SAugust 1987 Bulletin 872 (technical)
Electron Microscopy of Negatively Stained
Clarified Viral Concentrates
Obtained from Small Tissue Samples
Appendices on Negative Staining Techniques
Stephen R. Christie, Dan E. Purcifull,
Willie E. Crawford, and Nabila A. Ahmed
Agricultural Experiment Station
Institute of Food and Agricultural Sciences
University of Florida, Gainesville
J. M. Davidson, Dean for Research
Electron Microscopy of Negatively Stained
Clarified Viral Concentrates
Obtained from Small Tissue Samples
Appendices on Negative Staining Techniques
Stephen R. Christie, Dan E. Purcifull, Willie E. Crawford, and
Nabila A. Ahmed
Department of Plant Pathology, Plant Virus Laboratory,
University of Florida, Gainesville FL 32611
Table of Contents
Summary ................... .......................... i
Table of Abbreviations ................................ ii
Introduction ......................................... 1
Materials and Methods ................................ 4
Virus Cultures and Their Plant Hosts
Preparation of the Clarified Viral Concentrates ........ 4
Standard CVC Procedure ......................... 4
Variations on the Standard CVC Procedure ......... 6
Electron Microscopy ................................. 7
Results .................. ............................ 7
Electron Microscopy ................................ 7
Virion Exclusion: The Competition for Grid Sites .... 15
Discussion ......................................... 17
Literature Cited .................................... 19
Appendix 1. Preparation of Support Films
for Negative Staining ................................. 21
Appendix 2. Negative Staining Techniques ............. 31
Negative Stains and Wetting Agents ................ 31
Negative Staining of Virions ...................... 36
The Staining Procedure ............................. 37
Negative Staining of Immunosorbent Materials ....... 40
Problems with Negative Staining .................... 41
Appendix 3. Production of Electron Micrographs ........ 43
Selected List of Suppliers ............................ 44
A method has been developed to obtain clarified viral concentrates
(CVC) from extracts of small samples. The amount of starting tissue
may be as little as 1 g, and the procedure consumes minimal time and
material. In a typical CVC procedure, virus-infected tissue was tritu-
rated in a buffer with a mortar and pestle and expressed through
cheesecloth. The subsequent exudate was clarified with organic sol-
vent(s) and centrifuged in a microcentrifuge. Polyethylene glycol
(PEG) was added to the supernatant to precipitate the virions, and
the suspension was again centrifuged. The resultant pellet was re-
suspended in buffer and subjected once more to microcentrifugation,
and the virus-containing supernatant was retained for electron mi-
The CVC method was used successfully to extract, clarify, and
concentrate the virions of cowpea mosaic (comovirus), cucumber
mosaic (cucumovirus), nandina stem-pitting (closterovirus), peanut
stripe (potyvirus), pepper mottle (potyvirus), sonchus yellow net
(rhabdovirus), tobacco mosaic (tobamovirus), tobacco necrosis (tobac-
co necrosis virus group), and tobacco ringspot (nepovirus) viruses as
well as particles associated with a disease of lettuce and escarole,
which closely resemble ilarvirus virions (e.g. tobacco streak virus).
Negatively stained samples from the CVC procedure were
mounted on specimen grids, and the numbers of virions were deter-
mined using an electron microscope. The number of virions found in
the last step of the CVC preparations ranged from 6 to 114 times
greater than those found in their comparable starting exudates.
These increases of unexpected magnitude could have been only part-
ly due to physical concentration of the sap extract (actual concentra-
tion was five to six times the concentration of the starting extract). In
the course of the CVC procedure, many of those sap components that
compete with the virions for available grid adsorption sites were
eliminated, and this played a significant part in elevating the ratios
of virion concentration.
CVC has shown potential to aid in the identification of virus hosts
and to develop virus purification procedures by expediting the selec-
tion of buffers, solvent systems, reducing agents, and other additives
from a spectrum of candidates. CVC can also be used in many situa-
tions when the amount of starting tissue is limited and only small
amounts of virus preparation are needed. Another use for the tech-
nique is to permit the simultaneous production of a number of virus
preparations. Applications for the CVC technique could also include
the preparation of samples for serology, polyacrylamide gel elec-
trophoresis, and immunosorbent electron microscopy.
Appendices pertaining to the electron microscopy of negatively
stained virus preparations are included. These techniques are useful
for evaluation of the CVC preparations, microscopy of virions pre-
pared by other purification schemes, crude virus preparations (com-
monly known as leaf dips), and immunosorbent electron microscopy.
The routine production of quality electron micrographs is also dis-
Table of Abbreviations
BSA bovine serum albumin
CMV cucumber mosaic virus
CpMV cowpea mosaic virus
CVC clarified viral concentrates)
EDTA ethylenediaminetetraacetic acid
ELISA enzyme-linked immunosorbent assay
ISEM immunosorbent electron microscopy
NSPV nandina stem-pitting virus
PEG polyethylene glycol
PMV pepper mottle virus
PSV peanut stripe virus
PTA phosphotungstic acid
PVY potato virus Y
SP standard phosphate (buffer)
SYNV sonchus yellow net virus
TEV tobacco etch virus
TMV tobacco mosaic virus
TNV tobacco necrosis virus
TRSV tobacco ringspot virus
UA uranyl acetate
UF uranyl format
Transmission electron microscopy is frequently used in plant virus
research to detect virions in crude tissue extracts (leaf dips). The
discovery of virus-like particles in these extracts indicates viral in-
fection, and may even allow the microscopist to assign them tenta-
tively to known virus groups on the basis of particle morphology (19).
Although Kitajima (15), and Hitchborn and Hills (13) have been able
to detect isometric virions in leaf dips, the technique has usually been
applied to the detection of elongated virions. It takes considerable
skill and experience to recognize small isometric particles in the
background clutter associated with crude preparations unless the
virions can be found in arrays or if they are unusually numerous.
Immunosorbent electron microscopy (ISEM) (21) consists of var-
ious serological techniques that aid in the detection and identifica-
tion of virus diseases. The most important of the ISEM techniques is
Derrick's method of coating electron microscope grids with virus-
specific antisera, and subsequently exposing these grids to extracts of
known or suspected viruses (6). Derrick's technique is often used in
any of several variations, most of which have been devised to increase
the speed and ease of the technique (7,20), but at least one, the
protein A modification (9,26), is purported to improve the sensitivity
of the method.
Whenever Derrick's technique or a variation of it is used, there is
seldom any problem in recognizing virions, even the isometric ones,
because they are usually discovered in great concentrations. All of
the ISEM techniques require the use of antisera to the viruses being
studied. If these antisera are not available, or if the nature of the
viruses is unknown, it is appropriate to seek other techniques for
particle enhancement. The regimen described here, generically
termed CVC [clarified viral concentratess), comprises related proce-
dures that are nonserological, depending rather on various clarifica-
tion techniques and on the established use of polyethylene glycol
(PEG) to precipitate virions. The virion precipitation characteristics
of PEG were first noted by Hebert (12), and subsequently used for the
purification of papaya mosaic virus by de Bokx (3). PEG has now
become a primary tool for the purification of many viruses of diverse
The CVC method is first described as the "standard" procedure,
followed by descriptions of some variations. The CVC method can
greatly increase the numbers of virions discovered by electron mi-
croscopy compared to the numbers found in the starting crude ex-
tracts. The use of the CVC method can produce sufficient amounts of
virus suspensions to meet the requirements of electron microscopy
Virus source tissue
Grind tissue in buffer and ex-
press through cheesecloth (Step 1)
Combine extract with
solvent mixture (Step 2)
Centrifuge (Step 3)
1st pellet-discard 1st supernatant
Add polyethylene glycol (Step 4)
Centrifuge (Step 5)
2nd supernatant-discard 2nd pellet
Resuspend in buffer (Step 6)
Centrifuge (Step 7)
3rd pellet-discard 3rd supernatant
(usually final supernatant)
Figure 1. Generalized flowchart of the CVC procedure illustrating the basic
principles. The "standard" procedure and the variations are uti-
lizations based on this concept.
without the expenditures of time, equipment, and materials usually
needed for full scale purification procedures. The only extraordinary
equipment required for the CVC method is a rather inexpensive
microcentrifuge. A generic flowchart for CVC is presented in Figure
1, while the flowchart in Figure 2 outlines the standard procedure.
Despite the advances in electron microscopic techniques in recent
years, the techniques concerned with the negative staining of virions
have undergone little improvement since they were described by
Pease in 1964 (24). Suggested improvements can be found in the
literature, but they have seldom been actually incorporated into
current negative staining techniques. For a number of years we have
sought to combine techniques from a variety of sources into useful
integrated regimens. Some of these procedures are described in
The other appendices, while not directly concerned with negative
staining, are supportive of it in some way. Published sources for the
techniques described in the appendices are cited when known, but
unfortunately few of them are. Rather they have been taken from
uncredited data sheets, or have been passed to us by word-of-mouth.
We have adapted some of them from more specific uses to serve as
general techniques. As an example, the procedure for washing grids
was adapted with some modification from the ISEM procedure of
Derrick and Brlansky (7).
Some of the procedures outlined in the appendices represent origi-
nal work by the senior author (including the development of stable
uranyl format solutions), and their incorporation into useful proce-
dures. The techniques were tested by trial and error and many that
were unusable, inconvenient, or unhelpful were rejected. Sometimes
we have set aside procedures that held promise, simply because we
have been unable to incorporate them satisfactorily into our routines.
Especially disappointing to reject were methods to improve sub-
strates, either by making them stronger, as substituting Butvar for
1 g virus-infected tissue
Macerate in 2 ml 0.1 M potassium
phosphate buffer, pH 7.5 and
express through cheesecloth (Step 1)
Stir vol:vol with solvent
mixture of n-butanol and chloroform in
equal amounts (Step 2)
SCentrifuge at 12,200 x g for 5 min (Step 3)
(discard) Add polyethylene glycol, MW
6,000-8,000, and sodium chloride to final
concentrations of 6% and 0.125 M,
respectively (Step 4)
Centrifuge at 12,200 x g for 5 min (Step 5)
(discard) Resuspend in 5-6 drops 0.1 M
potassium phosphate buffer, pH 7.5 (Step 6)
Centrifuge at 12,000 x g for 5 min (Step 7)
Pellet Final supernatant
(discard) CVC fraction
Figure 2. Flowchart illustrating the "standard" CVC procedure. This has
been the CVC procedure most often used, and serves as a basic
model for the assorted variations.
Formvar (11), or to make Formvar substrates thinner, yet strong and
free of defects (5); but we have been unable to use them successfully
with our regimen.
The appendices are by no means intended to be the last word in
negative staining and related procedures, but rather provide work-
able starting techniques from which each user may develop routines
that are suitable for particular job requirements.
