The result is a warning for drug developers: blocking a promising enzyme in a laboratory dish does not necessarily translate into effective anti-inflammatory activity in an animal. The FELL sequence, along with the related peptides TDIFELL and TDIFELLK, was investigated in earlier work and was described as being derived from the human calcium-binding protein spermatid-specific 1, or CABS1. Previous studies examined these peptides in a mouse model of lipopolysaccharide-induced neutrophilia and an ex vivo rat model of antigen-induced intestinal anaphylaxis.

In the mouse model, all three peptides reduced white-blood-cell accumulation in bronchoalveolar lavage fluid more than the control compound AIM-102. The investigators proposed that the shortest FELL sequence might be a core motif responsible for the observed activity, motivating further testing of engineered analogues. The current study focuses on structural variants of that tetrapeptide scaffold and asks whether simple chemical edits can improve or alter its anti-inflammatory profile.

The motivation is connected to cAMP, an intracellular messenger that helps regulate immune-cell activity. Enzymes called phosphodiesterases break down cyclic nucleotides, and PDE4B is especially abundant in monocytes, neutrophils and other cells involved in inflammation. By inhibiting PDE4B, a compound might prevent cAMP degradation, maintain higher intracellular cAMP levels and reduce the activation of inflammatory cells.

The therapeutic appeal is considerable, but the relationship between enzyme inhibition and whole-animal efficacy is not straightforward. Peptides can differ in stability, tissue distribution, cellular uptake and interactions with pathways other than PDE4B. The study therefore combined an in vitro enzyme assay with an animal swelling model to see whether changes in the FELL sequence affected both forms of activity.

For the current work, the investigators synthesized a family of FELL analogues through standard solid-phase peptide synthesis and prepared both C-terminal amide and free-acid forms. The compounds underwent confirmation by mass spectrometry and high-performance liquid chromatography, with a minimum chromatographic purity of 95 percent. Structural changes included replacing leucine with hydrophobic amino acids such as nor-leucine, valine or isoleucine at the third and fourth positions, extending the peptide chain, and introducing D-amino-acid stereoisomers at the N-terminus.

These modifications allowed the researchers to examine how side-chain chemistry, peptide length and molecular handedness affected PDE4B inhibition and anti-inflammatory activity. The peptide structures are designated by codes such as BB1, BB3 and BB4, with the parent amide and free-acid compounds corresponding to FELL-NH2 and FELL-OH. Anti-inflammatory effects were tested in six-week-old male Wistar rats using carrageenan-induced paw edema.

Carrageenan was injected into the right hind paw to trigger acute swelling, after which animals received either saline, diclofenac or one of the peptides. Experimental peptides were administered at 2 milligrams per kilogram of body weight, while the positive-control drug diclofenac was given at 25 milligrams per kilogram. Paw volume was recorded with a plethysmometer at two, three and four hours, and the researchers measuring swelling were blinded to treatment allocation.

Each of the ten groups contained eight animals. Edema was calculated as the increase in paw volume from baseline, expressed as a percentage, and reductions relative to the untreated control were treated as evidence of activity in this acute model. Diclofenac produced the strongest and most consistent suppression of swelling, with edema remaining between about 13 and 15 percent across the observation period, compared with more than 72 percent in untreated animals.

The peptide response was much less uniform. BB3, a valine-containing analogue, and BB4, an isoleucine-containing analogue, did not significantly outperform controls at two hours. At three hours, however, BB3 reduced edema to 41.28 percent and BB4 to 43.21 percent, with statistically significant p-values of 0.034 and 0.029.

Their effects weakened by four hours, when neither result remained significant, indicating a moderate and time-limited action. In the untreated group, edema declined from 82.40 percent at two hours to 72.61 percent at four hours, showing that some natural resolution occurred even without treatment. Several other analogues produced no detectable anti-edematous benefit.

These included a nor-leucine-substituted peptide, compounds carrying D-phenylalanine at the first position and additional hydrophobic variants. BB2, for example, showed edema values from 89.23 to 92.23 percent across the three measurements and did not differ significantly from controls. BB5, BB5′, BB7 and BB8 also maintained edema levels close to the untreated group, with no significant inhibitory activity.

The enzyme assay told a different story. At a concentration of 5 millimolar, BB5 and BB5′ achieved PDE4B inhibition of 67 and 62 percent, respectively, yet neither reduced paw swelling. The C-terminal amide parent BB1 inhibited PDE4B by 49 percent, and the free-acid parent BB11 by 58 percent.

BB3, despite producing a significant three-hour reduction in edema, showed only 26 percent enzyme inhibition. BB4 was intermediate, with 50 percent PDE4B inhibition and a measurable, though temporary, anti-inflammatory response. Roflumilast produced 75 percent inhibition of PDE4B at 5 millimolar, while the nonspecific phosphodiesterase inhibitor IBMX produced 38 percent.

