Comparative chromatography of the chloroplast pigments

ANALYTICAL

BIOCHEMIETBY

24, 64-69

(1968)

Comparative Chromatography Chloroplast Pigment9 HAROLD AND

Chemistry

Division,

H. STRAIN, MARIAN

of the

JOSEPH SHERMA,2 GRANDOLFO

Argonne National

Laboratory,

Argonne, Illinois

SO.@9

Received July 18, 1967 Most chromatographic studies have shown that the major pigments of leaves are the chlorophylls a and b, the xanthophylls neoxanthin, violaxanthin, and lutein, and the carotenes, &carotene with or without a-carotene (l-4). The minor pigments include traces of the xanthophylls cryptoxanthin and zeaxanthin and possibly isolutein (5) or lutein epoxide (6-8) and zeaxanthin monoepoxide or antheraxanthin (6, 9) and perhaps others (5,6,8). There are few systematic and reliable chromatographic procedures for the complete resolution of the leaf pigments. Usually, the chromatographic separations must be carried out empirically with various chromategraphic systems until no further resolution can be achieved (2-5). As might be expected, the isolation and identification of the minor pigments is more difhcult than that of the major pigments. If the chromatographic systems are overloaded, the zones of the major pigments may overrun and obscure the zones of less-sorbed minor pigments. Moreover, the zones of the major pigments may trail so that the following zones of more-sorbed, minor pigments are also obscured. Usually, the lighter the loading of the sorptive media with the leaf extracts, the better is the separation of the zones from one another. With very light loading, however, the zones of the minor pigments may be so diffuse or so pale that they are not readily detectable, and the quantities isolated may not suffice for the determination of many significant physical and chemical properties (2). Occasionally, spurious zones may be formed from a single pigment. lBased on work performed under the auspices of the U. S. Atomic Energy Commission. *Present address: Department of Chemistry, Lafayette College, Easton, Pennsylvania 18042. 54

COMPARATIVE

CHROMATOGRAPHY

55

In paper, colorless polar substances in the plant extracts often distort the trailing portions of the pigment zones so that additional Pigments appear to be present (2). In columns as well as in paper, some xanthophylls, adsorbed from nonpolar solvents, are partially precipitated in the adsorbent by weakly polar wash liquids, thus yielding several zones even though a single pigment was adsorbed (2, 3). The chromatography of the pigment mixtures and the identification of the constituents require special care in the handling of the plant material, in the preparation of t.he extracts, in the selection of sorbents and solvents, and in the treatment of the individual pigments. Some of the pigments may be altered in the plant material, in the extracts, or on the adsorbent. products may The resultant hydrolytic, oxidation, and isomerization easily be mistaken for natural pigments (3,10-14). Chromatographic separations of the leaf pigments have been performed with a variety of sorbents in various forms, as in columns, thin layers, and sheets. These separations have been based upon adsorption and partition effects ut’ilized in one-way, two-way, and radial migrations. Many of t,hese applications of the chromatographic methods have been summarized in extensive reviews (l-3, 7, 8, 15, 16). For the same chromatograpbic system, the effectiveness of the separations should be similar with columns, thin layers, and sheets, but the separations are not always comparable. Variations of the separations may be due to the loading of the system, to the shape of the initial zone of the mixture, to variation of the migration dist,ance, and to alteration of the activity of the adsorbent upon formation of the thin layers from a slurry of the adsorbent with a polar liquid, such as water or methanol (2, 11). In order to provide more-comparable results on the chromatography of the leaf pigments, we have now studied the separations with various adsorbents and wash liquids. We have employed columnar methods and thin-layer methods under comparable conditions, and we have followed the separations in relation to the loading of the systems with the pigment mixtures. Our objectives have included the direct comparison of thin-layer and columnar techniques. We have sought to determine whether or not the major pigments may be resolved further, and we have also endeavored to confirm the identity of the minor pigments. Application of all the modifications of chromatography to an examination of the chloroplast pigments of all kinds of plants would be an impractical undertaking. For the present, the more promising chromategraphic systems have been utilized for the separation of the pigments of a single species (Xunthium or cocklebur) grown under reproducible, greenhouse conditions (2).

