Observations on the Fatty Constituents of Marine Plankton: II. General Character of the Plankton Oils

VERY little information is available regarding’ the chemical composition of the multitude of plant and animal organisms which collectively are termed plankton and which form the primary food supply on which all marine animal life directly or indirectly subsists. This is, of course, due to the difficulty of collecting sufficient material for adequate examination, and although the quantities available for our experiments were not as large as we would have desired, we think the results are sufficiently interesting to merit a report.

The material investigated by us consisted of a number of small samples of the oily constituents extracted from phyto- and zooplankton collected in the spring and summer of 1928 at Port Erin. Full facilities for the work of extraction were very kindly provided at the Sir William Dunn Institute of Biochemistry during the months of October to December, 1928, five months after the phytoplankton and three months after the zooplankton had been collected.

Since extraction of vitamins A and D formed an integral part of the work, precautions were taken to guard against their destruction through oxidation and the fats were separated from the plankton by methods protecting them from oxygen, strong heat and light, the agents most likely to have an injurious effect.

The plankton samples had been preserved partly by sodium chloride, partly by alcohol. The latter were deprived of the alcohol first by filtering⁵ and then by distilling at 60° C. under reduced pressure in a current of CO₂. At this stage the plankton was in a watery mush and subsequent processes of extraction were the same for samples preserved by either method. Excess moisture was first removed by filtering in a current of CO₂ by means of a Buchner funnel. The sample was next weighed and finally desiccated by grinding with 25 per cent, sand and 200 per cent. CaSO₄ in a mortar which was kept cool by ice and at the same time exposed to a continuous flow of CO₂.

The dried powders were extracted by light petrol (B.P. 50 − 60° C.) for about 30 hours in a Soxhlet apparatus shielded by a screen from direct light. Oxidation was guarded against by employing CO, as frequently as possible to displace air.

The extract from the zooplankton yielded deep red or brown-red oils after removal of the solvent in a current of CO₂. The oils from the phytoplankton were deep green or brown-green in colour. The final products were weighed, their volume recorded and sealed in glass containers in an atmosphere of CO₂.

A number of the extracted oils were examined by usual methods of analysis in order to gain a general idea of their chemical type. The results are given in Table I.

Samples of some of the individual oils were retained for vitamin tests (see Paper III), but as it was desired to obtain as much detailed information as possible regarding their chemical composition it was thought that the analytical data in Table I justified the mixing of the specimens so as to provide representative preparations of a phytoplankton and of a zooplankton oil. The results of the examination of the combined samples are given below.

Unfortunately, the quantity of phytoplankton oil was so small that the data derived from its restricted examination is insufficient to enable one to say very much about the composition.

The results of the analysis of the zooplankton oil are, however, of considerable interest as the fatty acids show definite resemblance to those obtained from fish-liver oils by Guha, Hilditch and Lovern (1930). For example, cod-liver oil fatty acids contain about 15 per cent, saturated acids, 35 − 40 per cent, of C₂₀ − C₂₂ highly unsaturated acids and yield about 35 − 47 per cent, of ether-insoluble bromides (ca. 70 per cent. Br).

Insufficient material for fractionation was available from the phytoplankton but nearly 11 gm. of non-saponifiable substances were provided from the hydrolysis of the zooplankton oil.

The material was a deep yellow-coloured soft wax and gave strong tests for cholesterol. Estimations of sterol precipitated by digitonin gave a value of 9 · 3 per cent., and the whole non-saponifiable fraction was then treated to remove these sterols quantitatively. Decomposition of the sterol digitonide in the usual manner with boiling xylene gave a good yield of a crystalline product melting between 131 and 135 ° C. and showing a rotation of ⁠. The material was repeatedly fractionated from methyl and ethyl alcohols, but the amount was in-sufficient for satisfactory separation. The least soluble fraction weighed 0·13 gm., was beautifully crystalline, melted at 139− 139·5° C. and showed an optical activity of ⁠. The acetate melted sharply at 113 · 5° C. and showed no marked depression on mixing with cholesterol acetate. It seems probable that this preparation was mainly cholesterol, but there was evidence of contamination with a small amount of another substance precipitated by digitonin.

The material accompanying the cholesterol was not isolated in anything approaching a state of purity. It was separated as a soft wax from the fractions showing the higher solubility in methyl and ethyl alcohol and appeared to represent about one-third of the material precipitated by digitonin. It was characterised by a much higher laevo-rotation than cholesterol. No clue as to its identity was obtained.

