Biological Chemistry
`
`
`
`
`6° The Journal of
`
` "fix
`Asia'iOiE
`
`4" “a.
`
`a?"
`
`American Society fur Emma! am MolecularW
`
`I:
`
`ARTICLE:
`
`A SIMPLE METHOD FOR THE
`
`ISOLATION AND PURIFICATION OF
`
`TOTAL LIPIDES FROM ANIMAL
`
`TISSUES
`
`Jordi Folch, M. Lees and G. H. Sloane Stanley
`J. Biol. Chem. 1957, 226:497—509.
`
`Access the most updated version of this article at
`http://www.jbc.org/content/226/1/497.citation
`
`Find articles, minireviews, Reflections and
`Classics on similar topics on the JBC Affinity
`Sites .
`
`Alerts:
`- When this article is cited
`
`- When a correction for this article is posted
`
`Click here to choose from all of JBC's e—mail
`alerts
`
`This article cites 0 references, 0 of which can be
`accessed free at
`http://www.jbc.org/content/226/1/497.citation.fu||.h
`tm|#ref-|ist-1
`
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`Petition for Inter Panes Review
`Of US. Patent 8,278,351
`Exhibit
`
`ENZYMOTEC - 1078
`
`000001
`
`

`

`A SIMPLE METHOD FOR THE ISOLATION AND PURIFICATION
`
`OF TOTAL LIPIDES FROM ANIMAL TISSUES“
`
`BY JORDI FOLCH, M. LEES,T AND G. H. SLOANE STANLEY:
`
`(From the McLean Hospital Research Laboratories, Waverley, and the Department
`of Biological Chemistry, Harvard Medical School, Boston, Massachusetts)
`
`(Received for publication, August 23, 1956)
`
`Work from this laboratory resulted in the development of a method for
`the preparation and purification of brain lipides (1) which involved two
`successive operations.
`In the first step, the lipides were extracted by
`homogenizing the tissue with 2 :1 chloroform-methanol (v/v), and filtering
`the homogenate.
`In the second step, the filtrate, which contained the
`tissue lipides accompanied by non-lipide substances, was freed from these
`substances by being placed in contact with at least 5-fold its volume of
`water. This water washing entailed the loss of about 1 per cent of the
`brain lipides.
`This paper describes a simplified version of the method and reports the
`results of a study of its application to different tissues, including the efli—
`ciency of the washing procedure in terms of the removal from tissue lipides
`of some non-lipide substances of special biochemical interest.
`It also re-
`ports some pertinent ancillary findings. The modifications introduced
`into the method pertain only to the washing procedure. A chloroform-
`methanol extract of the tissue, prepared as described in the original version
`of the method, is mixed with 02 its volume of water to which, for certain
`purposes, different mineral salts may be added. A biphasic system with
`out any interfacial fluff is obtained (2). The upper phase contains all of
`the non-lipide substances, most of the strandin, and only negligible amounts
`of the other lipides. The lower phase contains essentially all the tissue
`lipides other than strandin.
`In comparison with the original method, the
`present version has the advantage of being simpler, of being applicable to
`any scale desired, of substantially decreasing the losses of lipides incidental
`to the washing process, and, finally, of yielding a washed extract which
`can be taken to dryness Without foaming and without Splitting of the
`proteolipides (3).
`
`* This work has been aided by grant No. B—130 from the United States Public
`Health Service.
`
`TFellow of the American Cancer Society, 1951—53.
`IEli Lilly Traveling Fellow in Medicine, 195365.
`
`497
`
`000002
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`498
`
`ISOLATION OF TOTAL TISSUE LIPIDES
`
`Procedure
`
`Reagents-—
`Chloroform. Reagent grade.
`Methanol. Reagent grade. For use with tissues relatively poor in lip-
`ides, such as muscle or blood plasma, both the chloroform and methanol
`must be redistilled.
`
`Chloroform-methanol mixture. 2:1 by volume.
`Pure solvents upper phase and pure solvents lower phase. Chloroform,
`methanol, and water are mixed in a separatory funnel in the proportions
`824:3 by volume. When the mixture is allowed to stand, a biphasic sys-
`tem is obtained. The two phases are collected separately and stored in
`glass bottles.
`It has been found that the approximate proportions of chloro—
`form, methanol, and water in the upper phase are 3:48:47 by volume.
