Immunohistochemical localization of annexin V (CaBP33) in rat organs

1 Supported by MURST and CNR funds (RD).

2 Correspondence to: Prof. R. Donato, Section of Anatomy, Dept. of

Experimental Medicine and Biochemical Sciences, Univ. of Perugia, Cas.

Post. 81, 06100 Perugia Succ. 3, Italy.

1189

0022-1554/91/$3.30Thejournal of Histochemistry and Cytochemistry

Copyright © 1991 by The Histochemical Society, Inc.Vol. 39, No. 9, pp. 1189-1198, 1991

Printedin USA.

Original Article

Immunohistochemical Localization of Annexin V(CaBP33) in Rat Organs’

ILEANA GIAMBANCO, GRAZIA PULA, PAOLO CECCARELLI, ROBERTA BIANCHI,

and ROSARIO DONAlD2

Section ofAnatomy, Department ofE�cperimental Medicine and Biochemical Sciences, University ofPerugia, 06100 Perugia, Italy.

Received for publication November 20, 1990 and in revised form March 28, 1991; accepted April 5, 1991 (0A2166).

We investigated the cellular distribution of annexin V

(CaBP33) in rat tissues by immunohistochemistry. Severalcell types were shown to express the protein. Glial cells in thecerebellum and in the optic nerve, the cornea! epitheium,

the posterior epitheium in the iris, chondrocytes, skeletalmuscle cells and cardiomyocytes, the capillary endotheial

cells in many organs, the muscularis mucosae and the mus-cular layer in the intestinal tract, hepatocytes, Muller cellsin the retina, the lens fibers, Sertoi and Leydig cells in the

testis, and smooth muscle cells in the epididymis and bron-chi displayed intense immunostaining. In the adrenal gland,only the cortex showed inununoreaction product. In the kid-ney, no apparent staining ofrenal cells was observed, whereas

Introduction

Annexin is the name given to a multigene family of intracellular

Ca2� binding proteins unrelated to EF hand proteins, sharing the

ability to bind to acidic phospholipids and natural membranes in

the presence of Ca2� (for reviews see refs. 1-3). Individual annexins

are made of a definite number of repeat units which are about 70

residues long, each of which possesses a highly conserved motif,

the endonexin fold, which is 17 residues long and is thought to

coordinate the binding of both Ca2� and phospholipids (3). In-

dividual annexins differ from each other in a number of phys-

icochemical properties, including the length of their N-terminal

tails, which are believed to play an important role in the physiol-

ogy ofsingle proteins (1-3). Annexins have been implicated in the

regulation of such diverse activities as cell growth, the inflamma-

tory response, and exocytosis and endocytosis (1-3). Evidence in

favor of involvement of annexin II in the regulation of secretory

granule aggregation and fusion has recently been presented (4-6).

Some uncertainty remains as to whether or not annexins modulate

phospholipase A2 activity (7,8). Interaction of annexins with

F-actin, fodrin, and other unidentified cytoskeletal components

endotheial cells of perirubular capillaries were stained. Inthe heart, annexin V was found associated exdusively with

the sarcolemma and intercalated discs, as opposed to the dif-fuse distribution of the protein in skeletal muscle cells. Inthe spleen, only reticular elements in the white pulp andendotheial cells in the red pulp appeared to be immuno-

stained. The present data complement the biochemical workthus far done on annexin V and suggest that the protein isneither restricted to seaetory cells nor exdusively related to

exocytotic events in secretory cells. (JHistochem Cyrochem

39:1189-1198, 1991)

KEY WORDS: Annexin V (CaBP33); Calcium; Immunohistochemis-

try; Rat tissues.

(9-12) renders it plausible that these proteins might take part in

the Ca2�-dependent regulation of the structural organization of

membranes in relation to both exocytosis and endocytosis. Post-

translational modifications (e.g., phosphorylation) ofsome annexins

have been reported (1-3).

In search of novel Ca2� binding proteins in the nervous system,

we have recently purified from bovine brain two acidic proteins

by Ca2�-dependent affinity chromatography on phenyl-Sepharose

(13). These were termed CaBP33 and CaBP37 on the basis of their

Ca2� binding properties and their molecular weights, as estimated

by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-

PAGE) and gel filtration. Both proteins were shown to bind to acidic

phospholipids and to membrane skeleton preparations from bo-

vine brain in the presence ofCa2� and, once bound to brain mem-

branes, to resist extraction with detergents (12,14). Amino acid se-

quence analysis showed that CaBP33 is the brain form of annexin

V, whereas CaBP37, which thus far has been detected in bovine

but not in rat or porcine tissues (14,15), is an isoform of CaBP33

(annexin V) or, alternatively, the product of alternative splicing of

the same gene (manuscript in preparation). Strong immunologi-

cal crossreactivity was observed between the two proteins (14).

Although much biochemical work has been done on annexin

V (16-18) and data on its tissue distribution by an immunochemi-

cal approach have been presented (19), no thorough analysis of its

cellular and subcellular localization has been presented thus far.

As a part of a study aimed at elucidating this issue, we present

by guest on August 14, 2015jhc.sagepub.comDownloaded from

http://jhc.sagepub.com/

.�.. ...

B

... ‘.- *

� .4 ...� ‘-:‘:‘ �

_4�%=�� I. �Z

� � � �: �

M

�0

�

�: .

‘.5

#{182}

,, ..�.

