Notes on Lillie's (1945) Naphthol Green B Stain for Connective Tissue
During preparation for a class in histology at Victoria University a triple stain was sought for the distinction of muscle and connective tissue that would give consistently good results using stains that are usually readily available in a department of zoology. The material to be stained was sections of vertebrate tissue, in particular that of mouse and chick embryo. A staining method giving green connective tissue, pinkish cytoplasm and muscle, and blue nuclei was wanted. Gurr (1962) gives the following stains for obtaining a very similar staining reaction, viz. Weigert's haematoxylin, eosin yellowish, and naphthol green B. His method is as follows:
|1.||Paraffin sections are mounted on the slide and brought down to distilled water.|
|2.||Stain for 6 mins. in a freshly prepared mixture of equal volumes of Weigert's haematoxylin A and B.|
|3.||Wash thoroughly in tap water; then stain for 3 mins. in 1% aqueous eosin y.|
|4.||Wash in tap water; then immerse in a solution of 10% ferric chloride for 5 mins.|
|5.||Rinse well in distilled water; then stain for 5 mins. in a 1% aqueous solution of naphthol green B.|
|6.||Differentiate for 2-3 mins. in 1 % acetic acid.|
|7.||Drain well; then dehydrate with acetone, afterwards clearing in acetone-xylol; then mount.|
- connective tissue, green.
- muscle and cytoplasm, pink.
- nuclei, brown.
With the above procedure results were far from consistent and a clear differentiation of muscle and connective tissue was not obtained.
Testing the staining procedure consisted of working through the schedule using the times given by Gurr, and varying the times in the eosin, ferric chloride, naphthol green B, and acetic acid baths. Material used was F.A.A. fixed paraffin embedded sections of (a) mouse head, abdomen and duodenum, cut at 8μ on a rotary microtome, and (b) sections of frog stomach cut at 6μ on the same microtome. Staining procedures using alcoholic eosin as the plasma stain, and connective tissue stains of naphthol green B (George T. Gurr, 09506) in absolue alcohol and fast green FCF (Flatters and Garnett Ltd.) in absolute alcohol, clove oil, and cellosolve (ethylene glycol monoethyl ether) were also tried.
Results and Discussion
The main fault in staining was loss of the eosin from all but the strongest eosinophil cells (for example, red blood cells) on dehydration prior to mounting. This resulted in greenish muscle and cytoplasm. The suggested dehydrant, acetone, seemed too vigorous. Even the minimum dehydration time permissible caused a loss of eosin from some muscle and cytoplasm, causing uneven staining of sections. An attempt to correct the fault by dehydrating through a series of graded alcohols was not more page 2successful. Increasing the time in the eosin bath showed that overstaining was impossible, and did not help to keep the eosin in the tissues during dehydration.
Cutting the time in the naphthol green B to as little as 15 secs, still allowed the green to colour muscle and cytoplasm (the eosin being removed in dehydration). An attempt to stain first in the naphthol green B followed by differentiation in 1% acetic acid, then in eosin y in alcohol (0.5% soln. in 90% alcohol) was, again, unsuccessful. The eosin would not stain muscle tissue or cell cytoplasm which had already been stained with naphthol green B. Lengthening the time in the acetic acid bath had little effect.
Gurr (1962) gives no information as to the purpose of the 10% ferric chloride solution, or the reason for dehydrating in acetone and clearing in 50-50 acetonexylol. He gives a footnote that the original reference for the technique is Lillie (1945).
Lillie found that the staining method now under discussion was the most satisfactory using naphthol green B as a connective tissue stain. It was evolved during exploration for mordants for selective connective tissue staining, following a report that naphthol green B dyed wool green with an acid iron mordant. The ferric chloride bath in this staining method is described as a mordant by Lillie (p. 31), but it was found that omitting the ferric chloride prior to staining with naphthol green B resulted in deep green staining of all tissues. This suggests that the action of ferric chloride is not to mordant the tissues for the green dye, but to heighten the selectivity of the dye for connective tissue. Indeed, Lillie concludes in his paper that the selectivitiy of collagen stains is best at pH levels below 3.0, preferably below 2.0, and disappears at higher pH levels.
The best results using Lillie's triple stain were obtained by using the following adjusted staining schedule:
|1.||Sections are brought to water.|
|2.||Stain in freshly prepared Weigert's haematoxylin 6 mins.|
|3.||Wash in tap water, and stain in 1% aqueous eosin y for 5 mins.|
|4.||Rinse in tap water, and immerse in a 10% solution of ferric chloride for 3 mins.|
|5.||Rinse in distilled water and stain in naphthol green B (1% aqueous) for 15 seconds.|
|6.||Differentiate in 1% acetic acid 2 mins.|
|7.||Dehydrate as quickly as possible in acetone, clear in 50-50 xylol-acetone, and mount.|
Gurr (1962) states that many acid dyes devoid of sulphonic groups (—SO3—) can be completely extracted from tissues by ordinary solvents such as water and alcohol. He suggests this is because tissue proteins, already possessing carboxyl groups identical with those of carboxylated (acid) dyes, have little tendency for exchange of ions with the acid dye. The dye ion is thus not in chemical combination with the tissue protein.