`In the lower phase, the respective proportions are 86: 14: 1. Either of the
`phases may be prepared directly by making use of the above proportions.
`Pure solvents upper phase containing 0.02 per cent CaC'lz, 0.017 per cent
`MgClz, 0.29 per cent NaCl, or 0.37 per cent KCl. These solutions can be
`prepared in one of two ways. One is to shake the appropriate amount of
`salt with pure solvents upper phase in a glass—stoppered veSSel until solution
`is complete. The other is to proceed as forthe preparation of pure solvents
`upper and lower phases except that, instead of water, 004 per cent aqueous
`03.012, 0.034 per cent aqueous MgClz, 0.58 per cent aqueous NaCl, or 0.74
`per cent aqueous KCl is used.
`Extraction of Diptdes—For the purposes of this description, the volume
`of a tissue sample will be computed on the assumption that the tissue has
`the specific gravity of water; 2253., the volume of 1 gm. of tissue is1 ml. The
`tissue or tissue fraction is homogenized with 2:1 chloroform-methanol mix—
`ture (v/v) to a final dilution 20—fold the volume of the tissue sample; i.e.,
`the homogenate from 1 gm. of tissue should be diluted to a volume of 20
`ml. For amounts of tissue up to 1 gm., the homogenization is carried out
`in a Potter-Elvehjem type of homogenizer, the tube of which has been
`weighed, and calibrated at the volume of the final dilution of the particular
`tissue homogenate. Thus, the tissue sample can be weighed and the ho-
`mogenate diluted to volume without a transfer. For brain or tissues of
`similar consistency,
`3 minutes
`suffice for complete homogenization.
`Tougher tissues will require lengthier homogenization, and some organs
`rich in connective tissue, (2.9. peripheral nerve, may require special handling
`such as grinding with a mortar and pestle at the temperature of dry ice
`before homogenization with the solvent mixture. For amounts greater
`than 1 gm., the tissue is homogenized in an adequate blender with. about
`a 17-fold volume of solvent mixture; the balance of solvent mixture re-
`quired to dilute the homogenate to final volume is used to insure the quan-
`titative transfer of the homogenate into a volumetric flask. After tem—
`
`000003
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`J. FOLCH, M. LEEs, AND G. H. SLOANE STANLEY
`
`499
`
`the homogenate is
`perature equilibration and final volume adjustment,
`filtered through a fat-free paper into a glass-stoppered vessel. For the
`purposes of computation, this extract corresponds to 0.05 its volume of
`tissue; i.6., 1 ml. of extract corresponds to 0.05 gm. of tissue.
`Washing of Crude Extract—~The crude extract is mixed thoroughly with
`0.2 its volume of either water or an adequate salt solution (see “Experi-
`mental”), and the mixture is allowed to separate into two phases, without
`interfacial fluff, either by standing or by centrifugation. The volumes of
`the upper and lower phases are, respectively, 40 and 60 per cent of the total
`volume of the system. As much of the upper phase as possible is removed
`by siphoning, and removal of its solutes is completed by rinsing the inter—
`face three times with small amounts of pure solvents upper phase in such
`a way as not to disturb the lower phase. Finally, the lower phase and
`remaining rinsing fluid are made into one phase by the addition of meth-
`anol, and the resulting solution is diluted to any desired final volume by
`the addition of 2:1 chloroform-methanol mixture.
`
`The procedure can be run on any scale that is otherwise technically feasi-
`ble, and the actual details of operation will vary according to the amount
`of extract being washed. For instance, if 10 ml. of crude extract are to be
`washed, the extract is placed in a 15 ml. centrifuge tube. To it are added
`2 ml. of either water or salt solution, the two liquids are mixed with a stir-
`ring red, the rod is then rinsed into the tube with a minimal amount of pure
`sohmnflslowerphase,andthetubeiscappedxvfihadunnnuniflnlandtmntfi~
`fuged until complete separation of the system into two phases without any
`interfacial fluff is obtained. The duration of centrifugation varies from
`about 20 minutes at 2400 r.p.m. for white matter extracts to a very short
`time for blood plasma. The volumes of the upper and lower phases are
`4.8 and 7.2 ml., respectively. The upper phase is removed as completely
`as possible with a pipette or with a suction arrangement such as the one
`described by Van Slyke and Rieben (4). Next, the inside wall of the tube
`is rinsed with about 1.5 ml. of pure solvents upper phase, which are allowed
`to flow gently from a pipette so that the washing fluid collects on top of
`the lower phase Without any mixing of the two phases. The tube is ro—
`tated gently to insure mixing of the rinsing fluid with the remaining original
`upper phase, and the mixture is removed. This rinsing of the tube wall
`and interphase with pure solvents upper phase is repeated twice. Finally,
`the lower phase is diluted to a volume of 10 ml. as outlined above. With
`tissues poor in proteolipides, 6.9. muscle, plasma, and liver, or if time is no
`object, centrifugation may be omitted from the washing procedure.
