The Masked Lipids of Nuclei

For many years histologists have considered that some lipids may occur in such a form that they cannot be demonstrated until some pathological condition or special histological treatment reveals them. In order to explain this visualization of lipids it is necessary to assume either that the lipid is present in a form that is availabel for staining but is so finely dispersed that it cannot be seen, or that it is bound to other material in such a way that colouring agents for lipids cannot dissolve in it. Berg (1951), in a careful study of masked lipids of many types of tissues, tended to equate ‘masked lipid’ with that fatty matter which is present in very fine dispersion. During the fatty degeneration of the cells these finely divided particles were believed to coalesce to form resolvable droplets. Berg therefore considered that because of its fluorescence, 3:4-benzpyrene could be used to demonstrate the presence of this type of masked lipid and his results with this fluorochrome are impressive. In the opinion of the present authors, however, such fatty matter should be designated ‘finely dispersed lipid’ since it has no ‘mask’ or masking agent associated with it.

The possibility that lipids might be unavailabel to stain because of the presence of a true mask of protein, or other substance, has been suggested tentatively by Grundland and Bulliard (1938) (in Lovern, 1957). The purpose of the present investigation was to test the validity of the concept of ‘masked lipids’in histochemistry, particularly of phospholipids.

Phospholipids may be identified by Baker’s acid-haematein and pyridine extraction tests. Baker (1946, 1947), after a detailed and careful study, concluded that phospholipid was present in tissues if the acid-haematein test was positive under normal conditions, but negative after prior extraction with pyridine. Although such results were obtained in some parts of cells, others, notably the nuclei, stained more intensely after treatment with pyridine. Baker, although aware of the possibility that this might be due to the presence of masked lipids, was forced on reasonable grounds to conclude that, until the occurrence of such lipids was proved, it was not possible to deduce their presence from these data (also see Cain, 1950). Hence the only permissible conclusion that could be reached was that, where there was an increase in staining after the pyridine extraction test, nothing could be said about the presence or absence of phospholipids. Thus it was impossible to decide if these substances occurred on interphase or mitotic chromosomes. Moreover, since Baker (1946) had shown that nucleohistone from calf thymus yielded ‘spurious’ reactions for phospholipid with his tests, it seemed likely that this material was the ‘interfering substance’ present on the chromosomes.

Thus it was considered advisable to reinvestigate the phenomenon of masked lipids in histochemistry by means of the new methods which are now availabel (e.g. see Lovern, Olley, Hartree, and Mann, 1957) for the identification of very small quantities of phospholipids. Three questions had to be considered: First, is nucleohistone an ‘interfering substance’, or does it contain closely bound (masked) lipids ? If it were the former, then it would be impossible to show the presence of phospholipids in nuclei by histochemical means. If it contained bound lipids, however, these would be expected to occur also in intact nuclei. Secondly, does pyridine extract only fatty material or might it also remove a substance which could mask strongly bound lipid ? Thirdly, in a tissue in which lipid has been apparently unmasked, is there any evidence that fatty substances are present ?

Nucleohistone was prepared from calf thymus in two different ways. In the first (Mirsky and Pollister, 1947), the thymus was homogenized in a 1 M solution of sodium chloride in an M.S.E. homogenizer, filtered, and the dissolved deoxyribonucleohistone precipitated out of solution by diluting the salt concentration to about 0·14 M. It was redissolved in the stronger saline and then reprecipitated by dilution. The second method, that of Doty and Zubay (1956), is a modification of the procedure of Shooter and others (1954) for preparing nucleohistone of very high molecular weight. For this, 200 g of frozen calf thymus were placed in a Kenmix ‘55’ blender, after removal of most of the connective tissue. It was just covered with a saline-versene solution (0·075 M NaCl+0·024 M Na versenate at pH 8·o) containing 2 ml capryl alcohol (octan-2-ol) and homogenized for 3 min until an even homogenate was obtained. The volume was made up to 1 litre with the salineversene solution and homogenized for a further 10 min. The solution was filtered through gauze which had been washed with saline-versene, and spun in 250 ml buckets in an International Refrigerated Centrifuge at 2,800 r.p.m. for 20 min. At the end of the spinning the supernatant fluid WAS discarded. The sediment was resuspended in 500 ml of saline-versene and 1 ml of octan-2-ol, and spun for 10 min at 2,800 r.p.m. The supernatant was discarded and the sediment resuspended in 500 ml of saline-versene solution and 1 ml octan-2-ol. This was spun at 3,000 r.p.m. for 10 min. This last process was repeated three times more. The final precipitate was believed to be pure nucleohistone (58·68 g wet weight).