1190 GIAMBANCO, PUIA, CECCARELLI, BlANCH!, DONATO

-� � � �. �

abcdefgh I � a’b’c’d’e’f’g’h’i’j’

Figure 1. Immunochemical detection of annexin V (CaBP33) in rattissues. The

CaBP33/CaBP37 mixture from bovine brain (Lanes a,a!) and Ca�-precipttated,EGTA-solubilized proteins from lung (Lanes b,b’), kidney (Lanss c,c’), liver(Lansd,d’), testis(Lanes.,e’), brain(Lanesf,f’), heart(Lanssg,g’), eye(Larissh,h’), spleen (Lanes 1,1’), and skeletal muscle (Lanes Jj’) were subjected toSDS-PAGE. Gels were either stained with Coomassie blue(Lan.sa-j)ortrans-blotted onto nitrocellulose paper for immunostaining with the anti-annexin V(CaBP33) antiserum (Lane. a’-J’). Note that the rattissues examined appearedto express CaBP33 (annexin V) (arrow). The arrowhead points to the positionof CaBP37, and the asterisk to the position of annexin VI.

here immunohistochemical data on the cellular localization of an-

nexin V (CaBP33) in rat tissues.

Materials and Methods

Immunochemical Analyses. Organs were rapidly excised from decapi-

tated adult Wistar rats (250 g), rinsed in ice-cold saline, cut into small pieces,

and stored at - 20’C until use. On the day of experiments, individual or-

gans (3 g each) were thawed and subjected to the so-called Ca2� precipita-

tion method (20). Briefly, tissues were homogenized in 9 ml of2O mM Tris-

HCI, pH 7.5, 0.15 M NaCI, 5 mM EGTA, 0.25 mM phenylmethylsulfonyl

fluoride as a protease inhibitor (Buffrr A). The homogenates were centrifuged

at 40,000 x g for 20 mm in a Kontron (Centrikon H-405) centrifuge. In-

dividual supernatants were then brought to 6 mM CaCl2 with a 1 M solu-

tion of CaCl2, stirred for 15 mm, and centrifuged as above. Individual

pellets were re-suspended in 6 ml of Buffer A minus EGTA containing 1

mM CaCI2 (Buffer B) and centrifuged again. Individual pellets were re-

suspended in 6 ml of Buffer B minus NaCI and centrifuged. Individual

final pellets were re-suspended in 2 ml of 20 mM Tris-HCI, pH 7.5, 10

mM EGTA, 5 mM 2-mercaptoethanol, and centrifuged at 150,000 x g for

20 mm in a Kontron (Centrikon T-1055) centrifuge. All operations were

done at 4’C. The resultant supernatants (EGTA extracts) were subjected

to SDS-PAGE (10% acrylamide). Gels were stained with Coomassie blue.

In parallel experiments, electrophoretically separated proteins were trans-

blotted onto nitrocellulose paper for immunochemistry (immunoblotting).

linmunohistochemistry. Male Wistar rats(250 g)were anesthesized with

ether and intracardially perfused with 4% (w/v) paraformaldehyde plus

0.1% (w/v) glutaraldehyde in PBS for 30 mm. The organs to be examined

for the presence of annexin V (CaBP33) were rapidly excised and further

fixed by immersion in the above fixative for 6-8 hr at 4’C. The fixed or-

G,� - #{149}�

, - � #{149}:‘;;, � .‘ �:- �: - , .‘ #{149}��;:�.� � �

. E � �:,.‘;;:��.‘�‘ :�‘��‘ ‘ � :

Figure 2. Immunohistochemical localization of annexin V (CaBP33) in rat cerebellum. (A) Intense immunoreactivity is observed in the white matter, in the granulelayer (G), and around Purkinje cell bodies (long arrow). In the molecular layer (M), radial structures (short arrows), probably corresponding to processes of Berg-mann’s glia, are immunostained. Arrowhead points to the pial layer of meninges. (B) Control experiment done with the anti-annexin V (CaBP33) antiserum previ-ously absorbed with the bovine brain CaBP33/CaBP37 mixture. No immunostaining is visible. (C) Purkinje cell bodies (P) are surrounded by processes of glialelements nearby (arrows). (D) Immunostained astrocytes (arrows) in the granule layer send theirprocesses to surround granule cells, which are not immunoreac-tive. (E) Granule cells (G) not immunostained are surrounded by immunostained glial processes. Immune reaction product is seen in the white matter (WM), likelycorresponding to glial cells ensheathing axons. (F) Arrows point to Bergmann’s glial processes in the molecular layer. (G) Cerebellar molecular layer exposedto an anti-PGP 9.5 antiserum. Only the dendritic tree of Purkinje cells is immunostained. Original magnifications: A x 170; B x 65; C-E,G x 675; F x 1125.Bars: A a 40 tim; B - 96 tim; C-E,G = 10 tim; F - 6 tim.



IMMUNOHISTOCHEMISTRY OF ANNEXIN V (CaPB33) IN RAT 1191

gans were then rinsed in PBS over a period of6O mm with frequent changes,

stored in PBS overnight at 4C, and treated with graded alcohols to xylene

before embedding in paraffin. Sections 5 �im thick were finally processed

for immunohistochemistry. Briefly, tissue sections on glass slides were treated

with 1% (v/v) H2O2 in methanol for 15 mm in the dark to quench any

endogenous peroxidase activity, washed in PBS (three times for 5 mm each),and incubated with 3% (w/v) bovine serum albumin (BSA) in PBS for 5-6

hr in a humid chamber. At the end of this step, tissue sections were se-

quentially exposed to an anti-bovine brain annexin V (CaBP33) antiserum

(1:40 in 1% BSA in PBS at 4’C overnight), a sheep anti-rabbit IgG antise-

rum (1:40 in 1% BSA in PBS at room temperature for 1.5 hr), and a rabbit

peroxidase-anti-peroxidase complex (1:100 in 1% BSA in PBS at room tem-

perature for 1.5 hr). Washings between these steps were done in PBS (five

times for 5 mm each). To develop the immune reaction, tissue sections were

exposed to 0.3 mg ofdiaminebenzidine/10 ml ofSO mM Tris-HCI, pH 7.5,

containing 0.02% H2O2 for 20 mm in the dark. Tissue sections were then

mounted in permanent aqueous mounting medium and viewed in a LeitzLaborlux 12 light microscope. In control experiments, the anti-annexin V

