSCs became neural stem cells and then neurons. At the same time, relative levels of triacylglycerols, the cells’ storage fats, declined over the same trajectory. This reciprocal shift implies that developing neural cells actively break down their energy reserves and convert them into signaling-competent intermediates, possibly to fuel the demanding processes of proliferation, migration, and neurite extension that define neurodevelopment.

Among the glycerophospholipids, one change stood out with particular force: phosphatidylcholine, the most abundant phospholipid in most mammalian membranes, showed a drastic increase in relative abundance in mature neurons compared with pluripotent stem cells and neural stem cells. This enrichment likely supports the enormous expansion of membrane surface area required to build axons and dendrites, as well as the specialized electrical properties that neurons must acquire. In parallel, the ratio of phosphatidylcholine to phosphatidylethanolamine, two structural phospholipids whose balance influences membrane curvature and integrity, increased in both whole cells and mitochondria across the developmental sequence.

Yet when the researchers stepped back to compare the major lipid classes as a whole, a clear contrast emerged. The cellular lipid composition shifted significantly at every transition, while mitochondria maintained a relatively stable distribution of their principal lipids, including the glycerolipids and glycerophospholipids that form the inner and outer mitochondrial membranes. This stability makes functional sense: mitochondria depend on a tightly controlled lipid environment, particularly cardiolipin-rich inner membranes, to sustain oxidative phosphorylation, calcium handling, and the fission and fusion dynamics that developing cells rely on. Even as the cell around them reinvents itself, mitochondria appear to preserve their core lipid architecture.

Stability at the class level, however, did not mean molecular stagnation. More granular, species-level analyses of the mitochondrial lipidome revealed significant remodeling among individual lipid species, with differentially abundant molecular forms identified between PSCs, neural stem cells, and neurons through fold-change and univariate statistical comparisons. In other words, mitochondria fine-tune the precise fatty-acyl chains attached to their lipids while keeping the overall lipid class proportions constant, a strategy that could adjust membrane biophysics without disrupting core mitochondrial functions.

The study fills a genuine gap in developmental biology. Although mitochondria are known to be central players in neurodevelopment, powering the proliferation, differentiation, migration, and maturation of neural cells, the dynamics of their lipid composition during these critical windows had remained unexplored. Lipids are not passive structural fillers; they regulate membrane enzymes, transporters, signaling cascades, and the physical properties that determine whether membranes bend, fuse, or hold firm. Documenting their normal trajectory provides an essential reference point for understanding when development goes awry.

The clinical implications of this baseline dataset are considerable. Because many neurodevelopmental disorders, from autism spectrum conditions to intellectual disability syndromes, are now studied using patient-derived induced pluripotent stem cells, the lipid profiles established here offer a benchmark against which disease-associated lipid perturbations can be measured. If a patient’s neural lineage shows abnormal diacylglycerol accumulation, distorted phosphatidylcholine levels, or destabilized mitochondrial lipid proportions, researchers will now have the normal developmental template to compare against, potentially revealing metabolic vulnerabilities that underlie neurological disease.

The work also raises intriguing questions for future research. What mechanisms anchor the mitochondrial lipidome in place during development, and how do mitochondria communicate their lipid needs to the endoplasmic reticulum, the major site of phospholipid synthesis? Do specific lipid species emerging during the stem-cell-to-neuron transition serve as developmental signals in their own right? As lipidomics technology becomes faster and more accessible, the mitochondrial lipid landscape of the developing human brain, long an invisible dimension of neurobiology, is finally coming into focus, one fatty acid chain at a time.

Subject of Research: Lipidomic changes in cells and mitochondria during early human neurodevelopment

Article Title: Mitochondrial lipidome exhibits relatively higher stability than cell lipidome during early neurodevelopment

Article References: Hegde, S., Chandran, D., Tripathy, D. K., Ramesh, A. I., Prabhu, K., & Upadhya, D. (2026). Mitochondrial lipidome exhibits relatively higher stability than cell lipidome during early neurodevelopment. BMC Biology. https://doi.org/10.1186/s12915-026-02726-0

Image Credits: AI Generated

DOI: 10.1186/s12915-026-02726-0

Keywords: mitochondria, lipidomics, neurodevelopment, neural stem cells, pluripotent stem cells, phosphatidylcholine, diacylglycerols, triacylglycerols, neurons, BMC Biology, Mitochondrial, lipidome

A comprehensive review published in Discover Biotechnology by Ravi Narayan Venkatachalam of Jeneil Biotech Inc. synthesizes the rapidly growing field of microbial biosynthesis as an alternative route to human milk fat substitutes (HMFS). Conventional HMFS such as the commercial products Betapol and InFat are made by enzymatic interesterification of plant oils using sn-1,3-specific lipases. These processes work, and sn-2 palmitate-enriched formulas now command a substantial share of the premium infant nutrition market, but they are expensive, heavily dependent on palm oil feedstocks, and prone to acyl migration, a side reaction that scrambles the carefully engineered fatty acid positions. Microbial fermentation, the review argues, could bypass these limitations by building the desired molecular architecture from scratch inside living cells.