Materials and Methods
Virus Cultures and Their Plant Hosts
The viruses used in this study, the virus groups to which they are
assigned (19), and their maintenance hosts were: cowpea mosaic
virus (CpMV), comovirus, the Sb isolate obtained from G. Bruening,
Vigna unguiculata (L.) Walp. 'Knuckle Purple Hull'; cucumber
mosaic virus (CMV), cucumovirus, the winged bean isolate (17),
Vigna unguiculata (L.) Walp., 'Knuckle Purple Hull'; virus-like dis-
ease of lettuce (LEV), tentative ilarvirus, original isolate from
B. Falk, Lactuca sativa L., Cichorium endivia L., and Nicotiana x
edwardsonii Christie and D.W. Hall (4); nandina stem-pitting virus
(NSPV), proposed closterovirus, original isolate (1), Nandina domes-
tica Thunb., 'Nana Purpurea'; peanut stripe virus (PSV), potyvirus,
original isolate from J. Demski, Arachis hypogaea L.; pepper mottle
virus (PMV), potyvirus, the type isolate, Nicotiana tabacum L., 'Sam-
sun NN'; sonchus yellow net virus (SYNV), rhabdovirus, the type
isolate, Nicotiana x edwardsonii Christie and D.W. Hall, (4); tobacco
mosaic virus (TMV), tobamovirus, the common type isolate, Nico-
tiana tabacum L., 'Turkish'; tobacco necrosis virus (TNV), tobacco
necrosis virus group, isolate from D. Roberts, Nicotiana tabacum L.,
'Samsun NN'; and tobacco ringspot virus (TRSV), nepovirus, isolate
from D. Roberts, Nicotiana x edwardsonii Christie and D.W. Hall (4)
and Nicotiana tabacum L., 'Samsun NN'.
Preparation of the Clarified Viral Concentrates
The basic principles of the CVC procedures are illustrated by the
flowchart in Figure 1.
The first procedure listed is the one used most often and is referred
to as the "standard CVC procedure." Modifications, described as
"variations," follow the standard procedure.
Standard CVC Procedure (See Figure 2 for Flowchart). The
initial tissue extraction was done with a solvent system developed by
Steere (27). (Note: throughout the procedure, all virus fractions were
held in an ice bath except during manipulations and centrifugations.)
Step 1. One g of tissue was macerated in 2 ml of 0.1 M potassium
phosphate buffer, pH 7.5 (SP buffer). The resultant pulp was ex-
pressed through cheesecloth and the exudate was collected.
Step. 2. The exudate was clarified by stirring it (vol:vol) with a
solvent mixture made from equal volumes of chloroform and n-
Step 3. The mixture was centrifuged at 12,200 x g for 5 min. The
supernatant was retained.
Step 4. Polyethylene glycol (PEG) (Sigma, MW = 8000) and
sodium chloride were stirred into the supernatant to give final con-
centrations of 6% PEG and 0.125 M NaC1, respectively. The suspen-
sion was allowed to stand for 15 min. (Note: as CVC is essentially a
rapid, qualitative procedure, we have found it imperative to use drop
counting throughout, eliminating many tedious measurements, to
preserve the necessary simplicity. Thus, in the case of PEG/sodium
chloride, we prepared in advance a stock solution of 30% PEG with
0.6 M NaC1. This was then added to the supernatant at the rate of 1
drop PEG/NaC1 to 4 drops supernatant, giving a final concentration
of 6% PEG and 0.125 M NaC1.)
Step 5. The suspension was centrifuged at 12,200 x g for 5 min,
and the pellet was retained.
Step 6. The pellet was resuspended in SP buffer using frequent
agitation for a minimum of 30 min. The resuspension of the pellet
must be done with considerable care to obtain maximum virus yields.
(a) Following Step 5, the centrifuge tubes were first drained, and
then the sides were wiped dry of supernatant with the aid of a cotton
swab. Even small amounts of residual PEG may cause reprecipita-
tion of some of the virus, which would then be subsequently lost at
Step 7 (PEG traces may also interfere with the electron microscopy of
the virus under study).
(b) It is important to allow sufficient time for resuspension and it is
also important to provide agitation to aid in the disintegration of the
pellet. A Vortex Mixer can normally be used, but resuspension may
also be accomplished by teasing the pellet with the pointed end of a
bamboo splint. Normally we have used a combination of these two
procedures, and sometimes have resorted to using a Pasteur pipette
as a slush pump. If the latter device is used, particular care must be
exercised or the pellet(s) may become lodged in the pipette. Whatever
procedure is used, allow at least 30 min with frequent agitation to
ensure thorough resuspension of the virions.
(c) The amount of resuspension liquid gives a rough estimate of
final concentration. As an example, if buffer is added to the pellets in
Step 6 at a rate of 6 drops for each gram of starting tissue, then the
concentration would be five to six times the original sample. For
these purposes, 1 ml is estimated as 32 or 33 drops, and the starting
sample weight is taken as its liquid equivalent (1 g = 1 ml).
Step 7. The final virus-containing solution was obtained at this
step by centrifuging the slurry (12,200 x g for 5 min) and carefully
removing and retaining the supernatant.
The standard procedure was used for CpMV, PMV, TMV, and
Variations on the Standard CVC Procedure. The variations
differ from the standard procedure only at the specific steps indicated.
Consult the standard procedure description above for processing de-
Variation A. Extraction of tissue at Step 2 was made with SP
buffer containing 6.5% n-butanol at the rate of 1 ml extracting
solution to 1 g tissue. No other solvents were used. This variation was
used for CpMV and TNV.
Variation B. At Step 1, the tissue was macerated in 2 ml of 0.5 M
citrate buffer, pH 6.5, containing 5 mM ethylenediaminetetraacetic
acid (EDTA), and 0.5% thioglycolic acid. No solvents were used. At
Step 6, the pellet was resuspended in the same solution as was used in
Step 1. Step 2 was omitted. This variation was used for CMV.
Variation C. This was the same as Variation B except that, at Step
2, 1 volume of chloroform was stirred into each volume of exudate
with the aid of a Vortex Mixer. This variation was used for CMV.
VariationD. At Step 1, 0.1% sodium sulfite and 0.02 M EDTA were
added to the SP buffer. At Step 6, the pellet was resuspended in 0.05 M
Tris buffer, pH 8.2, containing 0.02 M EDTA. This variation was used
for lettuce virus.
Variation E. At Step 1, the tissue was triturated in liquid nitrogen
with a mortar and pestle. It was then extracted with SP buffer that
contained 1.5% sodium sulfite, at the rate of 2 ml/g tissue. At Step 6,
the pellet was resuspended in this same buffer solution. No solvents
were used, and Step 2 was omitted. This variation was used for
Variation F. All steps were the same as those in Variation E
except that Step 7 was followed by further treatment. The super-
natant from Step 7 was mixed vol:vol with 2% aqueous uranyl acetate
and then centrifuged for 5 min at 12,200 x g. The supernatant was
retained. This variation was used for NSPV. Note: electron micro-
scope grids that were prepared from this fraction were rinsed and
stained in the usual fashion, which is described in the following
section. The uranyl acetate added at this step was only added to
clarify the virus solution and not for its staining qualities. In fact,
most of the uranyl acetate is lost to the pellet during the subsequent
Variation G. At Step 1, 0.1% sodium sulfite and 0.01 M EDTA were
added to SP buffer. This solution was also used to resuspend the PEG
pellet at Step 6. This variation was used for PSV.
Variation H. The extraction solution used in Step 1 was 0.1 M
Tris-HC1, pH 8.4, that contained 0.1 M magnesium acetate, 0.04 M
sodium sulfite, and 0.001 M magnesium chloride (14). This solution
was also used at Step 6 to resuspend the pellet. Step 2 was omitted,
and no solvents were used. A subroutine was inserted following Step
3. A 10% (w/v) aqueous solution of activated charcoal was stirred into
the supernatant from Step 3. The charcoal mixture was allowed to
stand for 5 min and then centrifuged at 12,200 x g for 5 min. This
supernatant was used to continue the procedure at Step 4. This
variation was used for SYNV.
A droplet of test solution was applied to a carbon top-coated, Form-
var-clad electron microscope grid, which was then washed and
stained in the manner described in Appendix 2 ("Negative Staining
of Virions"). Bacitracin (10) was usually not needed in the washing
solutions, as there was sufficient protein present in the CVC prepara-
tions to provide good wetting. However, the negative stain, usually
uranyl acetate, always contained bacitracin. Grid preparation was as
described in Appendix 2.
The stained grids were examined in a Hitachi H-600 electron
microscope. The numbers of particles were determined at magnifica-
tions that presented a manageable number of particles in the viewing
field. Magnification was kept constant for any given experiment, but
varied from one experiment to another.
Electron micrographs were made on 35-mm film as described in
The data pertaining to the CVC trials are summarized in Table 1.
Utilization of the CVC procedure resulted in increased virion popula-
tions as determined by electron microscopy (Figure 3). Varying
amounts of clarification of the CVC extracts were realized through-
out the procedure, but only at Steps 4 through 6 was there actual
concentration of virions, achieved by resuspending the virus-
containing PEG pellet from Step 6 in a smaller volume than that of
the original extract. The degree of virus increase is given as a ratio
comparing the number of particles counted on specimen grids pre-
pared from the CVC preparations to the numbers seen on grids
Table 1. Increases in particle numbers resulting from procedures as determined by electron microscopy.
Total Mean Increase
Virus Experiment magnifi- No. of Total No. of in
name, No. and CVC cation fields No. of particles particle
group variation" (x) examined particles per field counts
Cowpea mosaic, Co-1
comovirus Control 400,000 50 19 0.4 0.6d
Standard 400,000 50 765 15 4.0 40:1
A 400,000 50 519 10 2.5 27:1
Control 400,000 100 49 0.5 0.8
Standard 400,000 100 2,691 27 5.2 55:1
Control 400,000 50 25 0.5 0.7
Standard 400,000 50 822 16 4.4 33:1
mosaic, Control 500,000 100 52 0.5 0.7
cucumovirus B 500,000 50 164 3.2 1.9 6:1
C 500,000 50 291 5.8 3.5 11:1
Lettuce virus, Le-1'
possible Control 500,000 50 30 0.6 0.9
ilarvirus D 500,000 50 372 7.4 3.1 12:1
Control 500,000 50 21 0.4 0.7
D 500,000 50 1,223 24 6.9 58:1
Nandina stem- Na-1
pitting, Control 100,000 50 8 0.2 0.5
closterovirus E 100,000 50 119 2.4 2.2 15:1
F 100,000 50 420 8.4 2.8 52:1
Peanut stripe, Ps-1
potyvirus Control 50,000 50 46 0.9 1.1
A* *-' A An An non 5 5Ap 879' 18 +5.8 19i.
Table 1. (cont.)