(CaBP33) antiserum was either omitted or absorbed with the bovine brain

CaBP3 3/CaBP3 7 mixture before use. Identical results were obtained with

either procedure.

Other Procedures. The CaBP33/CaBP37 mixture was purified from bo-

vine brain as reported (13). The anti-CaBP33/CaBP37 antiserum and the

anti-annexin V (CaBP33) antiserum were raised in rabbits and character-

ized as reported (14). They were used at a dilution of 1:1000 in immuno-

blot analyses. SDS-PAGE was done as described (21) and immunoblotting

as reported (22).

Results

Immunochemistry

The Ca2� precipitation method (20) is a convenient method to ob-

tam the complete set of annexins from a given organ. As we were

interested in annexin V (CaBP33) and had a specific antiserum

at hand, we did not attempt to purify individual annexins from

rat organs. All EGTA extracts obtained as described under Materials

and Methods turned out to contain a 33 KD polypeptide co-

migrating in SDS gels with the lower Mr species (annexin V,

CaBP33) in the bovine brain CaBP33/CaBP37 mixture (Figure 1).

Immunoblots of electrophoretically separated proteins (Figure 1)

showed that all EGTA extracts contained annexin V (CaBP33). The

intensity of the immune reaction product in individual lanes in

Figure 1 was not proportional to the corresponding Coomassie blue-

stained gel bands, probably because of the different proportions

of individual 32-37 KD annexins in single organs. We have shown

that our anti-annexin V (CaPB33) antiserum does not recognize

other annexins (14,15) and that rat brain does not express CaBP37

(14). The present data indicate that none of the rat tissues exam-

med express CaBP37. When the anti-bovine brain CaBP33/CaBP37

mixture antiserum was used, some immunoreactivity was observed

with EGTA extracts from rat lung, liver, and kidney at the level

of a 53 KD doublet (not shown). Owing to this, the immuno-

histochemical analyses to be described were done with the anti-

annexin V (CaBP33) antiserum.

Immunohistochemistry

Nervous system. In rat cerebellum, immune reaction product

was seen around the Purkinje cell bodies, in radial structures in

the molecular layer, in the white matter, and in pial elements (Fig-

ure 2A). Strong immunoreactivity was observed in the granule layer

at the periphery ofcells therein (Figure 2A). At higher magnifica-

tions, it was possible to establish that the immunoreactivity around

Purkinje cell bodies was due to staining of glial cell bodies and

processes (Figure 2C). The latter appeared to be responsible for

the immunostaining of the granule layer as well (Figures 2D and

2E). Figure 2G shows the staining pattern of the molecular layer

in sections exposed to an antiserum raised against the neuron-

specific protein gene product 9.5 (PGP 9.5) (23), whereas Figure

2F shows the same cerebellar layer treated with the anti-annexin

V (CaBP33) antiserum. It is clearly evident that PGP 9.5 is found

in the dendritic tree ofPurkinje cells, whereas annexin V (CaBP33)

is restricted to fine cell processes belonging to the Bergmann’s glia.

In control experiments (Figure 2B) no immunostaining was ob-

served.

Eye. Several eye structures displayed annexin V (CaBP33) im-

mune reaction product (Figure 3). Intense staining was seen in basal

elements of the cornea) epithelium (Figure 3A), the posterior epi-

thelium of the iris (Figure 3C), the striated fibers of eye muscles

(Figure 3E), and the lens fibers (Figures 3G and 31). Less intense

staining was observed in cells of the corneal endothelium (Figure

3A) and in elements of the iris stroma (Figure 3C). A certain per-

centage of striated muscle fibers were poorly stained (Figure 3C).

At present we have no explanation for this. Faint staining was ob-

served in the lacrimal gland, which appeared to be restricted to

myoepithelial cells at the periphery of acinar cells (Figure 3J). In

the retina, neurons were completely devoid of immune reaction

product (Figures 3L, 3N, and 30), whereas the poles and, less in-

tensely, the cell bodies of MUller cells displayed the presence of

the protein (Figures 3L, 3N, and 30). Annexin V (CaBP33) was

found in the optic nerve (Figure 3P), where it was associated with

scattered glial cells and with the periphery of axons. No immune

reaction product was found in control experiments (Figures 3B, 3D,

3F, 3H, 3K, 3M, and 3Q).

Heart. In the heart, strong annexin V (CaBP33) immunoreac-

tivity was detected on intercalated discs and on the sarcolemma

(Figures 4A and 4B). The sarcoplasm and the contractile elements

appeared devoid of immune reaction product. The capillary en-

dothelial cells also appeared immunostained. Control sections were

negative (Figure 4C).

Respiratory System. In the trachea, annexin V (CaBP33) im-

mune reaction product was found in chondrocytes of the tracheal

cartilage with no staining of other cellular elements (Figure 5A).

