Other Marine Natural Products

A large part of the marine chemical space sits outside the familiar ingredient families. These are compounds of a few hundred to roughly a thousand daltons, built as chemical defences by organisms that cannot move away from a threat, and they carry the fingerprints of that purpose: dense ring systems, several stereocentres in a row, halogen atoms in unexpected places and a natural abundance that rarely favours the isolation chemist. The field keeps expanding in a very visible way. Reviews of the annual literature counted 1,256 new marine compounds reported in 336 papers for 2024, against 1,220 in 340 papers for 2023 and 1,417 in 384 papers for 2022, and in 2024 alone 48 known marine structures were revised, which says something about how difficult these molecules are to pin down.

Eata Marisource works across four overlapping groups. First, polyketide and polyether scaffolds, including the phosphatase and proteasome probes that laboratories use as reference inhibitors. Second, macrolides, the large lactone rings that bind tubulin, protein kinase C and actin in ways nothing else does at the same concentration. Third, phenolic and halogenated metabolites, supplied as assigned single compounds rather than as broad extracts. Fourth, nucleoside analogues and the smaller heterocyclic metabolites, together with the reference materials that analytical teams need when they calibrate an instrument or validate a method. Material moves from grams of enriched fraction down to a milligram of an assigned solid, and the paperwork follows the same path.

Mapping the Territory

It helps to start from biosynthesis rather than from taxonomy, because the building programme determines nearly everything that follows, including how stable the compound is and how easily it separates. Polyketide synthases assemble carbon chains from acetate and malonate units and then fold them; the same machinery, decorated differently, produces both a linear inhibitor of a phosphatase and a rigid polyether built from fused and spiro rings. Macrolides come from the same assembly lines but close with a lactone bond, and the size of that ring turns out to be the most useful single descriptor of what the molecule will do. Phenolic metabolites in brown algae arise instead from polymerisation of phloroglucinol, a route that generates families of oligomers rather than one clean product. Nucleosides and small heterocycles are modified salvage intermediates, which is why they look so familiar and behave so unexpectedly.

Where the Material Comes From

The source organism decides the scale of the problem long before any chemistry begins. A bryozoan colony that grows as a feathery mat on a harbour piling puts its defensive chemistry together with a bacterial partner, and the compound can sit at a few parts per million of wet biomass. A sponge that filters hundreds of litres of seawater a day concentrates its metabolites into a fleshy body that is awkward to dry and easy to spoil. A dinoflagellate can be grown in a tank, but the yield depends on the strain, the light cycle and the day it is harvested, so two batches from the same culture collection can differ by a factor of several. Actinomycetes recovered from sediment grow happily in a fermenter once the medium contains the right sea salts, and that is the route that turned a compound found in deep Atlantic sediment into a material available in kilogram amounts.

Dried bryozoan colonies, one of the invertebrate groups that yield macrolide metabolites.

We keep material from several starting points in parallel, because no single one solves the supply question on its own. Wild-collected invertebrate biomass gives the widest chemistry but the smallest quantities. Aquaculture and mariculture give repeatable biomass of a few target species. Microbial fermentation gives scale where the producing organism has been identified and cultured. Semisynthesis and total synthesis remain the answer for compounds whose natural concentration makes isolation pointless, and for the labelled or modified variants that laboratories ask for later.

Polyketides: What the Assembly Line Produces

Because a polyketide synthase can be reprogrammed by swapping domains, the class covers a remarkable spread of behaviour. At one end sits okadaic acid, a polyether derived from a C38 fatty acid backbone, with the formula C44H68O13 and a molecular weight of 805.0. It is the standard tool for shutting down serine and threonine phosphatases in a controlled way: it inhibits protein phosphatase 1 with an IC50 of 10 to 15 nM, protein phosphatase 2A with an IC50 near 0.1 nM, and has only a marginal effect on the calcium-dependent phosphatase 2B at 5 micromolar. Laboratories order it precisely because that window is well documented. The solid is a glassy film, soluble in dimethyl sulfoxide, ethanol and methanol, and it melts between 171 and 175 degrees Celsius.

At the other end sits salinosporamide A, a compact bicyclic structure in which a gamma-lactam is fused to a strained beta-lactone. It was isolated from Salinispora tropica and Salinispora arenicola in ocean sediment, and its biosynthesis runs through a hybrid assembly line combining a polyketide synthase with a non-ribosomal peptide synthetase, starting from acetyl-CoA, chloroethylmalonyl-CoA and a cyclohexenylalanine unit. The chlorine substituent is not decoration: remove it and the potency collapses. The mechanism is covalent, with the lactone acylating the catalytic threonine of the proteasome, and the follow-on chloride displacement closes a tetrahydrofuran ring that keeps the enzyme blocked far longer than a simple acylation would.