56

STRAIN,

SHERMA,

MATERIALS

AND

GRANDOLFO

AND PROCEDURES

1. Leaf Extracts

The pigments were extracted from 6 gm of the leaves of cocklebur. Extraction was performed rapidly with acetone, 120 ml, in a chilled blender. The mixture was centrifuged, and all the pigments in the supernatant were transferred to petroleum ether, b.p. 65-llO“, equal to the volume of the acetone, by the addition of this solvent and aqueous salt solution. The residual leaf material was extracted a second time and the pigments transferred to petroleum ether. The pigments were recovered by evaporation of the petroleum ether (in vacuum below 35”) and were redissolved in petroleum ether, 3 ml. On this basis, each microliter contained all the pigments in 2 mg of the fresh leaves. For isolation of the carotenoid pigments free of chlorophylls and fatty substances, the centrifuged acetone extracts were treated with KOH, 3 gm, dissolved in methanol, 30 ml. After 20 min, the carotenoid pigments were transferred to ether/petroleum ether, l/l by vol, by the addition of these solvents and salt solutions. They were recovered by evaporation of the solvents at reduced pressure in a rotary evaporator and redissolved in ether/petroleum ether, l/l, 3 ml, so that each microliter contained the carotenoid pigments in 2 mg of leaf material. 2. Columns

and Loadi72g

Chromatographic tubes, 1 cm id. by 25 cm, were filled to a height of 20 cm with the adsorbent, which was packed dry. The resultant chromatographic columns were loaded with 50 to 400 ~1 of the pigment solutions and were washed, usually without suction, until the wash liquid had penetrated to the bottom, a distance of 20 cm. With suction, about 0.5 atm, the percolation time was reduced to about one-third of that without suction. 3. Thin Layers

and Loading

Thin layers were formed on glass plates, 20 X 20 cm, with a Desaga spreader, usually set for a thickness of 0.25 mm, which was slightly thicker than Whatman No. 1 paper (about 0.15 mm). The spreading and drying procedures are described below for each adsorbent. The initial zones of the mixtures, 0.25 to 20 ~1 of the pigment solutions, were placed 2.5 cm from the “lower” edge of the plate. Narrow lineal zones, nearly 20 cm long, were streaked onto the adsorbent with a RadinPelick TLC streaker. The wash liquid, in a closed “equilibrated” container, was allowed to migrate upward a distance of 15 cm through the thin layer. The migration distance of the wash liquid in plates was, therefore, only three-fourths of that in columns.

COMPARATIVE

.J. Comparison

CHROMATOGRAPHY

of Loading

57

in Columns and in Thin Layers

For equal lengths of a packed aorbent, comparable amounts of the sorbent are present in equal cross-sectional regions. A zone in a column 1 cm i.d. moves through a cross-sectional area of 78.5 mm2. By contrast, a zone from a spot 1 cm wide in a 0.25 mm layer moves through a croassectional area of 2.5 mm”, and a long (20 cm) narrow streak in the thin layer movea through a cross-sectional area of 50 mm2. For equivalent separations and with the aame distance of migration, the loading of the 1 cm column should be about 31 times that of a spot in the thin layer and about 1.5 times that of a line in the thin layer. For the same degree of separation, columns 10 cm in diameter, as used in preparative work (14), could be loaded with 1000 times more pigment than that in a spot in a plate. Because columns are often much longer than plates and because they may be overloaded and subjected to extensive washing with successively more polar wash liquids, a column 10 cm in diameter might separate nearly 10,000 times more pigment than that separable from a spot in a thin layer. With a narrow lineal streak of the pigments in paper or in thin layers, a single sheet or layer is required to teat the effects of each loading and of each solvent, as is necessary with columns. Usually the zones of pigments formed from an initial streak were uniform and their separation decreased with increased loading, whereas with spots of a mixture of the green and yellow pigments, double tailing of the zones often obscured

oc

0

0

-0a

c = carotenes V = Violaxonthin L = Lutein N = Neoxanthin

pt Saponified

Leaf

Extract

Fro. 1. Effect of loading on separation of saponified leaf pigments in a thin layer of magnesia plus Celite 545, l/l by wt, formed from a slurry with water and dried at 20” for 0 hr.

