Tuesday, January 15, 2019

Not the Collagen, but the Carnitine?

Note that I'm using the tradition Zooko and I developed on Ketotic.org of end-to-end citations, so check the references for supporting statements, not merely links.

In trying to understand why meat cures scurvy, and why long term carnivores don't develop it, I've previously written about how I discovered that the USDA database misleadingly lists beef as containing no vitamin C, which is inaccurate. Beef contains enough vitamin C that carnivore-diet-level intakes should provide just enough vitamin C to stave off scurvy, provided the meat isn't overcooked. I've also written about my hypothesis that less vitamin C is needed on a low carb diet due to increased endogenous antioxidants and less competition from glucose. The availability of vitamin C in meat, combined with the hypothetical lower requirement, could explain the observations.


However, these explanations have always seemed just barely adequate to me, and to others. Some have suggested meat supplies us with pre-formed collagen, and that this would relieve the burden of vitamin C to synthesise it. However, as far as I can tell, the collagen in meat would not be digested intact but rather broken down into its precursor amino acids during digestion.

It's not immediately obvious that this wouldn't still work, because those amino acid precursors include hydroxyproline, the very form that vitamin C is needed to create, by hydroxylating proline. However, the problem is that when collagen is formed, the hydroxylation is the last step, a post-translational step, after assembling the rest of the protein into procollagen out of mainly glycine and proline [Alb2017]. So collagen creation doesn't actually use hydroxyproline, it uses proline. This is borne out in studies. In vitro, preformed hydroxyproline isn't used to make collagen [Gre1959]. In vivo studies in rats and guinea pigs show that not only is labelled dietary hydroxproline not incorporated into tissues [Ste1949], but injected hydroxyproline doesn't help with wound healing in animals made deficient in vitamin C [Pea1960].

Collagen hydrolysate

Collagen / gelatin hydrolysate is a recent industrial by-product innovation [Góm2011]. Controlled enzymatic hydrolysis of collagen or gelatin is used to create bioactive peptides (chains of amino acids with biochemical activity) that contain hydroxyproline . The peptides appear to have better absorption than food-derived peptides [Iwa2005], [Liu2009] and even seem to be incorporated into animal tissues, at least in some studies [Oes1999], [Wat2010]. It is difficult to evaluate the claims, because the papers all appear to be motivated by product development, and may be overstated.

The product is marketed as useful due to the charge that aging animals, including humans, apparently lose the ability to synthesise collagen well, because of impairments in post-translational modifications to the proteins [Dan2015]. This is said to account for weakened joints and wrinkles.

The general assumption is that most actual food is broken down into amino acids during digestion and so these products are considered to be different from food, in that they are broken down into peptides that are small enough to be absorbed and somehow escape being further broken down before they are absorbed.

Therefore, while this is interesting from a pharmacological, supplement perspective, it doesn't seem to indicate that a meat-based diet would have this effect. To the contrary, if it were known that collagen peptides from meat provided these peptides, then the product would not be innovative and the process would be unnecessary! I've yet to see a direct comparison, though. Moreover, since none of these experiments involved vitamin C depletion, it's also unclear whether the proposed incorporation of the peptides is itself vitamin C dependent.

For these reasons, I conclude that the potential therapeutic value of collagen hydrolysate does not support collagen from meat as a significant way to spare vitamin C, and more likely contributes to refuting it.


Nonetheless, there is a second aspect of scurvy that eating meat does seem able to help with: carnitine. Vitamin C is not only needed to synthesise collagen, but to synthesise carnitine [Str2010]. Some of the symptoms of scurvy, including fatigue, which is the first sign of it [Hug1988], likely come from a deficiency of carnitine. Guinea pigs deprived of vitamin C live longer when given carnitine [Hug1981].

Unlike collagen, significant levels of carnitine do make it through the digestive system [Eva2003]. Therefore it stands to reason that eating meat spares vitamin C that would normally be used for carnitine synthesis, and then more of the vitamin C it provides can be used for collagen. I don't know how much of the 6-10 mg/day of vitamin C that's believed to be needed to prevent scurvy is accounted for by carnitine synthesis, but it could well be enough to tip the balance from barely adequate to easily adequate.