In the lung (Figure SC), elements of the respiratory epithelium

and smooth muscles of respiratory bronchi were stained. By light

microscopy, no evidence for the presence of annexin V (CaBP33)

in pneumocytes could be obtained. Control sections were negative

(Figures SB and SD).

Digestive Tract. The stroma of intestinal villi and that around

the crypts ofthe jejunum appeared intensely stained, whereas epi-

thelial and glandular cells were devoid of annexin V (CaBP33)

(Figures 6A and 6C). Most of the immunoreactivity was associated

with the muscularis mucosae. The muscular layer also displayed

annexin V (CaBP33) immune reaction product (Figures 6A and

6C). Controls were negative (Figures 6B and 6D). In the liver, he-



‘- � ‘.

. .,�

�C

�. : ;

F -�D

H-

�T��#{149}:�#{149}

I-� � ______________

�i�=_�- :�

I � �

j.�7s. K

A

#{149}�: � � � ., H.�.:

- �\‘ � � � �: � � � � _-� �-_

� .i�i-!a�� .,�j �- *� #{149}1

�. �

-,� : � ��j#{231}�k �

�t’ � � �

‘,� � 4’P ‘�

N - 0

1192

I �: ��#).

� �1�

�i

. .

. . cli

GIAMBANCO, PULA, CECCARELLI, BIANCHI, DONATO

Figure a Immunohistochemical localization of annexin V (CaBP33) in rat orbital structures. The structures examined were (A,B) the cornea, (CD) iris, (E,F) orbitalskeletal muscles, (G-I) lens, (J,K) lacrimal gland, (L-O) retina, and (P0) optic nerve. Control experiments (B,D,F,H,K,M,Q), done as described in the legendto Figure 2, are negative. (A) Intense immunostaining is observed in the corneal epithelium, particularly in basal elements. The corneal endothelium (arrow) isless intensely stained(A). No immunostaining is visible in the substantia propria(A). (C)The posterior epithelium ofthe iris is intensely immunostained. Scattered



�s � ,� .

‘�b ‘.� � � -I

;�. ,, �

�

�; �

.-D

.�. - B

Figure 5. Immunohistochemical localization of annexin V (CaBP33) in rat re-spiratory system. (A) Trachea: chondrocytes appear to be the only immuno-stained elements. The immune reaction product seems to be associated withplasma membranes in chondrocytes. By the present experimental approachit is not possible to accertain whether or not the cytoplasm of chondrocytesalso is immunostained. (C) Lung: elements of the respiratory epithelium andsmooth muscle cells (arrows) display immune reaction product. Pneumocytesare unstained. (CD) Control experiments done on trachea (B) and lung (D) asdescribed in the legend to Figure 2 do not show immune reaction product. Origi-nal magnification x 170. Bars = 40 tim.


C ..

Figure 4. Immunohistochemical localization ofannexin V(CaBP33) in rat heart.(A) Intense immunostaining is observed on intercalated discs. (B)At higher mag-nification, it can be seen that not only intercalated discs (arrowhead), but alsothe sarcolemma (short arrow) and capillary endothelial cells (long arrow) areimmunostained. (C) Control experiments done as described in the legend toFigure 2 are negative. Original magnifications: A,C x 170; B x 675. Bars: A,C

= 40 tim; B = 10 tim.

-. w�

�*�“�Th-)c_wf, �i’� , . �..

�- ---�

4�.

)

�-6

patocytes showed immune reaction product, whereas endothelial

cells ofsinusoidal capillaries did not appear to be stained (Figures

7A and 7B). In hepatocytes the immunostaining was diffuse in the

cytoplasm. At the light microscopic level it was not possible to ascer-

tan whether or not plasma membranes in these cells also were posi-

tive. Control sections were negative (Figure 7).

Kidney. No staining was observed within renal cells in either

cortex or medulla (Figures 8A and 8C). Instead, endothelial cells

ofcapillaries in both regions were intensely stained with the anti-

annexin V (CaBP33) antiserum (Figures 8A and 8C). By the pres-

ent experimental approach it was not possible to ascertain whether

or not basal membranes of tubular cells were immunostained. Con-

trol sections were unstained (Figure 8B).

Adrenal Gland. At low magnification (Figure 9A), only the

adrenal cortex displayed annexin V (CaBP33) immune reaction

etements in the stroma also are stained (C). (E) Diffuse immunostaining of some but not all fibers of orbital skeletal muscles is observable. (G) The lens fibersand, to some extent, the periphery of cell bodies of lens cells (arrows) giving rise to fibers are immunostained. (I) Most of the immune reaction product seemsto be associated with the plasma membranes of lens fibers. (J) Faint immunostaining is seen in myoepithelial cells of the lacrimal gland (arrows). The acinarcells appear largely devoid of immune reaction product (J). (L) The external limiting membrane (arrowhead), cell bodies of Muller cells (arrows), and the internallimiting membrane of the retina are immunostained, whereas retinal neurons are not immunoreactive. (N) The processes of retinal Muller cells that constitutethe internal limiting membrane are intensely stained. (0) Cell bodies of MUller cells (long arrows) and the internal limiting membrane (short arrows) in the retinaare immunostained. (P) Immune reaction product is seen in cellular (glial) elements and in their processes in the optic nerve. Original magnifications: A,B,E-Gx 280; CD x 525; H,I x 450; J,O x 675; K x 270; L,M x 300; N x 1680; RO x 170. Bars: A,B,E-G - 25 tim; C,D - 13 tim; H,I = 15 tim; JO - 10 tim;K = 24 tim; L,M - 23 tim; N = 4 tim; P�Q - 40 tim.



