STUDIES ON SEEDS II. Origin and Degradation of Lipid ...

STUDIES ON SEEDS

II. Origin and Degradation of Lipid Vesicles in

Pea and Bean Cotyledons

HILTON H . MOLLENHAUER and CLARA TOTTEN

From the Charles F . Kettering Research Laboratory, Yellow Springs, Ohio 45387

ABSTRACT

At least two kinds of lipid vesicles are present in pea and bean cotyledons which can berecognized at seed maturity on the basis of whether they do or do not interassociate intolipid vesicle sheets . Those that do interassociate into sheets are also characterized by (a)their association with plastids or plasma membranes during dormancy, and (b) the uniquetransformation into flattened saccules that they undergo during the first few days of seedgermination . These interassociated (or composite) lipid vesicles have been found in only afew seeds and may be restricted to certain classes of plants and/or certain states of cellulardevelopment .

Lipid vesicle-to-saccule transformation is predominantly confined to the germinatingseed. However, some lipid vesicle-derived saccules are already present in some cells evenbefore the seed reaches maturity . These partially transformed vesicles and saccules remainunchanged over dormancy, and then resume their transformation when the seed is germi-nated. This suggests that some stages of seed germination are already underway beforethe seed reaches maturity and are only resumed at seed germination .

The lipid vesicles that do not interassociate into sheets (i.e ., the simple lipid vesicles) arepresent in all tissues at all states of cellular development . These vesicles do not undergoany conspicuous structural changes during development .

INTRODUCTION

This report is concerned with the reserve lipids(principally triglycerides) of seeds, which areeasily recognized as distinct protoplasmic inclu-sions occurring as fluid droplets of various sizeseither dispersed in the cytoplasm or aggregatedinto larger masses . Reserve lipids of this type arecommon in seeds (4, 6, 10), spores, and embryos,but are also found to a lesser extent in meristematiccells and in differentiated vegetative cells (3, 6, 13) .

"Typical" droplets or vesicles of reserve lipidsare spherical in form and about 0 .1-2 .0 µ in diam-eter . They are bounded by an interfacial structure(possibly a membrane) which is clearly visible ingood electron microscope preparations (9, 16) .

Occasionally, lipid vesicles are closely associatedwith mitochondria (5, 7, 24), endoplasmic retic-ulum (22, 23, 24), or other lipid vesicles (13), butmost often they appear free in the cell cytoplasm .Some lipid vesicles may be equivalent to thespherosomes seen by light microscopy (2, 8, 11,19-21, 25), thus implying that these vesicles havea lysosomal function as well as acting simply as apool of reserve lipid . Previous reports have il-lustrated lipid vesicle form in various tissues (7,9, 13, 15) .

Reserve lipids are thought to be elaborateddirectly by the cytoplasm (6) or by portions of theendoplasmic reticulum (22-24) . Time-sequence

THE JOURNAL OF CELL BIOLOGY . VOLUME 48, 1971 . pages 3 9 5-405

395

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by PubMed Central

studies of intracellular movement of labeled lipidand lipid precursors have been carried out inseveral animal tissues (22-24) . The results indicatethat uptake and esterification of amino acids occurvery rapidly (in 1-5 min) in association with bothrough and smooth elements of the endoplasmicreticulum (22-24), and possibly also mitochondria(24) . In these tissues (liver, heart, and mammaryglands) the esterified lipids are primarily tri-glycerides which appear in the cytoplasm as smalldroplets bounded by an interfacial structure ormembrane (23) . The membrane is presumablyderived from the endoplasmic reticulum (23) .

The pathway of triglyceride synthesis is not aswell documented in plant tissues as it is in animaltissues, although it seems unlikely that the mech-anisms would be significantly different from thosein animal tissues. An example of a plant lipid-synthesizing system was described by Frey-Wys-sling and his collaborators in relation to the forma-tion of spherosomes (8) . These highly refractilebodies (as viewed by phase-contrast microscopy)are reactive with lipid stains (2, 8, 11, 19-21, 25)and with chemicals that demonstrate varioushydrolytic enzymes (2, 8, 11, 19-21, 25) . Frey-Wyssling et al . (8) proposed that spherosomes areformed from the endoplasmic reticulum in an im-mature state, containing enzymes for lipid syn-thesis but little or no reserve lipid. As lipid syn-thesis takes place, these immature bodies evolvefirst into spherosomes and then into vesicles ofreserve lipid . At maturity, these vesicles of reservelipid probably contain few or no enzymes (8, 9) .

