ELECTRON MICROSCOPY OF PLASMOLYSIS IN ESCHERICHIA COLI

EUGENE H. COTA-ROBLESDivision of Life Sciences, University of California, Riverside, California

Received for publication 30 August 1962

ABSTRACT

COTA-ROBLES, EUGENE H. (University ofCalifornia, Riverside). Electron microscopy ofplasmolysis in Esche7ichia coli. J. Bacteriol.85:499-503. 1963.-Escherichia coli cells plasmo-lyzed in 0.35 M sucrose reveal plasmolysis at onetip of a cell or in the center of dividing cells inwhich protoplast partition has been complete.Central plasmolysis reveals that protol)lastseparation can be completed before the in-vagination of the cell wall is complete. Thesestudies support the concept that these cellsdivide by constriction. The strength of the unionbetween cell wall and cytoplasm is not uniformaround the entire cell. It is strongest along thesides of these rod-shaped cells and weakest atone til) of the single cell. Thus, a single cellgenerally forms one cup-shaped vacuole inwhich the cytoplasm has collapsed away fromone tip of the cell.

Robinow's (1960) recent review of bacterialstructure succinctly summarizes our currentunderstanding of the anatomy of plasmolysis ofbacterial cells. He clearly points out that theunion between the cell wall and the cytoplasmicmembrane may be so strong that the separationof these two structures may be incompleteduring plasmolysis. In the observations de-scribed below, this point is underscored andelaborated upon by the examination of ultrathinsections of plasmolyzed cells of Escherichia coliB by electron microscopy.

MIATERIALS AND MNIETHODS

E. coli B was routinely cultured aerobically inFraser and Jerrel's (1953) glycerol medium at37 C for 16 hr. The culture was diluted 100-foldinto fresh medium and further incubated aerobi-cally at 37 C for 2 hr. The cells at this time are inthe early log phase of growth. The cells werethen harvested from the growth medium bycentrifugation and suspended in 0.35 M sucrose

to effect plasmolysis. To maintain the plasmo-lyzed state of the cells for any appreciable time,it was necessary to subject the cells to at leastone wash with distilled water prior to suspensionin the plasmolyzing medium. Plasmolyzed cellswere maintained at room temperature for 5 to 20min prior to addition of fixative.

Electron microscopy. Several fixative procedureswere used throughout these investigations. Themost successful one in our hands has been acombination of formalin fixation with Kellen-berger, Ryter, and Sechaud's (1958) procedureof OS04 fixation in the presence of Tryptone(Difco). To facilitate fixation with formalin, cellswere plasmolyzed in buffered 0.35 M sucrose.The buffer contained Na2HP04, 7.0 g; KH2PO4,3.0 g; NaCl, 4.0 g; MgSO4, 0.2 g; water, 1,000ml; pH 6.8. Plasmolyzed cells were fixed in 10Cc/)formalin for 1 hr at 24 C. The cells were thencentrifuged, and the pellet was resuspended in 1ml of 1 % buffered OS04 (Kellenberger et al.,1958) plus 0.1 ml of 1% Tryptone. This mixturewas maintained at room temperature for 16 hr.The cells were post-stained with 0.5 c0 uranylacetate for 2 hr. After uranyl acetate staining,the cells were centrifuged, and the pellets wereprepared for embedding in Epon 812 (Luft,1961) or Vestopal WV (Kellenberger et al., 1958).

Ultrathin sections were cut on a Porter-Pllummicrotome by use of glass knives. Occasionally,the sections were further stained with leadhydroxide (Watson, 1958). Electron micrographswere taken with an RCA EMU-3B electronmicroscope equipped with an objective aperture.

RESULTS AND DISCUSSION

The plasmolysis demonstrated by activelydividing cells of E. coli upon suspension in 0.35M sucrose after removal from the growth mediumis short-lived but characteristic. More long-lasting plasmolysis can be demonstrated after apreliminary wash in distilled water prior tosuspension in 0.35 M sucrose. I have not variedthe plasmolyzing conditions greatly, since our

499

on July 15, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

COTA-ROBLES

A

5.Oju~~~.

FIG. 1. Phase photomicrograph of cells of Escher-ichia coli B suspended in 0.35 M sucrose buffer.Note cell with central plasmolysis vacuole.

original observations of plasmolyzed cells were

made in conjunction with the studies of D-

amino acid-induced spheroplasts of this organism(Cota-Robles and Duncan, 1962). Consequently,I utilized 0.35 M sucrose in the observationsdescribed herein.The pattern of plasmolysis in 0.35 M sucrose

is quite striking when viewed by phase micros-copy (Fig. 1). Here, one can see that the separa-

tion of the protoplast from the cell wall is notuniform. One cell demonstrates a plasmolysisvacuole at one tip of the cell. A cell with a

central vacuole can also be seen in this figure.This pattern is more clearly demonstrated in

the series of electron micrographs containedherein. Figure 2, although not a particularlynew, unusual, or technically refined electron

micrograph, shows a normal cell of E. coli Bthat is in the process of cell division. Here, onecan discern that the cell wall adheres closely tothe protoplast. Nuclear division appears com-plete, even though the central invagination of thecell wall appears to have just commenced. Otherobservations (Fig. 3) reveal that nuclear divisionneed not be completed prior to invagination ofthe cell wall. The dividing cell depicted in thisfigure is one that was in the process of recoveryfrom plasmolysis. Hints of separation of theprotoplast from the tips of the cell can be ob-served.The central vacuole is well depicted in Fig. 4.