Both examples illustrate a rule worth repeating to anyone planning an assay: the value of these compounds lies in specificity, not in raw potency, and specificity is measured against a panel rather than against a single target.

Macrolides: The Ring Size Decides Where It Binds

A macrolide is a large lactone ring, and the ring is not a passive scaffold. Bryostatin 1, a 20-membered macrolactone carrying three tetrahydropyran rings, binds the C1 domain of protein kinase C isoenzymes with dissociation constants between 0.8 and 3.0 nM, and it also engages Munc13-1, with a reported Ki of about 8 nM for the isolated C1 domain and under 1 nM for the full-length protein. Its supply history is the reason we keep a semi-synthetic route alongside isolation: thirteen tonnes of the bryozoan Bugula neritina yielded 18 grams of the compound, which is roughly one and a half parts per million of wet material. The producing partner is now understood to be a bacterial symbiont rather than the animal itself, and that knowledge is what makes alternative routes worth pursuing.

Ring size, and the position of the oxygens around it, also explain why laulimalide and peloruside A behave alike without sharing a skeleton. Laulimalide, formula C30H42O7 with a molecular weight of 514.65, and peloruside A, a 16-membered lactone with a pyranose ring, both stabilise microtubules from a site on beta-tubulin that is distinct from the taxoid pocket, and each is a competitive inhibitor of the other. The practical consequence for a laboratory is that compounds binding outside the taxoid site often retain activity in cell lines that expel taxanes, which is exactly why this pair keeps appearing in resistance work. Larger rings go further: swinholide A and misakinolide A, both from Theonella swinhoei, act on the actin cytoskeleton rather than on tubulin, while calyculin A from Discodermia calyx suppresses the same phosphatases as okadaic acid and is built from a pyrophosphate protoxin that a sponge symbiont releases only after it is triggered.

From Biomass to Crude Extract

The first step is almost always drying, because water competes for the solvent and enzymes in fresh tissue keep working while the material sits in a bucket. Freeze-drying followed by grinding gives a powder with a large surface area, and the powder is then extracted by maceration or percolation in methanol, in a methanol and dichloromethane mixture, or in hexane and ethanol. Filtration and evaporation under reduced pressure give a dark crude extract, and this is the point at which the material becomes useful for screening and awkward for everything else: it contains the target, its isomers, pigments, lipids and a good deal of material nobody has named yet.

A dark resinous crude extract after solvent removal, the starting point for chromatographic work.

Liquid-liquid partitioning then sorts the extract into broad polarity bands. Hexane removes the fats, ethyl acetate takes the mid-polarity metabolites where most polyketides and macrolides live, and n-butanol picks up the more polar glycosides and phenolic material, leaving sugars and salts in the aqueous layer. Where the compound of interest is known to be thermally fragile, the newer extraction routes earn their place. Pressurised liquid extraction keeps the sample out of light and air and uses far less solvent than a soaking step; microwave-assisted extraction shortens the time considerably; ultrasound-assisted extraction works at low temperature, which matters for thermolabile material; and supercritical carbon dioxide handles non-polar targets without leaving solvent residues. To give a sense of the spread, one brown alga extracted for fucoidan gave 5.2 percent by dry weight with a conventional soak, 6.17 percent with microwave assistance and 13.15 percent under pressurised liquid extraction, so the technique can matter more than the raw material.

Phenolics and Halogenated Metabolites as Assigned Compounds

Brown algal phenolics are usually discussed as a total content figure, which is fine for a bulk extract and useless for anyone testing a mechanism. We supply the defined members of the family instead. Phloroglucinol itself is the 126-dalton monomer; eckol, a trimer with a dibenzodioxin core, has the formula C18H12O9 and a molecular weight of 372.3; dieckol, a hexamer of the same type, has the formula C36H22O18 and a molecular weight of 742.5, and 8,8-bieckol sits alongside it with the two halves joined directly. More than 150 phlorotannins have been described, spread across six subgroups defined by how the rings are linked, and the linkage type rather than the ring count is what governs both the redox behaviour and the chromatographic behaviour.

