http://jhc.sagepub.com/Journal of Histochemistry & Cytochemistry

http://jhc.sagepub.com/content/38/5/625The online version of this article can be found at:

 DOI: 10.1177/38.5.2332623

1990 38: 625J Histochem CytochemS L Erlandsen, W J Bemrick, D E Schupp, J M Shields, E L Jarroll, J F Sauch and J B Pawley

secondary and backscatter electron imaging.High-resolution immunogold localization of Giardia cyst wall antigens using field emission SEM with

  

Published by:

http://www.sagepublications.com

On behalf of: 

  Official Journal of The Histochemical Society

can be found at:Journal of Histochemistry & CytochemistryAdditional services and information for     

  http://jhc.sagepub.com/cgi/alertsEmail Alerts:

 

http://jhc.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

by guest on July 10, 2011jhc.sagepub.comDownloaded from

625

0022-1554/90/$3.30

The Journal of Histochemistry and Cytochemistry

Copyright © 1990 by The Histochemical Society. Inc.Vol. 38. No. 5, pp. 625-632, 1990

Printed in USA.

Rapid Communication

High-resolution Immunogold Localization of GiardiaCyst Wall Antigens Using Field Emission SEM withSecondary and Backscatter Electron Imaging’

STANLEY L. ERLANDSEN,2 WILLIAM J. BEMRICK, DANIEL E. SCHUPP,

JANIEL M. SHIELDS, EDWARD L. JARROLL, JUDITH F. SAUCH, and JAMES B. PAWLEY

Department of Cell Biology & Neuroanatomy (SLE,DES) and Department of Veterinary Pathobiology (WJB), University of

Minnesota, Minneapolis, Minnesota 55455; Department ofBiology UMS,ELJ), Cleveland State University, Cleveland, Ohio 44115;

Environmental Monitoring Systems Laboratory (fF5), US Environmental Protection Agency, Cincinnati, Ohio, 45268; and

Integrated Microscopy Resource (JBP), University of Wisconsin, Madison, Wisconsin 53 706.

Received for publication February 9, 1990; accepted February 14, 1990 (0C1902).

We describe here the ultrastructural localization of Giardiacyst antigens in the filaments associated with the outer par-tion of intact cysts and on developing cyst wall filaments inencysting trophozoites. Post-embedding immunogold label-

ing of thin sections of intact Giard.ia cysts with polyclonaland monodonal antibodies specific for cyst wall antigens (ma-jor protein bands of approximately 29, 75, 88, and 102 KDon Western blots) showed strong labeling of the filamen-tous cyst wall, whereas no labeling was seen on the mem-branous portion. High-resolution field emission scanning

electron microscopy (FESEM) of Giardia cysts revealed thatthe cyst wall-specific polyclonal rabbit antisera and mono-clonal mouse antibody produced gold labeling of2O-nm fda-ments in the cyst wall as detected with secondary electron

Introduction

Giardiasis is a gastrointestinal infection produced by the parasitic

protozoan Giardia. This disease is transmitted fecal-orally via per-

son-to-person contact (19), contamination offood (5), and by water-

borne transmission through ingestion of cyst-contaminated water

(6,7).

The diagnosis of Giardia in fecal specimens from humans and

in those from other animal species, as well as in environmental wa-

ter samples, has been primarily dependent on the successful isola-

tion and identification of the cyst stage in the life cycle. The sensi-

tivity of detection of Gi.ardia cysts in feces or in filtered water samples

has been greatly improved with the development of immunofluo-

I Supported in part by CR-814622 from the US Environmental Pro-

tection Agency, DRR-570 from National Institutes of Health, Ohio Board

ofRegents academic and research challenge grant, and the Minnesota Medical

Foundation.

2 Correspondence to: Dr. Stanley Erlandsen, Dept. Cell Biology & Neu-

roanatomy, 4-13 5 Jackson Hall, University of Minnesota, Minneapolis, MN

5 545 5,

imaging (SE!) and backscatter electron imaging (BEI) at 10kV, despite coating of the cells with platinum by ion sput-tering. FESEM studies of encysting Giardia trophozoitesdemonstrated that immunostaining with antibodies to cystwall antigens produced colloidal gold labeling of develop-ing cyst wall filaments on the cell surface; however, the in-tervening membrane domains were unlabeled. Substitutionof normal serum for cyst wall-specific antibodies, or pre-absorption of specific antibodies with Giardia cysts, elimi-nated immunolabeling of the filaments. (J Histochem

Cywchem 38:625-632, 1990)

KEY WORDS: Giardia; Cyst wall; High resolution; Immunocytochem-

istry; Colloidal gold; Field emission SEM.

rescence methods (24,26,30), and the immunoreactivity detected

appeared to be localized in the cyst wall. Using transmission dcc-

tron microscopy (TEM) and high-resolution, low-voltage field emis-

sion scanning electron microscopy (FESEM), the cyst wall of Giardia

has been shown to consist of distinct membranous and filamen-

tous components (9). However, the portion of the cyst wall con-

taming the antigenic sites recognized in immunofluorescence studies

has not been determined.