Many seeds store substantial quantities of re-serve lipids which can be isolated as intact vesiclesby relatively simple procedures . By and large,these lipids are synthesized during the last fewweeks of seed development and then utilized bythe embryo during the first few days of germina-tion. Because of this separation of metabolicevents, cellular changes accompanying lipid syn-thesis and utilization can be followed accurately .Pea and bean cotyledons were chosen for these

studies because they are similar morphologically,biochemically, and developmentally and, there-fore, can be used more or less interchangeably. Inpractice, however, most of the information onisolated lipids has come from bean cotyledon andmost of the information on structure has comefrom pea cotyledon, because bean cotyledons havegiven the most uniform isolates, and pea cotyledonshave been the most amenable for ultrastructuralanalysis . In all of this work, however, both cellular

396

THE JOURNAL OF CELL BIOLOGY . VOLUME 48, 1971

systems have been compared to ascertain that theinformation presented is characteristic of bothseed types .Two other points regarding the choice of a

cellular system for study should be mentioned .They have to do with the amount and the kindsof lipids that the seed accumulates . Pea and beanare particularly useful for developmental studiesbecause the changes accompanying lipid vesicleformation and degradation are well defined, andbecause the number of lipid vesicles formed issufficiently small so that structural changes canbe easily followed . Both pea and bean contain atleast two structurally identifiable kinds of lipidvesicles . This diversity adds markedly to the valueof these studies since it may lead eventually to anunderstanding of the mechanisms which controllipid synthesis and form .

For clarity, the work reported in this paper, inthe accompanying paper, and in two papers yet tobe published has been divided into four parts :techniques for fixing seeds, changes accompanyingseed development and germination, isolation andform of lipid vesicles, and chemical compositionof lipid vesicles at one stage during germination .Most of the information relates to the lipid vesicleswhich we have called "composite" because thesevesicles have not been described before and be-cause they are unique enough that changes in theirform can be related to stages in development .

MATERIALS AND METHODS

For developmental studies, peas (var. Alaska) andbeans (var. Topcrop) were grown in the garden inspring and early summer . The seeds were harvestedat various developmental stages until they weremature and the seed pods were brown in color.

For the germination studies, seeds from com-mercial sources were used . Care was taken to obtainuntreated seeds, since even slight residues of fungi-cides affect the structure of the cells during earlystages of germination . For short germination times(1-4 hr), the seeds were soaked in shallow dishes ofwater at room temperature and in the light . The seedswere not covered with water since this inhibitsgermination . For longer germination times, the seedswere presoaked as above and then transferred tomoist vermiculite in an incubator and germinatedin the dark at 27 °C. For test purposes, a few seedswere germinated in the light, but this did not seem toalter the sequence of events reported here .

In their preparation, fixation, and embedding, theseeds were handled as described in the accompanyingpaper (16) . In most instances, this involved prefixa-

FIGURE 1 Micrograph of pea cotyledon late in seed development . Protein bodies (PB) are fully formedat this developmental stage but lipid vesicles have not yet accumulated at the cell surface or adjacentto the plastids . Newly formed lipid vesicles are spherical and single (i .e ., not interassociated), and may beclosely appressed to the outer surface of the endoplasmic reticulum (see arrows) . X 30,000 .

397

tion in a mixture of acrolein, paraformaldehyde, andglutaraldehyde in 0.05 M collidine buffer, and post-fixation in collidine-buffered OSO4 . The tissues weredehydrated in a series of acetone solutions, embeddedin an Epon-Araldite epoxy resin mixture (12), andviewed with a Philips EM-200 electron microscope .

RESULTS

Kinds of Lipid Vesicles

Lipid vesicles (see Figs . 1-10) are present inmost cells of the cotyledon throughout develop-ment, dormancy, and germination, but they donot constitute a major cellular constituent untilthe seed approaches maturity. Near maturity theseed synthesizes great quantities of reserve lipids(Fig. 1) which are stored as small vesicles overdormancy (Fig . 2) . These vesicles become a con-spicuous component of the cell because they arepresent in such large numbers and because someof them occupy a unique position in the cell (Figs .2, 3) .

At seed maturity, the lipid vesicles can bedivided into two general classes according tostructural and developmental differences (see Fig .3) . One class of lipid vesicles is called "simple"because they are similar to those in most otherplant and animal cells (see arrow, Fig . 3) . Simplelipid vesicles do not form close associations with oneanother and do not appear to undergo any con-spicuous transformations . The other class of lipidvesicles is called "composite" because they forminto sheets as the seed approaches dormancy (Fig .2) . These sheets of lipid vesicles are characteristi-cally located near the cell surface or adjacent to theplastids (Figs . 2, 3, 5) . The composite lipid vesiclesundergo a unique kind of transformation duringthe first several days of dormancy (see Figs . 6-9) .

