the clara cell - thorax · clara cells are numerous andproject into the lumen. some clara cells...

Thorax (1974), 29, 147.

The Clara cellP. SMITH, D. HEATH, and H. MOOSAVI

The Department of Pathology, University of Liverpool

Smith, P., Heath, D., and Moosavi, H. (1974). Thorax, 29, 147-163. The Clara cell. Theultrastructure of the bronchiolar Clara cell was examined in normal adult and neonatalrats, in similar animals which had been acutely exposed to hypoxia, and in adult ratsfollowing repeated doses of the anorexigen chlorphentermine. The Clara cell has all thefeatures of a secretory cell, the product of secretion accumulating within smoothcisternae at the apex of the cell. The apical region is then extruded into the bronchiolarlumen in a process of apocrine secretion. Acute hypoxia accelerates this secretion inadult rats but has little effect upon neonatal rats. Administration of chlorphentermineinduces a hyperplasia of Clara cells which is associated with large quantities ofphospholipid free within the alveolar spaces and in macrophages. The Clara cells alsocontain accumulations of what appear to be phospholipid. This suggests that the Claracell secretes the phospholipid pulmonary surfactant.

The bronchial tree has functions additional to theconduction of air into the alveoli. As well as thefamiliar ciliated cells the bronchial epitheliumincludes goblet cells, Feyrter cells, which we havedescribed in a previous paper (Moosavi, Smith,and Heath, 1973), and Clara cells. Unlike gobletand Feyrter cells, which are located mostly in thebronchi, Clara cells are more numerous in theterminal bronchioles. Kolliker in 1881 noted themfirst but Clara (1937), after whom the cells arenamed, described them in detail (Azzopardi andThurlbeck, 1969). Their ultrastructure was firstdescribed in mice by Karrer (1956) and in rabbitsby Kisch (1958). Many Clara cells have apicalclub-shaped cytoplasmic processes (von Hayek,1962; Azzopardi and Thurlbeck, 1969; Corrin,Etherton, and Conning, 1973) which most workersinterpret as indicating a secretory function. VonHayek (1962) thought the secretion to be mucusbut Niden (1967) suggested it might be pulmonarysurfactant. This latter idea has been subsequentlysupported by other workers (Etherton and Con-ning, 1971; Corrin, Etherton, and Conning, 1973;Etherton, Conning, and Corrin, 1973).We have studied the ultrastructure of Clara

cells in rats of different ages and in bronchiolesof different calibre and the changes induced byexposure to acute hypoxia. If Clara cells areresponsible for the synthesis of pulmonary sur-factant, one might expect alterations in theirstructure and frequency in conditions involvingexcessive production of pulmonary phospholipid.

Accordingly we examined the Clara cells fromrats given daily injections of chlorphentermine. atreatment which has been shown to produce largequantities of phospholipid in the alveolar spaces(Heath, Smith, and Hasleton, 1973; Smith, Heath,and Hasleton, 1973).

MATERIALS AND METHODS

The bronchioles from five groups of rats wereexamined (Table). The first group consisted of 10adult albino rats, of which six were females of theWistar strain and four were males of the SpragueDawley strain. The second group comprised 12neonatal Wistar rats which were 4 days old. Noexperimental procedures were performed on these twogroups.The third group comprised four adult female Wistar

TABLEAGE, STRAIN, AND TREATMENT OF RATS IN THE FIVE

GROUPS

No. ofGroup Strain Age Rats Treatment

1 Wistar Adult 6 Control1 Sprague Adult 4 Control

Dawley2 Wistar Neonate 12 Control3 Wistar Adult 4 Subatmospheric

pressure of265 mmHg for12hr

4 Wistar Neonate 14 Subatmosphericpressure of380 mmHg for24 hr

5 Sprague Adult 20 Daily injections ofDawley chlorphentermine

Ii1

147

on August 27, 2020 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.29.2.147 on 1 M

arch 1974. Dow

nloaded from

http://thorax.bmj.com/

P. Smith, D. Heath, and H. Moosavi

rats which were kept at a subatmospheric pressure of265 mmHg for 12 hours. This pressure simulated analtitude of approximately 29,000 ft (8,840 m) whichwas roughly equivalent to the summit of MountEverest. The fourth group consisted of 14 neonatalWistar rats of 4 days of age. These rats were kept for24 hours at a pressure of 380 mmHg, simulating analtitude of 18,000 ft (5,500 m).The conditions of hypoxia were achieved by placing

the rats in a decompression chamber. Air wasremoved from the chamber with a vacuum pump whilea spring-loaded inlet valve maintained the desiredpressure within the chamber as well as replenishingthe air supply. The mothers of the neonatal ratswere also placed in the chamber to suckle their off-spring but they were not used further in theexperiments.At the termination of each experiment the rats