Total Mean Increase
Virus Experiment magnifi- No. of Total No. of in
name, No. and CVC cation fields No. of particles particle
group variation' (x) examined particles per field' counts'
Pepper mottle, Pm-1
yellow net, Control 35,000 100 11 0.1 0.2
rhabdovirus H 35,000 100 201 2.0 1.3 18:1
Control 35,000 200 9 <0.1 0.2
H 35,000 200 353 1.8 1.5 39:1
mosaic, Control 150,000 100 16 0.2 0.5
tobamovirus Standard 150,000 100 1,140 11 3.5 71:1
Control 100,000 50 88 1.8 1.2
Standard 100,000 50 2,378 48 13 27:1
necrosis, Control 1,000,000 25 18 0.7 0.7
tobacco A 1,000,000 10 584 58 8.7 81:1
necrosis group Tn-2
Control 1,000,000 25 9 0.4 0.6
A 1,000,000 25 605 24 4.6 67:1
Table 1. (cont.)
Total Mean Increase
Virus Experiment magnifi- No. of Total No. of in
name, No. and CVC cation fields No. of particles particle
group variation" (x) examined particles per field counts'
ringspot, Control 500,000 100 56 0.6 0.7
nepovirus Standard 500,000 100 3,262 33 6.0 58:1
Control 500,000 100 13 0.1 0.3
Standard 500,000 100 1,484 15 3.7 114:1
Summary: Controls 1,200 470 0.4
Experiments 1,285 18,272 14 35:1
"The controls consisted of aliquots of the exudates used for subsequent CVC procedures. The variations of the CVC procedures, indicated
alphabetically, are as they are described in the text.
bNumbers smaller than 10 are rounded to the nearest 0.1. Numbers greater than 10 are rounded to the nearest whole number.
"Ratios of numbers of particles detected in CVC preparations to numbers of particles detected in controls. These numbers are rounded down to the
nearest whole number.
dStandard sample deviation (n 1).
"Le-1 was purified from lettuce. Le-2 was purified from escarole.
fVirions were considerably more numerous on CVC than on control grids, but counts were not made due to extensive linear aggregation in the
CVC preparation (see Figure ID).
Figure 3. Electron micrographs of virions in exudates (A and C) of virus-
infected leaf tissue, and virions in CVC preparations (B and D)
made from these same exudates. A and B: tobacco necrosis virus.
C and D: pepper mottle virus. All of the micrographs are at the
same magnification. The preparations were stained with 2%
uranyl acetate using the washing and staining technique de-
scribed in Appendix 2. Bar = 500 nm.
prepared from samples of the original extracts of these preparations.
The samples of the original extracts were held on ice throughout each
CVC procedure, and all of the grids for each comparison were then
made up at the same time. The CVC ratios (final supernatant: start-
ing exudate) obtained in the trials presented here ranged from 114:1
to 6:1, with an average increase of 39:1. All told, over 2400 viewing
fields were examined and 18,700 particles were counted.
Cowpea Mosaic. In Experiment 1, the two procedural variations
were processed simultaneously using leaves from a pooled source so
that the 50% higher counts of particles obtained from the standard
procedure as compared to Variation A (Table 1) reflect a real differ-
ence in particle yields. It was also noted that the Step 7 supernatant
for the standard procedure was clear and colorless, while that from
Variation A was also clear, but yellow in color.
Cucumber Mosaic. A smaller increase in virion numbers was
realized by the CVC procedure for this virus than any other studied
here. Still, Variations B and C resulted in increases of 6- and 11-fold,
respectively, as compared to the number of virions found in the
original extracts. No virions could be found in CVC preparations
when the standard procedure was attempted.
Lettuce Virus. The virus-like particles associated with a disease
of Florida lettuce (Figure 4A) closely resemble those depicted in
micrographs of negatively stained ilarvirus particles that may be
found in numerous publications (e.g. 8). These ilarvirus-like particles
were unstable and were found only when sodium sulfite and EDTA
were added to the buffer solutions.
Nandina Stem-pitting. Greater numbers of particles were
obtained with the insertion of a simple step into the CVC procedure:
the addition of uranyl acetate to the supernatant at Step 7 (Variation
F). Although the addition of uranyl acetate dramatically lowers the
pH value of the sample, it is unlikely that this factor alone accounted
for the additional clarification that occurred, as buffers of similarly
low pH values failed to produce the same effect when they were
substituted for uranyl acetate. Unfortunately, this technique cannot
be universally applied, as viruses typically precipitate at low pH
values. Indeed, NSPV was the only virus in this study that was
amenable to this treatment, although several others were tested. In
the other cases, either no virions were found on the specimen grids or
they were found to be aggregated. The technique is simple, however,
and may be worth trying when all other attempted methods fail to
yield the degree of clarification desired. No negative staining effect
was obtained after the use ofuranyl acetate for clarification. Because
all of the yellow color of the uranyl acetate sedimented with the pellet
following centrifugation, it seems likely that the uranyl ions were
A a B
Figure 4. Electron micrographs of CVC preparations of viruses represent-
ing special problems (A-C) in obtaining successful results. A
and B represent viruses that require stabilization by additives
such as sodium sulfite. No particles were found in CVC prepara-
tions that did not contain the stabilizing additives. A: sonchus
yellow net, rhabdovirus, from stabilized CVC preparation. View
contains several bacilliform particles (arrows). Bar = 1000 nm.
Inset shows intact virion, with the helical core typical of rhabdovi-
ruses, from a CVC preparation. Bar = 100 nm. B: particles associ-
ated with LV that in size, shape, and texture indicate it is an
ilarvirus. CVC preparation from a stabilized extract. Bar = 100
nm. C: peanut stripe, potyvirus, processed from peanut, a host that
contains slime elements. Compare C with D, also peanut stripe,
but processed from white lupine, a host that is more amenable to
purification procedures than peanut. C and D are at the same
magnification: bar = 1000 nm. All micrographs were stained with
2% uranyl acetate using the washing and staining procedure de-
scribed in Appendix 2.
bound to host material. Because no staining was obtained, it was
necessary to wash and stain the specimen grids with uranyl acetate
in the usual manner for electron microscopy.
Peanut Stripe. Even though the PSV virions could be found more
readily in other hosts (Figure 4C), we felt that it was important to be
able to concentrate this virus directly from peanut, an important
natural host, and because the slime elements present in peanut sap
made that host more of a challenge to obtain successful results via the
CVC procedure. The results were satisfactory in that both cleaner
grid backgrounds and a respectable virion concentration were
obtained (Figure 4D). However, traces of the slime elements did
persist in the final supernatant.
Pepper Mottle. The standard procedure worked well with pepper
mottle for concentration and clarification. However, extensive linear
aggregation made counts of particles impractical. The extent of the
effectiveness of the procedure may be judged by comparing Figure 3,
C and D.
Sonchus Yellow Net (Figure 4B). The total number of particles
from CVC preparations of this virus were the lowest for any of the
viruses covered here. However, SYNV particles were so rare in the
starting extracts that the ratio of enhancement, as determined by
numbers of particles, was 18:1 and 39:1 in two separate experiments.
Virions were never found in preparations that lacked stabilizing
additives such as sodium sulfite.
Tobacco Necrosis. Relatively few counts were made from prepa-
rations of this virus. Even at the high magnification used, virions
found on grids prepared from the CVC-processed samples were so
numerous that it was likely they were in competition for the avail-
able grid sites (see Figure 3, A and B). Thus, the magnitude of
enhancement was probably considerably greater than the counts
indicated. The numbers of particles/field, considering the magnifica-
tion used, was the highest of any virus in this study.
Tobacco Ringspot. This virus had the highest ratios of particle
increases in the CVC preparations of any listed here, and the total
numbers of virions were only exceeded by TNV. The standard proce-
dure worked exceptionally well for this virus, not surprising con-
sidering that the solvent system used in the procedure was originally
devised for the purification of TRSV (27). The numbers of virions
observed on the specimen grids were so great that once again, as with
TNV, particles may have been forced to compete with each other for
grid sites, and the true ratios of enhancement obtained via the CVC
method would have been greater than the counts reflect.
Virion Exclusion: The Competition for Grid Sites
Virion numbers discovered by electron microscopy were often con-
siderably higher for the CVC final preparations when compared to
the starting exudates than could be expected on the basis of an
increase in virus concentration alone. Thus, it seemed probable that
these elevated counts owed, at least in part, to the removal of some of
the sap constituents which otherwise compete with virions for the
available grid sites (18,22). Virion exclusion was therefore investi-
gated to determine the magnitude of this phenomenon and to assess
its effect on the counts of particles made from the CVC preparations.
Purified CpMV was obtained from E. Hiebert. Following prelimi-
nary clarification steps, the virus had been subjected to approach to
equilibrium centrifugation on cesium chloride gradients, and the
virus-containing zones were collected. The virions were then precipi-
tated by PEG and resuspended in SP buffer. This suspension was
water-clear and free of any detectable impurities. When examined
with the electron microscope, virions were seen to be intact and
monodisperse and the background was devoid of debris.
The preparation was then diluted with SP buffer to make a stock
solution containing 2.5 mg/ml as determined by its optical density at
260 nm. Samples of this stock solution were mixed vol:vol with (I) SP
buffer (control), (II) the supernatant from a centrifuged extract of
healthy cowpea (1 g of leaf tissue was triturated in 2 ml SP buffer,
expressed through cheesecloth, and the expressate was centrifuged
at 12,200 x g for 5 min), (III) the supernatant from a healthy extract
of tobacco that had been prepared in the same manner as II, and (IV)
which was the same as II except that the expressate was stirred
vol:vol with n-butanol just prior to centrifugation and the aqueous
phase was collected and used in the mixture. Specimen grid mounts
were made with these products, and were washed and stained with
uranyl acetate as outlined in Appendix 2. Particles were counted and
the results of these counts are presented in Table 2, Experiment 1.
Addition of either cowpea (II) or tobacco (III) sap to the stock
solution of virus reduced the observed number of virions to less than
10% of control (I). However, when cowpea sap was clarified with
n-butanol prior to adding it to the virus solution (IV), the number of
virions tabulated was 80% as great as control (I) (see Figure 5).
In another test, cowpea mosaic virus was concentrated from in-
fected cowpea tissue using the standard CVC procedure. Samples
were then taken and used to prepare the following mixtures: (a) a
sample from the Step 7 (final) supernatant was mixed vol:vol with SP
buffer (control); (b) a sample from Step 1 (expressate) was centrifuged
as in Step 3, but the solvent treatment at Step 2 was omitted, and the
supernatant was then mixed vol:vol with SP buffer; and (c) a sample
from the Step 7 (final) supernatant was mixed vol:vol with centri-
fuged healthy cowpea sap prepared as in (b) with the solvent treat-
ment omitted. Specimen grids were prepared and the particles were
counted as in the above experiment, and the results are presented in
Table 2, Experiment 2.
The results of this test reinforced the findings of the preceding test:
The first supernatant (b), which presumably retains most of the
site-competing elements found in sap, had only 4% of the number of
virions found in the solvent-treated control mixed with buffer (a).
