Tier 3 — preclinical

Comparison of quantity and structures of hydroxyproline-containing peptides in human blood after oral ingestion of gelatin hydrolysates from different sources

Hiroki Ohara, Hitoshi Matsumoto, Kyoko Ito, Koji Iwai, Kenji Sato

Journal of Agricultural and Food Chemistry 2007 55(4):1532–1535

Bibliography

DOI: 10.1021/jf062834s
PubMed: PMID 17253720
Funding: No external grant funding statement disclosed in the paper. Acknowledgment thanks Masahiro Nomoto and Makoto Aoki at the Pharmaceutics Research Lab of Meiji Seika Kaisha for pharmacokinetic analysis assistance.
Competing interests: No competing interests statement disclosed in the paper. Three co-authors (Hiroki Ohara, Hitoshi Matsumoto, Kyoko Ito) were employees of Meiji Seika Kaisha Ltd., Food and Health R&D Laboratories — a commercial food ingredient manufacturer whose product lines include collagen. The fish scale and porcine skin gelatin hydrolysates were gifted by Nitta Gelatin (Osaka), a commercial gelatin supplier; the fish skin hydrolysate was purchased commercially. Remaining co-authors (Iwai, Sato) were affiliated with Kyoto Prefectural University. Substantial industry involvement despite the absence of a formal declaration.

Study snapshot

Design	Single-blind, three-arm crossover human bioavailability study with 1-week washout between arms
Model	Healthy Japanese adult males
Sample	5 male volunteers, mean age 33.0 ± 5.6 years, mean weight 69.8 ± 7.4 kg
Intervention	Single oral dose of type I gelatin hydrolysate at 0.385 g per kg body weight in 20% w/v aqueous solution, after 12 h fast, from three sources tested sequentially with 1-week washouts: fish scale, fish skin, or porcine skin
Duration	Single-dose kinetics per arm; blood sampling at 0.5, 1, 2, 4, 7 and 24 h post-ingestion
Endpoints	Area under the concentration–time curve (AUC0–24h) of free-form hydroxyproline in plasma; AUC0–24h of hydroxyproline-containing peptides in plasma; Identification of peptide sequences by Edman degradation; Semi-quantitative peptide composition per gelatin source

What the study showed, in plain terms

Not all collagen supplements are made from the same animal. Some come from cow hide, some from pig skin, some from fish scales, some from fish skin. This raises a practical question: does the source actually matter for what ends up in your bloodstream? By 2007 it was already known — from the same research group's earlier work — that some collagen peptides do reach the blood after ingestion. What was not known was whether the amount and the kinds of peptides depended on where the collagen came from.

Five healthy male volunteers each took, on three separate occasions, a single dose of gelatin hydrolysate prepared from either fish scale, fish skin, or porcine skin. Each session was separated by a week to let the previous dose fully clear. Blood was sampled repeatedly over the following 24 hours and analysed for two things: how much collagen-derived peptide showed up, and which specific peptides were present.

The results were clear. Fish scale hydrolysate delivered noticeably more collagen peptide into the bloodstream than the other two sources, and it delivered a broader variety of peptides — several small peptide fragments were unique to the fish sources and were not detected at all after porcine ingestion. Porcine skin, by contrast, produced almost pure Pro-Hyp and very little else. Because different peptides have been shown to have different biological effects in laboratory experiments, this paper is the earliest clean evidence that the source of a collagen supplement is not a cosmetic detail — it changes what actually reaches the blood.

Key findings

AUC0–24h of hydroxyproline-containing peptides in plasma was significantly higher after fish scale hydrolysate (522 ± 62 h·nmol/mL) than after porcine skin (345 ± 66) or fish skin (315 ± 67) (p<0.05). Fish scale was approximately 1.5-fold higher than fish skin.
AUC0–24h of free-form hydroxyproline was highest for fish scale (1238 ± 116) and porcine skin (1194 ± 305), and significantly lower for fish skin (799 ± 99) (p<0.05).
Across all three sources, hydroxyproline in peptide form accounted for approximately 20–30% of total detected hydroxyproline in plasma, with the balance in free amino acid form.
Porcine skin hydrolysate produced an almost pure Pro-Hyp profile: Pro-Hyp accounted for 95% of detected peptides, with only trace amounts of Ile-Hyp, Leu-Hyp and Phe-Hyp.
Fish skin hydrolysate produced Pro-Hyp (42%), Leu-Hyp (27%), Ala-Hyp (15%), Ile-Hyp (7%), Phe-Hyp (7%) and Pro-Hyp-Gly (3%).
Fish scale hydrolysate produced the most diverse profile: Pro-Hyp (39%), Ala-Hyp-Gly (16%), Ala-Hyp (15%), Ser-Hyp-Gly (12%), Leu-Hyp (10%), Pro-Hyp-Gly (5%), Phe-Hyp (3%), Ile-Hyp (2%).
Ala-Hyp-Gly and Ser-Hyp-Gly were detected only after fish scale hydrolysate ingestion — not after fish skin or porcine skin.
Peak plasma peptide concentrations occurred at approximately 1–2 hours post-ingestion across all three sources, and returned to baseline by 7 hours.

What this study can and cannot tell us

Sample size of only five male volunteers — inadequate for population-level inference and provides no data on female-specific pharmacokinetics.
All subjects were Japanese. Inter-ethnic differences in peptide transporter expression could affect generalisability.
Single-dose kinetics only. No data on whether the source-dependent differences persist under chronic daily dosing.
Only three sources compared. Bovine hide, chicken and eggshell collagen — all widely marketed — were not tested.
No dose–response within any source. The 0.385 g/kg dose was fixed; whether higher or lower doses would preserve the same relative differences is not established.
Substantial commercial conflict of interest — three co-authors employed by Meiji Seika Kaisha, and gelatin hydrolysates supplied by Nitta Gelatin. No competing interests statement.
The paper is purely analytical. It measures what appears in blood; it does not measure any skin, joint, bone or other clinical outcome. The suggestion that source-specific peptide profiles translate to source-specific clinical effects is a hypothesis, not a finding.
Peptide quantification was semi-quantitative, estimated from HPLC peak areas rather than absolute quantification against stable-isotope-labelled internal standards.

Tier 3

Identification of food-derived collagen peptides in human blood after oral ingestion of gelatin hydrolysates

Koji Iwai et al. · Journal of Agricultural and Food Chemistry · 2005