Vance L Albaugh, Kaushik Mukherjee, Adrian Barbul; Proline Precursors and Collagen Synthesis: Biochemical Challenges of Nutrient Supplementation and Wound Healing, The Journal of Nutrition, Volume 147, Issue 11, 1 November 2017, Pages 2011–2017, https://doi.org/10.3945/jn.117.256404

"Although the most straightforward approach to improve wound strength is to provide additional proline in the diet, neither proline nor hydroxyproline increase wound breaking strength (31). Ingested hydroxyproline is readily degraded and synthesis of hydroxyproline occurs only posttranslationally—not de novo—once proline has already been incorporated into collagen."


Daneault, Audrey, Véronique Coxam, and Yohann Wittrant. “Biological Effect of Hydrolyzed Collagen on Bone Metabolism.” Critical Reviews in Food Science and Nutrition, May 15, 2015, 00–00. https://doi.org/10.1080/10408398.2015.1038377.

"Regarding the mechanisms involved in ageing, Knott et al. highlighted an increase in the overall metabolism of collagen which may account for impaired post-translational modifications, leading to severe dysfunctions in the collagen network and a more fragile bone matrix (Knott & Bailey, 1998). Altered post-translational modifications hamper the formation of cross-links between collagen molecules based on aldehyde formation from specific telopeptide hydroxylysine or lysine residues (Knott & Bailey, 1998) and include an abnormal increase in lysyl hydroxylation or glycosylation, which are key to sustain the structural and mechanical integrity of the collagen network (M. Saito & Marumo, 2010; Yeowell & Pinnell, 1993 . These alterations lead to thinner fibrils and higher bone fragility."


Evans, Allan M, and Gianfranco Fornasini. “Pharmacokinetics of L-Carnitine:” Clinical Pharmacokinetics 42, no. 11 (2003): 941–67. https://doi.org/10.2165/00003088-200342110-00002.

"By comparing the amount of carnitine ingested per day with the amount recovered in urine and feces, it was found that subjects on a low-carnitine diet excreted more L-carnitine than they ingested, while those on the high-carnitine diet excreted less than they ingested. The results with the latter group suggested that humans do not absorb all of the L-carnitine that they consume.[99] In the same study, subjects on a low-carnitine diet excreted about 25% of an oral tracer of L-[methyl-3H]carnitine as metabolites of the compound (mainly trimethylamine-N-oxide and γ-butryobetaine, which appear to be formed within the gastrointestinal tract prior to absorption; see section 3.3). If it is assumed that there was negligible excretion of the tracer via expired air, the extent of absorption in the subjects fed the low-carnitine diet may have been as high as 75%, on average.[99] In those subjects on a high-carnitine diet, 37% of the dose could be accounted for as excreted metabolites, meaning that the extent of absorption might have been about 63%. Importantly, the results suggest that the efficiency of absorption tends to diminish as the carnitine content of the diet increases,[91,99] reflecting the involvement of specific transporters that can be saturated even with normal dietary intake. As described below (section 3.1), the bioavailability of supplemental or medicinal oral doses of L-carnitine tends to be even lower, at 5–18%."


Gómez-Guillén, M.C., B. Giménez, M.E. López-Caballero, and M.P. Montero. “Functional and Bioactive Properties of Collagen and Gelatin from Alternative Sources: A Review.” Food Hydrocolloids 25, no. 8 (December 2011): 1813–27. https://doi.org/10.1016/j.foodhyd.2011.02.007.

"Scientific literature about different alternative sources and new functionalities of collagen and gelatin has experienced a boom in the last 10e15 years, in part due to the growing interest in the economical valorisation of industrial by-products (from the meat and fish industry), the environmental friendly management of industrial wastes, and the search for innovative processing conditions as well as potential novel applications."


Green, N. M., and D. A. Lowther. “Formation of Collagen Hydroxyproline in Vitro.” Biochemical Journal 71, no. 1 (January 1959): 55–66. https://doi.org/10.1042/bj0710055.

"3. The addition of unlabelled L-hydroxyprohne to the incubation medium in the presence of L-[14C]proline had no effect on the ratio of the specific activities of collagen hydroxyproline and proline although the total radioactivity incorporated was reduced.

"4. Incubation of the tissue with L-[14C]hydroxyproline did not result in a significant incorporation of radioactivity into collagen.

"5. Radioactive free hydroxyproline was isolated from the slices and medium after incubation with L-[14C]proline but its specific activity was only half that of the neutral-salt-soluble collagen hydroxyproline and the total counts present were not increased when unlabelled hydroxyproline was present as a trapping agent.