58

STRAIN,

SHERMA,

AND

GRANDOLFO

the individual zones, especially in the tailing regions and at the higher loadings (2). With the saponified leaf extracts, the zones from spots were more uniform and without double tails; hence the separations of these carotenoid pigments obtained from spots and lineal zones in thin layers and in sheets were very similar to those obtained in columns. Under all these conditions, the effects due to variation of the loading were similar. The chromatographic systems were considered to be overloaded when the number of separated zones was less than that observed with lower loading, as indicated in Figure 1 (2). They were underloaded when the number of observable zones was less than that obtained with slightly higher loading. In the tables of results, the range of values for the loading represent approximately the range between underloading and overloading, although in some circumstances this range could only be approximated. 5. Location and Identification of the Separated Pigments Both in columns and in thin layers, the zones of the separated pigments were located visually. Color reactions produced by the vapors of concentrated HCl were employed as a supplemental basis for location of the carotenoids and as a property for description and identification of these pigments, especially for those separated in thin layers. Spectral absorption properties and isomerization with acids were also employed for the description and identification of the separated pigments (2, 11, 17). COMPARATIVE

CHROMATOGRAPHIC

SEPARATIONS

1. Powdered Sugar Commercial powdered sugar, already containing 3% starch, was rubbed through a coarse sieve with a pestle. Different lots showed but small variation of adsorption capacity. Columns (1 cm i.d.) were prepared from the sugar alone and from the sugar plus an equal weight of Celite 545. Thin layers (0.25 mm) were formed from a slurry of the sugar (40 g) in methanol (40 ml). These layers dried in air at 20” or at 45” yielded similar results. Typical chromatographic separations and sequences of the chloroplast pigments on powdered sugar are summarized in Table 1. The most adsorbed pigment is listed first. Abbreviations used in this table are explained in Table 2. Results similar to those reported for petroleum ether plus 0.5% n-propanol in Table 1 were also obtained with petroleum ether plus 1.0% n-propanol as the wash liquid. With 5% acetone in the petroleum ether, chlorophylls a and b formed contiguous zones that obscured

u See Table

Cellulose (paper)

Cellulose (powder)

Starch

Sugar

Adsorbent

Cal. TL (spot) TL (streak) Cal. TL (spot) Cal. TL (spot) Cal. TL (spot) Cal. TL (spot) Cal. TL (spot) Cal. TL (spot) Cal. TL (spot) Sheet (streak) Sheet (streak)

Form

Chromatography

2 for abbreviations

Comparative

used.

ext. (‘ L‘

ext. ‘<

ext. ct

ext. li

Plant ext. Sap. ext.

Sap. ext. I‘ ii

Plant L‘

Sap. ext. ,‘ Li

Plant (L

Sap. ext. “ (‘

Plant L‘

Sap. ext. (I LL

Plant c< (L

Pigments

of Chloroplast

0.25-0.75 50-400 0.25-5 GO-500 0.25-3 12-280 12-280

25-300

l-2 25-100 l-2 50-150 0.25-5 25-350 0.25-S

pl

TABLE 1” in Columns

50300 l-2 12-280 50300

Loading,

Pigments liquid

Thin

PI tc )/

Pr LL <‘ ‘) ‘.

Layers

PE + lyO PE+IG7,Pr

Pr

” Ii ” PE + 1% Pr t< “ I< << I( ‘L “ L‘ tc

PE + 0.50/, (‘ L‘ LL tt

[I
B7, “

PE + 0.57a L( (I ct ci [I (1
Wash

and min

go-100 40-50 40-50 4045 4045 ca. 300 ca. 60 ca. 180 ea. 60 3035 25-30 30-35 25-30 40-80 20-25 40-80 ‘LO-25 2540 2540

Time,

of Various

N, N, N, N, N, (4 @ N, N, N, N, N, N, N, N, N, N, N, N.

hdsorpt,ive and

sequence

W, b), L, a, C (V + b), L, a, C W, b), L, a, C V, L, C V, L, C N, a), V, LL, C + N), a, V, 12, C V, L, C V, L, C 0 + b), a, L, C W + b), a, L, C V, L, C V, L, C 0’ + b), CL, a), C W + b), CL, a), c: V, L, C V, L, C (V + b), (L, a), C W, T,>. C

Pigments

Saccharides

60

STRAIN,

SHERMA,

AND

TABLE Abbreviations

Used

GRANDOLFO

2

in Tables

AC, acetone Bz, benzene Cal, column Ext, extract Me, methanol PE, petroleum ether Pr, n-propanol Sap, saponified TL, thin layer

1 and 3

a, chlorophyll a chlorophyll b C, carotenes L, lutein N, neoxanthin V, violaxanthm ( ), contiguous zones (+), overlapping zone8 b,

the violaxanthin. With benzene as the wash liquid, separations in columns were slightly better than those in thin layers, as shown in Table 1. With the benzene, the chlorophylls were relatively much more sorbed than the xanthophylls. Similar results were obtained with powdered sugar mixed with Celite except that the percolation rate was faster than with the sugar alone. 3. Starch

