John Luther Mohr , Ph.D.
Department of Zoology
University of Southern California1
I. HISTORICAL REVIEW. Among the natural coloring matters, brazilin is the only one dating from the pre-Columbian era to survive the competition of the synthetic dyestuffs in microscopical technique, and although it has had staunch and distinguished advocates (Eisen, Schaudinn, Hickson, Champy, Belling), it is known to most of today's microscopists, if at all, vaguely as a red homologue of hematoxylin. Because it has had a particularly interesting history and because for a few specific purposes it offers advantages over other stains known to the writer, the following account is offered.
While the dyestuff is not so old in commerce as the reds from madder (known in Egypt even in pre-dynastic times), during the Middle Ages a red dyewood coloring matter was imported into Europe from southern Asia under the varying names, "brasilium", "bresillum", and " brezellem" (Oxford English Dictionary, I:1066). The name is of uncertain origin, several possible etymologies having been suggested, among them that it is a corruption of an Asiatic word, presumably now lost, for the redwood. According to Leggitt (1944, p. 50), the dyestuff appears as verzino, the usual Italian designation, in the taxlists of Ferrara as early as 1193 and of Barcelona, in Moslem Spain, in 1280. Sarton (II:1042) cites the use of brazilwood to make rose-colored letters in manuscripts as described by Abraham Ibn Hayyim working at Louie, Portugal, in 1262. Marco Polo (c. 1254-1324) reported the occurrence of brazil-wood forests in Sumatra, Siam, the Nicobars, Ceylon and along the Malabar coast of India and even tried (the seeds did not thrive) to grow brazil at Venice. He distinguished among the qualities of wood from the different sources. At this same period, according to Yule (1875), brazil-wood was a royal gift from Siam to China in customary interchanges between the two courts.
Most specific information available on the dyewood commerce of the early Renaissance is that in the veritable handbook of international trade (Libro di divisamenti paesi e di misure di mercatantie) compiled by the Florentine, Francesco Pegolotti, traveller and merchant of the international banking house of Bardi, about 1340 (Yule, Sarton). Pegolotti recognized several types of brazil of which verzino colomi or colombino was definitely that of Quilon on the Malabar coast (Yule), the kind held in highest esteem (i.e. commanding the highest price). Verzino Ameri is taken by Yule to have been of Lambri in northwest Sumatra, a grade of good quality, while verzino Seni (possibly for Sini) may have been that brought by Chinese merchants from Siam as far as India, a grade costing about a third as much as the Malabar variety. The fascinating Rihla (Journey) transcribed in 1356 by Ibn Juzayy from the verbal account of Ibn Battuta, the greatest traveler, "not excepting Marco Polo, in medieval times" (Sarton III:1614), contains a description of brazil-wood on the Malabar coast paralleling and confirming Polo's earlier account.
In England, whose textile and dyeing industries lagged behind those of the Italian city-states, of France and of the Low Countries, the red dyewood had a vernacular name by the fourteenth century as Chaucer's Nun's Priest (Canterbury Tales, ca. 1386) uses the phrase, "His colour for to dyghen with brasile". This "brasile" (a baker's dozen variant spellings are provided by the Oxford English Dictionary) or brazil-wood was the product of a leguminous tree, Caesalpinia, and probably mainly the Sappanwood, C. Sappan, although other caesalpinias of the Asiatic tropics may have been included. The adventurous Venetians were the major importers of the dyestuff across the land routes from India, supplying the dyers of the rest of Europe.
With the discovery by the Portuguese of an all-water route to India, the center of gravity of the spice and dyewood trade shifted from Italy to the Iberian peninsula. Soon thereafter a vast new supply of brazil-wood was found by the same Lusitanians in the Western Hemisphere as recounted proudly by Camoens (1572) in his epic of the Portuguese people, Os Lusiados:
"Mas ca onde mais se alarga, ali tereis
Parte tambem, co'o pau vermelho nota
De Santa Cruz o nome the poreis;"
"But where the land spreads broadest ye shall claim
The part that for its red wood is renowned
Of Santa Cruz ye shall bestow the name.