were killed with pentobarbitone sodium injectedintraperitoneally. The thoracic cavity was opened andthe thoracic organs were removed. Pieces of lungwere excised and cut with a razor blade into smallfragments measuring approximately 1 mm across. Inthe neonatal rats the tissue was taken from the hilumof the lung to maximize the chances of includinglarge airways in the sections. The fragments of lungwere transferred to bottles of ice-cold bufferedglutaraldehyde (6 25 %) and allowed to fix for 24 hoursat 2-40C. They were then post-fixed in aqueous 1 %osmium tetroxide, stained with alcoholic uranylacetate, and embedded in Araldite. Sections 1 ,mthick were cut from the blocks and stained withtoluidine blue for the selection of a suitable field.Those blocks which included a bronchiole were trim-med, and fine sections were cut with an Ultrotome IIIultramicrotome using glass knives. The sections werefloated onto copper grids, stained with lead citrate,and examined with an AEI EM 6B electron micro-scope using an acceleration voltage of 60 kV. Whereverpossible both proximal and distal regions ofbronchioles were examined from each rat. With thesmall pieces of tissue taken for electron microscopyit was not possible to determine the anatomicalrelationships of a given bronchiole, particularly whereonly a segment of the airway was available. In therat all the intrapulmonary bronchi and bronchiolesare lined by a single layer of epithelial cells.Bronchioles were classified as proximal or distalusing the following criteria. The epithelium ofproximal bronchioles was thrown into numerous foldswith bands of connective tissue supporting the crestof each fold. It consisted largely of ciliated cells whichwere markedly columnar in shape. The mucosa wassurrounded by a thick wall of smooth muscle amongwhich unmyelinated nerves were prominent. In distalbronchioles the epithelium was not folded, and thecells were cuboidal or no more than three times ashigh as they were wide, the majority being non-ciliated. There were few smooth muscle cells or nerves.Most of these characteristics could be ascertainedfrom the 1 ,m thick sections by light microscopy.

The fifth group of rats comprised 20 adult maleSprague Dawley rats. They were given intraperitonealinjections of a solution of chlorphentermine hydro-chloride dissolved in saline at a dose of 50 mg/kgbody weight. The injections were administered dailyfor six days in each week for a total period of 50days. On the fiftieth day of the experiment the ratswere killed with intraperitoneal injections ofamylobarbitone sodium. Small blocks of lung tissuewere taken from the superior lobe of the right lungfor electron microscopy as described above. Thebronchus supplying the superior lobe was ligated andthe lungs were distended with 10% formol salinethrough the trachea until the pleural surfaces weresmooth, and then immersed in a similar solution.After fixation, blocks of lung tissue were embeddedin paraffin wax and sections 5 ,um in thickness werecut and stained with haematoxylin and eosin, periodicacid Schiff's reagent, and the luxol fast blue stainingtechnique as described by Azzopardi and Thurlbeck(1969). Lung tissue from the other four groups ofrats was also fixed in formalin, sectioned, andsimilarly stained.

RESULTS

LIGHT MICROSCOPY OF GROUPS 1 TO 4 Clara cellswere easily identified in the terminal bronchiolesbecause they projected further into the lumen thanthe ciliated cells (Figs 1 and 2). In the moreproximal bronchioles Clara cells were seen as

*

II

I

FIG. 1. Adult control rat. Bronchiole from an adultSprague Dawley rat showing the low columnar epitheliumcomprising ciliated cells and Clara cells. The latter(arrowed) project further into the lumen than the ciliatedcells. (Haematoxylin and eosin x 265.)

148



j.com/


arch 1974. Dow

nloaded from


The Clara cell

I

I.=

A4

FIG. 2. Adult control rat. Part of a terminal bronchioleadjacent to a respiratory bronchiole. Clara cells arenumerous and project into the lumen. Some Clara cellspossess apical club-shaped cytoplasmic processes (arrowed).(H. andE. x575.)

occasional protrusions into the lumen (Fig. 1),but adjacent to respiratory bronchioles they werethe predominant epithelial cell (Fig. 2).Clara cells contained a central dark nucleus

above which was an apical region of amorphouseosinophilic cytoplasm. Often apical club-shapedcytoplasmic protrusions were present (Fig. 2). Inthe recesses of the bronchioles, detached islandsof cytoplasm identical with these cytoplasmicprotrusions were seen close to the Clara cells.There were few differences in the structure orfrequency of Clara cells between the four groupsof rats with the exception that apical cytoplasmicprotrusions were less common in the neonatalrats. Clara cells were not PAS positive and wewere unable to demonstrate convincingly that theystained more deeply with luxol fast blue than theciliated cells.