"6. It is concluded that free hydroxyproline is not an intermediate in the formation of the hydroxyproline of collagen.

"7. Both proline and hydroxyproline added to the medium were found to be concentrated intracellularly about 2-5 times. The lack of incorporation of free hydroxyproline cannot therefore be due to the impermeability of the cells towards hydroxyproline."


Hughes, “Recommended Daily Amounts and Biochemical Roles—The Vitamin C, Carnitine, Fatigue Relationship.” in Vitamin C (ascorbic acid) J. N. Counsell, D. H. Hornig. 1981 ISBN 0853341095, 9780853341093 http://www.mv.helsinki.fi/home/hemila/concepts/Hughes_1981.pdf

"Our studies have indicated that in guinea pigs such a relationship does in fact exist. By dietary means we produced tissue ascorbic acid concentrations of 12 % and 100 % saturation respectively in two groups of male guinea pigs. In the 'low ascorbic acid' group the mean concentration of skeletal muscle carnitine after 20 days was 0.5 ȝ g/g tissue and in the 'ascorbic-acid-sufficient' group it was l.15 ȝ g/g tissue. There was no concomitant emergence during this period of any of the symptoms customarily regarded as presaging the emergence of scurvy in guinea pigs— such as growth depression and kidney hypertrophy [38] (Table 1)."

"In a further study it was shown that administration of carnitine (10 mg per animal daily) prolonged significantly the life span of male guinea pigs given a scorbutogenic diet (Fig. 5). This could imply that carnitine may replace ascorbic acid in certain of its functions—a biochemically unlikely explanation. It is more likely that carnitine prolongs the life span by significantly 'sparing' ascorbic acid which would otherwise be used in the formation of endogenous carnitine."

"It would therefore appear that the involvement of ascorbic acid in carnitine biosynthesis is a nutritionally significant happening and that muscle carnitine is a highly sensitive indicator of ascorbic acid status."


Hughes, RE Ascorbic acid, carnitine and fatigue. Med. Sci..Res., 1988; 15, 721-723

"References to the early emergence in scurvy of fatigue and lassitude were ìnvariable features of the earliest clinical descriptions of the disease [23]. Eugalenus in 1658 spoke of "spontaneous debility" [24], Lister, in 1696, wrote of "weakness of limbs and considerable fatigue" [25] and Sydenham in 1742 of "spontaneous lassitude and difficulty of breathing after exercise" [26]. Naval surgeons with first hand experience of scurvy were equally clear in their descriptions: "The signes of the Scurvie are many, namely a general lazinesse ... shortnesse and difficultie of breathing, especially when they moove themselves" commented Woodall in 1639 [27] and Lind, over a century later, wrote: "... this lassitude, with a breathlessness upon motion, are observed to be among the most common concomitants of the distemper" [28]. Practising 'land physicians' in the last century made similar observations. Shapter, a careful clinical observer, describing an outbreak ofscurvy in Exeter in 1847, perhaps put the matter most clearly: "... the spongy and swollen gum appears to me to have been erroneously estimated as amongst the primary and most obvious manifestations of the scurvy ... I am inclined to say there is a class of well-marked symptoms preceding this... The first or initiatory stage ... has appeared to me to be characterised by ... debility ... weakness, listlessness and a disinclination to exercise" [29].

"More recent cases of scurvy have also underlined the early emergence of fatigue. In 1952 it was noted in a case history that the patient had, during the year before admission, "become increasingly weak and easily fatigued" [30] and reports of experimentally induced scurvy in human volunteers similarly drew attention to the early emergence of fatigue [3]-33]. Crandon, who placed himself on a scorbutogenic dief, commented that a feeling of fatigue developed from the beginning of the 3rd month of deficiency, a full 6 to 8 weeks before the emergence of the traditional 'overt' signs of scurvy such as perifollicular hyperkeratotic papules, petechiae, poor wound healing and softening of the gums [34].

"It will be noted that the fatigue of scurvy, like the fall in muscle carnitine in hypovitaminotic C guinea-pigs [7], evidences itself before the traditional overt signs of scurvy and it has been suggested that it reflects an impairment of the endogenous biosynthesis of camitine in the absence of adequate ascorbic acid [23]. The pathological features customarily associated with scurvy are all, theoretically, amenable to reductionist treatment in terms of the hydroxylation of lysyl and prolyl residues in the formation of collagen. Fatigue bears no identihable relationship to collagen formation, and this is possibly the reason why this fêature of incipient scurvy has been generally ignored by students ofthe disease."