The method of Folch and others (1951b) was followed. The crude matter was dissolved in a mixture of one volume of methanol to two or three of chloroform and this solution was pipetted, drop by drop, into a column of water in a 2-litre measuring cylinder. The drops fell into a small dish which acted as a false bottom to the cylinder. The liquid was allowed to stand, under the water, at least overnight in a cold room. The methanol passed rapidly into the water, both from the individual drops and from the accumulated fluid, causing severe currents by which water-soluble material escaped into the water. The residual solution in chloroform was then removed, three volumes of methanol were added, and the process was repeated with a fresh column of water. The matter which remained in the chloroform, plus any present at the chloroform-water boundary, was regarded as lipid or as substances probably combined with lipids. The efficiency of this procedure in removing ultra-violet-absorbing substances which were dissolved in methanol-chloroform can be seen in fig. 2.

Lecithins and kephalins, in the free state, can be hydrolysed by boiling for 2 h in a 0·5 N solution of potassium hydroxide in 96% ethanol under a reflux condenser or saponification column (see Lovern, 1957). The unsaponifiable matter is removed from the cooled hydrolysate by shaking with ether. The fatty acids are not removed because they are present as the water-soluble potassium soaps; they are freed from the soaps, therefore, by acidifying to about pH 5·5 and are then removed by shaking again with ether. The remaining aqueous hydrolysate contains such components as choline, glycerophosphate, serine, and ethanolamine. During such hydrolysis there is some conversion of α-glycerophosphate to theβ-form (Dawson, 1957).

The water-soluble components of the hydrolysate were concentrated and an ‘Agla’ micrometer syringe was used to place measured volumes of it on to a line on a sheet of Whatman No. 1 filter paper. Known amounts of the pure substance were placed between these ‘spots’ of the hydrolysate. It was sometimes preferable to use filter paper that had been washed with hydrochloric acid (Hanes and Isherwood, 1949), as this gave cleaner separations. Ascending chromatograms were then run, the solvent being 80% n-propanol.

When dry each chromatogram was studied to visualize a particular component of phospholipids. Choline was demonstrated by the method of Levene and Chargaff (1951), in which the substance is treated so as to produce an insoluble phosphomolybdate. The free phosphomolybdic acid is washed out of the paper and the insoluble material is reduced to yield an intense blue dye. The evidence that a particular coloration on the chromatogram represents choline is (a) that it produces an insoluble phosphomolybdate, and (b) that it has the same R_F value as the control spot for choline (see fig. 3), Glycerophosphate was visualized by the method of Burrows and others (1952), which demonstrates the presence of phosphoric acid. Again, the proof that the material which contains phosphate is α-or β-glycerophosphate is obtained by comparing its R_F value with that of the control, run simultaneously on the same piece of paper. Serine and ethanolamine were identified by spraying with 0·2% ninhydrin in absolute n-butanol and heating for 10 min at about 110° C. This reaction is given by any free amino group. Inositol was sought by the method of Trevelyan and others (1950).

When quantitative estimates were required, three control spots containing different amounts of the known substance, and between these two different concentrations of the hydrolysate, were run. The areas of the developed spots were measured by planimetry. There is frequently a logarithmic relation between the area of the spot and the amount of a particular material present (see Levene and Chargaff, 1951; also Fisher and others, 1948), so that the amounts of a substance present in the known volumes of hydrolysate were estimated from a graph of this relationship obtained from the controls.

In the present study sphingosine has not been estimated. The evidence concerning its occurrence, probably as sphingomyelin firmly bound to protein, in the nucleohistone used in the present investigation, has been presented by Chayen and Gahan (1958).

Lipids were demonstrated by the use of an alcoholic solution of Sudan black B, by Baker’s acid haematein test, and by the orange G aniline blue method described by La Cour and Chayen (1958).

The nucleohistone was prepared by Doty and Zubay’s method (1956). It was boiled, under a reflex condenser, for 2 h in ethanolic potash. Most of the material dissolved and the small sediment which remained was removed (C, see below). The hydrolysate was shaken with ether to remove the unsaponifiable matter and then acid was added to adjust the pH so as to free the fatty acids. As the pH approached pH 6, a pale yellow precipitate consisting of short adhesive threads began to form. The pH was brought to pH 5–6 and the precipitate was removed (B). The filtrate was shaken with ether to remove the fatty acids, the aqueous phase (material A) was concentrated and was examined by paper chromatography. This demonstrated the presence of a phosphate- and a choline-like moiety, both apparently bound to other substances since they did not run freely from the starting line. A spot which seemed to correspond to free β-glycerophosphate was found; no reducing sugars were observed.