All comparative tests were made with commercial corn starch (Argo brand, Corn Products Co.). Columns were prepared from starch alone and from a mixture of starch plus Celite 545, l/l by wt. Thin layers were prepared from a mixture of starch plus gypsum (13%)) which was slurried with twice its weight of water. The air-dried plates were finally dried at lOO-110’ for 30 min. Some of the separations are compared in Table 1. Leaf extracts were separated to about the same degree with starch and with starch plus Celite using various solvents. Petroleum ether plus 1.0% n-propanol provided somewhat better separations than those obtained with petroleum ether plus 0.5% n-propanol (Table 1). All these separations were better than those obtained with petroleum ether plus 1 to 3% acetone, which provided another sequence, namely, chlorophyll b plus neoxanthin, chlorophyll a, violaxanthin, lutein, and carotene. 3. Cellulose (as Powder)

Comparative tests were made with Camag “cellulose powder for thinlayer chromatography.” Columns were formed from the dry adsorbent. Thin layers were prepared from a slurry of the cellulose, 15 gm, with water, 97 ml. They were dried at room temperature and then at 110-115” for 15 min. Leaf extracts, separated in the cellulose columns with petroleum ether plus 1% n-propanol, yielded only five principal zones containing the six pigments found in sugar columns (Table 1). Somewhat poorer results

COMPARATIVE

CHROMATOGRAPHY

61

were obtained with petroleum ether plus 0.25% n-propanol and with petroleum ether/benzene/chloroform/acetone/isopropanol, 50/35/10/‘0.5/ 0.17, a mixture found to provide good separations in paper. In the cellulose plates, the green extracts were separated much as they were in columns and in paper, As in paper, zones of the green extracts added as spots formed double tails (2). 4. Cellulose

(as Paper)

Because of the double tailing of the pigments observed when the leaf extracts were absorbed as spots in paper (2), separations were repeated in paper sheets (Whatman No. 1) with the starting zone as a uniform, narrow streak. With petroleum ether plus lo/O n-propanol (or with 2% propanol) as the wash liquid, the pigments separated as indicated in Table 1. These separations are equivalent to those observed at the center regions of the migrating spots. Cellulose layers on film (E’astman Chromagram Sheets) provide results somewhat superior to those obtained with thin layers of cellulose and with paper in many systems (18). 5. Polyethylene

Two preparations of polyethylene were tested as sorptive media (12). One product was Type 6560 Powder S Dylan J-2 by Sinclair-Koppers, the other by Dow Chemical Co, These preparations were employed in columns and in thin layers, the latter made from 20 gm of the powder slurried with a mixture of water plus ethanol (l/l), 50 ml. With polyethylene, and with polypropylene as described in the next section, the sorptive properties depend upon distribution of the pigments between the polymeric hydrocarbon and an aqueous organic liquid. Leaf extracts were not sorbed by polyethylene with benzene alone or with petroleum ether plus 5 or 20% tetrahydrofuran as the wash liquids. With either acetone or methanol plus 10 to 25% water as the wash liquids, the chlorophylls were strongly bound, forming overlapping or contiguous zones, which usually obscured the carotene. The xanthophylls formed only one or two yellow zones below the chlorophylls. The loading limits were 5 to 20 ~1 for the saponified and unsaponified extracts on thin layers and about 100 ~1 for columns. 6. Polypropylene

A finely divided polypropylene (Hercoflat, from Hercules Powder Co.), 35 gm, and 80 ml of water provided uniform thin layers that flaked off rather easily. Separations of the leaf extracts and of the saponified ex-

62

STRAIN,

SHERMA,

AND

GRANDOLFO

tracts in columns and in thin layers of the polypropylene those obtained with polyethylene. 7. Kel-F A powdered preparation of this used in columns under the same and polypropylene. The sorption similar to those observed with the

resembled

300

chlorofluorocarbon (Analabs, Inc.) was conditions employed with polyethylene capacity and the separations were very polyethylene.