Iwai, Koji, Takanori Hasegawa, Yasuki Taguchi, Fumiki Morimatsu, Kenji Sato, Yasushi Nakamura, Akane Higashi, Yasuhiro Kido, Yukihiro Nakabo, and Kozo Ohtsuki. “Identification of Food-Derived Collagen Peptides in Human Blood after Oral Ingestion of Gelatin Hydrolysates.” Journal of Agricultural and Food Chemistry 53, no. 16 (August 2005): 6531–36. https://doi.org/10.1021/jf050206p.

"In the present study, we isolated and identified some food-derived collagen peptides in human serum and plasma as show in Table 2 . Among them, Pro-Hyp, which has been demonstrated to be present in urine ( 15 ), is a major constituent in any case. In the case of the oral ingestion of chicken type II gelatin hydrolysates, a significant amount of Pro-Hyp-Gly was detected in human plasma. This motif is also abundantly present in type I and II collagens. However, only a less amount of Pro-Hyp-Gly was observed in the blood from those who ingested type I gelatin hydrolysates. The chicken type II gelatin hydrolysate preparation contained a significant amount of mucopolysaccharide ( Table 1 ), which might affect digestion and absorption of collagen peptides. Another tripeptide, such as Gly-Pro-Hyp, could not be detected in all cases. Some dipeptides consisting of hydrophobic amino acids (Ile, Leu, and Phe) and Hyp are contained in human blood as minor constituents after loading of the gelatin hydrolysates. So far up to now, biological activities of them have not been described."


Liu, Chinfang, Kazuko Sugita, Ken-ichi Nihei, Koichi Yoneyama, and Hideyuki Tanaka. “Absorption of Hydroxyproline-Containing Peptides in Vascularly Perfused Rat Small Intestine in Situ.” Bioscience, Biotechnology, and Biochemistry 73, no. 8 (August 23, 2009): 1741–47. https://doi.org/10.1271/bbb.90050.

"It is generally assumed that, during absorption, proteins derived from foodstuffs are hydrolyzed, generating small peptides and amino acids in the lumen. These small peptides are then hydrolyzed by intracellular peptidases, leading to the appearance of digestive products, mainly as free amino acids, in the portal vein. In contrast, there is some evidence that the intestinal transport of peptides or macromolecules may give, to a small but significant extent, antigens or biologically active substances."


Oesser, S., M. Adam, W. Babel, and J. Seifert. “Oral Administration of (14)C Labeled Gelatin Hydrolysate Leads to an Accumulation of Radioactivity in Cartilage of Mice (C57/BL).” The Journal of Nutrition 129, no. 10 (October 1999): 1891–95. https://doi.org/10.1093/jn/129.10.1891.

"Several investigations showed a positive influence of orally administered gelatin on degenerative diseases of the musculo-skeletal system. Both the therapeutic mechanism and the absorption dynamics, however, remain unclear. Therefore, this study investigated the time course of gelatin hydrolysate absorption and its subsequent distribution in various tissues in mice (C57/BL). Absorption of (14)C labeled gelatin hydrolysate was compared to control mice administered (14)C labeled proline following intragastric application. Plasma and tissue radioactivity was measured over 192 h. Additional "gut sac" experiments were conducted to quantify the MW distribution of the absorbed gelatin using SDS-electrophoresis and HPLC. Ninety-five percent of enterally applied gelatin hydrolysate was absorbed within the first 12 h. The distribution of the labeled gelatin in the various tissues was similar to that of labeled proline with the exception of cartilage, where a pronounced and long-lasting accumulation of gelatin hydrolysate was observed. In cartilage, measured radioactivity was more than twice as high following gelatin administration compared to the control group. The absorption of gelatin hydrolysate in its high molecular form, with peptides of 2.5-15kD, was detected following intestinal passage. These results demonstrate intestinal absorption and cartilage tissue accumulation of gelatin hydrolysate and suggest a potential mechanism for previously observed clinical benefits of orally administered gelatin."


Peacock, E. E. “Effect of Dietary Proline and Hydroxyproline on Tensile Strength of Healing Wounds.” Experimental Biology and Medicine 105, no. 2 (November 1, 1960): 380–83. https://doi.org/10.3181/00379727-105-26117.