ULTRASTRUCTURE IN CONTROL ANIMALS

Adult rats (group 1) Under the electron micro-scope at low magnifications Clara cells were seenas non-ciliated cells extending further into thebronchiolar lumen than adjacent ciliated cells(Fig. 3). In the proximal bronchioles Clara cellswere uncommon except at the crests of folds inthe epithelium. In the distal bronchioles Claracells were numerous and formed the majority ofcells in the epithelium. In both 'classes of

bronchiole the ultrastructure of the Clara cellswas similar.The type and distribution of organelles varied

in different regions of the cells so that it waspossible to divide them into three zones compris-ing base, middle, and apex (Fig. 4). At the base ofeach cell there was a prominent basement mem-brane (Fig. 4) beneath which was the connectivetissue of the bronchiolar mucosa. The cytoplasmin the basal zone contained numerous elongatedmitochondria, scanty rough endoplasmic reticu-lum, smooth endoplasmic reticulum, and freeribosomes (Fig. 4). In most Clara cells there werethree or four large lysosomes (average diameter550 nm). These were usually circular or occa-sionally oval in section and were bounded by amembrane (Fig. 4).The middle zone of the Clara cell contained

the nucleus which was elongated and bilobed, thearms of the lobes pointing towards the base ofthe cell (Figs 3 and 4). This was in contrast to theciliated cells in which the nucleus was oval andnot divided into lobes. On either side of thenucleus there was almost invariably a Golgiapparatus (Figs 4 and 5). In the centre of theseGolgi bodies the cisternal spaces were clear butat either edge they contained a pale amorphousmaterial (Fig. 5). Rods and vesicles also contain-ing this material were seen apparently breakingoff, or fusing with, the membrane stacks of theGolgi apparatus (Fig. 5). Smooth endoplasmicreticulum was abundant in the middle zone andconsisted of short rods and vesicles which con-tained a pale amorphous substance identical withthat seen in the Golgi apparatus. Rough endo-plasmic reticulum was also abundant in the middlezone. It consisted of pairs of unbranched mem-branes with narrow cisternal spaces (Fig. 6). Therough endoplasmic reticulum was situated closeto the nucleus surrounding the apical half of it.Its orientation was usually parallel to the nuclearmembrane although near the apex of the nucleusit was sometimes arranged in the form of whorls.The cisternae of the rough endoplasmic reticulumbore scanty ribosomes and the cisternal spacecontained a pale amorphous substance. Ballooningof the cisternae at their edges was sometimes seen(Fig. 6), and the presence of these balloons freewithin the cytoplasm suggested that they hadbecome detached. Also present in the middlezone were numerous elongated mitochondria,occasional lysosomes, and free ribosomes.The apical zone of the Clara cells was charac-

terized by a profusion of smooth endoplasmicreticulum, far in excess of that seen elsewhere in

149

.:.....1

daimL.Aw.

IW.

. W,



j.com/


arch 1974. Dow

nloaded from


150

:ttt%O. C

So~~~~~~~A

XX^i4,...


-C>.

C ,-1

tt-

r ^S .... p..

/ 4..Sm

*~~~~~~~~~~~~~~~A- < :Z'

*...,: tAX i.g..

24ON

FIG. 3. Adult control rat. Electron micrograph ofpart of a terminal bronchiole showing ciliated cells (B) andClara cells (A). Note the bilobed nuclei ofthe Clara cells (N). The bronchiolar lumen contains detached islands ofcytoplasm (C) which have been extrudedfrom the apices ofthe Clara cells. ( x 7,500.)

A.



j.com/


arch 1974. Dow

nloaded from


The Clara cell

NA I nnl CI LJA

-

JLVJI IV LIL .LL-

FIG. 4. Adult control rat. Reconstruction montage ofelectron micrographs to demonstrate that Clara cells can be dividedinto three zones. The cell rests upon a basement membrane (B) beneath which arefibrocytes. The basal zone contains elon-gated mitochondria, rough endoplasmic reticulum, and a lysosome (L). The middle zone contains the lobed nucleus (N), aGolgi apparatus (G), and a pale cytoplasmic 'cleft' (C). In the apex there are spherical mitochondria and large quantitiesofsmooth endoplasmic reticulum. The boundary ofthe cell is indicatedby arrows. (Approx. x 8,000.)

FIG. 5. Adult controlrat. Golgiapparatus (G) with closely associatedmembranes ofsmooth endoplasmic reticulum(arro wed) containing apale amorphous substance. (Electron micrograph x 75,000.)

APEX

PAQ1I I

151



j.com/


arch 1974. Dow

nloaded from



4'~~~~~~~~~~~~~~~~~~~~~

FIG. 6. Adult control rat. Clara cell showingpart ofthe middle zone (to the left) and the apical zone (to the right).There is prominent rough endoplasmic reticulum (R), the edges of which show ballooning (arrowed). Many ofthese dilatations occurfreely in the cytoplasm. The apex ofthe cell isfilled with smooth endoplasmic reticulum (S).Mitochondria (M) contain few cristae and are invested by membranes of the smooth endoplasmic reticulum.(Electron micrograph x 37,500.)

the cell (Fig. 6). The smooth endoplasmic reti-culum consisted of vesicles and rods of varyinglength and these were commonly branched. Thespace enclosed by the membranes of the smoothendoplasmic reticulum contained a pale amor-phous material identical with that seen in otherzones of the cell. Mitochondria in the apical zonewere typically spherical and were unusual in thatthey were virtually devoid of cristae (Figs 6and 7). The smooth endoplasmic reticulum oftenclosely surrounded the mitochondria, giving thelatter the appearance of having a multilayeredwall (Figs 6 and 7). In the apical zone and some-times extending into the middle zone there werepale elongated 'clefts' in the cytoplasm (Fig. 4).At high magnification these structures were seento be bounded by a membrane and they containedthe same pale amorphous substance as the smoothendoplasmic reticulum (Fig. 7).