8. Calcium

Carbonate

A preparation of “calcium carbonate, low in alkalies, powder” (J. T. Baker Chemical Co.) served as the adsorbent. The thin layers, from this CaC03, 30 gm, plus CaSO,, 3.9 gm, plus water, 60 ml, were dried in air, activated at 150’ for 0.5 hr, and cooled for 5 min before use. With petroleum ether plus 20% acetone, the leaf pigments were separated in the sequence N, V, L, b, a, C (cf. Table 2). The loading range was 1 to 10 ~1 for thin layers and about 100 ~1 for columns. 9. Hydroxylapatite

A preparation of ‘Lhydroxyl calcium phosphate, dried Bio-Gel HTP” (Bio-Rad Laboratories), 25 gm, was made into thin layers with CaSO,, 3.3 gm, and water, 75 ml. The air-dried plates were heated at 150” for 1 hr. With petroleum ether plus 30% acetone, this adsorbent decomposed the chlorophylls in the green, leaf extracts but separated the yellow pigments, N, V, L, C. When the plates containing the separated carotenoid pigments were allowed to dry in air, the violaxanthin zone turned green-yellow. Elution of the pigment and determination of the spectral absorption maxima indicated that the violaxanthin had partially rearranged to luteoxanthin and auroxanthin, as occurs on many siliceous adsorbents (10, 11). 10. Alumina

Experiments were carried out with several preparations of adsorptive alumina. Aluminum oxide G (E. Merck; Brinkmann Instruments) was employed in thin layers, from 35 gm of the alumina plus 45 ml of water. These layers were used without activation and also after activation at 110’ for 30 min. Some of the alumina that had been made into thin layers and dried in air was scraped loose and used in columns. Chromagram Sheets (Eastman Kodak Co.) were employed as purchased and after having been treated with acidic (pH 5-6) and basic (pH ca. 9) buffers and dried in air and at 110” for 30 min. Gelman Type A aluminaimpregnated glass-fiber paper was also employed, both before and after drying at 110“.

COMPARATIVE

CHROMATOGRAPHY

m

Pigments in the green leaf extracts were usually not separated cleanly with most of the aluminas. A portion of the green pigments remained near the starting point, and the migrating green zones trailed through the zones of the more sorbed carotenoid pigments, an effect observed with a variety of solvents. With the basic and neutral Chromagram Sheets, there was a minimum alteration of the chlorophylls, which, however, were often poorly separated from each other. A few wash liquids, such as isooctane/acetone/ether (3/1/l) or isooctane/acetone/carbon tetrachloride (3/1/l), provided a separation of the chlorophylls and of the carotenoids on the Chromagram Sheets. With a variety of wash liquids, the yellow pigments in the saponified leaf extracts were readily separable with all the alumina preparations except the Type A paper and the acidic Chromagram Sheets (Table 3). None of these systems isomerized the neoxanthin or the violaxanthin, which were recovered unchanged from the columns, the thin layers, the papers, and the Chromagram Sheets. 11. Magnesia

Activated magnesia, introduced as Micron Brand Magnesium Oxide No. 2641 (1, 5)) is now sold as Sea Sorb 43 (Fisher Scientific Co.). This product is so very finely powdered that for use in columns and in thin layers it must be mixed with about an equal weight of a filter aid such as Celite 545. For the preparation of the thin layers, this magnesia/ Celite mixture (l/l by wt) was ground with 137% by weight of gypsum and twice its weight of water. Some of the air-dried layers were act.ivated by drying at 100-110” for 30 min. Magnesia alters adsorbed chlorophylls so that they cannot be washed along in the adsorbent and so that they cannot be recovered by elution (1, 4). For this reason, most of our chromatographic experiments with magnesia were limited to separations of the saponified pigments, as indicated in Table 3 and illustrated in Figure 1. In columns of the untreated magnesia the carotenes were much less sorbed than the xanthophylls. Weakly polar wash liquids, such as diethyl ether and petroleum ether plus 5 to 20% acetone, served for their separation. More polar wash liquids, such as petroleum ether plus 40 to 70% acetone or petroleum ether plus 5 to 10% propanol, were required for resolution of the mixtures of xanthophylls. A magnesia ~1~s Celite mixture that had been made into a slurry, dried in air for 16 to 24 hr, and packed into columns was less sorptive than the untreated magnesia, because the xanthophylls could be separated with petroleum ether PIUS about 20% acetone or 2 to 5~~ propan (Table 3).

Sheets

Magnesia + Celite 545 (l/l), slurried; dried, 20”, 16 hr; llO”, 0.5 hr

Magnesia + Celite 545 (l/l), dried, 20°, 16-24 hr

slurried;

Chromatography

Type A paper Magnesia + Celite 545 (l/l)

AlUmina

Alumina Chromagram

Alumina G

Adsorbent

Comparative

50-400 0.25-20 l-3 0.25-20 0.25-20 0.25-20 ca. 10 ca. 10 150-300 0.5-3 150-400 0.5-S ca. 400 ca. 400 ca 400 0.25-0.75 ca. 400 ca. 400 ea. 400 0.75-5