Here, separation of the daughter protoplasts iscomplete, and the large central vacuole is bor-dered by a cell wall that is stretched out. One ofthe newly formed protoplasts presents a concavesurface toward the center; the cytoplasm can beseen adhering to the cell wall to form a cup-shaped plasmolysis vacuole. The plasmolysisvacuole may be both central and at the tip, asdescribed in Fig. 5. This figure demonstratesthe beginning of an invagination of the cell wall.However, an unusual feature is that the invagina-tion appears to be spatially disoriented withrespect to the separation of the protoplasts. Thesame protoplast that appears to be out of placereveals a marked separation from the wall at thetip of the cell. Again, a cup-shaped vacuole isformed. It is my present belief that separation ofdaughter protoplasts occurs prior to completeinvagination of the cell wall. The misplacedprotoplast can be seen to possess a limiting mem-brane. This protoplast could have been drawnout of place as a consequence of the act ofplasmolysis. Figure 6 offers support to the con-tention that protoplasts are capable of separating,under the influence of plasmolyzing conditions,prior to complete invagination of cell wall. Here,one can see that the central plasmolysis vacuoleis completely formed, but that the separation ofdaughter protoplasts is uneven. A well-definedvesicle can be seen within the vacuole. A secondvesicle can also be discerned. However, thisstructure is still joined to the protoplast. Thenature of these vesicles is unknown, but theycould be portions of the cytoplasmic membranethat have been pinched off.

Figure 7 reveals the marked separation of walland protoplast that can occur at the tip of adividing cell in which the separation of proto-

500 J. BACTERIOL.

on July 15, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

VOL. 85, 1963 ELECTRON MICROSCOPY OF PLASMOLYSIS IN E. COJi

4'k . W

w s+,.. X S l . .

Pip- 1f

ss f *X ,_

'%#*:e

#

* 4L.*'

P ..

^ :w= l'X,: p[ ;w H)_" PF w%.

. X *

# w r,,.

4 O 5 X,.,,,.

3

x'.k

FIG. 2. Electron micrograph of a thin section of normal cells of Escherichia coli B. Fixed with osmiumtetroxide, embedded in Epon 812, and stained with lanthanum nitrate.

FIG. 3. Electron micrograph of a thin section of plasmolyzed cells of Escherichia coli B. Fixed with osmiiumtetroxide, embedded in Epon 812, and stained with lead hydroxide.

FIG. 4. Electron micrograph of plasmolyzed cells of Escherichia coli B. Fixed with formalin and osmiiiumtetroxide, embedded in Vestopal W, and stained with lead hydroxide. Note the cup-shaped plasmolysis vacuole.

FIG. 5. Electron micrograph of plasmolyzed cells of Escherichia coli B. Fixed with formalin and osmiumontetroxide, embedded in Vestopal W, and stained with lead hydroxide. Central plasmolysis vacuole is notcoordinated with the invaginating cell wall.

1:21.Unwrlm-',

LI 50 1

A, "I%,.,:,A

e 1

"i

Ir4.

:I 0;.

t,0

'o A'i O'

t

k&;': :l.,

11.... I

on July 15, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

rM.j X DJ: ;h

Rik"

6

v

7

*~~~~*0 I

.e *: .

,,,... ,.4.:

.'':

6 a::

4

0': 9e

FIG. 6. Electron micrograph of plasmolyzed cells of Escherichia coli B. Fixed with formalin and osmiumtetroxide, embedded in Vestopal W, and stained with lead hydroxide. Separation of protoplasts almost com-

pleted, permitting formation of central plasmolysis vacuole.FIG. 7. Electron micrograph of plasmolyzed cells of Escherichia coli B. Fixed with formalin and osmium

tetroxide, embedded in Vestopal W, and stained with lead hydroxide. Marked plasmolysis vacuole retaininga small vesicle at the tip of the cell.

FIG. 8. Electron micrograph of plasmolyzed cells of Escherichia coli B. Fixed with formalin and osmiumtetroxide, emibedded in Vestopal W, and stained with lead hydroxide. Note that cytoplasm is still conncctedto small vesicle within the plasmolysis vacuole.

FIG. 9. Electron micrograph of plasmolyzed cells of Escherichia coli B. Fixed with formalin and osmiumtetroxide, embedded in Vestopal W, and stained with lead hydroxide. The large cup-shaped plasmolysisvacuole emtiphasizes seciure attachment between the wall and the cytoplasm along the sides of the cell.