(Aubertin translation of 1884)
This Santa Cruz "co'o pau vermehlo", with the red dyewood tree and for it called alternatively Terra de Brasil, became quickly Brasil (on which the historian De Barros commented sourly "as if the name of a wood for coloring cloth were of more moment than that of the Wood which imbues the Sacrements with the tincture of Salvation" -- and attributed the change "to the suggestion of the Evil One" (Yule, op. cit., p. 368). The New World yielded also Campeachy wood or logwood, the source of hematoxylin. It is curious that both brazil and logwood suffered marked disfavor in England. As early as the thirteenth century the bylaws of the Painters' Guild of London forbade painting on gold or silver except with fine (mineral) colors " e nient de brasil, ne de inde de Baldas, ne de nul autre mauveise couleur" (Yule, op. cit., p. 371). Act 24, 1532-1533, of Henry VIII states, "Diers.......haue vsed deceyuable waies in dyeing with brasell and such other lyke subtilties" ("deceivable" here with a now defunct meaning "having the habit of deceiving"), while, Leggett (1944) tells us of logwood, "In 1580, an Act of Parliament forbade it to be used for dyeing, and large quantities thereof were burned."
Biologically decoctions of brazilwoods appear to have been used earliest by Reichel (1752, 1758) in his investigations of the vessels of plants. After boiling scrapings of Pernambuco wood, he dipped plant stems (Phaseolus, Lupinus) into the cooled brew and, because the colored fluid passed along the spiral vessels, he concluded that these carry sap and not air as maintained earlier by Malpighi. Reichel cut both cross and longitudinal sections of the stained stems. A little later Hedwig (1782) applied the method to tissues of squash and Mayer (1793) did further work on plant vessels with Pernambuco and brazilwood decoctions. Lewis (1942) remarks, "It was inevitable that logwood (Haematoxylon) should soon be used in these injections" and describes the first use of that tincture by a biologist, Knight (1803, 1808), in the same manner as were the brazil decoctions on similar plant materials.
Saefftigen (1884) of Heidelberg, a student of Otto Biitschli, was at least among the earlier workers using brazilin on animal tissues having employed unspecified brazilin and hematoxylin methods on sections of thorny-headed worms (Acanthocephala). Both Breglia and Flechsig in 1889 published neurological methods employing redwood extracts. Breglia stained sectioned pieces of central nervous system (mammalian) fixed in Muller's or Erlitzki's fluid and mordanted with lead, iron or copper salts in various aqueous solutions of Pernambuco wood extract. The extractions, like those he made from logwood for hematoxylin, were carried out with alcohol. Borax and ferrocyanide mixtures were in some cases used with the dyes.
Flechsig combined a Golgi sublimate-dichromate impregnation of human brain with his brazilin stain cutting his sections at 50 microns. For this he made a stock solution of one gram of pure extract of Japanese redwood in ten grams of absolute alcohol. Diluting this with 900 cc. of water and five of a saturated aqueous solution of Na2SO4 and tartaric acid, he stained the sections three to eight days at about 35 Deg-C. After washing, the sections were treated with potassium permanganate followed by an oxalic acid--potassium sulfite bath as with a Weigert-Pal stain. Rawitz (1895) mentioned the use of an alum brazilin employed as a nuclear stain but gave no particulars. Heimann (1898) used a "Delafield's brazilin" for staining ganglion cells and likewise provided no details about his procedure.
Eisen (1897) proceeded from Böhmer's alum hematoxylin formula apparently being the first to use brazilin in a really effective manner on a variety of objects. He allowed his solutions to ripen for a few weeks until they were a deep, brilliant red with numerous bluish flakes. These flakes, insoluble in water, he considered to be brazilein and these he filtered out, dissolving them in glycerine-alcohol for use. His critical judgement was that "for scientific investigations of (animal) cell structure and cell differentiation both hematoxylin and brazilin have but little use". He considered brazilin a satisfactory stain for classroom or pathological material and a very good dye with no tendency to overstain for plant materials. Of its metachromasy he noted "with some tissues it is a treble stain", e.g. with newt spermatozoa. He used weak solutions of nigrosin or indulin for counterstaining.
Wide use of brazilin began with Eisen, Schaudinn (1900) and Hickson ( 1901) just as Hickson was impelled to remark about brazil-wood, "of recent years it has been superseded by other colouring substances and practically driven out of the market". Major supplies of brazilin in the last 75 years have come from a few species of Caesalpinia, a tropicopolitan genus of leguminous trees, although possibly the same tinctorial principle could be extracted from many species of Caesalpinia and the closely related genus Peltophorum. These are known as the soluble redwoods or the brazil-woods in contrast to another group of of dyewoods, the insoluble redwoods (barwood, camwood, etc.) which do not yield their color content to aqueous extraction.