In the distal bronchioles the apical zone ofmany of the Clara cells projected from the rest ofthe cell in the form of an apical cap (Fig. 8). Thecytoplasm in this cap was less dense than that inthe rest of the cell and was sharply demarcatedfrom it. It contained large quantities of smooth

endoplasmic reticulum (Fig. 8). The bronchiolarlumen contained a variety of detached islands ofcytoplasm (Fig. 3). Some of these were structurallythe same as apical caps of Clara cells whereasothers lacked cytoplasmic organelles and weredisintegrating. There were many intergrades be-tween these two extremes in which there wasprogressive dissolution of the membranes of thesmooth endoplasmic reticulum (Fig. 9). It wassometimes possible to find an apical cap attachedto a Clara cell by a narrow isthmus suggesting thatthe islands of cytoplasm originated by extrusionof the apical caps into the bronchiolar lumen.

Apical caps were also present on some of theClara cells in the proximal bronchioles but theproportion of cells showing this feature was lessthan in the distal bronchioles.Neonatal rats (group 2) Clara cells were numer-ous in the distal bronchioles of neonatal rats butdiffered from their counterparts in the adult ratsin several respects. The most striking differencewas the presence in the Clara cells of neonatalrats of many spherical, membrane-bound,electron-dense bodies (Figs 10 and 11). Theseinclusions were most numerous in the apical half

152



j.com/


arch 1974. Dow

nloaded from


The Clara cell

FIG. 7. Adult control rat. Cytoplasmic 'cleft' (C) in the middlezone of a Clara cell. Note that the mitochondria lack cristae andare invested by smooth endoplasmic reticulum. (Electron micrographx 60,000.) _.....

...B

A-I'

FIG. 8. Adult control rat. Apex ofa Clara cellshowing a cytoplasmic cap consisting entirely ofsmooth endoplasmicreticulum, the cisternae ofwhich contain apale amorphous substance. There is a sharp division between the apicalcap and the rest ofthe cell (arrowed). (Electron micrograph x 18,750.)

153



j.com/


arch 1974. Dow

nloaded from


154 P. Smith, D. Heath, and H. Moosavi

M:~~I

/~~~~.~~~~~~~~~~~~~~~f3nFIG. 9. Adult control rat. Detail of a cytoplasmic extrusion from aClara cell showing the smooth endoplasmic reticulum containing apale amorphous substance. There is evidence that the membranes areundergoing dissolution (arrowed). (Electron micrograph x 75,000.)

V2

T~~~~~~~~~~~~~~~~~~~~~~~~~~M~~~~44K. 7~AB

5w. .., w,.q~.y ... .-u

FIG. 10. Neonatal control rat. Two Clara cells showing a nucleus (N), sparsely distributed rough endoplasmicreticulum (R), and dark inclusions (I). Note the almost total absence ofsmooth endoplasmic reticulum at the apex.(Electron micrograph x 12,500.)



j.com/


arch 1974. Dow

nloaded from


The Clara cell

T *r

FIG. 11. Neonatal control rat. Clara cell showing the apex caught in the process ofextrusion into the bronchiolarlumen. It contains little smooth endoplasmic reticulum. The main body of the cell contains rough endoplasmicreticulum, dark inclusions, and mitochondria. (Electron micrograph x 25,000.)

of the cells and were smaller than the lysosomesof Clara cells in the adults (average diameter255 nm).The Clara cells in neonatal rats contained

prominent but sparsely distributed rough endo-plasmic reticulum (Fig. 10). This consisted oflong, twisting, unbranched cisternae bearingscanty ribosomes. Unlike that in the adult rats,the rough endoplasmic reticulum was evenlydistributed throughout the cells and was notorientated parallel to the nuclear membrane. Ina minority of Clara cells the apical region con-tained an abundance of rough endoplasmicreticulum (Fig. 12) which consisted of narrowcisternae arranged in the form of loops and spirals.A few of these cisternae showed ballooning attheir edges. Mitochondria were numerous and inthe apical zone they were spherical and investedby smooth-surfaced membranes (Fig. 11). Al-though mitochondria in the apex appeared tocontain less cristae than in the basal region thisdifference was not as striking as in the adult rats.The profusion of smooth endoplasmic reticulumseen in the apex of Clara cells in adult rats wasnot a feature of the Clara cells in neonates (Figs10, 11, and 12). Clefts in the cytoplasm containinga pale amorphous substance were found in a

minority of cells but were not as common as inthe adult rats. Similarly, extrusions of the cyto-plasm into the bronchiolar lumen were uncom-mon and contained only a few smooth-surfacedmembranes (Fig. 11).

