Vitamin C and collagen synthesis

Obviously vitamin C is required for formation of scar tissue

Robertson and Schwartz (1953) investigated the formation of collagen in granuloma (scar) tissue from guinea pigs maintained at different levels of vitamin C nutrition; they induced granuloma formation by subcutaneous injection of the polysaccharide carrageenan. The table shows the percentage composition (g /100 g) of granuloma tissue 14 days after injection of carrageenan.

[From data reported by Robertson WvB & Schwartz B. Ascorbic acid and the formation of collagen. Journal of Biological Chemistry 201: 689-696 1953]

Many students have a problem interpreting this table – remember that the figures show percentage composition of scar tissue (ie per 100g). They key point is that there is very much less collagen in the tissue from deficient animals, and the missing percentage is made up by water.

Isolated fibroblasts can be cultured for a considerable time, and will synthesise and secrete collagen. However, if they are cultured under conditions of low vitamin C availability, they secrete only a very small amount of collagen, and a soluble protein accumulates in the rough endoplasmic reticulum.

Vitamin C is required for collagen synthesis – but the accumulation of a soluble protein in endoplasmic reticulum in deficient fibroblasts suggests that a precursor is accumulated. Students should know that proteins that are to be expected from the cell (and you should know that collagen is an extracellular protein secreted by fibroblasts) are secreted into the endoplasmic reticulum during synthesis, and then undergo post-synthetic modification.

Similarly, incubation of fibroblasts under anaerobic conditions, or in the presence of iron-chelating compounds, also prevents the secretion of normal collagen and leads to the accumulation of the soluble protein in the rough endoplasmic reticulum.

Both iron and oxygen are obviously essential for the post-synthetic modification of the precursor protein.

One of the characteristic features of collagen is its relatively high content of hydroxyproline (Hyp).

Green and Lowther (1959) investigated the origin of hydroxyproline in collagen by incubating slices of guinea pig granuloma tissue with [14C]proline or [14C]hydroxyproline. They incubated slices of guinea pig granuloma tissue with [14C]proline or [14C]hydroxyproline, then precipitated the protein synthesised by the tissue, hydrolysed it and measured the radioactivity in proline and hydroxyproline, after separating the amino acids by chromatography.

*In samples maintained at 4*C there was apparent incorporation of 2 - 2.5 dpm / µmol into hydroxyproline. (dpm = disintegrations per minute, a measure of radioactive material)

[From data reported by Green NM & Lowther DA. Formation of collagen hydroxyproline in vitro. Biochemical J 71: 55-66 1959]

Some students find this difficult to interpret. Tissue was incubated with either labelled proline or labelled hydroxyproline – but radioactivity in collagen was measured specifically in hydroxyproline. The second line shows that label from proline, but not from hydroxyproline was incorporated into hydroxyproline in proteins.

This might mean either that hydroxyproline cannot be incorporated as such (which is correct) or that hydroxyproline does not enter cells – but the upper row in the table shows that both proline and hydroxyproline enter the cells.

Peterkovsky & Udenfriend (1963) investigated the incorporation of [14C]proline into proline and hydroxyproline in proteins synthesised by a cell-free system from chick embryos. They incubated the samples with [14C]proline, then precipitated the protein synthesised, hydrolysed it and measured the radioactivity in proline and hydroxyproline, after separating the amino acids by chromatography. The results are shown on the right.

[From data reported by Peterkovsky B & Udenfriend S. Conversion of proline to collagen hydroxyproline in a cell-free system from chick embryo. Journal of Biological Chemistry 38: 3966-3977 1963]

Note the different scales for radioactivity in proline and hydroxyproline.

If you look at radioactivity in proline incorporated into protein it does not look sensible – it increases for 50 – 60 min as you would expect, but then it decreases on longer incubation.

The answer is in the appearance of radioactivity in hydroxyproline – initially little or none, and it only really becomes detectable after a considerable amount of proline has been incorporated. This is a classical precursor / product relationship.

The conclusion is that proline is incorporated into the protein, and then later undergoes reaction to yield hydroxyproline.

Stone and Meister (1962) investigated the incorporation of [3H]proline into hydroxyproline in collagen formed by granuloma tissue from control and vitamin C deficient (scorbutic) guinea pigs. The table shows radioactivity (dpm) in hydroxyproline in granuloma tissue:

[From data reported by Stone N & Meister A. Function of ascorbic acid in the conversion of proline to collagen hydroxyproline. Nature 194: 555-557 1962]

Again note that we are incubating with radioactive proline and measuring radioactivity in hydroxyproline. As you would expect, the control shows increasing radioactive hydroxyproline with increasing time of incubation – the effect of vitamin C deficiency suggests that the specific role of vitamin C in collagen synthesis is in the formation of (protein incorporated) hydroxyproline from proline – ie in proline hydroxylase.