`In-
`stead, the extract plus water mixture can be allowed to separate into two
`phases by prolonged standing.
`In that case, it is more convenient to carry
`out the washing in glass—stoppered cylinders.
`Permissible Departures from Procedures/1119 technique described can be
`
`000004
`Downloaded from http://www.jbcorg/ by guest on June 12, 2013
`
`

`

`500
`
`ISOLATION OF TOTAL TISSUE LIPIDES
`
`changed in many details if so indicated by the size and nature of the tissue
`sample or by the particular problem under study. Thus, if necessary, in
`the preparation of the crude extract, the tissue homogenate can be diluted
`to more than 20-fold the volume of tissue. Also, centrifugation can be
`used in preference to filtration as a means of obtaining a clear extract.
`Centrifugation of the homogenate itself is unsatisfactory because the spe-
`cific gravity of the solvent mixture is too close to the density of the sus-
`pended material. Therefore, if centrifugation is to be used, it is necessary
`to lower the specific gravity of the homogenate by the addition of methanol.
`Usually, the addition of 0.2 its volume of methanol suffices for the purpose.
`The amount of methanol added must be noted.
`
`In the washing procedure described, chloroform, methanol, and water are
`present in the system tissue extract plus water in the proportions 8:423
`by volume, as can be computed if account is taken of the fact that the
`extract contains all the water from the tissue. These proportions are
`critical and must be kept constant. Therefore, in cases in Which the tissue
`extraction has been substantially changed, it is necessary to modify the
`washing procedure in a way that will restore the required proportions of
`solvents. For instance, if the homogenate has been diluted to 40-fold the
`volume of tissue, the water contributed to the extract by the latter will be
`half as much as in the standard 20-fold dilution; 216., it will be 2 per cent
`of the extract as compared to the usual 4 per cent. Therefore, the amount
`of water added to the extract for washing should be 22 per cent instead of
`the usual 20 per cent.
`If methanol has been added to the extract, twice
`as much chloroform must also be added and the amount of water adjusted
`accordingly.
`
`EXPERIMENTAL
`
`Analytical AIethods—Most of the methods used in this study have been
`described elsewhere (3, 5).
`Degree of Completeness of Extraction of Tissue Lipides—Earlier work had
`shown that the extraction procedure removes all lipides from brain (1) and
`blood plasma (6), with the exception of a specific fraction of lipides which
`is combined to tissue proteins by a linkage which withstands the action
`of neutral solvents.
`In the present study, the completeness of extraction
`of lipides from liver and muscle was studied by reextracting the residue with
`hot solvent and determining the amount of lipides in the second extract.
`The original extraction can be considered complete if the second extract
`contains no more lipides than can be accounted for by the aliquot of first
`extract left wetting the residue. The experiment was carried out as fol-
`lows: The tissue was homogenized with chloroform-methanol as described,
`and the homogenate filtered through a previously weighed Bfichner funnel.
`
`000005
`Downloaded from http://www.jbcorg/ by guest on June 12, 2013
`
`

`

`J. FOLCH, M. LEES, AND G. H. SLOANE STANLEY
`
`501
`
`filtration being stopped before the insoluble residue became dry. The
`filter was weighed again and the weight of the wet residue it contained was
`computed by difference. Next, the residue was reextracted with a new
`portion of solvent mixture by boiling under reflux for 24 hours, the sus-
`pension was filtered, and the twice extracted residue collected and dried to
`constant weight. The amount of first extract left wetting the tissue residue
`could then be computed from the equation, ml. of extract in residue =
`(weight of wet residue after first extraction minus weight of dried residue) /—
`(specific gravity of first extract).