When the solvent-treated fraction was mixed with untreated sap (c),
the virion numbers were reduced to 16% of the same control (a).
Table 2. Effects of added healthy plant extracts on the numbers of cow-
pea mosaic virus particles detected by electron microscopy"
No. Total Average
of No. No. %
fields of parti- of
exam- parti- cles/ con-
Treatment' ined cles field trol
Experiment 1 Purified CpMV' plus
buffer (control) 200 3,987 20 100
Purified CpMV' plus
cowpea extract' 200 343 1.7 9
Purified CpMV' plus
tobacco extract 200 168 0.8 4
Purified CpMV' plus
extract' 200 3,194 16 80
Experiment 2 CVC-derived CpMV" plus
buffer' (control-1) 100 997 10 100
CpMV infected cowpea
buffer (control-2) 100 39 0.4 4
CVC-derived CpMV9 plus
cowpea extract' 100 161 1.6 16
"Magnification in all cases was 400,000 x.
"All mixtures were made vol:vol.
'Purified CpMV at a concentration of 2.5 mg/ml.
d0.1 M potassium phosphate, pH 7.5.
'The plant extracts were the supernatants from the following treatment: healthy
leaf tissues were triturated in 0.1 M potassium phosphate, pH 7.5, at the rate of
1 g/2 ml. The pulp was expressed through cheesecloth and centrifuged for 5 min at
12,200 x g.
SPlant extracts prepared as in footnote e above and clarified by the addition of 6.5%
n-butanol to the phosphate buffer.
'Supernatants from Step 7 of CVC variation B (see "Materials and Methods").
Figure 5. Electron micrographs of highly purified cowpea mosaic virus at a
starting concentration of 2.5 mg/ml. A: virus preparation mixed
vol/vol with buffer. B: virus preparation mixed vol/vol with a low
speed supernatant of healthy cowpea expressate. C: virus prepara-
tion mixed vol/vol with a low speed supernatant of healthy cowpea
expressate which had been clarified with n-butanol. These figures,
although typical of the grids from which they were selected, were
chosen to give graphic representation to the exact particle percent-
ages which are reported in Table 2. All of the micrographs are at
the same magnification. The preparations were stained with 2%
uranyl acetate using the washing and staining technique de-
scribed in Appendix 2. Bar = 500 nm.
The CVC procedures facilitate the detection of virions by electron
microscopy. Particles were considerably more numerous in the CVC
preparations than in the unprocessed controls, owing both to the
concentration of virions and to the removal of sap components, some
of which compete with the virions for the available adsorption sites
on specimen grids. Actual concentration of virus took place when
virions were precipitated by PEG, concentrated into a pellet by cen-
trifugation, and finally resuspended in a volume of buffer smaller
than the starting volume. Clarification of virus extracts with sol-
vents was proven to eliminate many of the virus-competing elements
that are found in crude preparations. CVC processing can take place
in less than 2 hours, and the small table model centrifuge used in the
procedure allows for the handling of multiple samples. The polypro-
pylene and polyethylene centrifuge tubes used in the procedure with-
stood the solvents used in the concentrations and times required by
the CVC procedures.
Although Table 1 lists numbers of particles and ratios of increase
realized for certain viruses, it can be seen from the same table that
these are not constants. As was the case for CpMV (ratios of 40:1,
55:1, and 33:1), SYNV (ratios of 18:1 and 39:1), TMV (ratios of 71:1
and 27:1), TNV (ratios of 81:1 and 67:1), and TRSV (ratios of 58:1 and
114:1), numbers can vary from one experiment to another even when
the same hosts and CVC variation are used.
Thus, the numbers listed in Table 1 are suggestive rather than
predictive of virion enhancement, especially in the case of the aver-
age ratio given for the sum of the experiments done here (ratio =
35:1). Rather, the conclusion to be drawn from Table 1 is that the
CVC procedures can often result in a considerable increase in actual
and apparent virus concentration. It was also our observation that
the clarified samples from the CVC preparations allowed micro-
graphs to be produced that were superior in the sharpness of delinea-
tion and reduction of background clutter to those produced from
Viruses have often been purified on a small scale by density gra-
dient centrifugation, and TMV has been purified in small amounts
using PEG (2). Our effort has been directed to producing small quan-
tities of semipurified virus suitable for use in detecting virions by
electron microscopy. We have endeavored to keep the procedure fast
and simple so that multiple samples from disparate virus groups may
be processed readily.
It should be mentioned that the CVC procedures may prove to be
valuable in areas of virus research other than electron microscopy.
Koenig observed that sap components interfered with virion adsorp-
tion when crude saps were used with the indirect enzyme-linked
immunosorbent assay (ELISA) (16). CVC methods might therefore
be useful to expedite indirect ELISA, as the CVC systems that use
clarifying solvents tend to reduce the sap constituents that compete
for adsorption sites with virions.
CVC procedures may be valuable as pilot programs to develop
full-scale virus purification procedures and should abbreviate the
time and diminish the quantity of materials normally required for
such pilot studies. Because CVC techniques expedite the processing
of multiple samples, the CVC protocols can be varied to test various
buffers, solvent systems, etc. The results may be compared directly
for yields and degrees of clarification. In fact, the CVC procedure has
been used successfully to devise a purification regimen for pepper
mild mosaic virus (E. Debrot, personal communication). After upscal-
ing a successful CVC technique to produce larger quantities of
purified virus, the procedure was refined by employing CVC to im-
prove further the yields and purity of the macro procedure, with a
considerable saving of time and materials as opposed to using a series
of macropurification runs for this purpose.
The use of PEG to enhance reactions in immunodiffusion tests
using small samples of leaf tissue infected with tobacco etch virus
(25) suggests that the CVC procedures may be useful for the serolog-
ical detection of viruses. Crude extracts have been used to detect
viruses by electrophoresis (23), which suggests another possible ap-
plication for CVC techniques.
1. Ahmed, N. A., Christie, S. R., and Zettler, F. W. 1983. Identification and
partial characterization of a closterovirus infecting Nandina domestic.
2. Bateman, J. G., and Chant, S. R. 1979. A modification of the polyethylene
glycol technique for the purification of small quantities of tobacco mosaic
virus. Microbios 25:33-43.
3. Bokx, J. A. de. 1965. Hosts and electron microscopy of two papaya
viruses. Plant Dis. Rep. 49:742-746.
4. Christie, S. R., and Hall, D. W. 1979. A new hybrid species of Nicotiana
(Solanaceae). Baileya 20:133-136.
5. Davison, E., and Colquhoun, W. 1985. Ultrathin Formvar support films
for transmission electron microscopy. J. Electron Microsc. Technique
6. Derrick, K. S. 1973. Quantitative assay for plant viruses using serologi-
cally specific electron microscopy. Virology 56:652-653.
7. Derrick, K. S., and Brlansky, R. H. 1976. Assay for viruses and mycoplas-
mas using serologically specific electron microscopy. Phytopathology
8. Fulton, R. W. 1981. Ilarviruses. In: Handbook of Plant Virus Infections
and Comparative Diagnosis (E. Kurstak, ed.), pp. 377-421. Elsevier/
North-Holland Biomedical Press, Amsterdam.
9. Gough, K. H., and Shukla, D. D. 1980. Further studies on the use of
protein A in immune electron microscopy for detecting virus particles. J.
Gen. Virol. 51:415-419.
10. Gregory, D. W., and Pirie, B. G. S. 1973. Wetting agents for biological
electron microscopy. I. General consideration and negative staining. J.
11. Handley, D. A., and Olsen, B. R. 1979. Butvar B-98 as a thin support film.
12. Hebert, T. T. 1963. Precipitation of plant viruses by polyethylene glycol.
13. Hitchborn, J. H., and Hills, G. J. 1965. The use of negative staining in the
electron microscope examination of plant viruses in crude extracts.
14. Jackson, A. 0., and Christie, S. R. 1977. Purification and some physico-
chemical properties of sonchus yellow net virus. Virology 77:344-355.
15. Kitajima, E. W. 1965. A rapid method to detect particles of some spheri-
cal plant viruses in fresh preparations. J. Electron Microsc. 14:119-121.
16. Koenig, R. 1981. Indirect ELISA methods for the broad specificity detec-
tion of plant viruses. J. Gen. Virol. 55:53-62.
17. Kuwite, C. A., and Purcifull, D. E. 1982. Some properties of a cucumber
mosaic virus strain isolated from winged bean in Florida. Plant Dis.
18. Lesemann, D. E., Bozarth, R. F., and Koenig, R. 1980. The trapping of
tymovirus particles on electron microscope grids by adsorption and sero-
logical binding. J. Gen. Virol. 48:257-264.
19. Matthews, R. E. F. 1982. Classification and nomenclature of viruses.
20. Milne, R. G. 1980. Some observations and experiments on immunosor-
bent electron microscopy of plant viruses. Acta Horticulturae 110:129-
21. Milne, R. G., and Luisoni, E. 1977. Rapid immune electron microscopy of
virus preparations. In: Methods in Virology, Vol. VI (K. Maramorosch
and H. Koprowski, eds.), pp. 265-281. Academic Press, New York.
22. Nicolaieff, A., and van Regenmortel, M. H. V. 1980. Specificity of trap-
ping of plant viruses on antibody-coated electron microscope grids. Ann.
Virol. (Inst. Pasteur) 131 E:95-110.
23. Paul, H. L. 1974. SDS polyacrylamide gel electrophoresis of virion pro-
teins as a tool for detecting the presence of viruses in plants. Phyto-
pathol. Z. 80:330-339.
24. Pease, D. C. 1964. Histological Techniques for Electron Microscopy, 2nd
edition. Academic Press, New York.
25. Purcifull, D. E. 1967. Physical and chemical nature of flexuous plant
viruses. Fla. Agric. Expt. Sta. Annu. Rept., pp. 161-162.
26. Shukla, D. D., and Gough, K.H. 1979. The use of protein A from Staphy-
lococcus aureus in immune electron microscopy for detecting plant virus
particles. J. Gen. Virol. 45:533-536.
27. Steere, R. L. 1956. Purification and properties of tobacco ringspot virus.
Preparation of Support Films for Negative Staining
There are a variety of methods for preparing support films for
attachment to electron microscope grids. Such films may be prepared
from Parlodion or Formvar using techniques that are very simple-
casting films on water by dropping a plastic solution onto the water
surface-or complex, such as the use of very elaborate film casting
chambers that precisely control draining and drying rates. The
method we use is fairly simple and straightforward, and yet it will
usually produce satisfactory support films. The films are generally
more uniform than water-cast films and the thickness of the film can
1. Ethylene dichloride (1,2-dichloroethane), reagent grade.
2. Formvar powder.
3. Grids of suitable mesh. Fine mesh grids are needed to give
good support to the plastic film, which is delicate. Rather than the
usual square mesh, we use 75 x 300 or 100 x 400 mesh grids that
have rectangular openings. When oriented properly in the electron
microscope, this type of grid will allow long scanning treks uninter-
rupted by grid bars, yet providing excellent specimen support.