Identification of Clara cells in the proximalbronchioles of neonatal rats presented difficultybecause a variety of non-ciliated cells was present.Some of these were immature ciliated cells whichbore microvilli but lacked cilia and basal bodies.Another type of cell resembled the Clara cell andcontained large apical granular inclusions. Therewere many intergrades of structure between thesebodies and the mucus droplets found in maturegoblet cells. Where positive identification of Claracells was possible in the proximal bronchioles,their ultrastructure was identical with that ofClara cells in the distal bronchioles.

ULTRASTRUCTURE IN HYPOXIC ANIMALS

Adult rats (group 3) Exposure to acute hypoxiainduced changes in the Clara cells which were anexaggeration of their normal structure. The pro-cess of extrusion of the apical caps into thebronchiolar lumen was greatly enhanced (Fig. 13).Most of the Clara cells possessed large bulbous

155



j.com/


arch 1974. Dow

nloaded from



FIG. 12. Neonatal control rat. Apex of Clara cell showing abundant rough endoplasmic reticulum arranged intheform ofloops and spirals. Note the absence ofsmooth endoplasmic reticulum. (Electron micrograph x 25,000.)

apical caps which were distended with smooth-surfaced membranes (Fig. 13) consisting of short,branched rods and vesicles. Many of these werefound lying freely in the bronchiolar lumina(Fig. 13) together with detached islands of cyto-plasm showing various stages in the dissolutionof the smooth endoplasmic reticulum (Fig. 14).Extracellular smooth-surfaced membranes werealso found scattered about the bronchiolar lumina(Fig. 14) together with cytoplasmic debris andmyelin figures.With the exception of an apparent increase in

frequency of cytoplasmic 'clefts' there were noother alterations in structure of the Clara cells inthe hypoxic rats.Neonatal rats (group 4) The response of theClara cells of neonatal rats to acute hypoxia wasless dramatic. The number of apical extrusionswas greater than in the control rats but these didnot contain the profuse smooth endoplasmicreticulum of the adult rats. Furthermore, mostof the membranes within the extrusions consistedof long, unbranched, narrow cisternae with onlya few rods or vesicles (Fig. 15). Large fluid-filledvacuoles were sometimes seen within the cyto-plasmic extrusions (Fig. 15).

Within the main body of the Clara cells smoothendoplasmic reticulum was as uncommon as in

control neonates. Cytoplasmic 'clefts' were en-countered much more frequently, whereas theelectron dense apical inclusions appeared to be lesscommon. Rarely, small lamellated structures wereencountered in Clara cells of hypoxic neonataland adult rats.

EFFECTS OF CHLORPHENTERMINE ON THE CLARACELL Light microscopy of the bronchioles of ratstreated with chlorphentermine revealed a strikinghyperplasia of Clara cells (Fig. 16). In addition tothis increase in their number, each Clara cellprojected further into the bronchiolar lumen thannormal. The cytoplasm of the Clara cells had avesiculated appearance and stained more deeplywith luxol fast blue than the ciliated cells. Thealveolar spaces contained large foamy macro-phages and debris which also stained faintly withluxol fast blue.

Electron microscopy showed the presence oflarge inclusions in the cytoplasm of the Claracells (Figs 17 and 18). These had an averagediameter of 4130 nm and were therefore man)times larger than any inclusions found in theClara cells in the other four groups of rats. Theseinclusions were of two types, dense homogeneousbodies and lamellated structures (Fig. 18). Alintermediate magnifications the dense bodies pre-

156



j.com/


arch 1974. Dow

nloaded from


The Clara cell

0~~~~

IG. 13. Adult hypoxic rat. Clara cellspossess large apical caps (A) distended with smooth endoplasmic reticulum.These caps have become extruded into the lumen toform islands ofcytoplasm (E) which are alsofilled with smoothendoplasmic reticulum. One such extrusion contains a small lamellated inclusion (arrowed). The bronchiolarlumen also contains cytoplasmic debris. (Electron micrograph x 12,500.)

157

-4."T

........

ge8J;- W



j.com/


arch 1974. Dow

nloaded from



or.3-WV.

.~~f

N~~~~~~~~~~~k

... '.. ...

FIG. 14. Adult hypoxic rat. Apical extrusions from Clara cells showing various stages in the dissolution of thesmooth endoplasmic reticulum. There are membranes of smooth endoplasmic reticulum free in the bronchiolarlumen. (Electron micrograph x 12,500.)*..: g .. <: . :.wy....--.;^.? , ;,-

FIG. 15. Neonatal hypoxic rat. Two apical extrusions from Clara cells. The smooth endoplasmic reticulumconsists largely oflong parallel membranes. Large vacuoles (V) appear to contain oedema fluid. (Electron micro-graph x 18,750.)