`In the case of liver, 40 gm. of tissue were
`extracted as outlined above in succession with 760 ml. and 400 ml. of solvent
`
`mixture. The first extract contained 2.46 mg. of lipides per ml., while
`
`28.8 ml. of extract with a computed total lipide content of 71 mg. were left
`in the residue. The second extract contained a total of 69 mg. of lipides;
`6.6., the amount to be expected from the aliquot of the first extract in the
`residue.
`In an identical experiment with muscle tissue, the first extract
`contained 0.743 mg. of lipides per ml., while 2-1.3 ml. of extract with a total
`lipide content of 18.1 mg. were left wetting the residue. The second ex-
`tract contained a total of 21.6 mg. of lipides; £13., 3.5 mg. more than were
`to be expected from the aliquot of the first extract remaining in the residue.
`This difference, which amounts to <0.5 per cent of total tissue lipides,
`cannot be considered significant.
`Study of Washing ProcedureflThe washing procedure has been studied
`by (a) determining the amount of lipides lost during the washing, (b) de—
`termining the amount of non-lipide substances remaining in the lower
`phase, (0) investigating an effect of certain non-lipide substances upon the
`distribution of lipides between the two phases formed during the washing
`procedure, (d) determining the effect of mineral salts on the distribution
`of lipides in this particular biphasic system, and finally (6) ascertaining the
`efficiency of the washing procedure in relation to some substances of im—
`portance in metabolic studies by the use of radioisotopes.
`Loss of I/ipides Incidental to Washing Procedure and Degree of Removal
`of Non-I/ipz'de Contaminants_8ince lipides are undialyzable, the amount
`of undialyzable substances in the upper phase would represent the maximal
`amount of lipides lost, and the dialyzable substances would, of necessity,
`represent non-lipide contaminants.
`In a typical experiment, 175 ml. of
`brain white matter extract were washed with 35 ml. of water. The upper
`phase, which had a volume of 84 ml., was collected quantitatively. The
`lower phase was equilibrated with 84 ml. of pure solvents upper phase, and
`the resulting second upper phase was collected. Both upper phases were
`concentrated to dryness by vacuum distillation of the solvents, the residues
`were each dissolved in 10 ml. of water, and the solutions were dialyzed ex-
`haustively. The dialyzable and undialyzable fractions thus obtained were
`
`000006
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`502
`
`ISOLA'I‘ION OF TOTAL TISSUE LIPIDES
`
`dried and analyzed. The solutes in the undialyzable fractions were com-
`pletely soluble in chloroform-methan 01, indicating that they were all lipides.
`The pertinent results are given in Table I. Thus, the values for the first
`upper phase show that no more than 0.3 per cent of the tissue lipides other
`than strandin was lost in the washing. Also, since the second upper phase
`contained only negligible amounts of dialyzable substances, the conclusion
`is warranted that, after one washing, the lower phase is essentially free
`from non-lipids substances. The same type of experiment has been carried
`out with white matter, gray matter, liver, and muscle, with the results
`given in Table II.
`In all the tissues studied, one washing was found suffi-
`
`Distribution of Salutes in CH Cl3:C'H 30H Extract of Brain White Matter
`between Subsequent Fractions
`
`TABLE I
`
`.
`Yield, mg.
`
`Yield as % total
`solutes in
`crude extract
`
`1. Total solutes in crude extract... ................
`
`2000.0
`
`4.79
`95.85
`2. 1st upper phase; total solutes .....................
`4.07
`81.5
`3. Dialyzable solutes ................................
`0.72
`14.35
`4. Undialyzable solutes (lipides + strandin) .........
`0.39
`7.75
`5. Strandin in undialyzable fraction .................
`0.33
`6.6
`6. Lipides other than strandin (4) — (5).... .. .
`.
`.
`.
`.
`.
`.
`2.33
`46.7
`7. 9nd upper phase; totalsolutes. .
`.
`.
`0.08
`1.7
`8. Dialyzable solutes ...............................
`2.25
`45.0
`9. Undialyzable solutes (lipides + strandin) .........
`0.23
`4.6
`10. Strandin in undialyzable fraction .................
`2.02
`40.4
`11. Lipides other than strandin (9) — (10). .
`.
`.
`.
`.
`.
`. ..
`92.75
`1855.0
`12. Final lower phase; total solutes ...................
`
`
`1914.413. Total lipides including strandin (4) + (9) + (12). 95.72
`
`cient for removing all the non-lipids contaminants from the crude extract.