.:‘: �.i.1 �� 1

::�

�‘ ‘1‘ . �‘,, (�. �: �

;‘�tj�� (!�..-‘�#{149}/‘� �

, I

��i�*%1L

A

r � � �

�� \� � � (� I

� � .. �

� � I� ,:.:q�r� � �

� 4�1fr��H 1

� D -

Figure 6. Immunohistochemical localization of annexin V (CaBP33) in rat jeju-num. (A) Smooth muscle elements in intestinal villi (arrow) and around crypts,belonging to the muscularis mucosae, and the muscular layer are immuno-stained. The intestinal epithelium is unstained. (C) Immunostaining is restrictedto the muscular layer, the muscularis mucosae, and elements, possibly glialcells, of the myoenteric plexus (arrow). (B,D) Control experiments done as de-scribed in the legend to Figure 2 do not show immune reaction product. Origi-nal magnifications: A,B x 65; C x 270; D x 600. Bars: A,B = 96 tim; C =

24 tim; D - 10 tim.

product. At a higher magnification (Figures 9C and 9D), it was

clear that the zona fasciculata and the zona reticularis were intensely

stained, whereas the zona glomerulosa was less intensely stained.

Chromaffin cells of the adrenal medulla were unstained (Figures

9A and 9D). Control sections were unstained (Figure 9B).

Testis and Epididymis. In the testis, Sertoli cells appeared to

display annexin V (CaBP33) immunoreactivity (Figures 1OA and

bC). In fact, the outermost cell layers in seminiferous tubules were

not immunostained (Figure bA), except for some scattered struc-

tures probably corresponding to the outer processes ofSertoli cells

(Figures bA and bC). Interstitial (Leydig) cells also appeared to

be immunostained (Figure 1OD). In the epididymis, epithelial cells

of ductuli efferentes and mature spermatozoa appeared devoid of

immune reaction product, whereas smooth muscle fibers at the pe-

riphery of ductuli displayed annexin V (CaBP33) immunoreactivity

(Figure 1OF). Control sections were unstained (Figures lOB and bE).

� ,�

..... 4___. � � . �

� .. . ‘- .,�‘.. ‘�:-� 1”

.‘5eAt- .�

I,. � � � � � � :. � � � -k: � ,� .‘ - � ?� .:‘ ‘ � � <)..

� � 1�.. � � .�

� ��jq�4i�kt, � ::.�i� �%L

tv �.. � � #{149}#{149}

� ...t,’t .4�,.: �

� � �:: �‘�i

C

Figure 7. Immunohistochemical localization of annexin V (CaBP33) in rat liver.(A,B) Hepatocytes display immune reaction product. (C) Control experimentsdone as described in the legend to Figure 2 do not show immune reaction prod-uct. Original magnifications: A x 65; B x 270; C x 170. Bars: A = 96 tim;B = 24 tim; C = 40 tim.

lbB) and capillary endothelial cells in the red pulp (Figure liD)

were immunostained. Controls were unstained (Figures 11C and

liE).

Table 1 summarizes semiquantitatively the results illustrated in

the figures.

Discussion

In the present work we have examined the distribution of annexin

V (CaBP33) in rat tissues by an immunochemical and an immuno-

histochemical approach. The bulk of data indicate that annexin

V (CaBP33) has a widespread distribution, in accordance with previ-

ous observations (19), yet it is not ubiquitous. In the cerebellum,

the overall picture is suggestive of a glial localization of the pro-

tein, in that Purkinje cells and neurons in the molecular layer are

not stained, whereas the white matter and radial structures in the

molecular layer probably corresponding to processes of Bergmann’s

glia are intensely stained. The presence of annexin V (CaBP33) in

Bergmann’s glial processes has been confirmed by immunoelec-

tron microscopy (submitted for publication). In the granule layer,

intense staining is observed in glial elements in the proximity of

Purkinje cell bodies and in cytoplasmic processes around granuleSpleen. Reticular elements in the white pulp (Figures hA and




-“1 �

� � .�

.,

,t�.1

� .

,,I. � .�

..

B il:’ti

Figure a Immunohistochemical localization of annexin V (CaBP33) in rat kidney. Cellular elements outside glomeruli and tubular cells, probably belonging toendothelial cells, are immunostained in both cortex (A) and medulla (C). Control experiments done as described in the legend to Figure 2 do not show immunereaction product (B). Original magnification x 25. Bars - 40 tim.

B

c; �

:�

.‘:P� � � �

� . ‘ .

� i�i

D

Figure 9. Immunohistochemical localization ofannexin V (CaBP33) in rat adre-nal gland. (A,C,D) Cells in the zona fasciculata (F) and zona reticularis (A) ofthe adrenal cortex (C) display immune reaction product. Cells in the zonaglomerularis (G) show a less intense immunoreactivity (C). Chromaffin cellsin the adrenal medulla (M) are unstained (A,C,D). (B) In control experimentsdone as described in the legend to Figure 2, no immune reaction product couldbe seen. Original magnifications: A x 65; B x 170; C,D x 675. Bars: A =

96 tim; B - 40 tim; C,D = 10 tim.


(:�,�J�t: �

“�3I � �‘, � .�.

k1 � t� \� �

cells. A glial localization of annexin V (CaBP33) is also suggested

by the immunostaining of the cytoplasm of cellular elements in

the optic nerve and of Muller cells in the retina, with unstained

retinal neurons. On the other hand, a glial localization of annexin

V in pig brain has been recently suggested (24). Preliminary data

__..---_1�-.. � . � �-� 5��=�% -

� � �- �.