158



j.com/


arch 1974. Dow

nloaded from


159The Clara cell

I I

:p.

f ' \.:1 \ t

^ ~ ~ L. ci'

FIG. 16. Adult rat after chiorphentermine administration.Terminal bronchiole showing hyperplasia of Clara cells.Compare with Fig. 1 which shows a normal terminalbronchiole ofsimilar calibre. (H. andE. x 275.)

-.. t4r. ,.tk..

FIG. 17. Adult rat after chiorphentermine administration. Three Clara cells showing nuclei (N) and the presenceof large, dense inclusion bodies. One of these (arrowed) shows the development oflamellae at its periphery. Notethe large numbers ofmitochondria. (Electron micrograph x 12,500.)



j.com/


arch 1974. Dow

nloaded from



V

4F

FIG. 18. Adult rat after chlorphentermine administration. A Clara cell showing large dense inclusions (A) andlamellated inclusions (B). Note the stippledappearance ofthe dense bodies. (Electron micrograph x 37,500.)

sented a stippled appearance (Fig. 18), which athigh magnification was shown to be due to aregular arrangement of minute hexagonal globulesor tubes. The tubular configuration was restrictedto the periphery of the bodies. The lamellatedinclusions, on the other hand, consisted of a densehomogeneous core surrounded by concentricelectron dense lamellae, the interlammellar dis-tance increasing towards the periphery (Fig. 18).Structures intermediate in appearance betweenthese two inclusions were common, and in thesethe dense bodies showed the formation of lamellaeat their periphery (Fig. 17).

Elsewhere in the Clara cells there was evidenceof high metabolic activity in which mitochondriaand smooth and rough endoplasmic reticulumwere abnormally abundant (Fig. 17). However,protuberant apical caps did not occur as innormal and hypoxic rats.

DISCUSSION

Our study has confirmed the opinion of otherworkers that the Clara cell is an active secretorycell. Thus it contains a prominent Golgi apparatusand abundant rough and smooth endoplasmicreticulum. The cisternal spaces of all three ofthese components contained a pale amorphous

substance which may have been a product ofsecretion. This suggests either that three differentsecretory products are produced by the cell or thatthe three cellular components are concerned withdifferent stages of the synthesis of a single pro-duct. Ballooning and detachment of the edges ofthe cisternae of the rough endoplasmic reticulumwas probably the mechanism whereby the syn-thetic product within the cisternal spaces wasreleased into the cytoplasm. Vesicles containingthe same pale amorphous substance were closelyassociated with the Golgi apparatus and wereeither fusing with, or breaking off, the Golgistacks. This observation is in agreement with thewidely held view that precursors of secretions inmany cells fuse with the Golgi body, are coupledtogether, and are then released as a finishedproduct. We suggest that, in the middle zone ofthe Clara cell, precursors of the secretion are pro-duced in the rough and smooth endoplasmicreticulum and are coupled in the Golgi apparatus.The final synthesized product contained withinrod-shaped membranes then detaches from theGolgi apparatus and accumulates at the apex ofthe cell. Here the membranes do not fuse to formsecretory granules, as occurs in many types ofsecretory cell. Instead the apical region of cyto-plasm bulges en masse into the bronchiolar lumen

160



j.com/


arch 1974. Dow

nloaded from


The Clara cell

in the form of an apical cap. These are the ultra-structural equivalent of the club-shaped cyto-plasmic processes seen on light microscopy. Theapical cap is then extruded into the bronchiolewhere the endoplasmic reticulum breaks down,presumably permitting the secretion to diffuse intothe surrounding cytoplasm. The amorphous islandof cytoplasm so produced then disintegrates,thereby releasing the secretory product into thelumen of the bronchiole. The mode of secretionof the Clara cell is therefore a type of apocrinesecretion. In reaching this conclusion we agreewith Corrin et al. (1973), who noted a similarprocess in the bronchioles of mice. However, theydescribed dissolution of the apical endoplasmicreticulum before the apical cap became extrudedinto the bronchiole. The presence of a pale amor-phous substance contained within the cytoplasmic'clefts' suggests that these also contain the productof secretion. They did not have the appearance ofsecretory granules and were probably distendedcisternae of smooth endoplasmic reticulum.The differences in the ultrastructure of Clara

cells between adult and neonatal rats are probablya reflection of the immaturity of the latter. Thusin the proximal bronchioles, in particular, identifi-cation of Clara cells was often rendered difficultowing to their similarity to differentiating gobletand ciliated cells. The apical region lacked theprofusion of smooth endoplasmic reticulum typicalof the adult, and cytoplasmic extrusions wereuncommon, small in size, and contained fewmembranes. Also, in all but a few neonatal Claracells there was less smooth and rough endoplasmicreticulum in the middle zone than occurred inadult rats. It seems likely, therefore, that the Claracells in neonatal rats are not fully developed assecretory cells.The Clara cells respond to the stimulus of