`In the case of gray matter, lipides other than strandin lost in the course of
`the first washing amounted to no more than 0.6 per cent of the tissue lip-
`ides; for liver and muscle, the values were somewhat higher, ranging up to
`2 per cent.
`Recognition of I/ipide Distribution—Altering Factor—It can be seen from
`Table II that the second upper phases contained more lipides than the
`corresponding first upper phases. This unexpected finding was investi-
`gated by preparing in duplicate six successive upper phases from aliquots
`of white and gray matter extracts, as described above. The lipides from
`each phase were recovered and analyzed (Table III).
`It was found in
`both cases that the amount of lipides increased markedly from the first to
`the second upper phase; it remained unchanged from the second to the
`
`000007
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`J. FOLCH, M. LEES, AND G. H. SLOANE STANLEY
`
`503
`
`third, and then decreased from the third through the sixth by a fairly con-
`stant factor which corresponded to the distribution of a group of lipides
`exhibiting a distribution coefficient of about 2.7 in favor of the lower phase.
`Thenegligible amount of lipides in the first upper phase could be explained
`only by assuming that some unknown “distribution coefficient altering”
`factor had been in operation in the washing of the original extract. The
`
`TABLE II
`
`Lipide and Non-Lipide Substances Removed by First and Second Washings of
`Total Lipide Extracts of Various Tissues
`
`Ist upper phase
`2nd upper phase
`
`
`Tissue
`
`White matter
`
`Gray matter
`
`Liver
`
`Muscle
`
`
`
`Non-lipide
`substances
`(dialyzable
`solutes)
`
`315155551127
`8.3
`7.8
`9.3
`9.2
`
`10.3
`10.2
`11.4
`11.9
`
`19.0
`18.8
`
`13.6
`13.9
`
`0.56
`0.57
`0.31
`0.65
`1.4
`1.2
`2.0*
`2.4*
`
`0.35
`0.27
`0.11
`0.05
`0.04
`0.11
`0.41
`0.39
`
`1.5
`1.7
`0.8
`0.9
`1.9*
`1.7*
`1.6*
`1.6"
`
`0.47
`0.47
`0.21
`0.44
`0.62
`0.54
`1.02*
`1.17*
`
`0.26*
`0.28*
`
`2.08*
`0.31*
`0.13
`1.8*
`2.41*
`0.36*
`0.17
`1.9*
`
`
`Lipides other than
`strandin
`
`Non-lipide
`substances
`(dialyzable
`solutes)
`
`Lipides other than
`strandin
`
`_
`.
`~
`mg. per
`35%;?35.532 p8’122752353“ 57,53,511 75.555.81.553;
`0.70
`0.31
`0.27
`4.1
`0.51
`0.24
`0.55
`3.5
`0.74
`0.34
`0.19
`4.6
`0.69
`0.32
`0.18
`4.7
`
`per cent
`$37,553;
`1.85
`1.58
`2.11
`2.18
`
`1.82
`2.06
`1.19
`1.34.
`4.38*
`3.92*
`325*
`3.25”“
`
`
`
`
`
`* Lipides, including strandin.
`
`effect of this factor was still evident in the second equilibration, most likely
`because of contamination of the system by first upper phase.
`The lipides from white matter upper phases 3 through 6 were pooled and
`analyzed, in per cent: S 1.4, P 1.9, N 1.5, NHg-N 0.54, a—amino acid N 0.54,
`carbohydrate, as galactose, 7.9, S + P/N atomic ratio 0.98, atoms S per
`moles of galactose 1.00, choline none. Thus, the lipides affected by the
`distribution-altering factor consisted of a mixture of 40 per cent sulfatides,
`35 per cent phosphatidyl serine, and 25 per cent other phosphatides; 1.6.,
`they Were mainly, if not exclusively, acidic lipides.
`Identification of Lipide Distribution-Altering Factors—The observed facts
`might be explained by assuming that the distribution of water, chloroform,
`
`000008
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`504
`
`ISOLATION OF TOTAL TISSUE LIPIDES
`
`and methanol between the two phases had changed significantly between
`the first and subsequent equilibrations. This possibility was investigated
`by determining the volume and the specific gravity of both phases through
`the procedure of preparation of six upper phases (see above). No changes
`were observed. Thus, it can safely be assumed that the composition of
`the phases had remained essentially constant.
