A route for diseased

J Clin Pathol 1983;36:161-166

A possible route for lymphocyte migration intodiseased tissuesAJ FREEMONT

From the Department ofPathology, University ofManchester Medical School, Oxford Road, Manchester MJ39PT

SUMMARY Vessels within the lymphocytic infiltrate in 161 examples of 24 different diseases havebeen studied using light and electron microscopy and histochemistry. Hitherto undescribed venuleshave been identified within lymphocyte aggregates which exhibit a structure, ultrastructure andcytochemistry previously believed to be restricted to the high endothelial venules of lymphoidtissues. The significance of these findings is discussed within the context of lymphocyte migrationinto abnormal tissue.

Tissue lymphocyte infiltration occurs in response tomany different pathological stimuli. Although theirfunction is often unclear, lymphocytes are believed toplay an important part in initiating and modifyingdisease processes. Despite their importance, little isknown of the factors stimulating and controllinglymphocyte migration from blood into the affectedtissue. '

Physiological lymphocyte migration, into lym-phoid tissue, has been intensively studied over thepast few years.23 In these organs lymphocyte trafficfrom blood occurs, almost exclusively, across thewalls of specialised venules4 which have a uniquemorphology and histochemistry.5 6 Vessels with thesecharacteristics have not been described outside thelymphoid organs in health, but vessels with a similarultrastructure have been found in experimentallyinduced chronic inflammation in laboratory animals,7raising the possibility that vascular specialisation foraccelerated lymphocyte diapedesis may not berestricted to the lymphoid organs.

In this study a number of human diseases with anassociated lymphocytic infiltrate have been examinedto determine if any of the local blood vessels exhibitfeatures suggestive of specialisation for selectivelymphocyte traffic from circulation to tissue.

Material and methods

The study material consisted of biopsies from 161patients with various pathological conditions associ-ated with lymphocytic infiltrates. In every case, tissue

Accepted for publication I September 1982

was processed for paraffin 'wax embedding and,where possible, samples were taken for histo-chemistry and electron microscopy (Table 1).

PARAFFIN WAX EMBEDDINGAfter fixation in 10% neutral buffered formalin,

Table 1 Those diseases with lymphocytic infiltratesshowing the numbers examined using various techniques.

Disease No ofcases

LM Histochem EM

Foreign body reactions 18 7 3Peptic ulceration 15 1 1Ulcerative colitis 12 2 2Rheumatoid synovium 9 1 8Parotitis 7 2 2Chronic cholecystitis 5 1 1Myositis 4 4 0Reiter's synovium 1 0 0Chronic pyelonephritis 3 0 0Oesophageal SCC 4 2 2Adenocarcinoma of thyroid 3 1 1Adenocarcinoma of stomach 5 2 2Medullary carcinoma of breast 6 1 1IDC breast 20 6 8Seminoma 9 4 2Basal cell carcinoma 7 4 4Transitional cell carcinoma of bladder 9 3 3Thyroid adenoma 4 3 3Large bowel adenoma 1 1 1Adenolymphoma 10 4 4Benign prostatic hyperplasia 5 1 2Coeliac disease 1 1 1Cystic hypertrophic mastopathy 1 1 1Lichen planus 1 0 0Hashimoto's thyroiditis 1 1 1Total 161 53 53

LM = paraffin embedding. Histochem non-specific esterase histochemistry.EM = electron microscopy.

161

on May 29, 2022 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.36.2.161 on 1 February 1983. D

ownloaded from

http://jcp.bmj.com/

162

tissue was embedded in paraplast, sectioned at 3 ,umand stained with haematoxylin, methyl green pyroninand periodic-acid Schiff (PAS) using standardtechniques. Metachromasia was detected with azureA employing the method described by Ball andJackson.8 For this, prior to staining, one section fromeach case was pretreated with crystalline ribonuclease(Sigma) at a concentration of 1 mg/ml in phosphatebuffer pH 6 for 3 h at 37°C and another was treatedwith phosphate buffer without ribonuclease underthe same conditions for the same time.

ENZYME HISTOCHEMISTRYCubes of unfixed tissue (5 mm3) were immersed inice-cold formol sucrose and after 18 h transferred toHolt's syrup.9 Cryostat sections (7 ,um) were cut thefollowing day and a simultaneous coupling azo-dyetechnique employed for localising the enzyme non-specific esterase (NSE). This em' oyed a methoddevised in this laboratory'0 wherety tissue sectionsare incubated at 37°C with a-naphthyl propionate for30 min, the position of the final reaction productbeing displayed with fast garnet GBC.