Stassen and coworkers (1973) investigated the effects of the protein synthesis inhibitor cycloheximide on the stimulation by ascorbate of proline hydroxylation in mouse fibroblasts in culture:

[From data reported by Stassen FLH, Cardinale GJ & Udenfriend S. Proc Nat Acad of Sciences 70: 1090 1973]

What the second row shows is that adding ascorbate to fibroblasts in culture permits formation of hydroxyproline in protein – and if you add a protein synthesis inhibitor, so that there is no new protein synthesis, then 15 min later add ascorbate, you get as much new hydroxyproline as before – ie there must be accumulated precursor protein in the cells waiting to be hydroxylated when ascorbate is added.

(Earlier experiments showed the accumulation of a soluble protein in the endoplasmic reticulum in vitamin C deficiency, so the you should be able to make the connection).

Prolyl hydroxylase has been purified. It requires ascorbate, molecular oxygen and iron (Fe2+ ) for activity. There is no change in the redox state of the iron during the reaction.

In addition, with the purified enzyme there is an absolute requirement for alpha-ketoglutarate for activity.

Rhoads & Udenfriend (1968) demonstrated the disappearance of alpha-ketoglutarate and appearance of hydroxyproline on incubation of purified prolyl hydroxylase:

[From data reported by Rhoads RE & Udenfriend S. Proceedings of the National Academy of Sciences 60: 1473 1968]

Within experimental error, 1 mol of alpha-ketoglutarate is utilised for each mol of hydroxyproline formed.

Alpha-ketoglutarate is oxidised to succinate and carbon dioxide. Overall the reaction is:

There is net oxidation of ascorbate during the hydroxylation of proline, but very much less than 1 mol of ascorbate is oxidised per mol of hydroxyproline or succinate formed, or of oxygen consumed.

Myllylä and coworkers (1978) investigated the role of ascorbate in proline hydroxylation. In their first study they measured the hydroxylation of proline-containing peptides by the purified enzyme in the presence and absence of ascorbate over a short time-course:

[From data reported by Myllylä R, Kuutti-Savolainen E-V & Kivirokko KI. The role of ascorbate in the prolyl hydroxylase reaction. Biochemical and Biophysical Research Communications 83: 441-448 1978]

From their results, they calculated that over the first 10 sec the enzyme catalyses hydroxylation of ~30 mol of proline / mol of enzyme in the absence of ascorbate.

In the second experiment they added ascorbate to the ascorbate-free incubation after 3 min. After 3 minutes incubation without added ascorbate they noted that the protein bound iron had been oxidised from Fe2+ to Fe3+ , and was reduced back to Fe2+ on addition of ascorbate.

Ascorbate does not act like a coenzyme of the reaction, since it is net consumed in the hydroxylation of proline, but it does not seem to be a substrate, since consumption is not stoichiometric with alpha-ketoglutarate consumed or proline hydroxylated.

With the purified enzyme the hydroxylation of proline does not seem to require ascorbate, although we know that ascorbate is essential in vivo. However, in the absence of ascorbate the enzyme is gradually inactivated – ascorbate appear to protect the enzyme against this inactivation

Ascorbate can reactivate the inactivated enzyme – apparently by reducing Fe3+ back to the active form of the enzyme. The interesting point is that the normal reaction of the enzyme does not involve oxidation of the iron bound at the catalytic site – this seems to be an accidental side reaction of the enzyme. It is hardly surprising that an enzyme that binds molecular oxygen to iron at the active site sometimes undergoes oxidation. There must be an ascorbate binding site on the enzyme, since only ascorbate, and not other reducing agents, will reduce the iron back to the active state.

Overall the reaction can be shown as follows:

Key points from this exercise:

• Vitamin C (ascorbate) is essential for collagen synthesis. In deficiency a soluble precursor protein is synthesised, lacking hydroxyproline.
• The enzyme that catalyses the synthesis of hydroxyproline from proline in the precursor protein, prolyl hydroxylase, contains iron (Fe2+) and catalyses hydroxylation of proline and oxidation and decarboxylation of alpha-ketoglutarate. One atom of oxygen goes into each product.
• Prolyl hydroxylase is active for a short time in the absence of vitamin C, but rapidly undergoes inactivation as a result of oxidation of the reactive iron to Fe3+. When this happens, ascorbate is required to reduce the iron back to Fe2+.
• Ascorbate is thus neither a coenzyme (because it is consumed in the reaction) nor a substrate (because it is not consumed stoichiometrically with the other substrates.