2. Redox relationships in the development of anthocyanin. Keeble and Armstrong, Wheldale-Onslow, Atkins, and others have presented evidence suggesting the presence of oxidase enzymes and an oxidation system associated with the development of anthocyanin. In repeating the studies made by these early workers it is possible, in the light of revised redox methods, to correct several of the interpretations of the use of oxidase indicators, and it now appears that the oxidase enzyme of the earlier workers is in fact a lipid absorptive and oxidative system. It became increasingly apparent during the course of the present study that there is a localized absorption of the oxidized form of the common redox indicators in unsaturated fats present in anthocyanin bearing cells. The oxidation of p-phenelenediamine, 0( -naphthol, leuco methylene blue and related indicators prior to their introduction into sections of rch and rg tissue will give, in uniform and comparably cut sections, a greater localization of colored indicator in rch tissue. An iodimetric method applied to this absorptive system, in appropriately prepared tissue, has made possible a qualitative study of differences between colored (rch) and colorless (rg) tissue and has given an exact iodine number for different tissues where weak anthocyanin development, dependent upon R alleles, is to be compared with more strongly colored rch tissue.

Iodine absorption is always greater in anthocyanin bearing cells; hence practicable microscopic qualitative observations may be compared with macroscopic anthocyanin distribution, and differences in intensity of pigmentation, by using the iodine number as a qualitative guide. The higher iodine absorption of anthocyanin bearing tissue may be seen to be localized in free plasmal lipids, in lipid material localized in "mitochondrial" or lipoclastic bodies in the cell, and in lipids impregnating cellulose walls. The lipids are highly unsaturated condensation aggregates and not true glycerides. They are not readily soluble in ordinary fat solvents but are soluble in petroleum ether after preliminary hydrolysis of the tissue and extraction with an alkaline/alcoholic mixture. The unsaturated lipids in colorless (rg) tissue have a higher peroxide number as determined by oxidation of ferrous ammonium sulphate. The extracted lipids from rch tissue have 40% greater absorptive capacity (Wij's Iodine Method) than comparable extracts from rg tissue. Presented in the table below are the iodine numbers of leaf tissue of rch and rg sib comparisons, as determined by halogen solutions of increasing concentration. The samples were hydrolyzed to prevent iodine addition to starch and to facilitate iodine addition to unsaturated bonds; they were dried under nitrogen to constant weight and a standard iodine method with thiosulphate titration was used and endpoints were determined galvanometrically in some cases. The samples used ranged in weight from 0.020 mg. to 0.155 mg. so that the method may be applied to small samples of tissue that are held in ethyl alcohol (not above 50%), in order to remove chlorophyll, anthocyanin, etc., with frequent changes of alcohol to facilitate elution. At all stages in the process storage under nitrogen prevents oxidative degradation and a drop in iodine values.

Halogen solutions of increasing concentration
rch 3.55 9.24 12.55 50.54
rg 2.21 6.91 10.34 44.22

Using the methods outlined above a study was made of the development of pigment in excised leaves in culture. It was found that additions of dilute emulsions of unsaturated fats (corn oil, soybean oil, linseed oil) and various terpenes (thujone, etc.) greatly increased the production of pigment, but only when sugar was also present. Glucose solutions (16 × 10-3 molar) were less effective than glucose (8 × 10-3 molar) plus unsaturated fat emulsions (.4%). Holding the cultures under anaerobic conditions (under nitrogen) for the first two days of a culture study inhibits production of anthocyanin but increases overall pigmentation after aerobic conditions are restored. In the table below are the iodine numbers from a typical sugar culture experiment. A marked decline in iodine number in rg and a final rise in rch with pigmentation is clearly demonstrated.

All tissue from same leaf

  rch rg
Fresh Tissue 45.95 (colorless) 51.90 (colorless)
Sugar/Anaerobic 41.70 " 44.22 "
Sugar/Same as above,
but exposed to air one day.
50.54 (Anthocyanin) 40.02 "

In vitro preparations of anthocyanin extracts and unsaturated fat emulsions reveal that anthocyanin is a hydrogen acceptor and acts to dehydrogenate and oxidize the fat, and the anthocyanin becomes partially reduced and in some cases irreversibly reduced. This dehydrogenation of fat emulsions by anthocyanin is stronger when water extracts of rch tissues are added to the emulsions. Microscopic sections of anthocyanin-bearing tissue held under anaerobic conditions and at a pH of 7.0 to 7.4 show a reduction (loss of color) of anthocyanin in lipid granules in the plasma under intense illumination and a restoration of color on diminishing the light. This is direct evidence of a reversible redox relationship between lipids and anthocyanin pigments.

It is generally true that anthocyanin bearing cells are epidermal, hypodermal or bundle sheath cells which have an excess of lipid material, and it is a general rule that cells low in lipids are lacking in anthocyanin. This fact may be determined by iodine staining in combination with extraction methods outlined above. It is illustrated in corn by the siliceous epidermal cell which, unlike its couplet partner, the fat-bearing suberized cell, lacks anthocyanin unless cultured in sugar/fat media under nitrogen followed by oxygen. Fatty and other organic acids, as revealed through the use of polychrome stains and direct acid value determinations are present in anthocyanin bearing cells before pigment is produced and there are apparently less free acids after pigment production.

Preliminary trials on B determined pigmentation indicate there is in lipid/pigment development a redox relationship similar to that obtaining in Rr alleles. Trials on the other higher plants (Andropogon, Coleus, Petunia, Acer, etc.) reveal a similar redox problem in floral and autumnal anthocyanin development.

In summation, it now appears that the oxidase system, believed by early workers to be causal in anthocyanin development, is in reality a reflection of the oxidized and dehydrogenated state of lipids which absorb and possibly oxidize redox indicators. The absorption of iodine by these dehydrogenated lipids reveals qualitative but not absolute quantitative differences between pigmented and non-pigmented tissues. Anthocyanin acts in vitro to bring about the dehydrogenation of fats, and wherever anthocyanin appears in the plant associated with a lipid system the fats are more dehydrogenated than in comparable non-pigmented tissue.

D. S. Van Fleet