4. Glass microscope slides. These should not be precleaned, as it
is difficult to secure the release of Formvar films from slides that have
been cleaned with a detergent. Neither should they be fire-polished
nor have beveled edges, as sharp edges are required to cut the plastic
film free of the slide. By the same token, these unfinished slides with
their sharp edges are likely to cause cuts if they are improperly
handled. Exercise caution when polishing them. Some lots of slides
may work better than others for preparing films, even though they
are from the same supplier and bear the same description and catalog
numbers. If slides from a particular lot work especially well, then
reserve an ample supply from that lot to be used solely for filmcast-
ing. The slides that work less well can be used for preparing leaf dips
and for other tasks not requiring special characteristics. Slides (most,
but not all lots) that have usually worked well for us for casting films
are Thomas Red Label Micro Slides, Economy Grade, Catalog No.
6685-H21, Arthur H. Thomas Company.
5. Bamboo splints (skewers).
6. Filter paper disks of assorted sizes and user-preferred grades.
It is easy to pick up grids from a creped type of paper such as Fisher
P8, Catalog No. 09-790-12C (for 9-cm disks).
7. Dishes that are large enough and deep enough to float and
manipulate the supporting films. We use a Stender dish, 100 x 50
mm, with a fitted cover (Arthur H. Thomas Catalog No. 3836-Q50).
Dissecting dishes of similar size will serve the purpose, but require
some sort of cover such as a watch glass or foil to protect the film and
water surface between manipulations.
8. Aerosol type duster (convenient but not essential).
9. Paper of index card weight.
10. Silica gel, large mesh, color indicator.
11. Glacial acetic acid.
12. High grade absolute ethanol.
13. Deionized and/or glass-distilled water.
14. Glass Petri dishes, for drying films and regenerating silica gel
(100 x 20 mm dishes are suitable).
15. Tweezers with very fine points. We favor the curved type
(Dumont No. 7), but straight types (such as Dumont No. 5) may be
used. Dumont stainless steel tweezers sharpened to especially fine
points are available from E. F. Fullam (No. 5, Catalog No. 11040; No.
7, Catalog No. 13140).
16. Miscellaneous laboratory items including various sizes of
Erlenmeyer flasks, small beakers, lintless wipers (tissues or cloths),
and aluminum foil.
1. A small drying or all-purpose oven that can be regulated to
within a few degrees of 55C.
2. A vacuum evaporator rigged with carbon electrodes.
3. A small adjustable fluorescent lamp, e.g. Bausch and Lomb No.
31336601 (available from laboratory supply companies).
4. A magnetic stirrer and some small stirring bars.
1. Preparation of Formvar Solution. Prepare a 0.4% solution
of Formvar in ethylene dichloride (dichloroethane), cover tightly
with aluminum foil, and stir slowly overnight on a magnetic stirrer.
This concentration is a starting point, and it is usually necessary to
further dilute the solution with solvent until films of the desired
thickness can be cast. Films that are gold, red, green, or other in-
terference colors are cast from solutions that are too concentrated and
require further dilution with ethylene dichloride. Uneven films with
swathes of interference colors suggest that the stirring has been
Solutions that have produced satisfactory films may produce un-
satisfactory ones on long standing without stirring. Most textbooks
on electron microscopy quote or paraphrase the injunction by Pease
(24) that only freshly made solutions of Formvar be used to make
support films as the plastic solutions deteriorate on standing. Con-
trarily, we suggest that freshly made solutions should never be used
without prolonged stirring (preferably overnight), and that a thor-
ough restirring is usually all that is needed to regenerate a previous-
ly useful solution.
2. Cleaning the Grids. The grids should be clean and free of
oxides so that the supporting film will adhere to them. Wash them
briefly in glacial acetic acid, rinse with distilled or deionized water,
two rinses in absolute ethanol, and finally dry them on filter paper.
The acetic acid wash should be brief, as the metals may be attacked if
exposure time is too long. This is especially true of rhodium-plated
3. Preparing the Formvar Film (See Figure 6, Steps a-d). A
suitable microscope slide is carefully polished for 2/ of its length with
a clean, lint-free wiper (the remainder of its length is reserved for
handling). (Note: although a film will be formed on both sides of the
slide, it is best to expect the production of only one film per slide.
Thus, attention can be concentrated on one side of the slide, desig-
nated the "film" side.) Place the slide, film side up, in a preheated
oven set at 55C. Prepare several slides in this manner and prewarm
them so they will be available as needed. A suitable glass container is
filled to a height not to exceed the length of the cleaned part of the
microscope slides. There are various slide-staining jars that are
acceptable for this purpose, but we have found that a 30-ml tall form
Griffin beaker will do as well. Keep the container tightly covered
(with aluminum foil if no regular cover is provided) to prevent sol-
vent evaporation, when it is not in use.
Remove a slide from the oven, and immediately immerse the
polished end in the Formvar solution. Count slowly to 5 (to approxi-
mate 5 seconds) and then, without hesitation, yet without jerking,
remove the slide from the solution and drain it vertically on filter
paper. When it has dried, the plastic film is cut free by stroking the
slide edges with a bamboo splint. Start the stroke above the film
meniscus and carry it to the corner, and then along the bottom in both
directions to the corners. Work the corners with care, as these are the
places most likely to pose problems when trying to float the film free
of the slide. There is no need to cut along the top edge of the film
(meniscus); it will usually release there with no difficulty. As this
stroking proceeds, advance the splint in minute increments. This will
minimize the splinters and dust that might attach to the film. After
the good areas of the splint are worked over, discard it for a new one.
4. Floating the Film (See Figure 6, Steps e and f). A Stender
Figure 6. Step-by-step procedure for the production of Formvar sub-
strates for electron microscope specimen grids (see Appendix 1
for details), a: place a microscope slide polished with a tissue in a
warming oven. b: after warming, place the slide briefly in Form-
var solution, c: remove the slide from the solution and drain on
filter paper while holding vertically. d: cut the film free by scrap-
ing along the sides of the slide with a bamboo splint. Cut the film
along the sharp edges with repeated strokes of the splint, making
sure it is cut along both sides from the meniscus to the corners and
along the bottom. Take special care to insure that corners are cut.
e: float the film free by slowly immersing the slide into a water
bath while holding it at a steep angle. (To help free the film from
the slide, moisturize by holding it close to the mouth and breath-
ing on it.) Dotted lines indicate an area of the bath shown enlarged
in f. S: the slide held at the proper angle and pushed down in the
direction indicated by the down arrows. F: the film floating free of
the slide. The action is continued until the film is completely free.
dish or other receptacle should be ready-nearly full of the highest
purity available distilled or deionized water. If dust is a problem, the
surface should have been swept, either with filter paper or by overfill-
ing to a positive meniscus and sweeping the surface with a straight
edge (cardboard, polyethylene, Teflon, etc.). A small fluorescent lamp
should be arranged so that the surface can be monitored for dust, and
so that the support films may be conveniently examined for defects.
Hold the film side of the slide close to the mouth, and gently
breathe on the film to help release it from the glass. While holding the
slide at a steep angle to the water surface (about 45') with the film
side up, push it slowly straight down while watching to see if the film
floats free. As soon as the bottom edge releases, continue pushing the
slide down until the film floats completely free. At this point, as you
pull the slide from the water, you must be very careful, for the surface
tension will tend to pull the film back onto the slide. Alternatively,
simply allow the slide to slip to the bottom of the dish.
Examine the film carefully. It should be uniform and silver or, even
better, gray in color and show no other interference colors except
along the edges. Discard any films that exhibit more than a hint of
gold or other colors. Some of the problems and their possible causes
that may be helpful in evaluating the films follow.
Horizontal bars. Usually caused by withdrawing the slide from the
plastic solution too slowly. (Don't attempt to withdraw the slide
smoothly. This will usually fail. Rather, just pull it out without
hesitation, just as you would pick up something from a table.)
One or more dome-like interference bands, full slide width, and
usually rather ragged. The slide was too cool when removed from the
solution (either it was not hot enough to begin or it remained too long
in the Formvar solution), or it was jerked from the solution too
Dome-like interference bands, smooth an'd less than full slide width.
Caused by an impurity on the slide surface or a flaw in the glass.
Often, imperfect films will have areas that may be used. Simply
avoid the bad spots when placing the grids on the film.
Swathes of interference colors. The plastic solution has remained
standing too long without stirring, or was insufficiently stirred ini-
Film shows interference colors other than gray or silver. If these
colors are confined to the edges of the film, they can be avoided when
placing the grids on the film. However, if they cover large areas of the
film, the Formvar solution is too concentrated and should be diluted
with ethylene dichloride. The best films are of such a light gray that
they are nearly invisible. These have barely enough mechanical
strength to span the grid openings, but when coated with carbon,
they will provide a very stable support for electron microscopy.
Figure 7. Proper spacing for electron microscope specimen grids (g) in
relation to the plastic supporting film (f). A: grids are spaced too
closely. The plastic support film spans the gaps between the grids,
and when the individual grids are lifted from the index card base
(c) for use they will disturb and probably damage the film on the
surrounding grids. B: grids are spaced far enough apart to allow
the plastic substrate to be anchored between them. Grids may be
lifted individually without disturbing those nearby.
However, they must be handled very carefully to avoid tearing the
(Note: if possible, do film casting on a rainy day. It is astonishing
how much more easily the films float free of the slides when the
atmosphere is saturated. This is true even in air-conditioned labora-
tories. On a dry day, the film seems to be glued to the slide. Placing
the slides in a high humidity container will help loosen the film, but
upon removal from the chamber you must float the film free very
quickly, before it dries, which is very difficult to do.)
5. Arranging Grids on the Film (See Figure 7). Arrange the
grids on the film with the aid of forceps. Do not crowd them, as the
film should sag between grids to make contact with the mounting
surface, insuring that later, when the grids are put to use, they may
be lifted individually without disturbing adjacent grids. Touch each
grid lightly with the closed tips of the forceps to ensure complete
contact with the film. Put either the dull or shiny side in contact with
the film, whichever suits you, but be consistent so that the film side
can be readily identified if the grid is overturned while being hand-
led. We use copper grids that have been plated with rhodium on one
side, contrasting with the copper color of the unplated side. If such
grids are to be used, then insure that exposure to acetic acid during
cleaning is especially brief, or the rhodium will be dissolved.
6. Picking up the Film (See Figure 8). When grid placement is
complete, prepare a piece of index card cut to a rectangle of slightly
larger dimensions than those of the floating film. Use one edge of the
card to trap a corresponding edge of the film, well clear of any grids,
but overlapping the edge of the film enough to "grab" it squarely.