The original "brazil" is very likely Caesalpinia Sappan Linnaeus ( from Malay "sapang" probably meaning Japan referring to supposed origin) indigenous to the Asiatic tropics from India to Malaya and present in the Philippine Islands at least from prehistoric times. The best of dye sources is Pernambuco wood from C. crista Linnaeus of Brazil and Jamaica (although the species is pantropical according to Merrill). Brazil-wood (in the current restricted commercial sense,) C. brasiliensis Linnaeus, which also grows in Brazil, is reported to yield about half the amount of pure dye per unit of wood obtainable from Pernambuco Wood. The fact that all three of these trees were known to Linnaeus and that he described and named them in his Species Plantarum of 1753 is some indication of their prominence.
Caesalpinia echinata Lam., of South and Central America is often called peachwood and is a dyewood of good yield. While the dyers' cant seems to be a little less than standardized, the foregoing equivalents (scientific and commercial) appear to be generally valid. The writer can give no reasonably precise equivalents for the following: All Souls' Wood, Bahamawood, Bahiawood, Bukkumwood, Jamalcawood, Japanwood, Limawood, Nicaraguawood, and St. Martha wood. They may refer to one or more of the species above or to other Caesalpinias or Peltophorums of minor commercial importance. Brasiletto (braziletto) and sobrazil have been variously applied, but somewhat more consistently to the species of low yield. Hypernic, seemingly confined to American dyers' cant, refers either to a soluble redwood or to raw extract. Dividivi, which may include Caesalpinia tinctoria, is referred to by Knecht (1911) as used like Pernambuco wood in textile printing, but the name applies more usually to species employed for their tannins.
As early as 1808 M. E. Chevreul crystallized pure brazilin from soluble redwood. In the following century particularly through the investigations of English and German chemists the empirical formula of the compound and its probable structural formula were elucidated being respectively C16H1405 and, according to Pfeiffer,
In comparison hematoxylin may be considered a brazilin oxidized one step further to C16H1406 with the structural formula:
A molecule of brazilin with one and a half molecules of water forms a light red crystal of reported bittersweet taste (the writer's powdered, presumably not hydrated supply, tasted unintentionally, is very bitter and not at all sweet). Brazilin itself is a leuco-compound, its solutions being without color or at least pale in pure form. On oxidation, which proceeds somewhat slowly in absolute alcohol and more rapidly in aqueous solution, two hydrogens are dislodged leaving brazilein, C16H12O5, which is reddish brown in solution precipitating in shiny, silvergray flecks reddish brown with incident light.
Resonance (isorrhopesis), an equipoise between two of the theoretically possible ionized forms, is thought to be responsible for the coloration.
RESONANCE OF TWO IONIZED BRAZILEIN FORMS
Brazilein is, then, a chromophore or color-bearing compound, but it lacks strong auxochrome groups to make, it a fast dye. Likewise hematoxylin is oxidizable to a homolog of brazilein, hematein, which is a chromophore. Industrially brazilin is commonly used either in the form of raw decoctions or in purer form with a variety of salts which promote oxidation and form lakes (with the brazilin) which are complete stains ( i.e. have color and fastness) the lakes being formed either before dyeing or in the thing to be dyed itself. Formerly brazilin in various forms was widely used in the dyeing of cotton and wool, in printing fabrics, in coloring leather, in the making of wall-paper, and in the compounding of red writing ink. Tin, chromium, aluminium and iron salts were used as mordants. Karrer states, "even today it is still used in cotton printing, and for dyeing cotton which has been mordanted with sumach or tin salt".
Use in microtechnique, although it follows the general lines of industrial methods in some respects, requires a pure product such as is well defined under Colour Index No. 1243 (or Schultz Farbstofftabellen Num. 1375). Biological staining methods employing the purified compound (as opposed to the decoctions of Reichel and others) may be considered as members of groups of varying importance. First of these was the methylene blue-brazilin combination originated by Schaudinn (1900) and employed by him on the shelled rhizopod, Trichosphaerium and later by Lucke on the foraminiferan, Saccammina. Sections were stained five minutes in saturated aqueous methylene blue and a day in brazilin ( solution not specified) and were then differentiated about an hour in 43% ethyl alcohol. By this method nuclei were stained clear red, inclusions blue and plasma pink.