hypoxia by an acceleration of the process ofextrusion of the apical caps. In the adult rats theapical caps contained an excess of smooth endo-plasmic reticulum, suggesting that the syntheticactivity of the cells was also increased. The pro-

duction of cytoplasmic extrusions from the Claracells of hypoxic neonatal rats was also acceleratedbut these contained a paucity of smooth endo-plasmic reticulum. A possible explanation as towhy acute hypoxia should result in hypersecretionof the Clara cell is provided by the work of Castilloand Johnson (1969). They found that the surfacetension in the lungs of mice was elevated45 minutes after exposure to hypoxia. By twohours there was a degree of compensation but thesurface tension was still higher than normal. This

suggested that acute hypoxia caused a deficiencyof pulmonary surfactant which was partiallyreplaced by two hours. If the Clara cell is respon-sible for the synthesis of pulmonary surfactantone might expect a shortage of this substance inacute hypoxia to act as a stimulus for increasedsecretion by the Clara cells.Examination of rats treated with chlorphenter-

mine has provided evidence as to the nature of theapocrine secretion of the Clara cell. Chronicadministration of chlorphentermine elicits theaccumulation of large foamy macrophages in thealveolar spaces (Heath et al., 1973). Electronmicroscopy has shown that the alveolar spacescontain large numbers of phospholipid latticesand that the macrophages appear to ingest thissubstance which accumulates in their cytoplasmin the form of irregular lamellated inclusions(Smith et al., 1973). The macrophages then disinte-grate, releasing cytoplasmic debris and lamellatedstructures into the alveolar spaces and ducts.Lamellated inclusions also find their way into typeI alveolar epithelial cells and the capillary endo-thelium (Smith et al., 1973). Because the type IIalveolar epithelial cells, or granular pneumocytes,were abnormally numerous, it was suggested thatthe alveolar phospholipid in chlorphenterminelung originated from hyperactivity of these cells.For many years the granular pneumocyte has beenregarded as the sole source of pulmonary surfac-tant which accumulated in its characteristic lamel-lar bodies. This hypothesis was adopted becauseduring embryological development a drop insurface tension in the lungs coincided with theappearance of lamellar bodies (Buckingham andAvery, 1962; Kikkawa, Motoyama, and Gluck,1968), because radioactive precursors of pul-monary surfactant are taken up by granularpneumocytes (Buckingham, Heinemann, Som-mers, and McNary, 1966; Faulkner, 1969) andbecause diseases involving a deficiency of surfac-tant also show a shortage of lamellar bodies (Balis,Delivora, and Conen, 1966). However, the strikinghyperplasia of Clara cells which we have demon-strated in chlorphentermine intoxication suggestsan alternative-that the Clara cell produces aphospholipid pulmonary surfactant.

Etherton et al. (1973) injected tritiated palmiticacid, a radio-active precursor of pulmonary sur-factant, into mice. Subsequent electron autoradi-ography of the lung showed maximal labelling inthe endoplasmic reticulum of Clara cells after3 minutes. Heavy labelling also occurred in thegranular pneumocytes but only after one hour.This observation, plus our own findings, strongly

161



j.com/


arch 1974. Dow

nloaded from



suggests that the Clara cell secretes pulmonarysurfactant.The large dense bodies and lamellar inclusions

which occurred in Clara cells in chlorphentermineintoxication were an abnormal feature. Suchinclusions were absent from the control rats andeven in the hypoxic animals lamellated inclusionswere extremely rare. However, the hexagonalsubstructure of the dense bodies more closelyresembles the physical state adopted by a mixtureof phospholipid and water than does the irregular,loose, spiral arrangement of the lamellar bodiesin granular pneumocytes (Luzzati and Husson,1962). It may be that the metabolism of the Claracells is altered by chlorphentermine so that theseabnormal bodies are formed. Alternatively, anexcessive synthesis of phospholipid may lead toits deposition in the cytoplasm of the Clara cells.Azzopardi and Thurlbeck (1969) found that the

Clara cells of mice stain more deeply with luxolfast blue than the surrounding cells. We wereunable to reproduce their results in rats exceptin those animals given chlorphentermine. Luxolfast blue is believed to demonstrate lipoprotein(Pearse, 1968). Azzopardi and Thurlbeck (1969),using a variety of histochemical stains, came tothe conclusion that the Clara cell contained lipo-proteins rich in choline-containing phospholipids.They took the view that this lipoprotein is thepulmonary surfactant.