Lipid Vesicle Development

It has always been difficult to demonstrate andcorrelate structural and developmental changes inlipid vesicles, a problem which was also encount-ered in the present study . This difficulty is il-lustrated by the simple lipid vesicles which arealways present in the cell and never seem toundergo any change except a gradual enlargementor diminution in size . Because these changes aregradual, it is very difficult to relate lipid vesicleform to cellular activity .

The composite lipid vesicles are more amenableto study because definite stages in development

398

THE JOURNAL OF CELL BIOLOGY • VOLUME 48, 1971

and/or degradation can be followed . These ves-icles appear to be synthesized in cytoplasmic re-gions containing tubular endoplasmic reticulumand Golgi apparatus (Fig. 1) . They first appear assmall, individual, spherical vesicles (Fig . 1) whichare free in the cytoplasm or close to the outersurface of an element of the endoplasmic reticulum .They do not appear to form in the lumen of theendoplasmic reticulum or in the Golgi apparatus .

It is important to point out that these lipidvesicles are not interassociated at this develop-mental stage; they are single vesicles that are freein the cytoplasm and look very much like thesimple lipid vesicles described above. They can berelated to the composite lipid vesicles only becausethey appear in such great numbers and becausethey are more uniform in size and are usuallysmaller than the simple lipid vesicles .As they mature, the composite lipid vesicles

increase in size, but otherwise do not changesignificantly in appearance . They seem to remainin the region in which they are formed and do notmigrate to the cell surface or to the plastids untilthe seed is almost mature .

Ultimately, the composite lipid vesicles come toreside at the cell surface or adjacent to the plastids .The sequence of their interassociation is not clear,but the evidence suggests that they migrate asindividual vesicles to the surface to which they willattach . When they reach this surface they elongateand interconnect and/or interassociate with eachother and with the membranes of the cell surfaceor the plastids (Figs . 2, 3, 5) . By the time the cellreaches dormancy, the composite lipid vesicleshave formed into sheets at the cell surface or ad-jacent to the plastids . They remain in one of thesepositions until the seed is germinated .

Degradation of Lipid Vesicles

Simple lipid vesicles are present during germina-tion but no special or unique transformations canyet be associated with them . Any structural changesthat might be associated with their turnover aretoo small to be defined .Changes in composite lipid vesicles are less

difficult to follow ; i .e ., these vesicles are transformedinto thin saccules and then disappear . This trans-formation takes only 2-4 days .

The most striking aspect of lipid vesicle trans-formation is the formation of saccules bounded bymembranes which often appear more dense andmore distinct than other membranes of the cell

FIGURE 2 Micrograph of dormant pea cotyledon showing the alignment of lipid vesicles along the cellsurface. The same alignment pattern is also found around the plastids but is not illustrated in this figure(however, see Fig. 5). X 32,000 .

HILTON H. MOLLENHAUER AND CLARA TOTTEN Studies on Seeds . II

399

FIGURE 3 Micrograph of pea cotyledon showing the form of composite lipid vesicles at an early stage ofseed germination. Multilayers of lipid vesicles are common only in regions where several plastids (P) arelocalized as they are here . A simple lipid vesicle is shown at the arrow. Protein body (PB) . X 28,000 .

FIGURE 4 Sheets of lipid vesicles remain intact even after homogenization and isolation, and have thesame form that they do in vivo. This preparation is from bean cotyledon after the seed had been soakedin water for 2h-3 hr . A simple lipid vesicle is shown at the arrow . X 24,000 .

4 0 0

(Figs . 6-8) . These saccules are probably flattenedsacs, since no tubular profiles of them have beenseen in thin section . They are always about 150 Ain over-all thickness, but may vary considerably inapparent length and/or width . It is not knownwhether these variations in size are the result ofdifferent planes of section through a saccule or ofreal differences in saccular dimensions .

Each composite lipid vesicle is eventually con-verted into a saccule, but the process is not syn-chronous . Many transformational states can beseen in each cell, and even in a single sheet of lipidvesicles (see Figs . 6 and 8) . The rate of transforma-tion seems to be proportional to the rate at whichthe seeds germinate. In pea and bean cotyledons,the transformations are usually complete in 2-4days. After they are formed, the saccules remaininterassociated for a while and then disappear fromthe cytoplasm (Fig. 9) .