Measurement deserves a word here, because the two routine colorimetric methods disagree in opposite directions. The Folin-Ciocalteu assay is calibrated against phloroglucinol and reports in phloroglucinol equivalents, but it responds to any reducing species present, so proteins, sugars and thiols inflate the number. The 2,4-dimethoxybenzaldehyde assay is far more selective for 1,3- and 1,3,5-substituted phenols, yet it under-reads oligomers whose reactive positions are blocked or branched. For an assigned compound neither assay is the right tool at all: quantitative nuclear magnetic resonance, integrating the aromatic region between roughly 7 and 5.5 parts per million, gives a mass fraction that does not depend on a colour response, and that is what we report for single compounds.

Halogenated metabolites need their own note. Bromine is common in sponge and algal chemistry, appearing in simple bromophenols, in brominated indole alkaloids and in polybrominated diphenyl ethers whose biosynthesis has been traced to the microbial community inside the host rather than to the animal. Because these compounds are lipophilic and bind readily to glassware and plastic, we supply them in solution where the project allows it, and we record the solvent and the concentration rather than leaving them to be inferred.

Concentration and Clean-Up Between Columns

Chromatography rarely finishes a purification on the first pass. A normal-phase run gives fractions that are mostly solvent, and those fractions have to be reduced before anyone can see whether the compound is in them. Nitrogen blow-down under a gentle stream at controlled temperature is the least damaging way to do that for a molecule with a fragile ring system, because it avoids the prolonged heating that a bath evaporation implies, and it can be stopped at any point. It is also the step where a lot of careful work is quietly lost, since a compound that concentrates onto the wall of a tube as an invisible film may not redissolve in the next solvent.

Nitrogen blow-down concentration of a set of column fractions before the next separation step.

Our routine between separation steps is deliberately conservative. Fractions are combined on the basis of thin-layer or liquid chromatography profiles rather than by fraction number, solvent is removed at the lowest workable temperature, and every residue is brought back into a known volume of a solvent that has been tested for it. Where a target is present at low concentration, the intermediate fraction is kept rather than pooled, because a compound that disappears during clean-up is far more expensive to recover than one that was never combined.

Turning Spectra into a Structure

Assigning a marine metabolite means answering three separate questions, and each one needs a different experiment. The first is connectivity: which atom is bonded to which. One-dimensional proton and carbon spectra give the functional groups, and two-dimensional experiments establish the framework, with correlation spectroscopy tracing the proton network, heteronuclear single-quantum correlation pairing each proton with the carbon it sits on, and heteronuclear multiple-bond correlation reaching across one, two and three bonds to tie the fragments together. The second question is composition, and high-resolution mass spectrometry answers it by giving the elemental formula directly, which is often what rules out the first proposed structure. The third question is stereochemistry, and this is where the class becomes difficult, because a compound with nine or ten stereocentres has thousands of possible arrangements that all fit the same connectivity.

Optical rotation measurement, used to compare an isolated compound against literature values.

We attack stereochemistry from several directions at once. Optical rotation at a defined wavelength, temperature and concentration is the oldest check and still the quickest, provided the literature value was measured under comparable conditions. X-ray diffraction on a suitable crystal settles the question absolutely, and it has been the arbiter in more than one reported revision of a marine structure. Chiral chromatography and nuclear magnetic resonance with chiral shift reagents extend the same logic to compounds that will not crystallise. Where a structure is genuinely ambiguous, we say so, describe the experiments that were run, and leave the assignment open rather than presenting a tidy answer that the data do not support.

Purity, Assignment and the Number You Actually Get

Chromatographic purity and mass purity are not the same quantity, and the gap between them is where most arguments about a marine standard begin. A peak area percentage from a diode-array detector tells you that the visible impurities are a small fraction of the total ultraviolet response; it does not tell you what fraction of the weighed mass is the compound, because water, residual solvent and inorganic material do not absorb at the wavelength you chose. Two routes are used to close that gap. Gravimetric preparation of a solution gives a known total mass but assumes the solid is dry. A quantitative nuclear magnetic resonance measurement with an internal standard gives an absolute mass fraction and does not care whether the solid is dry, and it is commonly combined with liquid chromatography and tandem mass spectrometry to confirm identity at the same time.

Melting point determination as a rapid identity check on a purified solid.

Certified marine reference solutions illustrate the level of detail that is achievable when a laboratory needs traceability. One widely distributed saxitoxin calibration solution carries an assigned concentration of 20.9 micrograms per gram with an expanded uncertainty of 1.6 micrograms per gram, a stated purity of at least 99 percent, and a delivery format of a measured mass of the toxin in an amber vial as a dilute hydrochloric acid solution at a stated pH. Every one of those numbers exists because someone downstream needs to know not just how much compound is in the vial, but how well that amount is known. We work to the same structure of reporting: identity, mass fraction, method used, and the uncertainty that goes with it.