If it is true that the Clara cell secretes thepulmonary surfactant, what then is the functionof the granular pneumocyte? It may also producepulmonary surfactant, perhaps of a differentchemical nature, since it too takes up labelledpalmitic acid. Niden (1967) has stated that thegranular pneumocyte is phagocytic but this hasbeen disputed by Corrin (1970). The periphery oflamellar bodies has been shown to contain acidphosphatase (Corrin, Clark, and Spencer, 1969;Kuhn, 1968) which suggests that they may bephagosomes. Also the granular pneumocytes showless oxidative enzyme activity than the bronchiolarepithelium (Azzopardi and Thurlbeck, 1967), andactive secretory activity in granular pneumocyteshas not yet been convincingly demonstrated. It isclear that the evidence in favour of the granularpneumocyte being the producer of pulmonarysurfactant has now been balanced by a weight ofopposing evidence. Attention must now be fo-cused on the Clara cell as a possible source of thissubstance. In particular, the pathology of the Claracell in diseases involving a deficiency of pulmonarysurfactant such as the idiopathic respiratorydistress syndrome should be examined.

Professor Heath acknowledges with gratitude thereceipt of a grant from the Nuffield Foundation tosupport this research.

REFERENCES

Azzopardi, A. and Thurlbeck, W. M. (1967). Theoxidative enzyme pattern in developing and adultmice and adult rabbits. Laboratory Investigation,16, 706.and (1969). The histochemistry of the

nonciliated bronchiolar epithelial cell. AmericanReview of Respiratory Disease, 99, 516.

Balis, J. U., Delivora, M., and Conen, P. E. (1966).Maturation of postnatal human lung and theidiopathic respiratory distress syndrome. Labora-tory Investigation, 15, 530.

Buckingham, S. and Avery, M. E. (1962). Time ofappearance of lung surfactant in the foetal mouse.Nature, 193, 688.-, Heinemann, H. O., Sommers, S. C., andMcNary, W. F. (1966). Phospholipid synthesis inthe large pulmonary alveolar cell. Its relation tolung surfactants. American Journal of Pathology,48, 1027.

Castillo, Y. and Johnson, F. B. (1969). Pulmonarysurfactant in acutely hypoxic mice. LaboratoryInvestigation, 21, 61.

Corrin, B. (1970). Phagocytic potential of pulmonaryalveolar epithelium with particular reference tosurfactant metabolism. Thorax, 25, 110., Clark, A. E., and Spencer, H. (1969). Ultra-structural localization of acid phosphatase in therat lung. Journal of Anatomy, 104, 65.

-, Etherton, J. E., and Conning, D. M. (1973).Apocrine secretory activity in the terminal bron-chiole of mouse lung: morphological evidence.Proceedings of the Pathological Society of GreatBritain and Ireland. Journal of Pathology, 109, Pv.

Etherton, J. E. and Conning, D. M. (1971). Earlyincorporation of labelled palmitate into mouselung. Experientia, 27, 554.

and Corrin, B. (1973). Apocrine secretoryactivity in the terminal bronchiole of mouse lung:ultrastructural autoradiographical evidence. Pro-ceedings of the Pathological Society of GreatBritain and Ireland. Journal of Pathology, 109, Pv.

Faulkner, C. S. (1969). The role of the granularpneumocyte in surfactant metabolism. A rchivesof Pathology, 87, 521.

Hayek, H. von (1962). Cellular structure and mucusactivity in the bronchial tree and alveoli. InPulmonary Structure and Function. Ciba Foun-dation Symposium. Ed. by A.V.S. de Reuck andM. O'Connor. Churchill, London.

Heath, D., Smith, P., and Hasleton, P. S. (1973). Theeffects of chlorphentermine on the rat lung.Thorax, 28, 551.

Karrer, H. E. (1956). Electron microscopic study ofbronchiolar epithelium of normal mouse lung.Experimental Cell Research, 10, 237.

Kikkawa, Y., Motoyama, E. K., and Gluck, L. (1968).Study of the lungs of fetal and newborn rabbits.American Journal of Pathology, 52, 177.

162



j.com/


arch 1974. Dow

nloaded from


The Clai

Kisch, B. (1958). Electron microscopy of the lungs inacute pulmonary edema. Experimental Medicineand Surgery, 16, 17.

Kuhn, C. (1968). Cytochemistry of pulmonary alveolarepithelial cells. American Journal of Pathology,53, 809.

Luzzati, V. and Husson, F. (1962). The structure ofthe liquid-crystalline phases of lipid-water systems.Journal of Cell Biology, 12, 207.

Moosavi, H., Smith, P., and Heath, D. (1973). TheFeyrter cell in hypoxia. Thorax, 28, 729.

Niden, A. H. (1967). Bronchiolar and large alveolar

163

cell in pulmonary phospholipid metabolism.Science, 158, 1323.

Pearse, A. G. E. (1968). Histochemistry, Theoreticaland Applied, 3rd ed., vol. 1. pp. 426-427.Churchill, London.

Smith, P., Heath, D., and Hasleton, P. S. (1973).Electron microscopy of chlorphentermine lung.Thorax, 28, 559.

Requests for reprints to: Professor Donald Heath,Department of Pathology, The University, Liverpool,L69, 3BX.



j.com/


arch 1974. Dow

nloaded from


the clara cell - thorax · clara cells are numerous andproject into the lumen. some clara cells...

Documents