The sterol-free material was a soft wax resembling a pale yellow butter. It possessed an iodine number of 50 · 8 and was optically inactive. In view of the possible relation between the composition of plankton oils and fish oils an attempt was made to ascertain whether the material contained the curious hydrocarbon, squalene, which is found in certain fish-liver oils. Unfortunately, only a small amount of the preparation could be spared for these tests. Bromination of 0 · 16 gm. dissolved in dry ether gave a minute yield of a white insoluble bromide, which on heating darkened at 120 ° C., but did not melt below 280 ° C. Treatment of a solution of 0 · 165 gm. in dry acetone with hydrochloric acid gas gave, on standing, a minute yield of a greyish white crystalline sediment. The crystalline form resembled that of squalene hydrochloride, but the melting-point (107 − 110° C.) was low for that substance. These tests indicate the presence of a very small proportion of a hydrocarbon, either squalene or a similar substance, in the zooplankton oils.

Efforts to separate crystalline fractions from the main bulk of the sterol-free material by the use of the usual solvents having failed, distillation was attempted. Working at approximately o-i mm. four fractions were separated.

Fractions 1a and 1b were recrystallised repeatedly from methyl alcohol and yielded several crops of a clean white crystalline solid M.P. 49·0−49·5° C. The phenylurethane melted at 72° C. These properties, together with the analytical results, suggest identity with cetyl alcohol.

4·910 mg. gave 14·225 mg. CO₂ and 6·04 mg. H₂O. C 79·1 %; H 13·7 %.

4·961 mg. gave 14·360 mg. CO₄ and 6·17 mg. H₂O. C 78·9 %; H 13·8 %.

C₁₈H₃₄O requires: C 79·25 %, H 14·15 %.

The mother liquors from the crops of cetyl alcohol gave soft yellow waxes from which no clean preparations of solid substances were separated by treatment with other solvents. Fractions 2a and 2b also failed to yield any solid materials, other than small crops of impure cetyl alcohol. After the removal of the majority of the solid alcohol the residues of these fractions were oily liquids, the properties of which suggested the presence of a considerable proportion of an unsaturated alcohol of the type of oleyl alcohol or eicoseneyl alcohol C₂₀H₄₀O; neither the elementary analysis nor the values for iodine number indicated alcohols of higher oxygen content such as selachyl alcohol. The preparations were, therefore, reduced by hydrogen in the presence of palladium catalyst. The reduction proceeded readily and the saturated product separated out as a clean white crystalline substance during the final stages of the treatment. On recrystallisation from methyl alcohol an alcohol melting sharply at 69 ° C. was separated. The properties and analysis of this compound corresponded with eicosyl alcohol.

5 · 55 mg. gave 16 · 42 mg. CO₂ and 6 · 90 mg. H₂O. C 80 · 57 % ;H 13 · 80 %.

5 · 048 mg. gave 14 · 85 mg. CO₂ and 6 · 29 mg. H₂O. C 80 · 20 % ; H 13 · 88 %. M.W. depression of F.P. of camphor 308, 316.

Calculated for C₂₀H₄₂O (M.W. 312): C 80-42 %, H 14 · 19 %.

The alcohol gave crystalline acetate, M.P. 49 ° C., and phenvl urethane, M.P. 85 ° c.

3·762 mg. gave 10·7 mg. C0₂ and 4·45 mg. H₂O. C 77·56 %, H 13·14 %. Calculated for C₂₂H₄₄O₂: C 77·64 %, H 12·94 %.

4·579 mg. gave 12·97 mg-CO₂ and 4·65 mg. H₂O.

3·513 mg-gave 0·098 c.c. nitrogen at 23 C. and 754 mm. C 77·31 %, H 11·28 %, N 3·19 %.

Calculated for C₂₇H₄₇O₂N : C 77·69 %, H 11·27 %> N 3·35 %.

Comparatively small quantities of C₂₀ unsaturated alcohols have been detected in the head oil of the Sperm whale (Hilditch and Lovern, 1929), but as far as we are aware there is no other previous record of an unsaturated C₂₀ alcohol of this type having been detected in marine oils and its occurrence in plankton is of considerable interest. It is to be expected that it would be found as a constituent of the non-saponifiable fraction of the liver oil of fish or other plankton-feeders.

No pure substance other than eicosyl alcohol was isolated from the hydro-genated material, although a small yield of an alcohol, melting at 60 − 65° C., and showing solubilities resembling those of batyl alcohol was obtained. Insufficient material for purification was available but the one elementary analysis made tended to support the view that the material was batyl alcohol.