Sap. ext. “ ‘( 6‘ ‘6 Sap. ext. ‘I CL 6‘ “ Sap. ext. I‘ ‘I Plant ext. I, “ Sap. ext. “ “ Sap. ext. ‘I “ ‘I “ I‘ ‘I Sap. ext. “ “ “ ‘I ‘I “

/.d

Loading,

Piiments

Bz Bz Bz PE ‘I “ Bz I‘ PE PE PE PE PE PE PE PE PE PE PE PE + + + + + + + + + + -t +

+

+ + + +

Wash

36% AC 22% AC 10% Pr 30‘& AC ‘I ‘I “ “ 17y0 AC “ I‘ 60% AC 30% AC 60% AC 300/o AC 2070 AC 30% AC 20/, Pr 30% AC 40% AC 5y0 Pr 10% Pr 300/, AC

liquid

and sequence

N, V, L, C N, V, L, C N, V, L, C N, V, L, C N, V, L, C N, (V, L), C (N + V + L, C) (N + V + L, C) (b + a), N, L, V, C (b + a), N, L, V, C N, L, V, C N, L, V, C N, (V + L), C N, (V + L), C N, (V, L), C N, (L, V), C N, L, V, C N, L, V, C N, L, V, C N, L, V, C

Pigmenta

and Magnesia min

90-105 ca. 30 ca. 30 ca. 75 ca. 80 ca. 75 ca. 20 ea. 20 ca. 60 ca. 82 ca. 60 ca. 80 ca. 35 ca. 40 ca. 34 ca,. 28 ca. 40 ca. 30 ca. 40 ca. 83

Time,

TABLE 3 Pigments in Columns and Thin Layers of Alumina

Cal. TL (spot) TL (spot) Neutral (spot) Basic (spot) Acidic (spot) Unheated (spot) Heated (spot) Cal. TL (spot) Cal. TL (spot) Cal. Cal. Cal. TL (spot) Cal. Cal. Cal. TL (spot)

FOCUl

of Chloroplsst

COMPARATIVE

CHROMATOGRAPHY

65

By contrast with these results in columns, thin layers of a slurry of magnesia plus Celite, dried in air for 2, 4, 6, 16, and 24 hr, provided rather weak adsorption of the pigments and poor separation of the lutein and violaxanthin (see Fig. 1). A mixture of untreated magnesia plus Celite that had been allowed to stand in a shallow layer exposed to the air for 3 days and then packed into a column exhibited adsorption capacity and seIectivity much like that of the original untreated magnesia. Thin layers from a slurry of magnesia plus Celite dried in air and reactivated at 110” for 0.5 hr exhibited properties similar to those of the original preparation (Table 3), but the capacity and the selectivity varied a great deal. 1% Siliceous

Adsorbents

Although silicic acid and various inorganic silicates have significant resolving power for mixtures of the chloroplast pigments, most of these adsorbents did not lend themselves to the comparative experiments because they decompose the chlorophylls (12) and isomerize the epoxyxanthophylls (10, 11). It is noteworthy, however, that one preparation of silica gel, stabilized in sheets of glass fibers (Gelman Instrument Co.), provided clean separations of the pigments (11). With petroleum ether plus 0.5 to 1% n-propanol, the zones from initial spots were uniform and in the sequence b, a, N, V, L, C. CHROMATOGRAPHY

PLUS RECHROMATOGRAPHY

1. Larger Scale Separations

In the first experiments in which zeaxanthin, cryptoxanthin, and isolutein (1, 5), probably lutein epoxide (7)) were encountered as minor constituents of leaf xanthophyll, the columns were heavily overloaded. The failure to detect these minor pigments, under the experimental conditions described in the preceding sections, may have been due, in part, to the small quantities of these constituents in the lightly loaded chromatographic systems. Consequently, we have isolated larger chromatographic fractions of t,he saponified pigments and have submitted these to further chromatographic fractionation. As an example, 25 gm portions of cocklebur leaves were extracted with chilied acetone or methanol, 250 ml, in a chilled blender. The mixture was centrifuged, and the residue was agitated with fresh solvent, 250 ml, and centrifuged again. The extracts from two 25 gm portions of leaves were combined and treated with KOH, 18 gm, dissolved in methanol, 180 ml. After 20 min, the carotenoid pigments were transferred to diethyl ether plus petroleum ether, l/l, 250 ml, by addition of t,hese solvents and excess salt solution. The ethereal solutioii was