`
`Ltp'ides Present in Successive Upper Phases of System 175 Ml. of
`CHClssCH30H Extract Plus 35 Ml. of Water
`
`TABLE III
`
`Upper phases
`
`White matter lipides other than
`Gray matter lipides other than
`strandin
`strandln
`
`
`___‘
`
`Yield
`
`P content
`
`Yield
`
`P content
`
`misfit: 3:037:17}.
`
`per ”m
`
`”lgsszireslriiatil.
`
`1’" cent
`
`
`
`
`
`
`Ia*
`1b*
`2:).
`21)
`
`3a
`3b
`4a
`
`4b
`5a
`5b
`6a
`6b
`
`6.1
`4.5
`35.8
`30.1
`
`38.5T
`33.2T
`25.6T
`
`24.8T
`20.0T
`18.0T
`11.0T
`13.0]L
`
`’
`
`.
`
`2.32
`2.82
`2.42
`2.69
`
`1.67
`1.98
`2.01
`
`2.01
`2.06
`2.14
`2.10
`1.98
`
`'
`
`;
`
`4.1
`4.1
`13.2
`14.7
`
`13.4T
`12.4T
`9.0T
`
`9.1T
`7.7T
`9.3T
`3.4T
`5.11“
`
`2.72
`2.96
`2.54
`2.54
`
`2.36
`2.50
`2.61
`
`2.67
`2.47
`2.54
`2.40
`2.70
`
`* Upper phases a and b refer to duplicate experiments.
`T Strandin estimations were not carried out on lipides from these phases. Thus
`values given for total lipides include strandin, but the amounts of strandin present
`are negligible.
`
`Another explanation could be that the factor was one or more of the
`solutes in the crude extract which would be removed by the washing pro-
`cedure, and therefore would be found in the first upper phase. This was
`shown to be the case by the following type of experiment. A stock of
`lower phase was prepared by washing crude white matter extract once
`with water. The solutes from the upper phase were recovered.
`Identical
`aliquots of lower phase were mixed with equal volumes of pure solvents
`upper phase. Different amounts of first upper phase solutes were added
`to some of the mixtures. After centrifugation, the upper phases were
`analyzed for P centent, which had been shown to be a reliable indicator of
`the total amount of lipides present.
`It was found that the amount of
`lipides in the upper phases was decreased by the presence of the added
`
`000009
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`J. FOLCH, M. Lnns, AND G. H. SLOANE STANLEY
`
`505
`
`solutes in proportion to the logarithm of the concentration (Fig. 1), an
`observation which provided a means for measuring the distribution—altering
`effect of any material. Thus, it was possible to trace this effect quantita—
`tively from the first upper phase solutes to their dialyzable fraction and to
`
`50
`
`7?
`
`w
`
`
`
`
`O
`20 406080100
`51‘L4b‘631gotldo
`°/o DECREASE IN AMOUNT
`"I° DECREASE IN AMOUNT
`OF LIPIDE IN UPPER PHASE
`OF LIPIDE IN UPPER PHASE
`FIG. 1
`FIG. 2
`
`FIG. 1. Effect of solutes from the upper phase of the biphasic system White matter
`chloroform—methanol extract plus 0.2 its volume of water on the distribution of lip-
`ides between the two phases of a system of identical solvent composition.
`FIG. 2. Effect of different salts on the distribution of lipides between the two
`phases of the same solvent composition as those obtained from the system White
`matter chloroform-methanol extract plus 0.2 its volume of water; 0, KCl; X, NaCl;
`0, MgClg; A, 08:012.
`
`the ash therefrom; ’i.6., the effect was caused by the mineral salts present
`in the crude extract.
`
`By the same procedure, the distribution-altering effect of different con-
`centrations of NaCl, KCI, 03012, and MgClg was determined (Fig. 2).
`It
`was found that virtual absence of lipides from the upper phase could be
`obtained by the addition to it of CaCh or MgCh at a concentration of
`0.003 N, or of NaCl or KCl at a concentration of 0.05 N.
`Comparison between Amounts of Lipides Lost upon Washing with Water or
`with Mtneral Salt SolutionsmThis comparison has been made by washing
`
`crude extracts of various tissues in parallel with either water or aqueous
`
`0000010
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`
`
`
`0SALTADDEDTOUPPERPHASEASMEQ/L
`
`5 U
`
`?