Yet neither reference result determines how long an anti-inflammatory effect might last in an animal. The investigators suggest that hydrophobic side chains can improve interaction with PDE4B, while changes in peptide conformation, stability and local bioavailability may strengthen or weaken activity elsewhere. The discrepancy between enzyme inhibition and swelling outcomes indicates that PDE4B inhibition may contribute to the anti-inflammatory profile but is not sufficient on its own to predict it.

BB3 and BB4 could also act through non-PDE4 inflammatory pathways, although that possibility was not directly tested in the reported experiments. The study has clear limits. PDE4B inhibition was measured at only one concentration, so the researchers could not calculate half-maximal inhibitory concentrations or compare potency quantitatively.

Peptide stability, pharmacokinetics and tissue distribution were not measured, even though enzymatic degradation could explain why some compounds lost activity in vivo. The animal observations covered only the first two to four hours after carrageenan, leaving later and more prolonged effects unknown. The experiment also involved one acute model, male rats and a single peptide dose, so it cannot establish effectiveness in human disease or in chronic inflammatory conditions.

The findings are therefore a structure–activity signal from a preclinical model, not evidence of a clinical therapy. Despite those limitations, the results offer a practical lesson for peptide-drug research. The most active variants, BB3 and BB4, contain valine or isoleucine in place of leucine at positions three and four, but their effect appears brief.

By contrast, the original FELL structures, BB1 and BB11, retain moderate PDE4B inhibition and a more balanced overall pharmacological profile, including previously reported analgesic activity. The authors therefore identify the parent compounds as favorable starting points and BB3 and BB4 as useful leads for further study. Follow-up work will need cytokine profiling, dose-response testing, stability measurements, pharmacokinetic characterization and evaluation in additional inflammatory models before these tetrapeptides can be considered therapeutic candidates.

Subject of Research: Structure–activity testing of FELL tetrapeptide analogues for PDE4B inhibition and anti-inflammatory effects

Article Title: Design, synthesis, and anti-inflammatory activity of analogs of tetrapeptide Phe-Glu-Leu-Leu

Article References: Borisova, B., Nemska, V., Zlatanova-Tenisheva, H., Laronze-Cochard, M., Gérard, S., Romanova, P., Moulas, A. N., & Danalev, D. (2026). Design, synthesis, and anti-inflammatory activity of analogs of tetrapeptide Phe-Glu-Leu-Leu. Pharmacia, 73, Article e198051. https://doi.org/10.3897/pharmacia.73.e198051

Image Credits: AI Generated

DOI: 10.3897/pharmacia.73.e198051

Keywords: FELL tetrapeptide, anti-inflammatory peptides, PDE4B inhibition, cAMP signaling, carrageenan paw edema, structure–activity relationship, peptide therapeutics, bioactive peptides, rat inflammation model, peptide synthesis, Design, synthesis

Bioactive peptides are short amino acid sequences that lie dormant within the parent proteins of everyday foods. They do nothing while locked inside the protein chain, but once the protein is broken down by digestive enzymes, laboratory hydrolysis, or microbial fermentation, these fragments are released and can exert measurable physiological effects, from scavenging free radicals to lowering blood pressure and modulating the immune system. Whey, the liquid left over from cheese production, is a particularly attractive source because it makes up 15 to 20 percent of milk protein and is packed with essential amino acids such as leucine and cysteine, the latter serving as a building block of the antioxidant glutathione.

There is a catch, however. Whey proteins are highly hydrophobic and notoriously resistant to enzymatic hydrolysis, meaning that conventional commercial enzymes often fail to cut them at the right places to liberate their embedded bioactive sequences. The research team at Dankook University, led by Cheolhyun Kim, turned to an alternative strategy: fermentation with lactic acid bacteria strains that carry their own powerful proteolytic enzyme systems. Because these bacteria evolved to digest milk proteins as a nitrogen source, their enzyme cocktails can slice whey proteins in ways that commercial enzymes cannot, generating novel peptides at low production cost.

The researchers isolated ten bacterial strains from homemade dongchimi samples collected across South Korea. Using 16S rRNA gene sequencing, they identified eight strains of Leuconostoc mesenteroides, one strain of Pediococcus pentosaceus, and one strain of Latilactobacillus sakei, which they designated DC10. Dongchimi-derived strains were deliberately chosen because kimchi bacteria used in dairy fermentation can sometimes impart undesirable flavors, so the team pre-screened for strains with both probiotic potential and technological compatibility with dairy products. All ten isolates proved Gram-positive and survived artificial gastric juice at pH values between 1.5 and 3.0 as well as 24 hours of exposure to 0.3 percent oxgall bile, a robustness profile considered essential for probiotic applications.