Colloidal gold markers have been used in SEM with a variety

ofligands, such as lectins and antibodies, for ultrastructural detec-

tion ofspecific surface antigens (13). Secondary electron imaging

(SEI), backscatter electron imaging (BEI), and even a mixture of

the two signals have been used to identify gold particles greater

than 20 nm (8,33). Gold probes smaller than this have been con-

sidered difficult to visualize, particularly when specimens have been

osmicated or coated with thin metallic films (8,13,23,29,35).

In this study, the high-resolution localization of Giardia cyst

antigens in the cyst wall was investigated by using colloidal gold

immunocytochemistry with SE! and BE! in a high-resolution FESEM

(21). Detection of 15-nm colloidalgold immunolabeling for Giardia

by guest on July 10, 2011jhc.sagepub.comDownloaded from

626 ERLANDSEN, BEMRICK, SCHUPP, SHIELDS, JARROLL, SAUCH, PAWLEY

cyst antigens on the filamentous layer of the cyst wall was readily

accomplished with both SEI and BE!, even though the cysts were

osmicated and stabilized with a layer of platinum to make them

conductive.

Materials and Methods

Giardia. Cysts of G. muris were isolated from feces by sucrose flotation

as previously described (25,27). Axenic cultures ofthe MR4 strain of G. duo-

denalis were maintained in our modification (28) ofthe TYI-S-33 medium

of Keister (16). In vitro encystment of trophozoites was performed using

the procedure of Schupp et al. (28).

Antibodies. Three polyclonal rabbit antisera were produced against

Giardia cysts or isolated cyst walls. One rabbit antiserum (R-AGLMB) was

produced against human cysts and has been shown by immunofluorescence

microscopy to primarily stain the cyst wall of Giardia (26). Two other rabbit

antisera were produced against G. muris cyst walls prepared by the method

ofJarroll et al. (15). After the collection of pre-immunization serum, one

antiserum was generated against the SDS-insoluble portion of purified cyst

walls (R-ACWsds) by intramuscular inoculation of 7 x i0� cyst walls in

complete Freund’s adjuvant and four weekly subcutaneous injections of ap-

proximately 5 x i0� cyst walls using incomplete Freund’s adjuvant. The

other was raised in rabbits, after collection of pre-immunization serum,

against a 9 M urea extract (containing 5 mM dithiothreitol) of 108 pun-

fled cyst walls (R-ACWurea) inoculated intramuscularly in complete Freund’s

adjuvant followed by three subcutaneous weekly boosters of urea extractsof 108 cyst walls in incomplete Freund’s adjuvant. The mouse monoclonal

antibody (mcln RD348) against Giardia cysts was obtained from MeridianDiagnostics (Cincinnati, OH). This monoclonal antibody was originally pro-

duced by Sterling et al. (31) against in vitro-derived G. lamb/ia cysts ob-

tamed from a Peruvian human isolate, and has been reported to stain the

cyst wall of Giardia.

Formation of the Giardia cyst wall and the presence of cyst wall antigen

within cultured trophozoites requires the presence ofbile (28). Therefore,

all ofthe above polyclonal antisera and the monoclonal antibody were im-

munoabsorbed with Giardia trophozoites grown in culture medium with-out bile, to remove any antibodies that might crossreact with trophozoite

antigens.Electrophoresis of G. muris cyst extracts was performed using the method

ofLaemmli (17). Cyst extracts were prepared in sample buffer and electro-

phoresed on 16% polyacrylamide gel in the presence ofSDS. The proteinswere then transferred to nitrocellulose according to the method of Towbin

et al. (34), and their presence was revealed with the peroxidase-antiperoxi-

dase method of Woodcock-Mitchell et a!. (36), using 4-chloro-1-naphthol

to visualize the peroxidase reaction.

Immunological Staining. Giardia cysts were attached at room temper-

ature to glass chips pre-coated with poly-L-lysine as described by Erlandsen

et al. (9). Gthrdia trophozoites of the MR4 strain [G. duodena/is type ac-

cording to the classification of Filice (12)1 were cultured in TYI medium

without bile, and were stimulated to undergo encystment by addition of

bovine bile (28). Encysting trophozoites were concentrated by centnifuga-

tion (275 x g) for 10 mm at 4C and then attached to poly-L-lysine-coated

glass chips by incubating them on a warming plate at 37’C for 2-3 mm.

Unfixed Giardi.a cysts and trophozoites were rinsed with PBS, pH 7.4, con-

taming 1% bovine serum albumin (PBS-BSA) and 0.1% Tween 20 (poly-

oxyethylenesorbitan monolaurate; Sigma, St Louis, MO). This was followedby incubation with primary antibodies (30 mm at room temperature for

cysts and 5 mm at 4�C for trophozoites stimulated to encyst) consisting

of either rabbit polyclonal or mouse monoclonal antibodies to the Giardia

cyst wall. The specimens were then rinsed with PBS-BSA and incubated

with 15-nm gold-conjugated anti-immunoglobulin antibodies [goat anti-

rabbit IgG or rabbit anti-mouse IgG/IgM antibodies(AunoProbe EM;Janssen

Life Sciences Division, Olen, Belgium)] for the same time period and at

the same temperature as described in the primary antibody step. Controls

employed included (a) substitution ofpre-immunization or normal rabbit

serum for specific rabbit polyclonal antisera to Giardi.a cyst walls and sub-

stitution of mouse ascites fluid to replace the specific mouse monoclonal

antibody to Giardia, and (b) absorption ofthe primary antibody with iso-

lated G. muris cysts. After the immunostaining, the specimens were rinsed

with PBS, pH 7.4, followed by fixation for electron microscopy.

