The strawberry-derived permeation enhancer pelargonidin enables oral protein delivery

Abstract

Although patients generally prefer oral drug delivery to injections, low permeability of the gastrointestinal tract makes this method impossible for most biomacromolecules. One potential solution is codelivery of macromolecules, including therapeutic proteins or nucleic acids, with intestinal permeation enhancers; however, enhancer use has been limited clinically by modest efficacy and toxicity concerns surrounding long-term administration. Here, we hypothesized that plant-based foods, which are well tolerated by the gastrointestinal tract, may contain compounds that enable oral macromolecular absorption without causing adverse effects. Upon testing more than 100 fruits, vegetables, and herbs, we identified strawberry and its red pigment, pelargonidin, as potent, well-tolerated enhancers of intestinal permeability. In mice, an oral capsule formulation comprising pelargonidin and a 1 U/kg dose of insulin reduced blood glucose levels for over 4 h, with bioactivity exceeding 100% relative to subcutaneous injection. Effects were reversible within 2 h and associated with actin and tight junction rearrangement. Furthermore, daily dosing of mice with pelargonidin for 1 mo resulted in no detectable side effects, including weight loss, tissue damage, or inflammatory responses. These data suggest that pelargonidin is an exceptionally effective enhancer of oral protein uptake that may be safe for routine pharmaceutical use.

Keywords: oral drug delivery; permeation enhancers; protein therapeutics.

Fig. 1.

Plant-derived foods are a novel source of nontoxic permeation enhancers that enable oral protein delivery. (A) A food-derived permeation enhancer is incorporated into a capsule containing a powdered macromolecular drug (e.g., insulin) for oral administration. The capsule is coated with a pH-sensitive polymer (enteric coating) that delays release of the enhancer and drug until the capsule has localized to the small intestine. There, the enhancer opens the tight junctions between intestinal cells, enabling the protein drug to enter circulation. (B) Food extracts screened at 15 mg/mL on Caco-2 cells exhibited a variety of behaviors, and well-tolerated food extracts differed in their effect on TEER (y axis), a surrogate measurement for intestinal permeability. Lower TEER typically corresponds to higher drug permeability. (C) The most effective TEER-reducing extracts increased the permeability of the model drug, calcein, across Caco-2 intestinal monolayers (y axis). Strawberry was chosen as our top candidate because it enabled full recovery of TEER values within 24 h of treatment removal (x axis). Error bars display SEM (n = 3 replicate wells for TEER and calcein permeability, n = 8 replicate wells for Presto viability).

Fig. 2.

Polyphenolic compound(s) in strawberry enhance intestinal permeability. (A) Only the red varieties of several foods enhanced the permeability of Caco-2 monolayers. (B) White Carolina Pineberry strawberries were grown for this study. (C) White strawberries, which lack polyphenolic pigments, were not effective permeation enhancers. (D) Polyphenols were extracted from strawberry via adsorption to Amberlite resin and a sequence of washing steps. (E) Polyphenolic compounds isolated from strawberries enhanced intestinal permeability by TEER and (F) by calcein permeability to the same extent as crude strawberry extract at one-third of the dose. (G) Treatment with strawberry polyphenols doubled the uptake of orally administered 4-kDa dextran (FITC-DX4) and (H) 40-kDa dextran (FITC-DX40) in mice. (I) One unit per kilogram of insulin was delivered by intestinal injection following oral delivery of either saline (control) or strawberry polyphenols and compared to subcutaneous injection of the same dose. Strawberry polyphenols induced sustained reductions in blood glucose levels. Integrated areas above the curves from H demonstrate that strawberry polyphenols enabled significantly better insulin bioactivity than subcutaneous injection. Error bars display SEM (n = 3 replicate wells for A, C, and E, n = 6 mice for panels F and G, and n = 5 mice for H and I). *P < 0.05 with respect to control by two-tailed t test with Welch’s correction.

Fig. 3.

Chromatographic separation of strawberry polyphenols identified pelargonidin as the primary, active permeation enhancer. (A) Of 22 fractions that resulted from MPLC separations of strawberry polyphenols, only the ε3 fraction enhanced the permeability of Caco-2 monolayers. “+” denotes fractions that were too small to examine at the otherwise standardized concentration of 1 mg/mL. (B) Similarly, of a large group of commercially purchased phenolic compounds known to occur in strawberries, only the pigment molecule pelargonidin significantly increased the permeability of calcein across cell monolayers. (C) Chemical structure of pelargonidin. (D) Pelargonidin permeabilized intestinal Caco-2 cells, as measured by TEER, and (E) improved calcein permeability in a dose-dependent manner. Error bars represent SEM (n = 3 or 4 replicate wells).

Fig. 4.

Pelargonidin, the active component of strawberry, enables oral delivery of functional proteins in mice. (A) Kinetic experiments of pelargonidin permeation enhancement in mice showed that intestinal permeability to 4-kDa FITC-dextran (FITC-DX4) peaks 1 h after treatment and returns to baseline within another hour. (B) Pelargonidin treatment improved uptake of 40-kDa dextran but increased permeability tapered off for larger dextrans. (C) The efficacy of intestinally injected insulin at a dose of 1 U/kg was dependent on the dose of oral pelargonidin pretreatment, with (D) higher pelargonidin doses leading to higher bioactivity of the insulin. (E) Oral insulin doses of 1 U/kg (maroon) and 5 U/kg (red) reduced blood sugar in healthy mice when administered with pelargonidin in capsules. (F) Pelargonidin–insulin capsules resulted in double the bioactivity of subcutaneously injected insulin for 1 U/kg oral insulin. Error bars represent SEM (n = 8 to 10 mice for B, n = 5 or 6 for all other experiments). *P < 0.050 with respect to control, unless otherwise denoted, by two-tailed t test with Welch’s correction.

Fig. 5.

Pelargonidin induces reversible, nontoxic opening of intestinal tight junctions. (A) Compared to untreated Caco-2 cells, (B) pelargonidin-treated monolayers showed no difference in morphology for nuclei or midcell actin. (C and D) However, actin at the apical surface, as well as the tight junction proteins (E and F) ZO-1 and (G and H) occludin rearranged into more punctate forms as a result of pelargonidin treatment. (I) During 1 mo of daily pelargonidin treatment, mice did not lose weight compared to control animals. Periodic weight loss was observed in both control and treated groups as a result of overnight fasting for weekly checkups. (J) Treated mice did not develop elevated levels of the inflammation markers LBP, (K) I-FABP, or (L) PCT. qRT-PCR revealed no statistical difference in mRNA expression of the tight junction proteins (M) Claudin 2, (N) Claudin 3, (O) ZO-1, or (P) JAMA in the small intestines of control and pelargonidin-treated mice. (Q) Representative histological images of duodenal sections of the small intestines from control mice and (R) pelargonidin-treated mice displayed no tissue damage resulting from treatment. (S) There were also no discernable histological differences between proximal colon samples from control and (T) treated mice. (White scale bars, 10 µm; black scale bars, 100 µm.) Error bars represent SEM (n = 4 to 12). For I–P, no comparisons between treated and control mice achieved statistical significance by two-tailed t test with Welch’s correction.