"In normal animals there is very little incorporation of dietary hydrosyproline in the general protein pool; nearly all of N15 labeled hydroxyproline can be recovered in the urine and stools (Stetten(9)) However, there is an extremely low turnover of the amino acids in a normal animal's collagen (Neuberger and Slack (8) ) , therefore Stetten's experiments do not necessarily mean that an animal which was actively at tempting to synthesize new collagen under the handicap of protein or scorbutic acid deficiency would not be able to by-pass the hydroxylation of proline and utilize free hydroxyproline for collagen synthesis.

"Green and Lowther (3) investigated the possibility of incorporating free Nl" labeled hydroxyproline into collagen which was being formed by an in vitro tissue slice from a carrageenin granuloma. Their results showed that practically all of the hydroxyproline was produced by hydroxylation of bound proline and that almost none of the free labeled hydroxyproline was incorporated in the saline extractable or new collagen fraction. They showed, however, that fibrolblasts' cell membranes were freely permeable to hydroxyproline, therefore we still wondered if free hydroxyproline could be utilized by fibroblasts in healing wounds."

"Protein depleted rats were given dietary supplements of .5% synthetic d-1 hydroxyproline and 1% synthetic d-1 proline. Guinea pigs on an ascorbic acid deficient diet were given a dietary supplement of 5% d-1 hydroxyproline. The animals were wounded by a standard technic and the tensile strength of their healing wounds was tested at 48-hour intervals between the 6th and 21st postoperative days. Neither hydroxyproline nor proline exerted a significant effect upon the rate of gain of tensile strength in the wounds of protein deficient rats or scorbutic guinea pigs."

"Depletion and ascorbic acid deficiency reveal that the results of Stetten and Green also apply to the wounded animal, and that impaired wound healing in pathological states cannot be overcome by administration of dietary hydroxyproline. "


Stetten, Marjorie R. “Some Aspects of the Metabolism of Hydroxyproline, Studied with the Aid of Isotopic Nitrogen.” Journal of Biological Chemistry 181, no. 1 (November 1, 1949): 31–37.

"Some of the isotope was found in the body proteins. The very low isotope concentration of the hydroxyproline isolated from the body proteins indicated that less than 0.1 percent of the hydroxyproline in these rats had been derived from the dietary hydroxyproline in 3 days. A higher concentration of N15 was found in the glutamic acid, aspartic acid, and arginine of the proteins and probably came indirectly from degradation products of the hydroxyproline.

"The body proline contained only traces of N16, indicating that little if any of the proline of the body is derived from dietary hydroxyproline.

"The hydroxyproline of the proteins is not derived to any appreciable extent from dietary hydroxyproline but rather from the oxidation of proline which is already bound, presumably in peptide linkage."


Strijbis, Karin, Frédéric M. Vaz, and Ben Distel. “Enzymology of the Carnitine Biosynthesis Pathway.” IUBMB Life 62, no. 5 (2010): 357–62. https://doi.org/10.1002/iub.323.

"The first enzyme of the carnitine biosynthesis pathway is TML dioxygenase (TMLD), which hydroxylates TML to yield 3‐hydroxy‐TML (HTML)."


"In addition to these cofactors, TMLD also requires the presence of ascorbate (vitamin C) for enzymatic activity, presumably to maintain the iron in the ferrous state."


Watanabe-Kamiyama, Mari, Muneshige Shimizu, Shin Kamiyama, Yasuki Taguchi, Hideyuki Sone, Fumiki Morimatsu, Hitoshi Shirakawa, Yuji Furukawa, and Michio Komai. “Absorption and Effectiveness of Orally Administered Low Molecular Weight Collagen Hydrolysate in Rats.” Journal of Agricultural and Food Chemistry 58, no. 2 (January 27, 2010): 835–41. https://doi.org/10.1021/jf9031487.

"Collagen, a major extracellular matrix macromolecule, is widely used for biomedical purposes. We investigated the absorption mechanism of low molecular weight collagen hydrolysate (LMW-CH) and its effects on osteoporosis in rats. When administered to Wistar rats with either [14C]proline (Pro group) or glycyl-[14C]prolyl-hydroxyproline (CTp group), LMW-CH rapidly increased plasma radioactivity. LMW-CH was absorbed into the blood of Wistar rats in the peptide form. Glycyl-prolyl-hydroxyproline tripeptide remained in the plasma and accumulated in the kidney. In both groups, radioactivity was retained at a high level in the skin until 14 days after administration."