The standout organism in the field is Rhodococcus opacus PD630, a soil bacterium that naturally accumulates large quantities of triacylglycerols. When fed mixtures of fatty acid ethyl esters, including ethyl palmitate, ethyl oleate and ethyl linoleate, the bacterium produced triacylglycerols closely resembling beta-OPL, a key human milk fat species, with palmitic acid concentrated at sn-2 and unsaturated fatty acids at the outer positions. Analytical verification using gas chromatography with flame ionization detection, pancreatic lipase digestion, carbon-13 nuclear magnetic resonance spectroscopy and ultra-performance liquid chromatography-mass spectrometry showed sn-2 palmitate enrichment exceeding 80 percent, and carbon-13 NMR confirmed more than 85 percent esterification of palmitate at sn-2 in the beta-OPL fraction. The bacterium also proved flexible with cheaper substrates: glycerol favored high triacylglycerol content with elevated palmitate, while soybean oil pushed the profile toward oleic and linoleic acids.

Notably, the Rhodococcus approach requires no genetic modification, relying instead on substrate engineering and fermentation conditions. However, the species is not currently listed as GRAS, or Generally Recognized as Safe, by regulatory agencies such as the FDA or EFSA, so its products would need rigorous purification to remove microbial biomass, or the pathways would need to be transferred into an approved host. That is where Corynebacterium glutamicum enters the picture. Long valued industrially for amino acid production and already GRAS-certified, C. glutamicum has been metabolically engineered to synthesize triacylglycerols de novo. Researchers introduced a complete biosynthesis pathway with heterologous diacylglycerol acyltransferases, a phosphatidic acid phosphatase and a lipid body assembly factor, while boosting precursor supply through a thioesterase and an acyl-CoA synthetase, deleting four cellular lipases and a diacylglycerol kinase to prevent lipid degradation, removing a transcriptional repressor of fatty acid synthesis, and disrupting organic acid byproduct pathways. The optimized strain produced roughly 2.38 grams per liter of intracellular fatty acids, about 17.8 percent of dry cell weight, dominated by palmitic and oleic acids, the two signature fatty acids of human milk.

The positional assembly of these molecules depends on three coordinated acyltransferases. Glycerol-3-phosphate acyltransferase initiates triacylglycerol construction at sn-1, lysophosphatidic acid acyltransferase, or LPAT, determines what enters sn-2, and diacylglycerol acyltransferase completes the molecule at sn-3. Because LPAT largely dictates the stereospecific identity of the lipid, most engineering efforts have targeted this enzyme, but the review highlights emerging evidence that co-modulating all three enzymes gives far greater control, allowing unsaturated fatty acids to be steered to the outer positions while palmitate is locked into sn-2, effectively reconstructing the USU architecture biosynthetically rather than through post-synthetic enzymatic rearrangement.

Yeasts add another layer of promise because of their GRAS status and industrial track record. Engineered strains of Yarrowia lipolytica expressing plant LPATs with sn-2 palmitate specificity, sourced from organisms such as Brassica napus and paired with thioesterases that enlarge the palmitoyl-ACP pool, achieved sn-2 palmitate enrichment of 20 to 40 percent depending on the enzyme and culture conditions. Under nitrogen-limited media, total lipid content rose and palmitate levels climbed, producing triacylglycerols that increasingly resembled the OPO and PPO species characteristic of human milk. In parallel, the non-conventional oleaginous yeast Trichosporon cutaneum accumulated more than 50 percent lipid per dry cell weight on glucose or xylose, with palmitic acid making up roughly 25 to 30 percent of total fatty acids, a profile well suited to further structured lipid engineering even though its stereospecific positioning has not yet been characterized.