Electron Microscopy. Giardiacysts and trophozoites were fixed in 2-3%

glutaraldehyde buffered with 0.1 M sodium cacodylate, pH 7.4, for trans-

mission EM as described byJanuschka et al. (14). For FESEM, trophozoites

and cysts were processed as previously described by Erlandsen et al. (10).

Briefly, the cysts were fixed in the same glutaraldehyde fixative described

above, while trophozoites were fixed in a fixative containing 3% glutaralde-

hyde, 3% paraformaldehyde, and 0.1% picric acid. All samples were post-

fixed in the osmium-ferricyanide fixative ofLangford and Coggeshall (18),

dehydrated in an ethanol series, and then processed by the CO2 critical

point-drying method ofAnderson (3). The samples were ion-sputtered to

minimize charging with a thin platinum coating (11), estimated at 1-2 nm

in thickness, using an Ion Tech, Ltd. (VCR Group; San Francisco, CA) sin-

gle gun, argon ion beam sputterer operated at 4 mA at an accelerating

voltage of 10 kV for 6-8 mm. After coating, the samples were examined

in the Integrated Microscopy Resource at the University of Wisconsin-Madi-

son with a modified Hitachi 5-900 FESEM operated at an accelerating volt-

age ofeither 1.5 kV (SEI) or 10 kV (SEI and BEI), using the standard SEI

and BE! detectors. Stereographic imaging of cell surfaces was performedat an 8’ tilt to produce stereo pairs with 4 x 5 Polaroid type 52 film.

Results

To characterize the rabbit polyclonal antisera used for immunocy-

tochemistry, Western blot analyses were performed against a SDS

extract of G. muris cysts. Each ofthe polyclonal antisera for FESEM

immunocytochemistry was made specific for cyst walls before use

by pre-absorption with non-bile-stimulated trophozoites (these ap-

peared to lack cyst wall proteins; )arroll, unpublished observation)

to remove antibodies reacting solely with trophozoite stage anti-

gens. A non-absorbed reference rabbit antiserum was produced

against sonicated G. muris cysts and contained antibodies to both

trophozoites and cyst wall antigens. In Western blots, as shown in

Figure 1, lane A, this antiserum revealed 18-27 bands ranging from

<29 KD to >200 KD. The cyst wall-specific R-ACWsds antiserum

reacted with two major bands (Mr 88 KD, Mr 75 KD) and, to a lesser

extent, with 8-10 minor bands (Figure 1, lane B). The cyst wall-spe-

cific R-ACWurea antiserum reacted prominently with one major

band (Mr 88 KD) and a series ofminor bands, all ofwhich appeared

identical to those recognized by R-ACWsds. The cyst wall-specific

R-AGLMB antiserum (Figure 1, lane D) produced strong staining

of one major band at Mr 29 KD. The R-AGLMB also produced two

slightly less intense bands at Mr 75 KD and Mr 102 KD, with the

latter being recognized by R-ACWsds and R-ACWurea.

Ultrastructural examination ofa typical G. muris cyst by TEM,

as seen in Figure 2, showed that cysts were polarized with nuclei

at one end and a cluster of flagella at the other. The peripheral

cyst wall was separated from the underlying undivided trophozoite

by the penitrophic space and was composed of two layers: a thin

layer of cytoplasm bordered by two cell membranes (the inner of

these bordered the peritrophic space), and another layer consisting

by guest on July 10, 2011jhc.sagepub.comDownloaded from

45.-.

HIGH-RESOLUTION SEM GOLD LABELING OF GIARDIA 627

In this study we utilized ultrastructural immunocytochemical

A BCDI-.-.- .�,_- ‘�“--“�h��

2O5.�---

116_� __ I

66-. 1

29- __

Figure 1 . Western blot analysis of rabbit polyclonal antisera prepared againstGiardia. An SOS extract of intact G. muris cysts (9 x 10�) was added to eachlane of a 16% SOS-PAGE gel, followed by electrophoresis and transblottingto nitrocellulose. The nitrocellulose blot was cut into strips and probed with thefollowing antisera at their optimal dilutions: lane A, 1 �2O0 rabbit anti-sonicatedG. muris cysts (containing antibodies to trophozoites and cysts); lane B, 1:500R-ACWsds; lane C, 1 500 R-ACWurea; lane D, 1 100 R-AGLMB. Molecularweights at the left are given in kilodaltons.

offilaments anchored to the outer cell membrane (Figure 3). Immu-

nocytochemical post-embedding staining ofthe G. muris cyst wall

with antibodies specific to the cyst wall (polyclonal sera R-AGLMB,

R-ACWsds, R-ACWurea; monoclonal antibody RD348) revealed spe-

cific immunolabeling ofthe filamentous layer (Figure 4). Specific

immunolabeling was not detected associated with the membra-

nous portion ofthe cyst wall, the peritrophic space, or the underly-

ing trophozoite. No specific immunolabeling of the filamentous

cyst wall was seen in control sections where pre-immunization or

normal serum had been substituted for cyst wall-specific antisera.