Push the card down vertically, far enough so that the plastic is
pressed against the card for the whole length of the film with the
grids sandwiched between card and film. Then lift the card straight
up until it clears the water surface. Hold the card squarely at arm's
length and flatten the film against the card with an aerosol duster
held close to the body, with the airstream directed at the card. If the
force is too great, or the card is not held perpendicular to the air-
stream, then the grids may be blown away.
If an aerosol duster is not used, then it is especially important to
insure that the grids are in good contact with the film by touching
each with the forceps or other pointed object individually, directly
after they have been placed on the floating plastic film. Blot any
water droplets from the card, avoiding the grid surfaces, and place it
in a Petri dish containing silica gel. Between grid-making sessions,
the gel should be regenerated by storing the dishes containing the gel
in the warming oven. However, the dishes should be removed to cool
before using them to dry grids, and in no case should the Formvar
Figure 8. Picking up the plastic support film and attached grids. A: grids
(g) are spaced on a floating Formvar film (f). A piece of index card
(c) is touched to the film (parallel and close to one edge of the plastic
film) and the card is pushed straight down (down arrows). B: the
downward thrust of the card is continued, and the grids are being
forced against the card. C: the entire film has been forced against
the card. At this point the card is withdrawn vertically.
clad grids be placed in the oven, or they will certainly be ruined.
When the grids are dry they are ready for carbon coating.
Note: Preliminary tests utilizing a method developed by H. C.
Aldrich (personal communication) for stripping Formvar films from
slides indicated it could be an acceptable alternative to the method
described above. The chief advantages of this alternative method are
that the Formvar films release more easily from the microscope
slides, and the slides do not have to be warmed as was necessary with
the preceding procedure. This alternative procedure is described
below with only slight deviation from Dr. Aldrich's outline.
Clean the grids, and prepare the Formvar solution as described
before. Prepare a solution from laboratory detergent powder (e.g.
Alcanox, Hemo-sol)-0.5 g/100 ml water should be sufficient. The
exact concentration does not seem to be critical. Clean a microscope
slide as before, but use some 95% ethanol to clean it more thoroughly,
and submerge it deeply enough in the detergent solution to cover the
area from which the film will be produced. Allow it to remain in the
detergent solution for at least 5 min. Drain the slide vertically on
filter paper, and prop it up to air dry. It will take 15 min or longer to
dry, but of course several slides can be processed together. When the
slide is dry, polish it with tissue to remove the powdery surface
residue. At this stage, the slide should look clean and shiny and have
a waxy surface. Submerge the slide into Formvar solution and pro-
ceed as described before, except that it will probably not be necessary
either to scrape the edges of the slide or to breathe on the film to
7. Carbon Coating the Grid Substrates. The exact set-up used
for carbon coating depends on the particular evaporator used, but
some general precautions may be useful.
Plastic substrates such as Formvar require a stabilizing coating to
keep them from squirming and tearing when bombarded by elec-
trons, as in the beam of an electron microscope. Carbon films will
disperse heat rapidly and are remarkably transparent to electrons, as
well as being chemically inert; thus, they are admirably suited as
coatings for stabilizing Formvar and other plastic substrates.
Although any substance in the beam path of an electron microscope
will reduce illumination and decrease contrast, the relative electron
transparency of carbon permits fairly heavy applications to be used.
Any loss of contrast and brightness due to carbon coating may be
recouped by keeping the thickness of the plastic substrate to the
absolute minimum required to provide enough mechanical support to
span the areas between the grid bars.
Grids used with negative stains generally require more carbon
than those used to support ultrathin sections, replicas, or metal-
shadowed particulate specimens. The extreme contrasts encountered
when illuminating negatively stained grids produce enormous dif-
ferentials that stress the plastic and cause distortion and tearing of
the substrate. Thin carbon coatings may be insufficient to deal with
this situation. For improved stability, we apply a carbon coating that
may be regarded as excessive by others. We expected to sacrifice some
resolution to achieve stability, but surprisingly, the resolution seems
as good as with thinner carbon coatings. Resolution in electron mi-
croscopy of negatively stained virions is severely limited by the
characteristics of negative stains; so other factors such as the electron
scattering caused by the density of the specimen and its supporting
medium are generally of lesser importance in achieving good resolu-
tion. This argues in favor of applying the carbon heavily enough to
achieve a substrate that is stable in the electron beam.
Avoid overheating the plastic substrate. Evaporate the carbon in
short bursts rather than in one continuous application. Overheated
Formvar may soften and consequently sag between the grid bars.
To gauge the thickness of the carbon coating, put a droplet of
diffusion pump oil on a small piece of white glazed porcelain and
place it in the evaporator near the grids to be coated. The white
"shadow" of the oil will contrast with the surrounding area and allow
the amount of carbon deposited to be gauged.
Negative Staining Techniques
Negative Stains and Wetting Agents
A satisfactory negative stain must dry without crystallization, be
reasonably inert chemically, be sufficiently dense to electrons to lend
good contrast to biological specimens (which are usually electron
transparent), be soluble in water or a solvent miscible with water,
and preferably be compatible with a suitable wetting agent (see
Figures 9 and 10 for a comparison of stains and wetting agents).
Negative staining results may be influenced by the source and age of
Phosphotungstic Acid (PTA). PTA is the most frequently used
negative stain for electron microscopy of biological extracts. Desir-
able features of PTA include high contrast, ready solubility in water
over a wide range ofpH values, and compatability with bovine serum
albumin (BSA), a commonly used wetting agent. The undesirable
features of this stain include an incompatability with some proteins,
a susceptibility to electron beam damage, a hygroscopic nature, and a
failure to delineate fine structure as well as other stains, such as the
uranyl stains (see Figures 9 and 10). The usual method of preparation
used for PTA is to adjust a 2% aqueous solution to pH 6.5 to 7.0, and
add Fraction V BSA to a concentration of 250 gg/ml. If bacitracin is
used instead of BSA, then resolution will be increased slightly, and
the background will be smoother and cleaner. PTA may also be used
at lower pH values to minimize its destructive effect on some pro-
Uranyl Acetate (UA). This stain is superior to PTA in many
respects, and if it is properly applied, it may be used in most situa-
tions as the stain of choice. UA has not usually been used for crude
extracts because of its extremely low pH value, which causes many
sap components to coagulate and cannot be altered without causing
the stain to precipitate. This low pH value also makes UA incompati-
ble with BSA. As plant saps curdle in the UA staining solution, it
cannot therefore be used to make leaf dips in the traditional way, i.e.
by extracting the tissue with the stain itself. However, if the tissue is
extracted with buffer, water, etc. before adding the stain and the
extract is applied to the grids, which are subsequently washed, then
the UA can be applied to these grids without causing damage, and
smooth, well stained grids will usually be produced, exhibiting good
contrast and specimen detail.
Grids prepared with UA usually have a more pleasing overall
aspect than do other stains. UA is far more stable in the electron
beam than other commonly used stains, including PTA, and unlike
PTA, UA is not hygroscopic, allowing grids to be stored for indefinite
periods without any special precautions. Grids prepared with PTA
should be stored in a desiccator if they are not used promptly.
Although UA is not compatible with BSA, it is compatible with
bacitracin, a wetting agent that is generally superior to BSA (8). We
normally use a 1% or 2% aqueous solution of UA containing 250
pg/ml bacitracin, prepared by carefully stirring in an equal volume of
500 Jg/ml bacitracin to a 4% solution of UA. Never add dry bacitracin
to UA solutions.
Uranyl Formate (UF). This stain is similar to UA except that it
is somewhat more susceptible to damage by electron bombardment
(but less so than PTA) and has a less immaculate appearance than
UA. Most importantly, UF solutions are unstable in most solvents,
including water. In many solvents, UF will begin to precipitate
within minutes after dissolving. Like UA, UF is incompatible with
BSA and it is stable with bacitracin for only a few minutes. In spite of
all of these defects, UF is unexcelled in delineating fine structural
detail, and should be used whenever it is necessary to study virion
Figures 9 (left)Effects of several negative stains and wetting agents upon
and 10 (right). the appearance and resolution of a purified preparation of
pepper mottle virus (potyvirus). Figures 9 and 10 are micro-
graphs taken from the same areas and differ only as to
magnification. A and B: stained with 2% PTA. C: stained
with 2% UA. D: stained with UF in water/methanol, as
described in Appendix 2. A: BSA (250 Lg/ml) used as a
wetting agent. B, C, and D: bacitracin (250 Jg/ml) used as a
wetting agent. In Figure 9, note the difference in texture in
the backgrounds of A and B. The BSA produces a distinctly
more pebbly background than does the bacitracin. In Figure
10, note the effect of the stain/wetting agent combinations on
particle substructure. The helical structure is not apparent
in A (PTA/BSA) and is only hinted at in B (PTA/bacitracin).
In C (UA/bacitracin), the helix is evident, and in D (UF/
bacitracin), it is quite distinct. The micrographs in Figure 9
are all at the same magnification: bar = 100 nm. The mag-
nifications of the micrographs in Figure 10 are the same: bar
= 50 nm.
ultrastructure. Common practice is to prepare fresh solutions of UF
whenever they are needed and to use them without a wetting agent.
Other techniques, such as exposure to a glow discharge plasma, may
be used to render the grids hydrophilic.
These difficulties tend to discourage the use of UF. We have found
however, that the principal limitations of UF can be overcome, and
that solutions of UF can be both stored and combined with a wetting
agent for negative staining applications. UF is soluble in methanol
(methyl alcohol) and stable in methanol solution for several months if
it is protected from light. Stir the desired quantity of methanol with
an amount of UF powder that exceeds the solubility limit of the
solvent. Arrange to do the stirring in darkness. If a magnetic stirrer
is used, it can be completely covered with a box, or the container can
be wrapped. Check occasionally to see if more UF powder should be
added, and continue stirring for about 1 hr. Absolute saturation is not
necessary, and probably not even desirable. The aim is to produce a
concentrated stock solution, and the solubility of UF in methanol is
probably at best 4% at 20C (solubility of UF powders varies from one
supplier to another).
When the stirring is completed, allow the solution to settle for a few
minutes, and then transfer aliquots to small test tubes without dis-
turbing the sediment. Seal the tubes tightly and store in a light-tight
container at room temperature until needed. This stock solution
should range in color from a deep canary yellow to orange. It will
remain free of precipitates for several months, and thereafter a white
precipitate will gradually form. But, since this precipitate will settle
as it forms, it will cause no problems, and the stock solution can be
kept in use until it becomes too weak to stain properly.
The negative staining solution should be prepared from the UF
stock solution just prior to use. Put a few drops of stock solution in a
small beaker, the well of a spot test plate, or a similar small container
and dilute it dropwise with water until the color is close to that of a
similar quantity of UA used for comparison. If a wetting agent is
needed, then after the correct color has been achieved, carefully swirl
in a drop of bacitracin solution (300 Lg/ml) for every 6 or 7 drops of
stain (the exact final concentration of bacitracin is not critical). If the
solution turns cloudy after the bacitracin has been mixed in, discard
it and start anew. Do not add the bacitracin solution to undiluted UF
stock solution, as it will surely precipitate. Use the stain within a few
minutes of preparation.