The material (B), which had been dissolved by the ethanolic potash but which precipitated at between pH 5 and 6, was hydrolysed by boiling for 3 h under a reflux condenser, with 6 N hydrochloric acid. After shaking with ether to remove fatty acids, the aqueous phase was evaporated to dryness and the residue was dissolved in water for chromatography. One chromatogram was sprayed with ninhydrin and showed many unidentified spots. β-Glycerophosphate, some bound phosphate, and free choline were also demonstrated; no reducing sugars were detected.

The sediment (C) which remained after the original hydrolysis in ethanolic potash was boiled in 6 N aqueous hydrochloric acid for about 2 h and this hydrolysate, treated in the same way as was that of B, was studied by chromatography. Choline and glycerophosphate, both bound and free-running with the correct R_p value, were found. There was much ninhydrin-staining matter, with strongly positive regions or spots which had R_F values similar to those of serine and ethanolamine. No reducing sugars were detected.

Substance C resembles a lecithin in containing glycerophosphate and choline, but the former is present in excess of what would be found in a normal phosphatidyl choline. It is possible that some of the glycerophosphate was combined with serine or ethanolamine in a kephalin-like arrangement. The astonishing fact, however, is that if kephalins and lecithins were present in this fraction, they were not even dissolved by boiling alcoholic potash, which dissolves and degrades normal glycerophosphatides to their component moieties. Substance B, on the other hand, dissolved in the ethanolic potash but was precipitated at a more acid pH. The results of its hydrolysis by acid suggest that it too contained either a lecithin with an excess of glycerophosphate or a mixture of lecithins and kephalins. The fact that they were not degraded by the alkaline hydrolysis and were precipitated on acidification, suggests that they may occur as a lipid-protein complex. The results obtained with ninhydrin support the view that protein may be present in this substance.

It would seem, therefore, that this nucleohistone contained four lipid-like components: those of fractions B and C; the matter in fraction A which yielded the free glycerophosphate after alkaline hydrolysis only; and the bound lipid-like matter of fraction A. It is not known whether the free glycerophosphate of A was derived from the lipid of the heterochromatic regions of the nuclei (Chayen and others, 1959), or from the bound matter. It is likely that the lipids of B and C were bound tightly to some other substance, probably protein, in such a way that they were protected from solution and from hydrolysis unless the protein itself were hydrolysed. This is of concern because of its significance in histochemistry, where a binding which is so resistant would be expected to protect the lipid from extraction in simple solvents.

It would seem desirable to make some estimate of the proportion of lipid-like matter in this nucleohistone. No precise value can be obtained from the present data since only the hydrolysis products of the lipids were determined and fatty acids have not been estimated. This is immaterial for our investigation, the purpose of which is not to identify the lipids exactly but rather to see whether lipid components occur in sufficient quantity to affect histochemical reactions.

The data in table 1 can be evaluated in at least two ways.

(i) According to Chayen and Gahan (1958), sphingomyelin could account for some 6% of this nucleohistone. Thus if their assumptions are correct, 168 mg of the present sample of nucleohistone (dry weight 2·81 g) could be sphingomyelin. If it is assumed that the fatty acid associated with this material is lignoceric acid (Lovern, 1957) the molecular weight of the sphingomyelin would be 844 and hence it follows that 168 mg would liberate about 24 mg of choline. The total amount of choline recovered was 122·7mg (table 1), so that some 98 mg remained to be accounted for.

If this residual choline was present in a phosphatidyl choline which contained C₂₂ acids as both fatty acid moieties, this choline plus about 173 mg of the glycerophosphate would have been derived from about 0·85 g of this lecithin. There is still an excess of 180 mg of glycerophosphate. If it is assumed that this is present only as glycerophosphate (perhaps bound to protein), the total lipid content of this nucleohistone would be 0·85 g of phosphatidyl choline, 0·17 g of sphingomyelin, and 0·18 g of glycerophosphate, totalling 1·2 g out of the original 2·81 g of ‘nucleohistone’. Thus lipid-like material would comprise some 46% of the ‘nucleohistone’.