DISCUSSION

Reserve lipids of most seeds are present in the formof spherical droplets which are clearly separatedfrom the cytoplasmic ground substance by a thininterfacial structure. In suitable electron micro-graphs, this interfacial structure appears relativelydense and, therefore, is thought to be a boundingmembrane. This interpretation is supported bythe following observations : (a) In tissue sections, amembrane, or at least a dense residue, is visible atthe lipid-cytoplasm interface even after lipids havebeen extracted from the tissues (9) . (b) Lipid drop-lets which become packed together, as they do inhigh-fat seeds and some meristematic tissues (13),do not fuse together. (c) In these seeds, the thininterfacial structures bounding the lipid dropletsare continuous with the more distinct membranesbounding the flattened saccules (Fig. 6) . (d) Innegatively stained preparations, the interfacialstructures of the lipid droplets are continuous withmembrane fragments (17) . Therefore, until furtherinformation is available, it is assumed that theselipid droplets are bounded by a membrane andthat they should be called lipid vesicles ratherthan lipid droplets .

The composite lipid vesicles appear to be syn-thesized in association with, but external to, theendoplasmic reticulum . We have not yet foundany instance in which pools of reserve lipids appearinside the endoplasmic reticulum, or in whichlipid droplets are continuous with, or appear to bebudding from, the endoplasmic reticulum . There-

fore, if reserve lipids are formed in the lumen ofthe endoplasmic reticulum, as suggested by thework of Stein and Stein (22-24), then they mustbe transported to lipid vesicles in packets whichare below the resolution limit of the microscope .At this time, it seems more reasonable to assumethat lipid synthesis occurs external to, or on thesurface of, the endoplasmic reticulum, or even inspherosomes as suggested by Frey-Wyssling et al .(8) .

Simple lipid vesicles seem to be present in allseeds at all developmental stages, but no con-spicuous transformations are associated with them .In many seeds (particularly the high-fat seedssuch as those of the peanut and the castor bean),simple lipid vesicles increase or decrease signifi-cantly in number (and occasionally also in size)during development and germination, respectively,but otherwise do not change in form, density, orgeneral appearance. However, in pea and beancotyledons most of the increase in lipid during seeddevelopment seems to come from composite, ratherthan simple, lipid vesicles. The simple lipidvesicles are never a conspicuous feature of thecytoplasm and, therefore, are not a very useful toolfor ultrastructural analysis .

Simple and composite lipid vesicles are bio-chemically and structurally similar (1, 17, andFig. 3) . However, it is not yet possible to determinewhat differences may actually exist between them .The factors responsible for interassociation and forthe vesicle-to-saccule transformation described inthese reports will be considered elsewhere . It hasbeen assumed that differences probably exist fromthe time of vesicle inception, but there is no realevidence to prove this . It will be necessary to findout first how the two classes of vesicles differ andthen to analyze the vesicles for these differences ateach developmental stage .

It was thought initially that the transformationof lipid vesicles into saccules might be a mechanismfor rapidly synthesizing smooth endoplasmic re-ticulum (14) . More recent work, however, hasshown that the saccular membranes are signifi-cantly different structurally from those of theendoplasmic reticulum or from those of othermembrane-bounded components of the cell . Itmay even be that, since they disappear shortlyafter they are formed, and since they do not appearto be common to all seeds (see below), the sacculesrepresent a breakdown product, or lipid vesicleresidue, with no metabolic function at all .

HILTON H. MOLLENHAUER AND CLARA TOTTEN Studies on Seeds . II 401

402

THE JOURNAL OF CELL BIOLOGY • VOLUME 48, 1971

It is common for lipid vesicles in seeds to bealigned with the cell surface (for example, see Fig .10) or around protein bodies (4, 10, 18, resultsunpublished), though the factors which cause thisare unknown. In most instances, however, thesealigned vesicles do not seem to be interconnectedor interassociated and, therefore, do not undergo atransformation of form. In the course of ourstudies, we have looked at soybean, peanut,pumpkin, watermelon, squash, castor bean, maize,various beans, and pea, and have found compositelipid vesicles and lipid vesicle-derived sacculesonly in bean and pea . Although this sampling isnot yet very large, it does indicate that the occur-rence of composite lipid vesicles may be a relativelyunique event that is restricted to certain tissues orstages of development . It also implies that onlycomposite lipid vesicles are capable of transforma-tion into saccules and that these conditions areprobably related .Membrane changes accompanying saccule