Absorption ofeach ofthe antibodies specific for the cyst wall with

isolated G. muris cysts completely abolished the immunogold label-

ing of the filamentous cyst wall in the thin section.

Visualization ofa G. muris cyst by low-voltage (1.5 kV) FESEM

revealed that the external surface had a filamentous composition

(Figure 5). Higher magnification of the external surface demon-

strated that several populations offilaments (7-20 nm in diameter)

(9) formed an interlacing meshwork (Figure 6).

When indirect immunolabeling with 15-nm colloidal gold as

the marker was coupled to an anti-immunoglobulin antibody, suc-

cessful immunogold labeling of filaments in the Giardia cyst wall

was obtained with either specific cyst wall polyclonal antisera

(R-GLMB, R-ACWsds, R-CWurea) or the monoclonal antibody (RD

348). Immunocytochemical labeling of the filaments in the cyst

wall with 15-nm gold was detected by topographical contrast, using

SE! at an accelerating voltage of 10 kV (Figure 7). Individual 15-

nm colloidal gold particles associated with the filaments were iden-

tified by their size and shape, in conjunction with their increased

secondary electron emission, as shown in Figure 7. The identity

ofthe 15-nm particles was confirmed by examining the atomic num-

ber contrast obtained by BE! (Figure 8). In comparison to the high-

resolution imaging of the filamentous layer obtained at low ac-

celerating voltage (1.5 kV in Figure 6), the use of 10 kV accelerating

voltage for BE! still permitted definition ofindividual cyst wall fila-

ments that facilitated the localization of the 15-nm immunogold

particles and their correlation with surface topography. A poten-

tial drawback noted in stereo imagining ofthe immunolabeled sur-

face was that the extended use of >10 kV accelerating voltage for

BE! produced severe radiation damage that distorted the cyst sur-

face, unless careful examination was employed.

!mmunolabeling of the filamentous surface of a G. muris cyst

with the monoclonal (RD 348) specific for Giardia cysts, and cx-

amination at high magnification by FESEM, clearly revealed the

labeling of individual filaments with 15-nm gold particles by

topographical contrast with both SE! (Figure 9) and BE! (Figure

10). The immunolabeling ofthe cyst wall appeared to be associated

with the population oflarger filaments (20 nm) that predominated

in the cyst wall. Owing to the 15 nm size of the gold particle, it

was not possible to determine ifthe population ofsmaller filaments

(7 nm) that have been reported to cross-link the larger filaments

(9) were immunolabeled. The correlation between 15-nm gold par-

tides detected by topographical and atomic number contrast was

almost one-to-one, as illustrated in Figures 7-10. Occasionally, cx-

amination of the immunolabeled filamentous surfaces of Giardia

cysts by topographical contrast revealed the presence of a few ad-

herent spherical particles resembling 15-nm gold particles, but these

could be easily differentiated from immunogold because of their

lack ofsignal when examined with BE! (Figures 7-10). Overall, the

BE! was slightly more useful than SE! for identification of the im-

munogold label, especially when the gold particles were associated

with or attached to surface debris that obscured the topographical

contrast of the gold particle (Figures 8 and 10).

Cultured Giardia trophozoites were induced with bile to un-

dergo encystment, and at 10-12 hr post induction began to form

vermiform projections over the entire surface of the cell. These

filamentous projections resulted from the growth of filament initi-

ation sites located on the extracellular surface ofthe plasmalemma of

the induced trophozoite (Figure 11). !mmunogold labeling of en-

cysting trophozoites with specific antibodies (R-AGLMB, R-ACWsds,

R-ACWurea, and mcln RD 348) to Giardia cyst wall antigens demon-

strated that the filamentous projections associated with filament

growth were labeled with 15-nm gold particles (Figures 12 and 13),

whereas the intervening membrane was not immunoreactive.

Discussion

by guest on July 10, 2011jhc.sagepub.comDownloaded from

628 ERLANDSEN, BEMRICK, SCHUPP, SHIELDS, JARROLL, SAUCH, PAWLEY

S

S

I

S

a

S

.5’

a

CS

is

S f

4.S �Ps

#{149}:.a

H/

)

,�‘-- ‘

�%�V � � .�o-.. � .‘

� � � ‘:,;�:��“�‘r “ � ,.d- �

� � � � . ‘p #{149}:‘� �‘� a ‘ ‘ � � - PpI �

�4:t � -. ‘.4. �a � .�“�#{149}C �

� .... ,t S

‘:�7/k’�: t *t�- 1#{149} .

.