(ii) This figure must be considered the upper limit and seems absurdly high. A lower limit can be calculated by assuming that sphingomyelin (168 mg) is present and that no fatty acids occur. The rest of the choline (98 mg) might then be combined with 173 mg of the glycerophosphate to yield 271 mg of glycerylphosphoryl choline (see Dawson, 1957); since this compound is water-soluble it would have to be linked to some other moiety, such as fatty acids or proteins. Similarly, the residual glycerophosphate (180 mg) must be assessed as glycerophosphate, which may be bound to protein. Hence the total amount of lipid-like material is 619 mg, which constitutes about 28% of the ‘nucleohistone’. This seems to be a more reasonable figure since no estimates were made of the fatty acid content. Although a lipid, biochemically, is defined as possessing fattyacids, the glycerophosphate (or the sphingosine) moieties may be taken as the histochemical markers for phospholipids.

In a somewhat preliminary study this nucleohistone was boiled in ethanolic potash for 2 h under a reflux condenser. Subsequently any unsaponifiable matter and fatty acids were removed and the hydrolysate was concentrated to a small volume for chromatography. Some glycerophosphate was found, possibly corresponding to the phospholipid of heterochromatic regions (Chayen and others, 1959). No free choline was found but much choline-like material, yielding insoluble deep blue salts with phosphomolybdic acid, was observed. Most of the choline-like matter had not moved from the starting line and correspondingly, phosphate-containing material occurred at the same points on the chromatogram. It seems likely, therefore, that even the reprecipitated nucleohistone contained the phospholipid of the heterochromatic regions plus other, alkali-stable lipids.

Hence it may be concluded that when nucleohistone stains positively for phospholipid, its reaction is probably a true one. Moreover, although it is possible that the pyridine extraction test may remove the surface phospholipid from the heterochromatic regions of nuclei, it may also unmask the tightly bound phospholipids (see below) which will then be free to give true positive reactions with the acid haematein test.

When plant roots were fixed in formaldehyde-calcium and then hardened either in dichromate or in Lewitsky’s (1931) fluid and embedded and sectioned, they stained very weakly with methods for phospholipids (see La Cour and others, 1958), but if after fixation they were treated with pyridine at 60°C for 24 h, all parts of the cells, but particularly the nuclei, stained vividly. A similar great increase in staining was obtained with calf thymus which had been treated with hot pyridine. If the pyridine removed only lipids, the increased staining was most probably due to a spurious reaction of the treated tissue, but if it extracted other material it was possible that this was the agent which masked a very stable lipid.

Pieces of calf thymus were placed in pyridine at 6o° C for 24 h, after which the pyridine was decanted, centrifuged to remove cell debris, and evaporated to dryness. The dried residue was weighed, shaken with methanol-chloroform, and left at 60°C for 2 h in a corked container to dissolve lipid-like substances. The solution was centrifuged to remove insoluble material and then subjected to the Folch procedure and left under water at about 4°C for 3 h. The insoluble matter did not seem to be lipoidal and did not dissolve in hot ethanol-ether. The material which remained soluble in the chloroform after the Folch procedure, plus insoluble matter floating on its surface, were considered as fatty substances and were collected together, dried, and weighed. They corresponded to only about 8% of the total residue obtained from the pyridine (see table 2).

It is very likely that, under the conditions of this test, the extraction of lipids by pyridine was incomplete. However, the experiments do answer the question posed, namely does pyridine extract only fatty material or might it also remove a substance which could be masking strongly bound lipid ?

‘Spurious’ reactions for phospholipids were obtained in plant root cells after fixation in formaldehyde-calcium and treatment with hot pyridine, with 5 % trichloroacetic acid at 90 ° C. for 15 min, or even with hot ethanol-ether. Each of these procedures greatly increased the staining for lipids throughout the cells but particularly in the nuclei. Since the action of trichloroacetic acid is fairly well known, it was thought advisable to examine the tissues after treatment with this liquid to see if the staining reaction was indeed spurious or whether lipid-like matter was present which might account for the reaction.

Roots of Trillium grandiflorum and Vicia faba were fixed in formaldehydecalcium, and those of Scilla campanulata in 45% acetic acid for 5 h. A few roots from each batch were removed for cytological examination to ensure that only the nucleoli stained appreciably for lipids. The rest of the roots were treated with 5% trichloroacetic acid at about 90° C for 15 min. Some were taken for cytological inspection to confirm that this procedure had induced strong lipid reactions in the nuclei and in the cytoplasm; these reactions were least marked in Vida roots. The remainder were put into alcoholic potash and boiled under a reflux condenser for 2 h, during which the roots and the liquid turned yellow. The hydrolysate was removed, concentrated, and examined by chromotographic methods.