66

STRAIN,

SHERMA,

AND

GRANDOLFO

washed with water, to remove the residual alcohol and the saponified green pigments, and evaporated at reduced pressure. The residue, dissolved in ether plus petroleum ether plus benzene, 2/2/l, was adsorbed in a column of powdered sugar, 8 X 30 cm, and washed with petroleum ether plus 0.5% n-propanol. The washing was continued until the carotene had been carried into the percolate and until the lutein extended to the bottom of the adsorbent. There were then three principal xanthophyll zones in the adsorbentneoxanthin, violaxanthin, and lutein. The very pale yellow portion of the column between the violaxanthin and the lutein was dug out separately, packed into a fresh tube, and the pigment eluted with ethanol and petroleum ether. Lutein remaining in the column was also eluted with petroleum ether plus ethanol. This procedure provided three fractions: (A) carotene plus xanthophylls, such as cryptoxanthin, which are less sorbed than lutein, (B) lutein plus zeaxanthin and other xanthophylls not separated from lutein, and (C) xanthophylls, such as lutein epoxide and zeaxanthin monoepoxide, which are more sorbed than lutein but less sorbed than violaxanthin. The pigments in each of the three fractions were transferred to ether plus petroleum ether, which was evaporated. The “carotene” residue (A) was dissolved in about 10 ml ether plus 10 ml petroleum ether, and the xanthophyll residues (B and C) were dissolved in ether plus petroleum ether plus benzene (5, 5, 2 ml), Each solution was added to separate columns (3 X 30 cm) of magnesia plus Celite (l/l). The carotene fraction (A) was washed with petroleum ether plus 20% acetone or with ether. The principal pigment was the all-truns ,&carotene. There were two more sorbed pigments, namely, some partial cis-p-carotene and the more sorbed cryptoxanthin, which yielded an orange zone on magnesia and a spectral absorption curve and chromatographic behavior identical with that of cryptoxanthin from the calyx of Physalis (5). A zone less sorbed than the p-carotene yielded a trace of a-carotene. Carotene epoxides were not detectable. The lutein fraction (B), chromatographed on magnesia plus Celite with petroleum ether plus 30 to 40% acetone as the wash liquid, yielded only a little of the strongly sorbed, orange zone of zeaxanthin (5). The quantity of the recovered zeaxanthin was about 6 times the quantity of the cryptoxanthin. The lutein was moderately sorbed. It was preceded by very small quantities of a yellow pigment with the properties of lutein epoxide (A,,,=, ethanol, 418, 442, 471 mJL; blue reaction with ether and concentrated HCl; complete conversion to the furanoid isomer by acetic acid in ethanol, X,,, 339, 423, 450 mp) (7, 11, 17). Minute traces of pigments between the lutein and the zeaxanthin could not be identified. The xanthophyll fraction (C), between lutein and violaxanthin in the

COMPARATIVE

67

CHROMATOGRAPHY

sugar column, yielded some lutein that had trailed from the principal lutein zone in the sugar. It also yielded traces of a less sorbed xanthophyll identical with the epoxylutein obtained from the lutein zone (B). Only minute traces of a pigment corresponding roughly to zeaxanthin monoepoxide were observed between the lutein and the lutein epoxide. PREFERRED

CHROMATOGRAPHIC

PROCEDURES

NO single chromatographic system provides a complete resolution of all the leaf pigments. For separation of the chlorophylls from each other and from most of the carotenoid pigments, mild adsorbents must be employed. Wit.h respect to selectivity, scale of operation, recovery of unaltered pigments, economy and convenience, commercial powdered sugar offers many advantages. The most useful solvent with respect to selectivity and convenience is petroleum ether plus 0.5 to 1.0% n-propan01 (1, 14). For the most extensive resolution of the carotenoid pigments of leaves, activated alumina and magnesia provide great adsorption capacity and highest selectivity. These separations may be made with the unsaponified leaf extracts, with the saponified extracts, or with mixtures eluted from zones in the sugar columns. For micro-scale separations and with equivalent loading, thin layers and columns provide comparable results. For macro-scale separations, however, there is no convenient substitute for large columns (14). Pigments isolated by chromatography may be concentrated and resolved further by elution and readsorption on the same or another adsorbent. This step is often essential in the search for minor pigments and in the removal of colorless plant constituents which may be distributed through the pigment zones after a single adsorption (14).