Storage: Keeping a Difficult Molecule Difficult in the Right Way

Marine metabolites fail in characteristic ways, and most of them are avoidable. Polyethers and macrolides are generally comfortable as dry solids in the dark at minus twenty, but they suffer from repeated freeze-thaw cycles, so the working solution should be divided into single-use portions rather than returned to the freezer. Phenolics are the opposite problem: they are stable when dry and cold but oxidise readily in solution, particularly in alkaline conditions, and they will discolour a clear glass vial within days. Brominated compounds adsorb onto surfaces and should stay in amber glass throughout. Nucleosides are hygroscopic, and a vial that has been opened warm will take up water that then changes the mass fraction you paid for.

The most common mistake we see is not a storage error at all but an arithmetic one: a solution is prepared from a hygroscopic solid, weighed on a balance calibrated for a far heavier load, and then used to build a standard curve. We would rather answer a question about how a lot was handled than replace material that was lost to a preventable cycle, so the handling conditions travel with the shipment.

Quantities, Formats and Packaging

This group of products is ordered in small numbers and the packaging has to respect that. A milligram of an assigned solid goes into an amber glass vial with an insert, under an inert headspace, with the assignment sheet attached to the lot rather than filed separately. Enriched fractions travel in larger amber bottles, because a customer who intends to run a second separation wants volume rather than purity. Gravimetric solutions travel in sealed ampoules or in crimp-cap vials with a septum, so that the material can be sampled without opening the container to the atmosphere each time. Where a compound is light sensitive, the outer packaging carries the instruction as well as the label, since a vial in a drawer is invisible to whoever opens it six months later.

Powder weighing and filling of a purified compound into an amber bottle.

Larger quantities are available for compounds where the source route supports them, and we will say plainly which route applies. Fermentation-derived material scales in a predictable way; a compound isolated from a slow-growing invertebrate does not, and a request for a gram of something that exists at two parts per million should be discussed before it is ordered rather than after.

Custom Isolation and Contract Work

Much of what we send out began as a customer question rather than a catalogue line. The work usually falls into a few recognisable shapes. Activity-guided isolation starts from a crude extract and a described assay, and follows the response through the separation until a single compound carries it. Targeted isolation starts from structural information, a retention time, or a spectrum supplied by the customer, and follows that signature. Dereplication runs a crude extract against known compound data to establish what is already described, so that effort is not spent rediscovering a familiar molecule. Structural re-confirmation takes material a customer already holds and rebuilds the assignment from fresh spectra, which is sometimes all that stands between a stalled project and a decision.

We also accept specification-driven development, where the request is a set of defined compounds at a defined purity, and we are free to choose the source and the route. Where a target has been described in the literature but is not commercially available at the quantity needed, that is precisely the case worth a conversation, because the answer may be fermentation, a semi-synthetic step, or a larger collection campaign.

When a Compound Refuses to Behave

Nobody who has worked with this class of molecules expects a clean run. The difficulties repeat often enough to be worth listing, along with what we do about them, because a customer forewarned can avoid most of the cost. A compound that will not dissolve may not be a solubility problem at all: the residue may be sitting on the glass as an invisible film, and a change of solvent is a cheaper experiment than a new order. A peak that tailed beautifully on the analytical column and spreads on the preparative one is usually interacting with free silanol groups, and an acid-modified mobile phase often fixes it. A material whose measured activity does not match its stated purity is usually a case of the two different purities described earlier, and it needs an assignment rather than a re-run.

What Helps Us Answer Quickly

A first enquiry does not need to be formal. Four pieces of information usually turn it into a costed proposal the same day. State the compound or family you are after, as specifically as you can. Say what quantity you need and over what period you expect to use it, because that decides whether the answer is a catalogue item or a separation campaign. Describe the intended use at a technical level, since an assay that requires a particular solvent or a defined counter-ion changes how the material should be prepared and shipped. Finally, tell us what has already been tried, including materials that failed, because a negative result narrows the search faster than a positive one.

Between the catalogued compounds and the contract work, most of the requests we receive in this part of the marine chemical space can be met from material already in hand, or through a defined piece of isolation work that we can scope honestly at the outset.

For Research or Industrial Raw Materials, Not For Personal Medical Use!