S.S

S

‘:� �

� 4� �� #{149}. #{149}:

‘I- � .4?aa�r’t�a �� � � #{149}�x�-- �

..,�. ‘a. � ‘ � . �� ‘ � #{149} S #{149} , � .l.--.*�

,-.- .-#{149}, �#{149}5� � . . ,, . .‘

� -�.�:‘�-4� , � S., - .�#{149} .#{149}‘,::�#{149}‘ . � , :��‘ #{149}� � ; - *

55 . - ‘ . . ‘�. - .. S t1� #{149}S..� 4- . a. #{149}

t:� � ...�.*s #{149}. ,�I- .�� � . ‘ 1�, �‘E �, � . � � ,

� � I� ._‘$ � “.‘i’.; b5�3�� #{149}‘#{149}5.. , S

�I. ‘‘�� -,� ,, �I t

�: .“ � -�:�::#{149}.�.‘ � �

� .‘:.�::,. .‘�t.,:s‘a’ #{149}� ,‘ #{149}0I

S � .‘#{149}� �_. a.,..

a. a#{149}. �*S S -.S C #{149}� �.

, . Sy. .. �. � . a � � ,.- . . ,�:‘ . �-

i S - -s’-”?’�”*

-�, � � :. : � - � �

by guest on July 10, 2011jhc.sagepub.comDownloaded from

HIGH-RESOLUTION SEM GOLD LABELING OF GIARDIA 629

methods, coupled with high-resolution FESEM, for detection of

colloidal gold labeling of Giardia cyst wall antigens on intact Giardia

cysts and encysting trophozoites. Immunogold labeling of the cx-

ternal surface of isolated cysts successfully demonstrated that the

cyst-specific polyclonal antisera and monoclonal antibody produced

gold labeling of the filamentous layer, but did not rule out the

possibility that the cell membrane possessed some immunoreac-

tivity. Experiments using either post-embedment staining of thin

sections from intact Giardia cysts or immunostaining of encysting

trophozoites revealed that (a) only the filamentous portion of the

cyst wall, not the membranous portion, was immunolabeled using

post-embedment immunocytochemistry, and (b) only the exposed

filament initiation sites on the extracellular surface of encysting

cells were labeled with 15-nm colloidal gold, whereas the interven-

ing plasmalemma between them was devoid of immunostaining.

Our results using Western blot analysis of the specific cyst wall an-

tisera, together with the ultrastructural localization of these anti-

bodies, have directly supported our interpretation that the cyst wall

antigens detected were a part of or were associated directly with

the filaments themselves, and not with the membranous portion

of the cyst wall. The exact chemical composition of the filaments

in the Giardia cyst wall has not been determined, but biochemical

analysis of purified filamentous cyst walls has shown them to con-

tam approximately 63 nmol of N-acetylgalactosamine per 106 cysts

(15), suggesting that the filaments may be composed of complex

arrangements of glycoproteins.

Treatment of cells with osmium fixation or the thin coating of

samples with materials having a high backscatter coefficient, such

as platinum, had been previously considered to be incompatible

with high-resolution localization of colloidal gold particles (8,13,

23,29,32). By using a high-resolution through-the-lens FESEM

equipped with a YAG type (cerium doped yttrium-aluminum gar-

net) scintillator, we were able to clearly resolve 15-nm colloidal gold

bound to cyst wall filaments, as demonstrated in Figures 7-10. Utiiz-

ing SE!, the shape and size of the 15-nm gold particles was easily

detected by both increased secondary electron emission and topo-

graphical contrast, while their identity was confirmed by atomic

number contrast with BE!. A good correlation was found between

the identification of 15-nm colloidal gold particles with SE! and

BE!, although occasional discrepancies between the two were found.

In previous studies using high-resolution BEI, the surface topogra-

phy of the cell surfaces labeled with colloidal gold particles was

often difficult, if not impossible, to visualize (8,29,35). Because

ofthis difficulty, some investigators have used a mixture ofSEI and

BE! signals to facilitate correlation of gold particles with surface

topography (8,29,32,33).

In this study, a thin platinum coating (estimated at <1 nm) was

applied by ion sputtering to minimize charging, and the thickness

ofthis layer was not sufficient to mask either the Z contrast signal

for the gold particle or the topographical features detected at 10 kV

with SE! and BE!, as seen in Figures 7-10. Despite the presence

of the platinum coating, examination of the filament size in Figures

7 and 8, and the particulate nature of the surface coating of the

gold particles in Figures 9 and 10, revealed a resolution of <5 nm.