The procedure is simple, and fresh UF stain can be made from the
stock solution in a few seconds. Often, satisfactory grids may be
prepared without the addition of a wetting agent, but the bacitracin
works well for the staining of those grids that are otherwise unwet-
table. Although the UF staining solution can be used in combination
with any of the procedures described in Appendix 2, it is most useful
with virus preparations that have undergone some degree of purifica-
tion. UA will delineate the gross morphology of virions better than
UF, but UF is peerless in the elucidation of virion ultrastructure.
Wetting Agents for Negative Stains. A great variety of addi-
tives have been suggested as wetting agents for negative staining,
but the mainstay for many years has been highly purified BSA
(Fraction V, 5 x crystallized). In spite of the popularity of BSA,
bacitracin is certainly superior to it in every respect. (a) Bacitracin is
a much smaller molecule than BSA and is not resolved at high
magnifications, while BSA produces a distinct and distracting pebbly
background. (b) Bacitracin is a more efficient wetting agent than
BSA, and may therefore be used effectively at lower concentrations.
(c) Bacitracin is compatible with all of the commonly used negative
stains, including the highly acidic UA, and to some degree with UF,
whereas BSA is incompatible with these stains.
Bacitracin may coagulate if added to acidic solutions in dry form,
and for that reason it is best to have aqueous stock solutions avail-
able. The degree to which proteinaceous wetting agents may compete
with virions for grid space has not been studied, but it seems a wise
precaution to abstain from adding either BSA or bacitracin directly
to virus solutions. For a thorough discussion on the use of bacitracin
as a wetting agent for negative stains, consult Gregory and Pirie (10).
Washing Grids with Bacitracin. Grids mounted with purified
virus preparations may be virtually unwettable, especially at high
dilutions, because the sap components that normally confer wettabil-
ity to the grid surface have been removed in the purification process.
If the washing solutions are repelled by the grid substrate, the final
results may be unacceptable, even when there is a wetting agent in
the stain and the stain seems to spread normally. The grid surface
may appear to have a large number of large, angular, dark objects
present, and the virus itself may stain poorly, or not at all. We have
found that, in these situations, excellently stained grids can be pro-
duced by adding bacitracin to the washing solutions, and especially
the final water wash. For washing purposes we have used concentra-
tions of bacitracin ranging from 100-300 pRg/ml successfully, and the
highest concentration has consistently produced the best results. For
reasons given in the previous section, we have not added the bacitra-
cin directly to the virus preparations.
Storage of Negative Stains and Wetting Agents. General prac-
tice is to prepare the stains complete with wetting agent, refrigerate
them, and to retain them either for a set period of time or until
contaminating growths appear in them. So, although the principal
problem with refrigeration is that storage time is limited because of
contamination with various organisms, some stains, such as UA,
tend to gradually precipitate at low temperatures, which results in a
decline in stain concentration. Some stains, such as PTA, can be
made up and adjusted for pH, complete with wetting agent, and then
be freeze-dried in aliquots; which will reduce waste and preparation
time. However, other stains are not amenable to such treatment.
An alternative storage technique, and one that has general ap-
plication to stains, wetting agents, buffers, and washes, is to pass
these solutions through sterilizing filters into sterile containers, and
to store them all at room temperature. We have found it convenient to
use membrane sterilizing filters, of the type that is mounted on a
hypodermic syringe, to prepare suitable quantities of these solutions.
If the quantities needed are anticipated, then there need be little
waste. As an example, we pass negative stains through filters with a
0.2-pm pore size, store aliquots of 2-3 ml in small sterile test tubes,
and keep them in a dark cabinet at room temperature. Adjustments
to pH and the addition of wetting agents are made prior to filtration.
Negative Staining of Virions
Materials Needed. Grids with fine mesh, 200-400, 75 x 300, or
100 x 400, for example. Copper grids are most commonly used, and
are available with one side rhodium-plated for purposes of orienta-
tion. The grids should be clad with a plastic substrate such as Form-
var that is coated with carbon.
Double-sided (adhesive on both sides) cellophane tape.
Blotting paper cut into squares of approximately 2.5 inches (6.5 cm).
This paper is available in large sheets or may be found as individual
Pasteur pipettes of the 5.75-inch (15-cm) length, with 1-ml rubber
Small beakers for mixing and diluting. Generally, disposable plas-
tic beakers of 5-, 20-, and 50-ml capacities are sufficient; although, of
course, glass beakers may be used.
Washing solutions. Quite often a buffer used for the final resus-
pension of the virus is used. Water, or an aqueous solution of bacitra-
cin is used as the final wash before staining.
The following items are needed only when crude extracts such as
leaf dips are to be mounted and stained: (a) glass microscope slides
(an economy grade will do); (b) razor blades; (c) wooden toothpicks
(the round type work best); and (d) extraction buffers.
If the virus to be mounted is known, a buffer known to be compati-
ble with that virus should be used. Otherwise, a general purpose
buffer or water should be used. Phosphate buffers are generally
suitable for a wide range of viruses. Molarities in the 0.05 to 0.1 range
with a pH range of 6.5 to 7.5 are suitable.
Negative stains. PTA, UA, and UF as prepared above are recom-
mended, but other stains such as ammonium molybdate may be used.
The Staining Procedure (See Figure 11)
Attach a length of the double-coated cellophane tape to a micro-
scope slide, parallel to the long axis of the slide and overlapping one
side (the right side for right-handed workers) by 2-3 mm. If the tape
is longer than the slide, it will conveniently anchor the slide in place.
We mount the slide on a large piece of filter paper so that spills will be
absorbed, and to provide a writing surface to label the grids. Position
the grids along the length of tape, slightly overlapping the bottom
surface of each grid on the upper surface of the tape, pressing it hard
enough to barely hold it in position. Too firm an attachment will
make the grid very difficult to remove without causing severe dam-
age to the grid or to the substrate. Do not crowd the grids along the
tape as cross-contamination with the contents of neighboring grids is
a distinct danger. Seven to eight grids/slide is reasonable, but experi-
ence and personal preference will dictate the exact number.
Mounting the Virions on the Grid. CVC preparations or
purified virus suspensions. Apply a droplet to the grid and allow it to
stand for 1 min. There is usually no benefit to be obtained by longer
standing, except perhaps in the case of very dilute suspensions.
Conversely, there are usually no problems associated with long
standing, as the grids are washed of excessive material before
Leaf dips. The term leaf dip originally applied to the practice of
dipping the cut edge of leaf into a negative stain droplet placed
directly on a specimen grid, and then removing all but a thin film.
Nowadays, however, it generally refers to any procedure that nega-
tively stains crude leaf extracts, no matter how they are derived. The
method we have normally used is to place a small piece of leaf blade
6-8 mm2 in 5-6 drops of buffer on a microscope slide which has been
polished clean, and chop it a number of times with a razor blade. The
idea is to expose a great number of cut edges to the buffer. To
facilitate this, the leaf piece may be chopped with a number of
parallel cuts while holding it in place with a toothpick, and then
rotating the tissue through 900 for further chopping. If gentle vertical
chopping is used rather than a slicing motion, the leaf tissue will
often hold together instead of yielding many tiny bits of tissue-
which will need to be coped with when drawing up the extract. (If, in
spite of precautions, bits of tissue are produced, then try to exclude
them when drawing the extract into a pipette.)
Gently roll a toothpick over the chopped tissue to expel the cell
contents. If the rolling action is too forceful, then unwanted materials
such as structural tissue will be extracted. Only in cases where
virions are unusually difficult to extract-certain monocotyledonous
hosts or phloem-specific viruses are good examples of difficult cases-
would this precaution be ignored.
Figure 11. Preparing grids for electron microscopy by washing and nega-
tive staining. A: specimen grids positioned for application of virus
suspension, washing, and negative staining, a: glass microscope
slide; b: double-sided cellophane tape; and c: grids attached to the
tape. One of them has a standing droplet of solution applied. B: a
folded piece of blotting paper is being positioned to adsorb, and a
Pasteur pipette is being positioned to deliver a solution to a grid
positioned on the tape. C: at the initial stage of solution applica-
tion, the blotting paper is positioned so that it touches the rim of
the grid, and the first droplets are applied. D: as the washing
proceeds, the blotting paper is slightly withdrawn from the grid
while the application of the solution continues, insuring a steady
flow of solution across the grid surface. The procedure is essen-
tially the same whether the grids are to be washed with buffer,
water, or bacitracin solution. Following the washing procedure
(the final wash should be with water or bacitracin solution),
several droplets of negative stain are applied, and the last droplet
is removed as completely as possible by the blotting paper. The
grid is then allowed to dry.
When the tissue extraction is complete, draw the extract into a
Pasteur pipette and examine it. If it is more than slightly colored, it
should be diluted with buffer until the color is just discernible. When
a satisfactory concentration is obtained, place a droplet on a grid that
has been mounted on the cellophane tape, and allow it to remain for 1
min. An extract that is too concentrated may present no problems,
but an extract of greater dilution is generally easier to work with. Of
course, the color of an extract is not a foolproof guide to its concentra-
tion, but in most cases will give acceptable results.
Remove the droplet from the specimen grid by touching the grid
rim where it is attached to the tape with a folded corner of a piece of
blotting paper folded in half. Immediately begin washing the grid
with a washing solution (usually a buffer) dropped from a Pasteur
pipette at a rate not exceeding the rate at which the blotting paper
absorbs it. As the washing proceeds, begin withdrawing the blotting
paper by pulling it away from the grid and across the tape for a
distance of 6-8 mm, meanwhile maintaining contact between the
blotting paper and the tape surface (see Figure 11). This tactic will
promote thorough washing by producing a vigorous flow of buffer
across the grid. Use at least 25-30 drops of washing solution (if you
use "1-ml" bulbs on Pasteur pipettes, then one dropperful is generally
about the right amount to use, and each piece of blotting paper can
easily absorb 50-60 drops of liquid). As the blotter becomes satu-
rated, switch ends, and change to fresh blotting paper as often as
necessary to maintain a fast flow.
Immediately wash the grid with at least 10 drops of distilled or
deionized water to remove all traces of the washing solution. This
step may be omitted if the previous washing solution consisted only of
water or of water containing bacitracin.
Wash the grid with 2 or 3 drops of negative stain. This will remove
any remaining water and insure a uniform distribution of stain. Use
a dry corner of the blotting paper to remove as much of the stain as
possible; leaving only a thin film. Allow the grid to dry before remov-
ing it from the tape. Exercise care in pulling the grid from the tape or
the substrate may be damaged (many parallel tears in the substrate
are usually indicative of this type of damage).