�a- :: � . :‘ ;�

- �- . �L�z: . . � . ,

. .. .� -�‘.-‘%a:::--. �- . .--.-

--.�‘- �- ‘ . a � � �‘.-

� S �::-‘..�.�-- �

‘_: c:�:T� �:

at both the light and the electron microscopic level indicate that

rat C6 glioma cells express annexin V (CaBP33)(manuscript in prep-

aration). Pial elements in the cerebellum also appear to express

the protein. Annexin II also has been reported to be localized to

glial cells in the brain (25), whereas annexin I is not expressed in

normal glial cells and is expressed in reactive astrocytes (26). The

possibility that our antiserum detects annexin II in glial cells can

be reasonably excluded, since it does not crossreact with polypep-

tides in Ca2�-precipitated, EGTA-solubilized proteins from bovine

brain that have not been retained by DEAE resins (not shown),

and since it does not bind to enterocytes (Figure 6A), which are

a rich source ofannexin II (27). In addition, we can exclude cross-

reaction between our antiserum and annexin VI on the basis of

immunoblot data (Figure 1) (15) and since this protein in cerebel-

lum is restricted to Purkinje cells (24).

In the eye, many other cell types in addition to retinal MUller

cells, including cells in the corneal epithelium and, to a lesser cx-

tent, in the cornea) endothelium, cells in the posterior epithelium

of the iris, the lens fibers, and striated muscle cells, are immuno-

stained. The lacrimal gland shows a faint staining, which is mostly

- evident in myoepithelial cells. The latter cell type in mammary gland

has been reported not to harbor annexins II and IV (28).

Intercalated discs and the sarcolemma, as well as capillary en-

dothelial cells, in the heart display an intense immunoreaction.

These localizations ofannexin V (CaBP33) in the heart strikingly

contrast with the diffuse localization of the protein in skeletal muscle

fibers. At present we have no explanation for these data.

Particularly evident is the immunostaining of chondrocytes in

the trachea. This result was somewhat expected given the strong

structural homology between annexin V and anchorin CII (29,30).

Anchorin CII is a protein purified from chondrocyte membranes

(29), displaying 82% sequence homology with the first, third, and

fourth repeat in annexin V (endonexin II), and only 8% sequence

homology with the second repeat in the same protein (30). By the

present experimental approach it is not possible to decide whether

the annexin V (CaBP33) immunostaining in tracheal chondrocytes

is restricted to the plasma membrane, as reported from annexin

V (anchorin CII) (29), or whether the protein is also present in the

cytoplasm. Immunoelectron microscopy will be used to discriminate

between the two possibilities. This seems an important issue, since

annexin V (anchorin CII) has been proposed to be embedded in

the extracellular face of chondrocyte membranes and to mediate



1196 GIAMBANCO, PUIA, CECCARELLI, BIANCHI, DONATO

I,

B

p

I

I. �

- � -k’z��..4,

I � a

,�y’ � -

/ I �.

� �L/ � -

E - F

� -

‘ � � .,�

, � �

, ,. .

� 1

..-,l

11’ -

.;

� i�.�;/l �

Figure 10. Immunohistochemical localiza-tion of annexin V (CaBP33) in rat testis andepididymis. (AC) Testis: spermatozoa andtheir precursors are not immunoreactive.The immune reaction product is found in theperipheral (arrows in C) and proximalprocesses of Sertoli cells (AC), and in in-terstitial(Leydig)cells(arrows in D). (B,E,F)Epididymis: the immune reaction productis restricted to smooth muscle cells at theperiphery of ductuli efferentes (F). Controlsections of seminiferous tubules (B) and in-terstitial cells (E) are unstained. Control ex-periments were done as described in thelegend to Figure 2. Original magnifications:A,B x 170; C,F x 675; D,E x 1125. Bars:A,B = 40 tim; C,F = 10 tim; D,E = 6 tim.

�. �_,. P.,: .-:�

,,.- ,--,� !-

.� . ,l-t�_ �

A�

I � � : � ::

- � �-

. -� � ‘. �

�, �: ‘:. � ‘r: � � �.

�(�/ . . ‘. 1.

�

,:‘��7.�’.” ‘

7� � -) . . �

4- 4 � ,‘q’. �‘ �,, . ‘,

�. , � #{149}. . .�,‘,‘.1- j,,,.

. � 0� � . , � . � -� 4,

. ‘I � ‘#{149}‘��I�� - E

Figure 11. Immunohistochemical localiza-tion ofannexin V(CaBP33)in ratspleen. Re-ticular elements in the white pulp (A,B) andendothelial cells in the red pulp (D) are im-munostained. Control sections of the white(C) and red (E) pulp are unstained. Controlexperiments were done as described in thelegendto Figure 2. Arrows in A and D pointto reticular elements and endothelial ele-ments in the white and red pulp, respec-

- tively. Original magnifications: A,C-E x 170;B x 675.Bars:A,C-E = 4Otim;B = lOtim.