The alcoholic iron brazilin technique of Hickson (1901) is certainly the most influential of brazilin methods yet devised. Hickson mordanted sections from one to three hours in a 1% solution of ferric alum in 70% ethyl alcohol, rinsed in 70% alcohol and stained in 0.5% brazilin in 70% alcohol from three to 16 hours. He differentiated in 70% alcohol, dehydrated, cleared and mounted. Hickson used the stain on a number of animal tissues including those of newt, dog and cat as well as on the suctorian, Dendrocometes. He held the advantages of alcoholic iron brazilin to be the avoidance of water and excellent metachromasy.
Hickson's regimen has been used and modified by various botanists. Miss Dale (1903), who used alcoholic iron brazilin on fungi, spoke of its results as "very certain" with no overstaining. Her preparations seemed equally good whether material was stained before or after sectioning. Cejka (1912), after fixation with mercuric chloride solutions followed by iodine treatment, used alcoholic safranin-brazilin stain regressively for a fungus of human hair.
It was on Hickson's stain that Belling (1928) based his very precise iron brazilin method for pollen mother cells. Following a chromic-formolacetic fixation, pollen mother cells were washed carefully, mordanted in fresh alcoholic ferric alum as much as three days, washed again, and stained for some hours in a well ripened solution of 0.5% brazilin in 70% alcohol. Belling made up his brazilin solution with absolute alcohol, differentiated his materials under the microscope (100-200 diameters magnification), and used only fresh, clear alcoholic ferric alum if destaining other than with alcohol alone was required. These refinements along with the deliberateness of his procedure may account for the fineness of his results. (Conversely, some of the criticism of brazilin for similar cytological work may spring from techniques less exacting). The schedules reported by Webber (1929), Capinpin (1930) and Sax ( 1931) are acknowledgedly derivative--and only that of Webber differs enough from Belling's to warrant review. Webber fixed anthers only t en minutes in a mixture of ten parts of glacial acetic acid to 25 parts o f absolute alcohol, washed in absolute alcohol and moved through a graded series to 60% alcohol from which he changed to an alcoholic ferric alum mordant. He mordanted only an hour and stained the anthers for a like period in 0.5% alcoholic brazilin differentiating in 70% alcohol followed by alcoholic ferric alum. This method, which is markedly shorter than Belling's, Webber considered good for matured chromosomes, cytokinesis, tetrads and some other structures, but he recommended the original Belling schedules for finer details of the thread stage. Both Belling and Webber recommended use of yellow-green filters for observation of brazilin-stained preparations.
Criticism of the alcholic iron brazilin and iron hematoxylin methods for pollen mother cells (characteristically milder for brazilin, however) has been voiced by Darlington (1933) and Darlington and La Cour (1947) and echoed by Grigg (1946) particularly on the score of density of cytoplasm as an obscuring factor in such preparations. Some precautions against such density will be discussed in part two of this paper.
It is not to be inferred from the foregoing account that the botanists preempted use of the iron brazilin method of Hickson. Other British zoologists employed the technique (inter alia Dendy, 1914, working on gametogenesis on the sponge, Grantia, at Plymouth); in the United States it appears to have been favored at Harvard (cf. Smallwood, 1904, on snail maturation stages and other workers on coelenterates, trematodes and chicks); in Bohemia at Prague it was employed extensively ( inter alia by Bilek, 1909, for ascarid cytology; Vejdovsky, 1912, and Vesely, 1913, for insect spermatogenesis). Gutherz (1922) commented favorably on the nuances achievable with the method as well as upon its permanence. These are references encountered in the course of reading and do not constitute a balanced account of the use of alcoholic iron brazilin in zoology.
Champy (1913) showed brazilin to have considerable possibilities for variation as a zoological stain. For rapid work he used it like a Weigert's hematoxylin (but with tones in red) and counterstained with light green. To achieve a preparation resembling somewhat one stained with the safranin and light green of Benda, he would fix in Flemming's solution and dye with ammonium alum brazilin (see Champy's plate II, figures 4, 5, 12). The alum brazilin he made by adding 5% of a saturated alcoholic solution of dye to a hot saturated solution of ammonium alum. With the supernatant portion of the cooled mixture he stained 20 to 25 minutes and counterstained with light green. Most striking was Champy's third method. Sections of amphibian testis fixed in Bouin's fluid he stained in iron hematoxylin and differentiated somewhat more than ordinarily. Then he stained the sections 24 hours in alum brazilin, differentiated slightly with alcohol, finishing with either light green or Congo red. Effects of this multiple stain are shown in Champy's plates VI and VII.