`
`0.5
`
`3:)U].—
`
` \\ -LOGGONG.0FSOLIDSADDEDTOUPPERPHASE
`
`

`

`500
`
`ISOLATION OF TOTAL TISSUE LIPIDES
`
`solutions of different salts at various concentrations, and determining the
`amount of lipides in the first upper phase. Table IV gives the comparative
`data for water and for 0.05 per cent CaClg which result in a concentration
`of Ca++ in the upper phase of 3.8 m.eq. per liter.
`It can be seen that the
`use of the latter decreases the loss of lipides incidental to the washing.
`In
`cases in which the use of CaCh is contraindicated, as when an insoluble
`Ca salt might be formed, a similar result can be obtained with 0.04 per cent
`
`TABLE IV
`
`Amounts of Lipides Removed from Lower Phase in
`Presence and Absence of Added Ca++
`
`Lipides in lst upper phase expressed as mg. per gm. fresh tissue
`
`
`Tissue
`
`
`In absence of added Ca”
`In presence of added Ca”
`
`Strandin
`glgagdsegggggg
`Straudin
`Eggdgfrgflg
`
`
`
`
`
`0.78
`0.74
`
`0.90
`0.91
`
`3.0
`2.7
`3.6
`0.19
`0.19
`
`White matter
`
`Gray matter
`
`Liver
`
`Muscle
`
`
`
`0.70
`0.51
`
`0.74
`0.69
`
`0.47
`0.47
`0.38
`0.62
`0.54
`
`1 .02*
`1. 17*
`
`026*
`0 . 28*
`
`0.44
`0.41
`
`0.32
`0.33
`
`1.95
`1.15
`1.90
`0.12
`0.09
`
`
`
`l
`‘
`
`.
`
`g
`'
`
`'
`'
`
`0.27
`0.28
`
`0.13
`0.18
`
`0.24
`0.13
`0.25
`0.28
`0.20
`
`0. 52*
`0. 54*
`
`0.14”“
`0. 04*
`
`* Lipides including strandin.
`
`MgClg, 0.73 per cent NaCl, or 0.88 per cent KCl. The procedure is ex-
`actly as described for water.
`Study of Efiiez'ency of Repeated Washing—While one washing is sufficient
`to purify lipides for the usual analytical purposes, in the case of metabolic
`studies involving the use of substances labeled with radioisotopes, it is often
`necessary to free lipides from non-lipide contaminants possessing specific
`activities 1000—f01d or more that of the lipides. Such a degree of purifica—
`tion can be reached by equilibrating the lower phase repeatedly with por-
`tions of pure solvents upper phase containing salt. The procedure is as
`follows: The crude extract is washed with water or with an appropriate
`salt solution, as already described. After quantitative removal of the
`upper phase, a portion of pure solvents upper phase containing the appro-
`
`000001 1
`Downloaded from http://www.jbcorg/ by guest on June 12, 2013
`
`

`

`J. FOLCH, M. LEEs, AND G. H. SLOANE STANLEY
`
`'
`
`507
`
`priate salt is added, the two phases are stirred, and the tube is capped.
`After centrifugation,
`the upper phase is removed quantitatively. The
`equilibration with fresh portions of pure solvents upper phase containing
`mineral salt is repeated as many times as is indicated by ad hoc experiments
`of the type reported below.
`A study of the efficiency of repeated washings has been carried out in
`collaboration with Dr. Manfred Karnovsky 0f the Biophysics Laboratory
`of Harvard Medical School.
`2 mg. samples of a C“- or P32-labeled com—
`pound were dissolved in 0.1 ml. of water and added to 10 ml. of a crude
`liver lipide extract. The level of activity ranged between 0.5 X 106 and
`1.0 X 106 c.p.m. per 10 ml. of extract. The extracts were washed repeat-
`edly as described above, aliquots of the lower phase were taken after each
`
`Extent of Removal of Added Substances by Repeated Washing of Ltpz‘de Extract
`
`TABLE V
`
`Amount remaining in lower phase after
`
`lst washing
`
`2nd washing
`
`3rd washing
`
`Per cent original added radioactivity
`
`
`
`
`
`Labeled substance added
`
`Glycerol .......................
`Glucose .....................