The grid is now ready for storage or examination in the electron
microscope. If the grid has been stained with a uranyl compound,
then it may be safely stored for extended periods with no special
precautions. However, as some stains such as PTA are hygroscopic, it
is sound procedure to routinely store all of the grids in a desiccator
(using Drierite, silica gel, or calcium chloride as the adsorbent). Grids
have been stored for years in this manner with no noticeable de-
Negative Staining of Immunosorbent Materials.
Immunosorbent electron microscopy is a term that covers several
different techniques that increase the number of virions adsorbed to
the surface of a specimen grid, and/or aid in the process of identifying
a particular virus infection. It is beyond the scope of this bulletin to
cover immunosorbent electron microscopy (ISEM) in depth; however,
it may be useful to briefly outline grid preparation procedures as we
have adapted them for use with ISEM. The materials that are needed
for ISEM are mostly the same as those that are needed for the regular
leaf dip procedure (see the materials in Appendix 1). Additionally,
appropriate antisera and a supply of protein A are needed.
1. The Derrick Method (Protein A Modification). Place a
droplet of an aqueous solution of protein A (10%, w/v) on a grid
attached to double-sided cellophane tape as described before, and
allow it to stand for 15 min. Wash the grid with 20 drops of SP buffer
(0.01 M KPO4, pH 7.2).
Immediately remove most of the buffer and wash the grid with 1
droplet of antiserum diluted with SP buffer. The precise amount of
dilution should not appreciably affect the number of antibodies
adsorbed to the grid as long as the dilution is not too great. We have
used dilutions ranging from 1:20 to 1:1000 with good results.
Washing the grid with a drop of the antiserum diluted to the proper
strength precludes any further diluting effect by the buffer. Apply a
second droplet of diluted antiserum and allow it to remain standing
on the grid for 1 hr. Make the droplet as large as practical without
forcing the droplet to spread, and cover it with the lid of a Petri dish.
Check the grid occasionally, adding buffer if necessary, to insure that
the droplet does not dry.
Wash the grid with 30-35 drops of SP buffer (if a Pasteur pipette
has a 1-ml bulb attached, this would be one full pipette).
Remove all but a thin film of buffer, wash the grid with 1 drop of
prepared antigen, and apply a large droplet of the antigen. Cover,
and allow the antigen to remain on the grid for 3 hr. Check frequent-
ly, and add buffer if it is necessary to prevent the grid from drying.
Wash the grid with 30-35 drops of SP buffer.
Wash the grid with 10 drops of distilled or deionized water.
Wash the grid with several drops of uranyl acetate, removing all
but a thin film of stain, and then allow the grid to dry.
2. The Decoration of Virions with Antibodies. Particles are
attached to the grid surface using any of the procedures above,
washing as called for, continuing with the steps below following the
final washing with buffer, but prior to the water wash and staining
steps. The Derrick procedure may be used to enhance virion numbers,
but the final resolution and contrast of the decorated virions will be
better if a nonserological method is used to attach the virus particles
to the grids. The best decoration will be realized using purified virus,
but virions obtained by one of the CVC procedures will often serve
nearly as well.
Apply a droplet of an appropriate antiserum, diluted 1:20 with SP
buffer, to the grid (virions have already been attached to the grid and
it has been washed with buffer) and allow it to remain for 5 min.
Wash the grid with 20 drops of SP buffer.
Wash the grid with 10 drops of water.
Wash the grid with 2 or 3 drops of UA, allowing it to dry after
having removed all but a thin film of the stain.
Problems with Negative Staining.
Most failures to obtain clean, uniform, well-stained specimen grids
result from inadequate washing. It is necessary to provide both a
vigorous flow and to use sufficient quantities of the washing solu-
tions. It is far better to wash the grids too much than to wash them too
little. Use at least 1 ml of washing solution (add the droplets so that
they fall on the grid rather than simply flow onto it). Change the
blotting paper frequently enough to insure that it is able to readily
absorb the solution, and draw the paper far enough away from the
grid to promote a vigorous flow of the washing solution across the
grid. Symptoms of poorly washed grids include densely stained back-
grounds and unsharp imaging. In general appearance they are dark,
murky, and obscure, and in extreme cases may have a cracked
appearance, much like a dry mud flat.
Crystals occur in some plant saps, but others seen on negatively
stained grids may result from incomplete removal of buffer prior to
staining because of the inadvertent omission of the water washing
step, or to crystals of stain that formed at the tip of a Pasteur pipette
on standing. These stain crystals are produced by evaporation and
suggest that the stain has been held in the pipette too long, usually
overnight. The remedy for this problem obviously is to replace both
stain and pipette. Stain crystals are generally very dense, but with
the exception of those produced from UF, they may not be immediate-
ly recognizable as such. UF crystals may often be elongated hex-
Another factor preventing the uniform staining of specimen grids
is poor wettability of the grid surface. Although the incorporation of a
wetting agent such as bacitracin into the final staining solution will
usually insure a uniform stain deposit, alone it may be insufficient to
produce good wettability. Grids that contain artifacts, those that
have little or no background stain, or those on which are discovered
fewer virions than are expected may suffer from this problem. The
best cure we have found for this problem is to incorporate bacitracin
into the washing solutions (at a rate of 300 jLg/ml).
However, bacitracin or other proteinaceous wetting agents prob-
ably should not be added at an earlier stage because of possible grid
site competition with virions. Grids mounted with crude prepara-
tions such as from leaf dips will only suffer from poor wettability in
unusual cases; routine washing with bacitracin solutions need only
be used for purified virus preparations or for CVC preparations of
outstanding clarity. This problem of grid wettability is most severe
when the samples are highly dilute.
A problem with similar symptoms is not caused by poor wettabil-
ity, but rather by the presence of substances that interfere with
staining. If these are used in the procedure they must be sufficiently
removed to prevent unwanted side effects. Examples of such sub-
stances are PEG, sucrose, and Triton X-100. Indeed, Triton could
prove to be very useful in the CVC procedure if subsequently it could
be removed. When Triton X-100 is used for full scale purifications,
then centrifugation on density gradients seems to remove it suf-
ficiently for electron microscopy. Other contaminants, such as cesium
chloride, solvent traces, and buffers, may present no staining prob-
lems whatsoever, if the grids are sufficiently washed prior to
Production of Electron Micrographs
Although our Hitachi H-600 microscope is equipped with a sheet
film camera, we use a 35-mm camera exclusively for micrograph
production. The argument that the image must be recorded at the
greatest possible size in order to achieve maximum micrograph res-
olution is less convincing in view of the extremely fine grained
emulsions that are available on roll film and the high resolution that
modern microscopes are capable of obtaining.
At present, the primary factor limiting resolution in the electron
microscopy of stained biological specimens appears to be the charac-
teristics of the stain itself: grain size, penetration of tissues or parti-
cles, inherent contrast, etc. The Philips Company has recognized that
large film formats do not automatically convey greater image resolu-
tion and have supplied 35-mm cameras, as well as the conventional
large format sheet/plate cameras, as standard equipment on the
Philips 201C electron microscope. The sales brochure for that micro-
scope presents comparative micrographs of the same area of a nega-
tively stained specimen of TMV on both large format and 35-mm
films. The micrographs, enlarged to the same final magnification
from both film sizes, are virtually indistinguishable.
The savings in time and materials of the 35-mm format over the cut
film/glass plate format are substantial, and filing is more compact.
The amount of space that would be required for filing if the nearly
40,000 micrographs that we have taken had been preserved on glass
plates rather than 35-mm filmstrips, would have been prohibitive.
The ease of recording and processing 35-mm film allows the operator
to take numerous photographs, and spares him the bother of deciding
if a certain image is worth preserving on an expensive sheet or plate.
Although there are roll films made specifically for electron micros-
copy, we find Kodak Technical Pan 2415 (TP2415) is nearly ideal for
micrograph production. The Kodak company does not at this time
suggest the use of TP2415 for electron micrographs; but it has ex-
tremely fine grain and resolving power, allowing it to be enlarged to
magnifications of 50 x or better, and the contrast of TP2415 can be
controlled by choosing an appropriate developer (this capability
makes the film useful for many laboratory applications other than
electron microscopy). A tendency for this film to fog when used for
high contrast work can be controlled by the addition of an antifogging
agent to the developer.
A procedure for developing Kodak Technical Pan 2415 35-mm roll
film to a contrast suitable for electron microscopy follows.
Mix 1 part Kodak HRP developer with 4 parts deionized/distilled
water. Add and thoroughly dissolve 0.75% (w/v) benzotriazole. Pre-
pare the developing solution just prior to use, and discard it after use.
(Benzotriazole is the generic name for Kodak Antifog A. The use of
Antifog A tablets precludes weighing, but the tablets are tedious to
dissolve, while benzotriazole powder dissolves readily.) Cool the mix-
ture to 21C and pour it into a developing tank. In darkness, sub-
merge the spooled film quickly into the developer and rap the tank
sharply to dislodge any bubbles. Cover the tank tightly and develop
the film for 5 min, agitating for 5 sec each minute. Drain the tank and
quickly rinse the film with 2 changes of water cooled to 21C. Fix with
continuous agitation (3 min if using Kodak Rapid Fixer), and rinse in
running water for 30 minutes. Rinse finally with two changes of
deionized/distilled water, drain the film, and allow it to air dry.
For further information consult the Kodak Technical Pan 2415
data in Kodak Pamphlet No. P255.
Selected List of Suppliers
Note: there are other suppliers for most of the materials that are
listed here. However, the sources listed have provided satisfactory
materials and thus may be expected to do so on future orders. No
further endorsement is either given or implied. Such a list is only
necessary for those workers who otherwise have no reliable sources
for these materials.
Eastman Kodak Company
343 State Street
Rochester, NY 14650
(photographic film, blotting paper, antifog, uranyl format and
Fisher Scientific Company
711 Forbes Avenue
Pittsburgh, PA 15219
Fisher Scientific Company, International Division
101 Thomson Road
No. 21-04 Goldhill Square
(crepe filter paper, general laboratory supplies)
Ernest F. Fullam, Inc.
900 Albany Shaker Road
Latham, NY 12110
(rhodium-plated copper grids, extra sharp forceps, other electron
Ladd Research Industries, Inc.
P.O. Box 1005
Burlington, VT 05402
(electron microscope supplies)
Ted Pella, Inc.
P.O. Box 150
Tustin, CA 92681
(electron microscope supplies)
Sigma Chemical Company
P.O. Box 14508
St. Louis, MO 63178
(bacitracin, polyethylene glycol, protein A, general biochemicals)
99 High Hill Road at 1-295
P.O. Box 99
Swedesboro, NJ 08085-0099
(microscope slides, Stender dishes, general laboratory supplies)
UNIVERSITY OF FLORIDA
This publication was promulgated at a cost of $3795.00, or 75.9
cents per copy, to explain a method developed to obtain
clarified viral concentrates from extracts of small samples.
All programs and related activities sponsored or assisted by the Florida
Agricultural Experiment Station are open to all persons regardless of race,
color, national origin, age, sex, or handicap.