Eosin y is a carboxylated but not a sulphonated dye. It therefore seems reasonable to expect a better result with a sulphonated dye for the plasma stain in Lillie's triple stain. Most of the plasma stains used by Lillie in his investigations were sulphonated, e.g. acid fuchsin, Biebrich scarlet, aniline blue, and many others. Eosin, however, gives excellent contrast as a plasma stain. It was decided to use the eosin in 90% alcohol and to follow it with the connective tissue stain in absolute alcohol or a clearing agent, rather than substitute a sulphonated dye for the plasma stain.
Gurr (1962, p. 59) gives the solubility of naphthol green B in absolute alcohol as 3% at 15°C. I found, however, that it is insoluble in absolute isopropyl alcohol, and only soluble in absolute ethanol to 0.25%. This 0.25% soln. failed to give selective page 3connective tissue staining following a plasma stain of 0.5% eosin y in 90% alcohol. Therefore replacement of naphthol green B with some other green connective tissue stain seemed advisable.
Fast green FCF was used by Lillie (1945) as one of a number of dyes that gave good connective tissue staining. He used it in aqueous solution, but in the present instance the fast green FCF was used in clove oil, and with eosin y in 90% alcohol gave excellent differential staining of muscle and connective tissue. Weigert's haematoxylin was replaced with Delafield's to give the nuclear stain, thus eliminating a source of waste as Delafield's haematoxylin need not be freshly made every few days. (Histologists will attest to the fact that Delafield's haematoxylin becomes better with age.)
The staining procedure for the triple stain using Delafield's haematoxylin, eosin y and fast green FCF is as follows:
|1.||Bring paraffin sections to water.|
|2.||Stain in Delafield's haematoxylin 6 mins. to give overstain.|
|3.||Differentiate in acid/alcohol until nuclei only remain coloured.|
|4.||Wash in tap water, and blue nuclei in tap water or in lithium carbonate solution if tap water is not sufficiently alkaline.|
|5.||Dehydrate to 70% alcohol, and stain in 0.5% eosin y in 90% alcohol for 40 seconds.|
|6.||Rinse in 95% alcohol, and differentiate in 100% alcohol 1 minute.|
|7.||Stain in a saturated solution of fast green FCF in clove oil for 50 seconds-1 minute.|
|8.||Wash off the excess stain thoroughly in xylol, and mount in DPX mountant.|
- connective tissue, green.
- muscle and cytoplasm, pink.
- nuclei, deep blue.
The above staining schedule was found to be very effective for mouse embryo tissue, giving very good results each time. The same procedure was used for sections of frog stomach, and gave excellent results when the time in the fast green was increased to 3 minutes.
The fast green was also used in cellosolve and in absolute alcohol. With cellosolve as the dye solvent the best results for mouse embryo tissue were obtained with a 0.25% solution and a staining time of 12 seconds. The excess stain must be washed from the slides by moving them in cellosolve for a few seconds, prior to clearing in xylol and mounting. Washing the slides in alcohol or xylol does not remove the stain quickly enough, causing the staining time to be increased and resutling in some muscle being coloured green. Because of the short staining time when cellosolve is used as the dye solvent the method does not allow the degree of control of connective tissue staining that is possible when the fast green is dissolved in clove oil. A 0.25% solution of fast green in cellosolve gave excellent differential staining of muscle and connetcive tissue in sections of frog stomach with a staining time of 35 seconds. Again, excess stain was removed in cellosolve.
A 0.026% solution of fast green in ethyl alcohol was also successful as a connective tissue stain.* With mouse embryo sections the best staining time was 20 seconds, and with frog stomach 1 minute 50 seconds. In both cases excess stain was removed from slides by moving them for a few seconds in absolute alcohol.
The above method using Delafield's haematoxylin, eosin y in alcoholic solution and fast green FCF in clove oil, and its variations give reliable results. It could be described as being a derivative of the method of Lendrum (cited in Lillie, 1945). page 4He used a nuclear stain of alum haematoxylin followed by a plasma stain of either gallic acid-eosin-erythrosin-phloxine in alcoholic solution, or formol-eosin in alcoholic solution, followed by a connective tissue stain with tartrazine, fast green FGF, aniline blue, or methyl blue, dissolved in cellosolve.
|1.||Lilie's triple stain of Weigert's haematoxylin, aqueous eosin y, naphthol green B is unreliable as a differential stain for connective tissue.|
|2.||The main fault with the method is rapid loss of eosin from sections during dehydration.|
|3.||Excellent differential connective tissue staining can be obtained using a plasma stain of eosin in 90% alcohol with the connective tissue stain in a clearing agent or in cellosolve.|
|4.||The recommended technique employs Delafield's haematoxylin as nuclear stain, eosin y in 90% alcohol as plasma stain, and fast green FCF in clove oil or in cellosolve as the connective tissue stain.|
Gurr, E., 1960: Encyclopaedia of Microscopic Stains . Leonard Hill, London.
Gurr, E., 1962: Staining Animal Tissues, Practical and Theoretical . Leonard Hill, London.
Gurr, E., 1965: The Rational Use of Dyes in Biology . Leonard Hill, London.
Johansen, D. A., 1940: Plant Microtechnique . McGraw-Hill, New York, London.
Lillie, R. D., 1945: Studies on selective staining of collagen with acid aniline dyes. J. Tech. Meth. , 25, pp. 1-47.
Baker, J. R., 1958: Principles of Biological Microtechnique . Methuen, London.
* Gurr (1960, 1962, 1965) gives the solubility of fast green FCF in absolute alcohol as 9% at 15°C. Johansen (1940) states the solubility of the dye to be 0.33% in absolute alcohol. My own findings were more in line with those of Johansen.