Microalgae contribute a different asset: native production of long-chain polyunsaturated fatty acids such as EPA and DHA, which are crucial for infant brain and eye development. Oils from Nannochloropsis oculata and Isochrysis galbana were subjected to acidolysis with palmitic acid using the 1,3-regioselective immobilized lipase Lipozyme RM IM, repositioning palmitate while preserving the native polyunsaturates. Optimized conditions of 60 degrees Celsius for six hours yielded sn-2 palmitate enrichment of 21.3 percent for Nannochloropsis and 24.6 percent for Isochrysis. Pushing the concept further, a microfluidic reactor packed with immobilized Lipozyme RM IM processed DHA-rich oil from Schizochytrium combined with tripalmitin in continuous flow, where superior heat and mass transfer maintained enzyme activity and suppressed acyl migration. At 50 degrees Celsius and a 1:2.5 tripalmitin-to-oil ratio, the system retained more than 70 percent of palmitate at sn-2, incorporated over 55 percent DHA at the outer positions, and delivered up to 38 percent yield of human-milk-style structured triacylglycerols resembling OPO-DHA and PPO-DHA species.

Verifying that these microbial lipids truly replicate human milk fat demands a layered analytical workflow. GC-FID quantifies the overall fatty acid profile but reveals nothing about positional distribution. Pancreatic lipase digestion followed by chromatography measures sn-2 enrichment, UPLC-MS resolves intact triacylglycerol molecular species such as OPO, OPL and PPO, and carbon-13 NMR confirms positional assignment without digestion artifacts. Emerging tools such as MALDI-MS for rapid triacylglycerol fingerprinting and FTIR spectroscopy for real-time fermentation monitoring could eventually streamline quality control, though they are not yet standard and must earn regulatory acceptance before displacing the gold-standard techniques.

The commercial stakes are considerable. The global infant formula market exceeds 55 billion dollars, and the HMFS segment, valued at 1.42 billion dollars in 2024, is projected to grow at a compound annual rate of 8.3 percent to more than 2.9 billion dollars by 2033. Microbial production aligns squarely with sustainability goals, using renewable feedstocks such as glycerol, glucose and agricultural byproducts, reducing dependence on palm oil, and offering the land, water and geographic advantages of closed bioreactor systems over oilseed agriculture. Significant obstacles remain, however: sn-2 specificity is not universal across strains, engineered plant LPATs often express poorly in microbial hosts, titers below 1 to 2 grams per liter remain far from cost-competitive even though engineered Yarrowia strains have demonstrated bulk lipid titers above 66 grams per liter, and regulatory definitions of natural, identical or bioequivalent lipids vary across jurisdictions while consumer perceptions of engineered organisms loom over infant products. The review points toward future advances in extremophile and metagenomic acyltransferase discovery, modular enzyme-swapping platforms, machine learning-guided strain design, and hybrid processes combining microbial biosynthesis with lipase-catalyzed finishing, arguing that interdisciplinary integration will determine whether microbially produced human milk fat substitutes move from laboratory proof-of-concept to the next generation of infant formula.

Beyond the technical achievements, the physiological rationale for sn-2 palmitate deserves emphasis. Infant digestive physiology differs markedly from that of adults: pancreatic lipase output is immature, and bile salt concentrations are low, making the neonatal gut unusually sensitive to the form in which fatty acids arrive. Because 2-monoacylglycerols are absorbed efficiently through intestinal enterocytes and re-esterified into chylomicron triacylglycerols, palmitate delivered at sn-2 effectively piggybacks on an absorption pathway that remains functional even when free fatty acid uptake is compromised. This explains why calcium soap formation, and the associated constipation, bone mineral loss and reduced calcium retention seen with conventional formula fats, is largely avoided when the USU configuration is preserved.

The review’s framing of microbial production as a sustainability strategy also merits context. Palm oil cultivation, the dominant feedstock for current HMFS synthesis, is associated with deforestation and biodiversity loss in tropical regions, and supply chains remain vulnerable to price volatility. Fermentation, by contrast, is decoupled from arable land and climate, and oleaginous microbes can valorize industrial byproducts such as crude glycerol from biodiesel manufacture. Whether these advantages translate into commercial viability will depend on closing the yield gap between laboratory strains and the multi-gram-per-liter titers that industrial lipid fermentation already achieves for other products, a benchmark the review identifies as the decisive next milestone.

Subject of Research: Microbial biosynthesis of human milk fat substitutes with sn-2 palmitate-enriched triacylglycerols for infant nutrition

Article Title: Microbial production of human milk fat substitutes: a review

Article References: Venkatachalam, R. N. (2026). Microbial production of human milk fat substitutes: a review. Discover Biotechnology, 3(1), Article 6. https://doi.org/10.1007/s44340-026-00054-1

Image Credits: AI Generated

DOI: 10.1007/s44340-026-00054-1

Keywords: human milk fat substitutes, microbial biosynthesis, triacylglycerols, sn-2 palmitate, Rhodococcus opacus, Yarrowia lipolytica, Corynebacterium glutamicum, microalgae, infant formula, metabolic engineering, structured lipids, sustainability