formation (see Fig . 6) can be correlated withsedimentation properties and protein-lipid ratiosof the composite lipid vesicles at various trans-formational stages (see references 1 and 17 fordetails) . In brief, these data show that the buoyantdensity and the protein-lipid ratio of the sacculesare greater than those of lipid vesicles, thus sug-gesting that there is either an increase in proteinor a decrease in lipid (or both) during conversionof the composite lipid vesicles into saccules . Webelieve that at least some part of these transforma-tions must be associated with membrane changesif the data are to fit the ultrastructural observa-tions . Thus, it seems reasonable to suggest that theincreased thickness and electron opacity of the

saccular membranes compared with the lipidvesicle membranes are due to a net accumulationof protein on, or in, the membrane of the saccule .

The reasons why lipid vesicles accumulate closeto the plastid and plasma membranes are notclear. The data only suggest either that plasmaand plastid membranes become similar (in thesense that they bind or attract the composite lipidvesicles), or that two classes of composite lipidvesicles exist such that one class binds to the plasmamembrane and the other class binds to the plastidmembrane. We presently favor the first hypothesissince all of the composite lipid vesicles are similarin form and development and do not appear to besubdivided into classes .

We have implied in the previous discussions thatlipid vesicle development and interassociation, andlipid vesicle transformation into saccules, areneatly restricted to seed development and germina-tion, respectively . This is not entirely true sincelipid fractions from dormant seeds contain somesaccules as well as the more usual spherical formof lipid vesicles. This situation is significantly moremarked in bean preparations than in pea prepara-tions, and is interpreted to mean that lipid vesicletransformation begins shortly after the vesiclesbecome interassociated, even if the seed has notyet reached dormancy. Insofar as we can deter-mine, these partially transformed vesicles andsaccules remain essentially unchanged over dor-mancy and then resume their transformation whenthe seed is germinated . This sequence of eventssuggests that some stages of seed germination arealready underway before the seed reaches maturityand are only resumed at seed germination .Various other data regarding composite lipid

FIGURE 5 If homogenization is gentle, then the composite lipid vesicles remain attached to the plasmamembrane (17) or to the plastids . X 10,000.

FIGURE 6 Micrograph of a pea cotyledon from a germinating seed showing several stages in the formationof lipid vesicle-derived saccules . These vesicles and saccules are next to the plasma membrane, but thesame transformational stages also take place in lipid vesicles adjacent to the plastids . Note that the mem-branes of the saccules are often thicker and more electron opaque than those of the lipid vesicles. X 100,000 .

FIGURE 7 Micrograph of pea cotyledon during seed germination showing a region where all of the com-posite lipid vesicles have been converted into saccules. In pea cotyledon, the complete transformationof composite lipid vesicles takes between 2 and 4 days, depending upon how fast the seed germinates . X80,000.

FIGURE 8 A micrograph of an isolated lipid vesicle-saccule sheet from bean cotyledon showing that thetwo structural forms are still interconnected . X 80,000.

HILTON H. MOLLENHAUER AND CLARA TOTTEN Studies on Seeds . II

403

FIGURE 9 Micrograph of pea cotyledon after the seed has germinated for about 8 days . Composite lipidvesicles and/or saccules are no longer present . Only simple lipid vesicles remain in the cytoplasm .X 35,000 .

FIGURE 10 This micrograph from the embryo of germinating maize is included to show that lipid vesiclesin these tissues also line the cell surface . These vesicles, however, do not appear to be interassociatedand are not transformed into saccules . These embryos were prefixed in glutaraldehyde-paraformaldehyde(16) and postfixed in KMnO4. X 8500 .

404

vesicles are given in subsequent reports (1, 17) and

will not be discussed here . The over-all picture,

however, reveals that a unique class of lipidvesicles exists in these tissues which can be recog-nized at seed maturity. These vesicles are capableof interassociating into sheets and of being trans-

formed into other structural entities .

Contribution No. 393 from the Charles F . KetteringResearch Laboratory, Yellow Springs, Ohio 45387 .Supported in part by United States Public HealthService Grant GM 15492 .

Received for publication 4 April 1970, and in revised form29 September 1970 .

REFERENCES

1. ALLEN, C . F ., P. GOOD, H. H. MOLLENHAUER,and C. TOTTEN . 1971 . Studies on seeds . IV .Lipid composition of bean cotyledon vesicles .J . Cell Biol. (In press) .