ELECTRON MICROSCOPYSmall portions of tissue were fixed for 4 h in 2-5%glutaraldehyde in 0 1 M sodium cacodylate buffer pH7-4 at room temperature. The specimens were thenwashed in 0 1 M cacodylate buffer and minced into 1mm3 cubes. This was followed by postfixation in 1%osmium tetroxide in cacodylate buffer pH 7- 4 at 4°C

Table 2 The normal and diseased tissues (withoutlymphocytic infiltrates) usedfor comparison with thosediseases in Table 1. The column headings are the same as inTable 1

Tissue No ofsamples

LM Histochem EM

Normal tissue 43 17 12Skin 15 3 1Breast 8 5 5Oesophagus 5 1 1Prostate 1 0 1Large bowel 6 1 1Synovium 2 2 2Thyroid 2 1 0Muscle 4 4 1Diseased tissue 66 29 31Granulation tissue 26 13 18Synovium in septic arthritis 3 0 0Breast carcinoma 12 5 6Oesophagus SCC 2 0 1Prostatic adenocarcinoma 1 1 1Transitional cell carcinoma of bladder 2 1 2Parathyroid adenoma 6 1 0Benign prostatic hyperplasia 4 1 2Lipoma 1 1 1Myocardial infarction 3 0 0Muscular dystrophy 6 6 0

Freemont

for 90 min, dehydration in graded alcohols andpropylene oxide and embedding in Emix resin.Sections (0 5 ,um) were cut and stained with 1%toluidine blue in borax and suitable areas selected forultrathin sectioning. Grids were double-stained usinguranyl acetate and Reynold's lead citrate andexamined in a Philips' 301 electron microscope.For comparison, samples of normal tissues and

diseased tissues without lymphocytic infiltrates werealso examined by light and electron microscopy usingidentical processing protocols (Table 2).

Results

Although the majority of small blood vessels in thediseased tissues had an identical morphology andhistochemistry to those flat endothelium-lined vesselsfound in normal tissue, areas of lymphocyte aggre-gation also contained another small, but easilydistinguishable, population of vessels. These vesselshad the general appearances of venules (Fig. 1),varied from 20 ,um to 100 pLm in diameter, were linedby between three and 10 cells in transverse sectionand, whilst distributed throughout areas of lympho-cytic infiltration, were found most commonly at theirperiphery.The number and size of the vessels varied with the

lymphocyte density. They were readily seen when thenumbers of lymphocytes exceeded 150/mm2 but semi-serial sectioning was often required to demonstratetheir presence when the density was lower, and none

Fig. 1 A high endothelium lined vessel with lymphocytes inits lumen and wall (H) and a residualfollicle (F) from a caseofHashimoto's thyroiditis. Haematoxylin and eosinx 300.



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

A possible route for lymphocyte migration into diseased tissues

was seen in the absence of lymphocytes in eithernormal or diseased tissues.When stained with haematoxylin and eosin the

endothelial cells were shown to differ from endo-thelium elsewhere. The cells were much plumper, thenuclei open and ovoid with a single, central violetnucleolus and a peripheral rim of dark blue chroma-tin. The cytoplasm was more abundant than in otherendothelial cells but tended to vary from one cell tothe next, occasionally producing a cuboidal profile(Fig. 2). The cells lay on a thick densely eosinophilic

Fig. 2 A plump endothelium lined vessel (P) showing thenuclear morphology, cytoplasmic volume, thick basementmnembrane atid intramural lymphocytes (A & B). From a

ca.se ol ttzedallarv carcinoma of breast x /100.

basement membrane with an irregular outer border.Within the lumen, between the endothelial cells,within the basement membrane and surrounding thevessels, often in concentric circles, were largenumbers of lymphocytes.

Cytoplasmic metachromasia could be demons-trated with azure A and was abolished by pretreatingthe sections with ribonuclease. The cytoplasm was

also pyroninophilic, unlike that of the flattenedendothelial cells and the basement membrane stainedstrongly with PAS.

Non-specific esterase was demonstrated in thevessels giving an intense diffuse cytoplasmic reactionin the endothelial cells (Fig. 3). The only other typesof vessels with any degree of non-specific esterasepositivity were arterioles, but here the stain was weakand restricted to a few perinuclear granules.

Although in many respects the plump endotheliumdiffered from the more common flat endotheliumonly in a quantitative manner ultrastructurally,certain features found in the former were absent fromthe latter. The plump endothelial cells were charac-

teristically squamous or cuboidal (Fig. 4), thecytoplasm having the appearance of ground glass.The nuclear morphology reflected the appearancesseen with the light microscope with the exception thatin preparations for electron microscopy, some nucleishowed surface irregularities and occasionally deepnarrow clefts (Fig. 4).