`
`Sodium acetate .................... j
`Choline, no CaClg added .......... i
`“
`03012 added .............
`Serine ...........................
`
`1.3
`0.8
`
`1.1
`16.2
`2.4
`0.2
`
`0.7
`0.16
`
`0. 12
`7.3
`0.
`0.006
`
`0.08
`
`E
`i
`
`0.4
`0.09
`
`0.09
`4.5
`0.1
`0.007
`
`0.09
`
`Sodium phosphate, monobasic ......
`
`0.26
`
`4th washing
`
`0.31
`0.07
`
`0.06
`2.0
`0.1
`0. 007
`
`
`
`washing, and the amount of radioactive test substance remaining in the
`lower phase was estimated by counting in a gas flow counter in the pro-
`portional range (7, 8). Glycerol, glucose, acetate, choline, serine, and
`phosphate have been studied in this way.
`It can be seen from Table V
`that, while the repeated washing procedure is highly effective, the rate of
`removal of the different substances in the successive washings does not
`follow a theoretical decrement line, with the possible exception of serine
`and glycerol in the first two washings. The difference in the behavior of
`choline in the presence and absence of added CaClg suggests that choline
`forms salts with acidic lipides in amounts determined by competition with
`other bases present.
`Behavior of Strandt'n in This Procedure—The distribution of strandin
`between the two phases is affected by the addition of mineral salts to the
`upper phase, especially by CaC12 (Table IV). The effect of KCl is much
`less marked, and even at KCl concentrations that result in the essential
`
`0000012
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`508
`
`ISOLATION OF TOTAL TISSUE LIPIDES
`
`absence Of acidic lipides from the upper phase the bulk of strandin is present
`in the first upper phase. The observations Of Svennerholm (9) on the
`effect of NaCl on the distribution of gangliosides, which most likely in—
`cluded strandin, suggest that the action Of NaCl is similar tO that of KCl.
`In summary, the use of KCl or NaCl in this procedure afiects the distribu—
`tion of strandin only slightly. To eliminate strandin from the lower phase
`completely, three washings with the appropriate salt solutions should suf-
`fice. To isolate strandin, the three washings are combined, concentrated
`almost to dryness, and dialyzed. Strandin will be found quantitatively
`in the undialyzable fraction.
`
`DISCUSSION
`
`The present work started as an attempt to modify the original procedure
`of washing crude lipide extracts with water.
`In a survey of possible alter-
`natives, crude brain white matter extract and water were mixed in var-
`ious proportions. Most mixtures resulted in emulsions which were hard
`to separate or were inseparable. The exception was a mixture Obtained
`by adding to the extract 0.2 its volume Of water, which, upon standing or
`by centrifugation, separated into two clear phases without the persistence
`Of any interfacial fluff.
`Investigation Of the two phases showed that the
`upper phase contained practically all of the non—lipide substances and only
`negligible amounts Of lipides, the lower phase thus representing a solution
`of essentially pure total tissue lipides. Further study revealed that the
`high efficiency of the washing procedure depended upon the presence, in
`the system crude extract plus water, of chlorides of Na, K, Ca, and Mg,
`which had been extracted from the tissue by the chloroform—methanol
`mixture and which altered the distribution of acidic lipides between the
`two phases of the system and practically eliminated them from the first
`upper phase.
`A possible explanation for the lipide distribution~a1tering effect Of the
`mineral salts is that the acidic lipides, which are extracted from the tissue
`as salts of Na, K, Ca, and Mg, are present in the upper phase partly in
`the dissociated forms and in the lower phase only as undissociated salts.
`The addition Of mineral salts containing the above cations would decrease
`the dissociation of the acidic lipides by a mass action effect with a conse-
`quent shift of lipides to the lower phase, the mineral salts remaining quan-
`titatively in the upper phase. A necessary corollary to this hypothesis
`would be that the various cations displace one another from combination
`with the lipides. A study of the interaction between these lipides and
`mineral salts, the details Of which will be published elsewhere, supports
`the above hypothesis.
`It will be noted that the addition of Ca012 to the water used in the first
`
`0000013
`Downloaded from http://www.jbc.org/ by guest on June 12, 2013
`
`

`

`J. FOLCH, M. LEES, AND G. H. SLOANE STANLEY
`
`509
`
`washing significantly reduces but does not completely eliminate the loss
`of lipides (Table IV); in subsequ