When the strains were tested for proteolytic activity on skim milk agar, L. sakei DC10 stood out, producing a clear digestion zone wider than 23 millimeters, the largest among all ten isolates. The team then inoculated DC10, along with two L. mesenteroides strains, into a 10 percent whey protein concentrate solution and fermented it at 37 degrees Celsius for 20 hours. Using the TNBS colorimetric method to track free amino groups, they observed that the degree of hydrolysis climbed rapidly during the first four hours and plateaued after about 16 hours, confirming that the bacteria were actively cleaving peptide bonds throughout the fermentation.

Sodium dodecyl sulfate polyacrylamide gel electrophoresis provided a visual confirmation of the protein breakdown. Bands corresponding to the major whey components, including lactoferrin, bovine serum albumin, immunoglobulins, beta-lactoglobulin, and alpha-lactalbumin, appeared between 14 and 100 kilodaltons in the unfermented sample. After 16 hours of fermentation with DC10, the higher molecular weight proteins had largely disappeared, degraded into smaller fragments. The fermented solution was then passed through a preparative gel filtration column to separate the resulting peptides into distinct fractions by molecular weight.

Fraction 3 emerged as the star of the study. In calcium solubilization assays, it achieved roughly 85 percent calcium solubility, below the 93.67 percent of casein phosphopeptides, the industry-standard calcium carrier, but well above unhydrolyzed whey protein. Because casein phosphopeptides are relatively expensive, the researchers argue that whey-derived calcium-binding peptides could serve as a cost-effective alternative for calcium-fortified foods and nutraceuticals. Peptides typically chelate calcium through electron pairs on their amino and carboxyl groups, with side chains of leucine, aspartic acid, and glutamic acid stabilizing the complex, and amino acid profiling of Fraction 3 revealed high levels of branched-chain amino acids, which previous work links to enhanced calcium absorption and vitamin D receptor expression.

The antioxidant results were equally striking. In the ABTS radical scavenging assay, most fractions outperformed non-fermented whey protein, with Fraction 3 posting the highest value. Fractions 2 and 3 matched casein phosphopeptides in DPPH radical scavenging, and Fraction 3 actually exceeded the commercial standard in the FRAP iron-reduction assay, which measures the capacity to donate electrons and break free radical chain reactions. Mass spectrometry of Fraction 3 identified four fully sequenced peptides with molecular masses between 658 and 905 daltons, including TVQVTSTAV and SLVYPFPGPIHNSLPQ, sequences previously reported in the literature as antioxidative, antimicrobial, and immunomodulatory.

To probe anti-inflammatory activity, the team treated RAW 264.7 mouse macrophage cells with the peptide fractions and then provoked inflammation with lipopolysaccharide. MTT assays confirmed that all fractions preserved more than 80 percent cell viability, ruling out cytotoxicity as a confounding factor. The fractions inhibited nitric oxide production in a concentration-dependent manner, and enzyme-linked immunosorbent assays showed significant suppression of the pro-inflammatory cytokines IL-1 alpha, IL-6, and TNF-alpha compared with LPS-stimulated controls. Because TNF-alpha drives the activation of other inflammatory mediators, damping its expression may cascade into a broader reduction of the inflammatory response.

The authors caution that their findings rest on in vitro experiments, and they call for future studies to clarify the underlying molecular mechanisms, assess bioavailability in living organisms, and explore clinical applications. Even so, the work illustrates an elegant convergence of traditional food heritage and modern protein chemistry: bacteria from a centuries-old Korean radish ferment, selected for dairy compatibility, unlocking health-promoting fragments from a cheese-industry byproduct. If subsequent animal and human studies bear out the calcium, antioxidant, and immune benefits observed in the laboratory, dongchimi-derived fermentation could offer the food industry a low-cost route to multifunctional whey ingredients that simultaneously support bone health, oxidative balance, and immune regulation.

Subject of Research: Bioactive peptide production from whey protein fermented with lactic acid bacteria isolated from Korean dongchimi

Article Title: Functional characteristics of peptides from whey proteins fermented with lactic acid bacteria isolated from Dongchimi

Article References: Yoo, J., Lee, S., Song, S., & Kim, C. (2026). Functional characteristics of peptides from whey proteins fermented with lactic acid bacteria isolated from Dongchimi. Food Science of Animal Resources, 46(1), Article 50. https://doi.org/10.1007/s44463-025-00053-6

Image Credits: AI Generated

DOI: 10.1007/s44463-025-00053-6

Keywords: bioactive peptides, whey protein, lactic acid bacteria, dongchimi, fermentation, antioxidant activity, anti-inflammatory, calcium solubilization, Latilactobacillus sakei, functional foods, probiotics, food science