++++

++++

++

Epididymis

Epithelial cells

Digestive tract

Small intestine

Epithelium

Liver

Hepatocytes

Endothelium

Kidney

GlomeruliTubular cellsPeritubular endothelium +

Lens

Lacrimal gland

Secretory cells

Myoepithelial cells

Muscles

Cardiac

Skeletal

Smooth

Respiratory tract

Tracheal cartilageBronchiolar epithlium

Alveolar epithelium

Testis

Sertoli cells

SpermatozoaLeydig cells

±

++

++


Table 1 . Distribution ofannexin V (CaBP33) in normal tissues

Nervous system

Astrocytes

Oligodendrocytes

Schwann cells

Ocular structures

Corneal epithelium

Corneal endotheliumIris ++++

++

SpleenWhite pulp

Reticular elementsRed pulp

Endothelial cells

Adrenal gland

Cortex

Zona glomerulosa

Zona fasciculata

Zona reticularis

Medulla

cell adhesion to extracellular matrix collagen (29,30). In the lung,

elements of the respiratory epithelium �.nd smooth muscle cells

in the respiratory bronchioli display immune reaction product,

whereas pneumocytes do not appear to be stained. To some extent,

this distribution of annexin V (CaBP33) in lung parallels that of

annexin I (30). We can exclude crossreactivity between our antise-

rum and annexin I, since the antiserum does not bind to Ca2�-

precipitated, EGTA-solubilized proteins from rat lung that have

not bound to DEAE resins or have bound to these resins but have

been detached by up to 0.1 M NaCI (not shown). In addition, our

antiserum does not bind to structures within glomeruli or to col-

lecting tubules in the kidney (Figures 8A and 8C), which instead

appear to express large amounts of annexin I (31).

Hepatocytes contain annexin V (CaBP33) in the cytoplasm. The

muscularis mucosae and the muscular layer in the jejunum are

stained, whereas epithelial and glandular cells are not.

In the kidney, by the present experimental approach only en-

dothelial cells of capillaries appear to be stained.

The zona fasciculata, the zona reticularis and, to a lesser ex-

tent, the zona glomerularis of the adrenal gland are stained. No

immunoreactivity is observed in cell elements in the adrenal

medulla. Our results on the adrenocortical zones producing ste-

roid hormones are in agreement with the presence of annexin V

immunoreactivity in luteal cells in the ovary (19) and in Leydig cells

in the testis (Figure 1OD). The lack ofimmunostaining ofthe adrenal

medulla is in accordance with the apparent absence of annexin V

(CaBP33) from neurons in the nervous system. Instead, annexin

Iv has been reported to be present in the adrenal medulla and

absent from the adrenal cortex (32).

In the testis, the overall picture is suggestive of the presence

ofannexin V (CaBP33) in Sertoli and Leydig cells, whereas in the

epididymis only smooth muscle cells appear to contain the protein.

Finally, in the spleen reticular elements in the white pulp - a

finding in accordance with the previous data (19)- and endothelial

cells in the red pulp show some immune reaction product.

It therefore appears that cells of both neuroectodermal and non-

neuroectodermal origin express annexin V(CaBP33). Certainly, dcc-

+ + + tron microscopic analyses ofthe subcellular localization ofthis pro-- tein in cells shown to express it would better help obtain information

on its function(s). As a Ca2�-dependent phospholipid and mem-

- brane binding protein, annexin V (CaBP33) is expected to be in-

- volved in Ca2�-regulated events at the membrane level. The pres-

+ + ent findings indicate that, with a few exceptions, the protein isabundant in the cytoplasm, suggesting that it might be involved

in the regulation of still unknown activities as a cytoplasmic pro-

4. tein and/or membrane-associated activities on Ca2�-dependenttranslocation to membranes. Our findings also indicate that the

+ protein is present in both secretory and non-secretory cells, andthat chromaffin cells in the adrenal medulla, and glandular cells

in the lacrimal gland and in the intestinal crypts, do not seem to

± express it. Therefore, annexin V (CaBP33) seems not to be a candi-+ + + date for a role in the regulation of exocytosis. On the other hand,+ + + its definite localization on sarcolemma and intercalated discs in

- cardiomyocytes and on the plasma membrane oflens fibers strongly

suggests that the protein might be involved in the regulation of

membrane-associated activities. In support of this suggestion are

our recent findings on Ca2�-inducible binding of annexin V

(CaBP33) to brain and heart membranes, as well as its distribution

in two distinct membrane pools, one EGTA-resistant and Triton

X-l00-extractable, and one Triton X-l00-resistant and EGTA-

extractable (12,14,15).

Literature Cited1. Klee CB. Ca2�-dependent phospholipid- (and membrane-) binding

proteins. Biochemistry 1988;27:6645

2. Crompton MR, Moss SE, Crumpton MJ. Diversity in the lipocor-

tin/calpactin family. Cell 1988;55:l

3. Burgoyne RD, Geisow MJ. The annexin family ofcalcium-binding pro-teins. Cell Calcium 1989;10:1

4. Drust DS, Creutz CE. Aggregation of chromaffin granules by calpac-tin at micromolar levels of calcium. Nature i988;331:88

5. Ali SM, Geisow MJ, Burgoyne RD. A role for calpactin in calcium-dependent exocytosis in adrenal chromaffin cells. Nature 1989;340:313

6. Nakata T, Sobue K, Hirokawa N. Conformational change and local-ization ofcalpactin Icomplex involved in exocytosis as revealed by quick-freeze, deep-etch electron microscopy and immunocytochemistry. 3 Cell

Biol 1990;110:13

7. Huang K, Wallner BP, Mattaliano RJ, Tizard R, Burne C, Frey A, Hes-sion C, McGray P, Sinclair LK, Chow EP, BrowingJL, RamachandranKL, ‘lang J, Smart JE, Pepinsky RB. Two human 35 kd inhibitors ofphospholipase A2 are related to substrates of pp6ov/src and the epider-mal growth factor receptor/kinase. Cell 1986;46:19i