2. BALZ, H. P. 1966. Intrazelluläre Lokalisationund Funktion von Hydrolytischen Enzymenbei Tabak . Planta . 70 :207 .

3. BoucK, B . 1963. Stratification and subsequentbehavior of plant cell organelles . J. Cell Biol .18 :441 .

4. BUTTROSE, M. S. 1963. Ultrastructure of the de-veloping aleurone cells of wheat grain . Aust . J.Biol . Sci. 16:768.

5. DE ROBERTIS, E . D. P., W. W. NowsNSKI, andF. A. SAEZ . 1965. The cell . In Cell Biology .W. B. Saunders Company, Philadelphia, Pa .

6. ESAU, K. 1965. Plant Anatomy. John Wiley andSons Inc ., New York .

7. FAWCETT, D . W. 1966 . An Atlas of Fine Struc-ture : The Cell, its Organelles and Inclusions .W. B. Saunders Company, Philadelphia, Pa .

8 . FREY-WYSSLING, A., E. GRIESHABER, and K . MUH-LETHALER . 1963 . Origin of spherosomes in plantcells . J. Ultrastruct. Res. 8 :506 .

9 . JACKS, T . J ., L . Y. YATSU, and A. M. ALTSCHUL .1967 . Isolation and characterization of peanutspherosomes. Plant Physiol. 42 :585 .

10. JONES, R . L. 1969 . The fine structure of barleyaleurone cells . Planta . 85 :359.

11 . MATILE, P. A., J . P . BALZ, E. SEMADENI, and M .JosT. 1965. Isolation of spherosomes withlysosome characteristics from seedlings . Z .Naturforsch . 20:693 .

12. MOLLENHAUER, H. H. 1964 . Plastic embedding

mixtures for use in electron microscopy . StainTechnol . 39 :111 .

13. MOLLENHAUER, H. H. 1967. A comparison ofroot cap cells of epiphytic, terrestrial andaquatic plants. Amer . J. Bot. 54:1249 .

14. MOLLENHAUER, H. H . 1967 . Formation of smoothmembranes during seed germination . J. Cell

Biol. 35(2, Pt . 2) :96 A . (Abstr.)15. MOLLENHAUER, H. H. 1969. The ultrastructure

and chemistry of some seed lipids . J. Cell Biol.43(2, Pt . 2) :94 a. (Abstr .)

16. MOLLENHAUER, H. H., and C . TOTTEN . 1971 .

Studies on seeds. I . Fixation of seeds . J. CellBiol. 48 :387 .

17. MOLLENHAUER, H. H., and C . TOTTEN . 1971 .

Studies on seeds. III. Isolation and structureof lipid-containing vesicles from seeds . J. CellBiol . (In press) .

18. NIEUWDORP, P. J. 1963 . Electron microscopicstructure of the epithelial cells of the scutellumof barley . Acta Bot . Neer. 12 :295 .

19. RoDKIEwICZ, B., and M. KWIATKOWSKA. 1965 .Enzymy hydrolityczne w rozwijajacym sicworeczku Zalazkowym lilii . Acta Soc. Bot. Pol .34:235.

20. SEMADENI, E . G. 1967 . Enzymatische charakte-risierung der Lysosomenä quivalente (Sphä-rosomen) von Maiskeimlingen . Planta . 72 :91 .

21. SOROKIN, H. P., and S . SOROKIN. 1966. Thespherosomes of Companula persicifolia L. A lightand electron microscope study . Protoplasma. 62 :216 .

22. STEIN, O., and Y. STEIN, 1967. Lipid synthesis,intracellular transport, storage, and secretion .I . Electron microscopic radioautographicstudy of liver after injection of tritiated palmi-tate or glycerol in fasted and ethanol-treatedrats. J. Cell Biol. 33 :319 .

23. STEIN, O., and Y . STEIN. 1967. Lipid synthesis,intracellular transport, and secretion . II .Electron microscopic radioautographic studyof the mouse lactating mammary gland . J. CellBiol . 34 :251 .

24. STEIN, O., and Y . STEIN 1968 . Lipid synthesis,intracellular transport, and storage. III . Elec-tron microscopic radioautographic study of therat heart perfused with tritiated oleic acid . J.Cell Biol. 36 :63 .

25. WALEK-CZERNECKA, A. 1965. Histochemicaldemonstration of some hydrolytic enzymes inthe spherosomes of plant cells . Acta Soc . Bot .Pol . 34 :573 .

HILTON H. MOLLENHAUER AND CLARA TOTTEN Studies on Seeds . II 405