Better known in the United States is Bensley's "Brasilin-Wasser-blau" combination (Bensley, 1916; Bensley and Bensley, 1938). Published color plates show effective use of this technique on thyroid and on pancreas. Endocrine tissues preserved in Zenker's or Zenker-formol fixative are stained for one to several hours in freshly prepared phosphotungstic acid - brazilin solution:
Phosphotungstic acid 1 gram
Distilled water 100 cc.
Brazilin 0.05 gram
(ripened with a few drops of barium carbonate filtered hydrogen
peroxide or of molybdic acid).
Sections are counterstained in phosphomolybdic-Wasserblau solution ( phosphomolybdic acid, 1.0 gram; aniline blue, water soluble, 0.2 gram; water, 100 cc.) one to five minutes, washed rapidly in distilled water, dehydrated in several changes of absolute alcohol, cleared in toluene and covered with mountant. Brief mordanting in fresh stannic chloride solution improves contrast, "but will detract greatly from the transparency and beauty of the preparation". By this method chromatin stains red, secretion antecedent in pale blue droplets; mitochondria, reddish purple; connective tissue, bright blue; and erythrocytes, orange red.
Mawas (1919), utilizing the fact that brazilin by itself is easily removed from tissue while with metals it forms fast lakes, devised a useful histochemical test for iron. Tissues fixed with non-metallic, non-mordanting fluids were imbedded, sectioned, decerated, hydrated and stained with 0.5 to 1.0% aqueous or alcoholic brazilin coloring tissue iron dark brown and chromatin red violet. Chloroform-alcohol would, moreover, extract the dye from the chromatin while the brazilin lakes of the tissue iron were not affected.
O'Leary's brazilin method for myelin sheaths (Cowdry, 1948) proceeds from nervous tissue fixed and mordanted with Miiller's fluid up to a day and sectioned in either paraffin or celloidin. The staining fluid, compounded of 10 cc. of well ripened 10% brazilin (Grubler) in absolute alcohol and five drops of glacial acetic acid in 100 cc. o f distilled water is applied for an unspecified period. Sections are washed in distilled water and differentiated briefly (up to five minutes) in 0.25% potassium permanganate. When gray matter appears pink and white matter is brilliant red, differentiation is complete. Excessive action of the permanganate is checked with Weil's solution (weak oxalic acid-sodium bisulfite solution). Incomplete differentiation can be remedied by reapplication of permanganate followed again by Weil's solution. Further treatment depends on the mountant used.
There remain a few miscellaneous applications of brazilin. Bensley and Bensley (op. cit., p. 107) recommended brazilin as a background stain for muchematein preparations, Reichert (1909) used aqueous brazilin (or hematoxylin) with added hydrogen peroxide for demonstration of flagella of bacteria. Mencl (1911) used a conventional Hickson alcoholic iron brazilin in his study of the "nuclear equivalents" of Azotobacter and for Guarnieri bodies, remarking its superiority over Heidenhain's hematoxylin for these objects. Maire and Tison (1909) are the only workers, to the writer's knowledge, to use a brazilin-eosin combination. These investigators employed an iron brazilin with eosin (after fixation in Maire's aqueous picro-formol preservative) on the plasmodiophoran, Sorosphaera veronicae.
Yamaha (1937) used a number of dyes, among them brazilin and hematoxylin, in paraffin oil as vital stains for algae, apparently in the case of these two with unsatisfactory results.
In summary extracts of brazil-wood have been used for dyeing fabrics from medieval times. Soluble redwood extracts have been used in biology since the eighteenth century and from the latter part of the last century various techniques employing pure brazilin have been applied. Of the various combinations devised the most generally valuable has been the alcoholic ferric alum followed by alcoholic brazilin of Hickson but various mordants other than ferric alum have been used. Methylene blue, Wasserblau, light green, safranin, eosin and Congo red have been used as counterstains. It is reported to be effective after a variety of fixatives including the picro-formol and picro-formol acetic, craf, osmium tetroxide containing and mercury-dichromate containing types.
BIBLIOGRAPHY
To be continued
1. A portion of this work was carried out while the writer was Research Associate of the Allan Hancock Foundation, the support of which is gratefully acknowledged.