Using the same modified FESEM employed in this study, Albrecht

et al. (1,2) have shown that 18-nm immunogold particles can be

visualized with both SE! and BE! at 15 kV, even though the surface

was ion-sputtered with platinum coatings ranging from 4.5 to 9 nm

in thickness. These investigators have also demonstrated that even

4-nm immunogold particles on platelet microtubule preparations

could be demonstrated using BE! at 15-20 kV on specimens coated

with carbon. As discussed by Pawley and Erlandsen (22), the metallic

coating of complex biological surfaces, such as the filamentous cyst

wall of Giardia, should not be thought ofas being uniform in thick-

ness. The useful resolution obtained when viewing a biological sam-

plc may have resulted from a number of variables, including not

only accelerating voltage (probe size) in the FESEM but also

beam-specimen interaction volume, contamination, delocalization

phenomena, and radiation damage. The latter has been shown to

produce significant distortion of the surface of Giardia cysts, which

was much greater at high accelerating voltages (>10 kV) used for

BEI than at low accelerating voltages (1.5 kV) used with SE! for

improved surface topographical resolution (22). The development

of methods in low-voltage scanning electron microscopy permit-

Figure 2. Transmission electron micrograph of a G. muris cyst fixed in the presence of 1% BSA. The cyst wall is composed of filamentous (c) and membranousportions (arrowheads) and is separated from the cytoplasm ofthe trophozoite by the peritrophic space (p5). N, nucleus; f, flagellum. Original magnification x 16,000.Bar = 0.5 �sm.

Figure a Transmission electron micrograph of the filamentous (c) Giardia cyst wall showing the course of individual filaments. The membranous portion of thecyst wall (arrowheads) separates the filamentous portion from the underlying peritrophic space (p5). f, flagellum. Original magnification x 80,000. Bar = 0.1 �tm.

Figure 4. Thin section of G. muris cyst wall, comparable morphologically to that seen in Figure 3 but immunostained with rabbit polyclonal antiserum (R-AGLMB)and goat anti-rabbit lgG labeled with 15-nm colloidal gold. Specific staining with immunogold is detected over the filamentous portion of the cyst wall, and nolabeling is seen on the membranous portion ofthe cyst wall (arrowheads) or in the peritrophic space (ps). f, flagellum. Original magnification x 80,000. Bar � 0.1 �tm.

Figure 5. Low-voltage field emission SEM of G. muris cyst taken at 1.5 kV and illustrating the filamentous nature of the cyst wall. Original magnification x 9500.Bar = 1 �tm.

Figure 6. Higher magnification ofthe filamentous cyst wall ofthe G. muris cyst seen in Figure 5. Individual filament populations, ranging from 7-20 nm, are easilydiscerned and appear to form a tightly interwoven mesh. Original magnification x 39,500. Bar = 0.1 �tm.

Figures 7 and 8. SEI (FIgure 7) and BEI (Figure 8) of the filamentous cyst wall of G. muris immunocytochemically labeled with rabbit antiserum to the Giardiacyst wall (R-AGLMB) and goat anti-rabbit IgG coupled to 15-nm colloidal gold. A comparison ofthe SEI and BEI taken at 10 kV by FESEM reveals the one-to-onecorrespondence (open arrows) between the 15-nm immunogold complexes associated with the filamentous cyst wall as seen by surface topography (Figure 7)or by atomic number contrast (Figure 8). In some instances, 15-nm immunogold particles were not obvious using SEI (small solid arrow), but were easily detectedby BEI, as seen in Figure 8. Other particles detected by SEI (arrowheads)were shown not to be immunogold by BEI. Original magnification x 80,000. Bar = 0.05 sm.

by guest on July 10, 2011jhc.sagepub.comDownloaded from

Figures 9 and 10. High-magnification SEI (Figure 9) and BEI (FIgure 10) images of a G. duodenalis cyst immunostained with monoclonal antibody (RD 348)to Giardia cyst wall antigens and goat anti-mouse lgG antibodies coupled to 15-nm colloidal gold and photographed at 10 kV. Colloidal gold molecules can beseen in individual filaments (open arrows) in the cyst wall as round spheres of greater secondary electron emission by topographical contrast (FIgure 9) or bytheir gold content using atomic number contrast with backscatter electron detection (Figure 10). On occasion, topographical features resembling 15-nm colloidalgold particles (small solid arrow) were seen; however, examination ofthe BEI showed thatthey did not contain gold. Original magnification x 232,000. Bar = 0.05 �tm.

630 ERLANDSEN, BEMRICK, SCHUPP, SHIELDS, JARROLL, SAUCH, PAWLEY

�-

fr �:

,�,- .. ..�

i2�

�#{149}#{149}�j....“

ii�.....,:1#{149} .,,;L1i�dJ �... , .,

by guest on July 10, 2011jhc.sagepub.comDownloaded from

HIGH-RESOLUTION SEM GOLD LABELING OF GIARDIA 631

ting SEI detection of 5-20-nm immunogold labels (20), together

with improved single crystal scintillator detectors for BE! (4), may

permit the use of low accelerating voltages (<3 kV) in FESEM for

simultaneous detection of<5-nm gold particles with both SE! and

BE!.

In summary, the development of new FESEM instrumentation

has permitted the accurate localization of immunogold labels on

cyst wall filaments of metal-coated Giardia cysts and trophozoites

with both SE! and BE!. In the future, the detection of immunogold

labels on Giardia or other cell types with either SE! or BE! in FESEM

at low accelerating voltage should offer increased potential for de-

tection of specific surface molecules, since spatial resolution may

be limited only by the size of the immunogold complex and the

technical aspects of specimen preparation.