502

w

...

p

* a o - 5 IL -.. .. -:i *174

Moll.

A N

JK

on July 15, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

VOL. 85, 1963 ELECTRON MICROSCOPY OF PLASMOLYSIS IN E. COLI

plasts has not progressed sufficiently to permit acentral vacuole to form. The great separation atthe sides, I believe, is a result of the preparativeprocedure which, in this instance, involved rapiddehydration. Discrete bridges of cytoplasmicmaterial can be seen joining the cell wall alongthe side of the cell. The ill-defined vesicular struc-ture that appears in the plasmolysis vacuole is ofunknown origin. It could be of a similar origin tothe structure that can be seen within the plas-molysis vacuole of the cell pictured in Fig. 8.Here, a portion of the protoplast remains in thevacuole, but it is connected by a thin bridge ofcytoplasm to the main mass of protoplasm. It canbe noted that the portion of cytoplasm remainingwithin the vacuole is not homogeneous. Theextreme upper part of the plasmolysis vacuole ofthis cell contains a second structure whose originis vague but which may be related to an ancillaryaspect of these investigations: namely, the infec-tion of plasmolyzed cells of E. coli with coliphageT2. Although the nuclear material is not as wellresolved in this figure as in others, there seems tobe a clear indication that the nuclear materialhas been completely partitioned prior to extensiveinvagination of the cell wall. Figure 9 is a moreextreme demonstration of the plasmolysisvacuole which presents a concave surface to thetip of the cell. It appears as though the separationof the cytoplasm from the wall at the tip hasbeen so forceful that the cytoplasm has collapsed,leaving a cup-shaped plasmolysis vacuole.From the foregoing observation, I believe that

the protoplast of a growing cell of E. coli is notuniformly bonded to the cell wall. In fact, itappears to be bonded tightly along the sides of thecell, less tightly at one tip of the cell, and poorly,if at all, at the center of dividing cells in whichprotoplast separation has been completed.

If the cell has completed protoplast partition-ing, a plasmolysis vacuole may occur centrally.It appears abundantly clear that protoplastseparation may be complete prior to any exten-sive invagination of the cell wall between daugh-ter protoplasts. From these observations, itappears that E. coli divides by constriction, i.e.,by the invagination of cell wall between partiallyor completely partitioned protoplasts. There isno doubt that the partitioning of nuclear mate-rial is completed prior to any plasmolyticallysensitive separation of the daughter protoplasts.These conclusions should be contrasted withthose of Conti and Gettner's (1962) recentdescription of cell division in E. coli.

One feature yet to explain is why the plas-molytic separation of protoplast from cell wallshould be limited to one tip of the cell (Fig. 1and 9). Aside from invoking uneven physicalstrains, the only plausible explanation whichcomes to mind is that the association betweenwall and membrane may become more securewith age. The two tips of a single cell wereformed at different times; perhaps the more re-cently formed tip has a more weakly bonded walland cytoplasm, whereas the older tip has amore strongly joined wall and cytoplasm. Thus,newly partitioned protoplasts in a dividing cellwould show weak bonding between wall andcytoplasm, permitting the formation of a markedcentral plasmolysis vacuole.

ACKNOWLEDGMENTS

The author gratefully acknowledges theskilled technical assistance of Dawn Coffmanand Barbara Raymond. Thanks are also duePaul Desjardins for use of the electron micro-scope.

This work was supported in part by researchgrant E-3525 of the National Institutes ofHealth, U.S. Public Health Service.

LITERATURE CITED

CONTI, S. F., AND M. E. GETTNER. 1962. Electronmicroscopy of cellular division in Escherichiacoli. J. Bacteriol. 83:544-550.

COTA-ROBLES, E. H., AND P. H. DUNCAN. 1962.The effect of D-glutamic acid upon spheroplastformation in Escherichia coli B. Exptl. CellRes. 28:342-349.

FRASER, D., AND E. A. JERREL. 1953. The aminoacid composition of T3 bacteriophage. J. Biol.Chem. 205:291-295.

KELLENBERGER, E., A. RYTER, AND J. SE]CHAUD.1958. Electron microscope study of DNA-con-taining plasms. II. Vegetative and maturephage DNA as compared with normal bacterialnucleoids in different physiological states. J.Biophys. Biochem. Cytol. 4:671-679.

LUFT, J. H. 1961. Improvements in epoxy resinembedding methods. J. Biophys. Biochem.Cytol. 9:409-414.

ROBINOW, C. F. 1960. Outline of the visible organi-zation of bacteria, p. 45-108. In J. Brachet andA. E. Mirsky [ed.], The cell, vol. 4. AcademicPress, Inc., New York.

WATSON, M. L. 1958. Staining of tissue sections forelectron microscopy with heavy metals.II. Application of solutions containing leadand barium. J. Biophys. Biochem. Cytol.4 :727-730.

503

on July 15, 2018 by guesthttp://jb.asm

.org/D

ownloaded from