The chromatograms run from the hydrolysate of S’, campanulata roots showed the presence of substances corresponding to α- and β-glycerophosphate (the β-form is produced readily from its isomer by hydrolysis), inositol, serine, and possibly arginine; a small amount of material resembling choline was present on the starting line. The hydrolysate from Trillium roots contained material which did not move from the starting line and which gave the colour reactions for choline, phosphate, and reducing sugars; α-glycerophosphate and possibly serine and arginine were also detected. These results were similar to those reported for reprecipitated nucleohistone (above). From Vida roots, α-glycerophosphate, some reducing sugars, and traces of amino-acids resembling serine and possibly arginine were obtained.

It seems likely, therefore, that lipoidal material was present in the roots after treatment with trichloroacetic acid and that this matter, or at least some of it, was rendered soluble in the potash but was not hydrolysed to its components. Hence it behaved similarly to the bound lipids of the calf thymus nucleohistone.

There have been two major obstacles to the histochemical study of the phospholipids, the first being that frequently reactions are intensified by treatment with reagents which would be expected to remove these substances, and the second that nucleohistone is an ‘interfering substance’, that is it stains as if it were a phospholipid. The data presented in this communication suggest that the solution to these problems lies in the possibility that true masked lipids occur which are so tightly bound that the usual lipid-solvents are unable to remove them. This binding may be so firm that boiling alkali, which hydrolyses free lecithin-like phospholipids completely, may not even dissolve these lipid-protein complexes, as was found with the sediment C. Moreover, not more than about 10% of the material extracted by hot pyridine from calf thymus was lipid (also see Olley and Lovern, 1954). Thus it seems likely that hot pyridine removes non-lipid matter, possibly protein, which normally masks tightly bound phospholipid; this lipid can be visualized also by treatment with trichloroacetic acid. Hence it would appear likely that in both the nucleus and cytoplasm, truly masked lipids are present which, without special treatment, are not availabel for staining by virtue of their association with some other substance such as protein and nucleic acid.

Some idea of the manner in which this protection is conferred was suggested by the study of phospholipids in nucleohistone. Since nucleohistone, prepared by two different methods, contained phospholipid-like matter, perhaps in relatively high concentration, it is not surprising that nuclei and chromosomes give positive reactions for phospholipids. The binding of these lipids to the nucleohistone, however, is so strong that they may not be availabel to the staining reagents before the bonds linking them are broken. This would explain the increase in staining for phospholipid by nucleohistone and by nuclei after treatment with pyridine, trichloroacetic acid, and other solvents.

It has been shown that after extraction with trichloroacetic acid, when an increase in such staining was observed, lipid-like matter was present in appreciable quantity in the tissues.

In the authors’ opinion it is advisable to restrict the term ‘masked lipids’ to those lipids which are truly ‘masked’, namely rendered unavailabel to the staining reaction by virtue of their binding or close physical association with other substances such as protein. Lipids which cannot be resolved easily because of their colloidal state may often be visualized by the use of a more sensitive method, such as Berg’s benzpyrene technique (Berg, 1951). Such lipids, in the authors’ view, should be called ‘finely dispersed’; they can be distinguished readily from masked lipids by their solubility.

That extraction with pyridine enhances lipid reactions has been observed in many tissues and raises the problem of whether this close association of phospholipid and protein may not be very widespread. The more recent biochemical work of Folch and Lees (1951), Folch and others (1951 a), Spiro and McKibbin (1956), Smith and others (1957), and Bruemmer and Thomas (1957), as well as the histochemical studies of Berenbaum (1954), Serra (1958), and ourselves, all suggest that frequently protoplasm may be a lipid-protein complex.

Some of the difficulties concerning the possible presence of phospholipids on chromosomes would seem to have been clarified, and defined more sharply. Mitotic chromosomes in plants and the heterochromatic segments of many interphase nuclei contain a phospholipid-like substance which is bound to protein, the whole complex being soluble in formalin (Chayen and others, 1959). It is likely that the lipid moiety would be removed from tissues by ethanol-ether or by pyridine. These solvents do not remove the staining properties of chromosomes and nuclei, however, but increase them. It is suggested that this effect is due to the increased availability of the tightly bound phospholipids of the nucleoprotein of chromosomes and interphase nuclei.

We are grateful to Professor J. T. Randall, F.R.S., for allowing us facilities in his Laboratory. We are very much indebted to Dr. H. B. Fell, F.R.S., and Dr. C. Long for helpful advice. One of us (J. Chayen) also wishes to acknowledge his debt to Dr. J. A. Lovern and Dr. J. Olley for instruction and help with the biochemical methods and for their kind interest. We are also indebted to Dr. G. Zubay for his help in preparing the unprecipitated nucleohistone.