Acknowledgments

We wish to express ourappreciation to Ms. Susan Spence and Mary Januschka

for expert technica/assistance andto Ms Iveta Dinbergsfor the gift of rab.

bit antiserum to sonicatedG. muris cysts. We thank Ms Miche/e E. Nicko/

ofMeridian Diagnostics, Inc., for theirgenerosity in supplying the mouse

monoc/onafantibody to Giardia anda/so Mr JVa/terJakubowskifor his critical

review ofthe manuscript. The FESEM in the Integrated Microscopy Re-

source at the University oflVisconsin-Madison is a national biotechnology

facility whichprovides advancedmicroscopic instrumentation to qualified

scientists.

Literature Cited

1. Albrecht RM, PrudentJ, Simmons SR. PawleyJ, ChoateJJ: Observa-

tions ofcolloidal gold labelled platelet microtubules: high voltage dcc-

tron microscopy and low voltage-high resolution scanning electron mi-croscopy. Scan Microsc 3:273, 1989

2. Albrecht RM, Simmons SR. PrudentJR, Erickson CM: High-resolution

SEM of colloidal gold labels. EMSA 46:214, 1988

3. Anderson TF: Techniques for the preservation of three-dimensional

structure in preparing specimens for the electron microscope. Trans NYAcad Sci 13:130, 1951

4. Autrata R: Backscattered electron imaging using single crystal scintil-

lator detectors. Scan Microsc 3:739, 1989

5. Barnard RJ, Jackson GJ: Giardia lamb/ia- the transfer of human in-fections by foods. In Erlandsen SL, Meyer EA, eds. Giard:a and giardi-asis. Plenum, New York, 1984, 365

6. Craun GF: Waterborne outbreaks ofgiardiasis: why they happen, how

to prevent them. Health Environ Digest 2:3, 1988

7. Craun GF: Waterborne outbreaks ofgiardiasis. InJakubowski W, Hoff

J C, eds. Waterbome transmission ofgiardiasis. Cincinnati, US Environ-mental Protection Agency (600/7-79-001), 1979, 127

8. de Harven E, Soligo D: Backscatter electron imaging of the colloidal

gold marker on cell surfaces. In Hayat MA, ed. Colloidal gold. Vol 1.

New York, Academic Press, 1989, 230

9. Erlandsen SL, Bemrick WJ, PawleyJ: High-resolution electron micro-

scopic evidence for the filamentous structure of the cyst wall in Giardiamuris and Giardia duodenalis. J Parasitol 75:787, 1989

10. Erlandsen SL, Gould RP, Frethem C, Wells CL, Pawley J, HamiltonDW: Membrane fixation for high-resolution low-voltage SEM: studieson Giardia, rat spermatozoa, and mouse macrophages. Scan 1 1:169, 1989

11. Evans AC, FranksJ: Specimen coating for high resolution scanning dcc-tron microscopy. Scan 4:169, 1981

12. Filice FP: Studies on the cytology and life history of a Giardia from

the laboratory rat. Univ Calif Pub Zool 57:53, 1952

13. Hodges GM, SouthgateJ, Toulson EC: Colloidal gold -a powerful tool

in scanning electron microscope immunocytochemistry: an overview

of bioapplications. Scan Microsc 1:301, 1987

14. Januschka MM, Erlandsen SL, Bemrick WJ, Schupp DG, Feeley DE:

A comparison of Giardia micmti and Spironucleus muris in the vole:

an immunocytochemical, light, and electron microscopic study. JParasitol 74:452, 1988

15. Jarroll EL Manning P, Lindmark DG, CogginsJR, Erlandsen SL: Giardi.a

cyst wall-specific carbohydrate: evidence for the presence of galactosa-

mine. Mol Biochem Parasitol 32:121, 1989

16. Keister DB: Axenic culture of Giardia lamb/ia in TYI-S33 medium

supplemented with bile. Trans R Soc Trop Med Hyg 77:487, 1983

17. Laemmli UK: Cleavage of structural proteins during the assembly of

the head of bacteriophage T4. Nature 227:680, 1970

18. Langford LA, Coggeshall RE: The use ofpotassium ferricyanide in neural

fixation. Anat Rec 197:297, 1980

19. Owen RL: Direct fecal-oral transmission ofgiardiasis. In Erlandsen SL,

Meyer EA, eds. Giardia and giardiasis. Plenum, New York, 1984, 329

20. Pawley J, Albrecht R: Imaging colloidal gold labels in LVSEM. Scan

10: 184, 1988

21. PawleyJB: Low voltage scanning electron microscopy. EMSA Bull 18:61,

1988

22. PawleyJB, Erlandsen SL: The case for low voltage high resolution scan-ning electron microscopy ofbiological samples. In Albrecht RM, Orn-

berg RL, eds. The science of biological specimen preparation formicroscopy and microanalysis 1988. Chicago, Scanning Microscopy In-

ternational, 1989, 163

23. Peters K-R: Working at higher magnifications in scanning electron mi-

croscopy with secondary and backscattered electrons on metal coatedbiological specimens and imaging macromolecular cell membrane struc-

tures. In Muller M, Becker RP, Boyde A, WolosewickJJ, eds. The scienceof biological specimen preparation for microscopy and microanalysis