SUMMARY

Many conditions influence the separability of the chloropbst pigments by chromatography. These conditions include the nature and treatment of the sorbent, the solvent or wash liquid, the form in which the sorbent is employed, and the loading of the sorbent with the mixture. Various adsorbents exhibit different capacities and selcctivities for the pigments, and they often provide different sequences of the separated pigments. Some adsorbents, such as magnesia, exhibit variation of capacity and selectivity when made into a thin layer from a slurry with water. Very finely divided adsorbents, such as activated magnesia, may be employed both in columns and in thin layers if mixed with a nons0rptix.e filter aid. With all the chromatographic systems t.hat were tested, the lower the

68

STRAIN,

SHERMA,

AND

GRANDOLFO

loading with the mixture, the greater is the resolution. Conversely, the higher the loading, the poorer is the resolution. Both with underloading and overloading, minor constituents may not be detectable. Readsorption of large fractions of the pigments obtained from one chromatographic system improves the separation and the detection of the minor leaf pigments. Readsorption with a different chromatographic system permits the separation of pigments, such as lutein and zeaxanthin, and a- and &carotene, that were inseparable in the first system. For these chromatographic and rechromatographic procedures, columns are much more adaptable than thin layers and sheets of the adsorbent. For the preparation of quantities of the leaf pigments, columnar \methods are more convenient than thin-layer and paper techniques. The separations in columns are faster and more adaptable than separations in thin layers or in paper. The major pigments of cocklebur (Xunthium) are chlorophylls a and b, neoxanthin, violaxanthin, lutein, and p-carotene. The minor pigments are cryptoxanthin, zeaxanthin, trace amounts of a-carotene and Iutein epoxide, and possibly minute traces of zeaxanthin monoepoxide (antheraxanthin) . initial

ACKNOWLEDGMENTS We are indebted to Mr. William Chorney for a supply of cocklebur plants and to Miss Maureen Beck for observations with paper chromatography. Dr. Kurt Aitzetmiiller and Mr. Walter A. Svec have generously provided many helpful suggestions, REFERENCES H. H., Ann. Priestley Lectures 32, (19.58). . H. H., SHERMA, J., BENTON, F. L., AND KATZ, J. J., Biochim. Biophys. 109, 1,16,23 (1966). H. H., in “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 1, p. 387. Academic Press, New York, 1966. MALLAMS, A. K., WAIGHT, E. S., WEE~ON, B. C. L., CHOLNOKY, L., GY~RQYFY, K., SZABOLES, J., KRINSKY, N. I., SCHIMMER, B. P., CHICHESTER, C. O., KATAYAMA, T., LOWRY, L., AND YOKOYAMA, H., Chem. Commun. 1967,484. STRAIN, H. H., “Leaf Xanthophyils,” Carnegie Inst. Wash., Publ. 490, (1938). HAGER, A., AND MEYER-BERTENRATH, T., Planta 69, 198 (1966). GOODWIN, T. W., in “Modern Methods of Plant Analysis” (K. Paech and M. V. Tracey, eds.), Vol. 3, p. 272. Springer, Berlin, 1955. SAAKOV, V. S., AND SHIRYAEVA, G. A., Akad. Nauk SSSR. Trans. Botan. Inst. V. L. Komarova. Ser. IV, Expt. Botan. Zssue 18, 151 (1967). YAMAMOTO, H. Y., NAKAYAMA, T. 0. M., AND CHICHESTER, C. O., Arch. Biochem. Biophys. 97, 168 (1962). KATAYAMA, T., Bull. Japan. Sot. Sci. Fisheries 30, 440 (1964). STRAIN, H. H., SHERMA, J., AND GRANWLFO, M., Anal. Chem. 39, 926 (1967). BACON, M. F., AND HOLDEN, M., Phytochemistry 6, 193 (1967).

1. STRAIN, 2. STRAIN, Acta 3. STRAIN, 4.

5.

6. 7. 8. 9.

10. 11. 12.

COMPARATIVE

13. 14. 15. 16. 17. 18.

CHROMATOGRAPHY

69

F. C., STRAIN, H. H., SVEC, W. A., AND KATZ, J. J., .I. Am. Chem. 8oc. 89, 3875 (1967). STRAIN, H. H., AND SVEC, W. A., in “The Chlorophylls” (I,. I-‘. Vernon and G. R. Seely, eds.), p. 21. Academic Press, New York, 1966. SEST~K, Z., Chromatog. Rev. 1, 193 (1959) ; 7, 65 (1965). SESTAK, Z., J. Chromatog. 1, 293 (1958). STRAIN, H. H., Arch. Biochw. Biophys. 48,458 (1954). SHERMA, J. AND ZWEIG, G., J. Chromatog. 31, 439,589 (1967). PENNINGTON,