8. Davidson FF, Dennis EA, Powell M, GlenneyJR. Inhibition of phos-pholipase A2 by “lipocortins” and calpactins. J Biol Chem 1987;262:1698

9. Gerke V, Weber K. Identity of p36K phosphorylated upon Rous sar-coma virus transformation with a protein purified from brush borders;calcium-binding to non-erythroid spectrin and actin. EMBO J 1984;

3:227




10. Edwards HC, Booth AG. Calcium-sensitive, lipid-binding cytoskeletal

proteins ofhuman placental microvilla region.J Cell Biol 1987;105:303

ii. Zokas L, Glenney)R. The calpactin light chain is tightly linked to the

cytoskeletal form of calpactin I: studies using monoclonal antibodies

to calpactin subunits. 3 Cell Biol 1987;105:2i11

12. Giambanco I, Pula G, Bianchi R, Donato R. Interaction of two brain

annexins, CaBP33 and CaBP37, with membrane-skeleton proteins. FEBSLets 1990;267:i71

13. Donato R, Giambanco I, Aisa MC, Ceccaelli P, Di Geronimo G. Iden-

tification oftwo novel acidic, calcium-binding proteins in bovine brain.Cell Biol Int Rep 1988;12:565

14. Donato R, Giambanco I, Pula G, Bianchi R. Two novel brain proteins,

CaBP33 and CaBP37, are calcium-dependent phospholipid- andmembrane-binding proteins. FEBS Lett 1990;262:72

15. Pula G, Bianchi R, Ceccarelli P. Giambanco I, Donato R. Character-ization ofmammalian heat annexins with special reference to CaBP33

(annexin V). FEBS Lett l990;277:53

16. Haigler HT, Schlaepfer DD, Burgess WH. Characterization of lipocor-tin I and an immunologically unrelated 33-Wa protein as epidermal

growth factor receptor/kinase substrates and phospholipase A2 inhib-

itors. J Biol Chem 1987;262:6921

17. Funakoshi T, Heimak RI, Hendrickson LE, McMullen BA, FujikawaK. Human placental anticoagulant protein: isolation and character-ization. Biochemistry 1987;26:5572

18. Pepinsky RB, Tizard R, Mattaliano R), Sinclair IX, Miller GT, Brown-

ing)L, Chow EP, Burne C, Huang K-S, Pratt D, Wachter L, HessionC, Frey A, Wallner BP. Five distinct calcium and phospholipid bind-ing proteins share homology with lipocortin I. J Biol Chem 1988;

263:10799

19. Kaetzel MA, Hazarika P, Dedman JR. Differential tissue expression

ofthree 35-kDa annexin calcium-dependent phospholipid-binding pro-

teins. J Biol Chem 1989;264:14463

20. Boustead C, WalkerJH, Geisow MJ. Isolation and characterization oftwo novel calcium-dependent phospholipid binding proteins from bo-

vine lung. FEBS Lett 1988;233:233

21. Laemmli UK. Cleavage of structural proteins during assembly of the

head of bacteriophage T4. Nature 1970;227:680

22. Towbin H, Staehelin T, Gordon I. Electrophoretic transfer of proteins

from polyacrylamide gels to nitrocellulose sheets: procedure and someapplications. Proc Nail Acad Sci USA 1979;76:4350

23. Thompson RJ, Doran JF, Jackson P, Dhillon AP, Rode J. PGP 9.5, anew maker for vertebrate neurons and neuroendocrine cells. Brain Resi983;278;224

24. WoolgaJA, Boustead CM, WalkerJH. Characterization of annexins

in mammalian brain. J Neurochem 1990;54:62

25. Burgoyne RB, Cambray-Deakin MA, Norman K-M. Developmental

regulation of tyrosine kinase substrate p36 (calpactin heavy chain) inrat cerebellum. J Mol Neurosci 1989;1:47

26. Johnson MD, Kamso-PrattJM, Whetsell WO, Pepinsky RB. Lipocortin-1immunoreactivity in the normal human central nervous system andlesions with astrocytosis. Am J Clin Pathol 1989;92:424

27. Gerke V, Weber K. Calcium dependent conformational changes in the36-kDa subunit of intestinal protein I related to the cellular 36-kDatarget ofRous sarcoma virus tyrosine kinase.J Biol Chem 1985;260:1688

28. LozanoJL, Silberstein GB, Hwang S-I, Haindl AH, Rocha V. Develop-

mental regulation of calcium-binding proteins (calelectrins and cal-pactin I) in mammary gland. J Cell Physiol 1989;138:503

29. Mollenhauer’.J, BeejA, Lizarbe MA, Von Der Mark K. Role of ancho-

rin CII, a 31,000-mol-wt membrane protein, in the interaction of chon-

drocytes with type II collagen. J Cell Biol 1984;98:1572

30. Haigler HT, FitchJM, JonesJM, Schlaepefer DD. Two lipocortin-likeproteins, endonexin II and anchorin CII, may be alternate splices of

the same gene. Trends Biol Sci 1989;14:48

31. Fava RA, McKannaJ, Cohen S. Lipocortin I (p35) is abundant in a

restricted number of differentiated cell types in adult organs. J CellPhysiol 1989;141:284

32. WalkerJH, Obrocki J, Sudhof it. Calelectrin, a calcium-dependent

membrane-binding protein associated with secretory granules in Torpedocholinergic electromotor nerve endings and rat adrenal medulla. J Neu-rochem 1983;41:139



Immunohistochemical localization of annexin V (CaBP33) in rat organs

Documents