1985. AMF O’Hare, IL, Scanning Electron Microscopy mc, 1986, 257

24. RiggsJL, Dupuis KW, Nakamura K, Spath DP: Detection of Giardia

lamb/ia by immunofluorescence. AppI Environ Microbiol 45:698, 1983

25. Roberts-Thomson IC, Stevens DP, Mahmoud AAF, Warren KS: Giardi-

asis in the mouse: an animal model. Gastroenterology 71:57, 1976

26. SauchJF: Use of immunofluorescence and phase-contrast microscopy

for detection and identification of Giardia cysts in water samples. AppI

Environ Microbiol 50:1434, 1985

Figure 11. SEI ofthe ventral surface ofa G. duodenalis (MR4 strain)trophozoite 12 hr after stimulation of in vitro encystment. The trophozoite was immunostainedwith monoclonaJ antibody (RD 348) to Giardia cyst wall antigens and goat anti-mouse IgG antibodies coupled to 15-nm colloidal gold. Developing individual fila-ments were detected over the entire surface of the trophozoite, including the ventral attachment disc. Developing cyst wall filaments within the framed area areshown at higher magnification in Figures 12 and ia Original magnification x 8800. Bar � I run.

Figures 12 and 13. Higher magnification of the posterior edge of ventral attachment disc of the immunostained Giardia trophozoite shown in Figure 11, as seenby topographical contrast generated by SEI(Flgur. 12) and atomic number contrast obtained by BEI(Flgure 13). Immunogold particles(15 nm) could be detectedover the newly forming filaments by topographical contrast (FIgure 12), but were more clearly revealed by use of atomic number contrast (Figure 13). The trophozoitemembrane area between the filament initiation sites was essentially devoid of specific immunogold labeling except at sites of new filament generation. f, flagellum.Original magnification x 54.800. Bar - 0.1 �tm.

by guest on July 10, 2011jhc.sagepub.comDownloaded from

632 ERLANDSEN, BEMRICK, SCHUPP, SHIELDS, JARROLL, SAUCH, PAWLEY

32. Studer D, Hermann R: Colloidal gold particles detected on highly struc-

27. Sauch JF: Purification of Giardia muris cysts by velocity sedimenta-

tion. Appl Environ Microbiol 48:454, 1984

28. Schupp DG,Januschka MM, Sherlock LAF, Stibbs HH, MeyerEA, Bem-

rick WJ, Erlandsen SL: Production of viable Giardia cysts in vitro: de-

termination by fluorogenic dye staining, excystation, and animal in-

fectivity in the mouse and mongolian gerbil. Gastroenterology 95:1, 1988

29. Soligo D, de Harven E, Nava MT. Lambertenghi-Deliliers G: Immuno-

cytochemistry with backscattered electrons. In MUller M, Becker RP,

Boyde A, WolosewickJJ, eds. The science of biological specimen prep-aration for microscopy and microanalysis 1985. AMF O’Hare, IL, Scan-

ning Electron Microscopy, Inc, 1986, 289

30. Sorenson 5K, RiggsJL, Dileanis PD, Suk TJ: Isolation and detection

of Giardia cysts form water using direct immunofluorescence. Water

Res Bull 22:843, 1986

31. Sterling CR, Kutob RM, Gizinski MJ, Verastegui M, Stetzenback L:

Giardia detection using monoclonal antibodies recognizing deter-

minants of in vitro derived cysts. In Wallis PM, Hammond BR, eds.

Advances in Gi.ardi.a research. Calgary, University ofCalgary Press, 1988,

219

tured surfaces oflarge samples by backscatter electrons in the scanningelectron microscope. In MOller M, Becker RP, Boyde A, WolosewickJJ, eds. The science of biological specimen preparation for microscopyand microanalysis 1985. AMF O’Hare, IL, Scanning Electron Micros-

copy Inc. 1986, 203

33. Stump RF, Pfeiffer JR. Seagrave J, Oliver JM: Mapping gold-labeledIgE receptors on mast cells by scanning electron microscopy: receptordistributions revealed by silver enhancement, backscatter electron im-

aging, and digital image analysis. J Histochem Cytochem 36:493, 1988

34. Towbin H, Staehelin T, GordonJ: Electrophoretic transfer of proteins

from polyacrylamide gels to nitrocellulose sheets: procedure and ap-plications. Proc Nail Acad Sci USA 76:4350, 1979

35. Walther P. MUller M: Detection ofsmall (5-15 nm) gold labelled sur-

face antigens using backscatter electrons. In Muller M, Becker RP, BoydeA, WolosewickJJ, eds. The science of biological specimen preparationfor microscopy and microanalysis 1985. AMF O’Hare, IL, Scanning Elec-

tron Microscopy Inc. 1986, 195

36. Woodcock-MitchellJ, Eichner R, Nelson WG, Sun T-T: Immunolocali-

zation ofkeratin polypeptides in human epidermis using monoclonalantibodies. J Cell Biol 95:580, 1982

by guest on July 10, 2011jhc.sagepub.comDownloaded from