Fig. 3 A non-specific esterase positive vessel (N) within thechronic inflammatory infiltratefrom a case ofmyositis.Macrophages (M) are also densely NSE + ve and Tlymphocytes (T) have esterase positive cytoplasmic granules.Other lymphocytes (B) are negative x 400.

Many mitochondria were present in the cytoplasmas were frequent cistemae of rough endoplasmicreticulum often in conjunction with Golgi vesicles.The major cytoplasmic components were largenumbers of free ribosomes between which ran finemyofilaments. Several different membrane boundvesicular structures were seen in the cytoplasmincluding residual bodies with heterogeneous con-tents, Wiebel-Palade bodies and multivesicularbodies with up to 20 vesicles of variable size in a densematrix (Fig. 5). Absence of intercellular junctionswas noticeable although at the bases of the cells bluntlateral interdigitations were seen.The thickened basement membrane described at

light microscopy was found to consist of a complex ofbasement membrane, collagen fibres, and pericyteprocesses.Lymphocytes were found within the lumen of the

vessels, between the endothelial cells and within thebasement membrane complex (Fig. 4) where theywere often compressed circumferentially about thevessel. These cells were almost invariably smalllymphocytes but very occasional transformedlymphocytes were also seen in the vessel wall.

163



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Freemont

.- II'I'pm7

* i '@,9 ' o _ +r

Discussion

Lymphocytic infiltrates have been described as partof many different diseases, both inflammatory andneoplastic, but the numbers of lymphocytes are

frequently variable, in otherwise identical lesions.Despite reports of important prognostic implications

Fig. 4 A transverse sectionthrough a high endothelium-likevenulefor the synovium ofa patientwith Reiter's disease. Clefts in thenuclear envelope (N) and anorganelle cluster (C) in the cytoplasmadjacent to an intraluminallymphocyte (L) can be seen. Alymphocyte process (1) lies betweentwo endothelial cells and two otherlymphocytes are present one (A)between an endothelial cell and thebasement membrane complex andanother (B) within thecomplex x 4000.

Fig. 5 Parts of three endothelial cellsshowing multivesicular bodies (M),an organelle cluster containingGolgi vesicles (G), roughendoplasmic reticulum (R) andmitochondria and nuclearfeaturessuch as the prominent nucleolus (N)and an intranuclear body (1) x 9100.

for this variability," 12 little is known of its cause andeven the factors responsible for the initiation andcontrol of lymphocyte migration into the tissue areonly poorly understood.

In experimental models of chronic inflammationinfiltrating lymphocytes have been shown to comefrom the circulating pool through local blood vessel

164

,.,- 40 ;",. ... :. ..,ik :- iW .,.,

1. ..... -, .,4.:,..0 OW.,



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

A possible route for lymphocyte migration into diseased tissues

walls;'3 a situation analogous to that normallyoperating in lymph nodes. Here circulatinglymphocytes selectively migrate into the internallabyrinth of the node by passage across the walls ofspecialised venules in the paracortex known as highendothelial venules'4 (HEV) because of the unusualplumpness of their endothelial cells.HEV were first described in 1898 ' and subsequent

studies in laboratory animals have shown, not onlythat they represent the sole site at which circulatinglymphocytes enter the node, but also that they have aunique structure,'6 ultrastructure' ' 1 and histo-chemistry,'9 which distinguishes them from all otherblood vessels.

In a recent description of human HEV, severalmorphological and cytochemical features werereported to be peculiar to these vessels.20 Thesefeatures include some which may be demonstrated inparaffin-embedded tissue using routine stains. HEVare recognisable as venules lined by large endothelialcells, frequently cuboidal in shape, containing paleopen nuclei, with a single central nucleolus, copiouscytoplasm which displays pyroninophilia and ribo-nuclease labile metachromasia and a thick basementmembrane staining strongly with eosin and PAS.Other distinguishing characteristics are the intense

granular cytoplasmic reaction for non-specificesterase in the endothelial cell cytoplasm and theultrastructural appearances. These latter include bio-synthetic organelle clusters containing Golgi vesicles,rough endoplasmic reticulum and mitochondria,much cytoplasmic RNA in the form of- freeribosomes, numerous cytoplasmic myofilaments,frequent membrane-bound lysosomal structures(residual bodies and multivesicular bodies), anabsence of cell junctions and a basement membrane,interwoven collagen fibres and pericytes. Ofparticular note are the large numbers of lymphocytesin the vessel lumina and within the various layers oftheir walls.

Several investigators have suggested the special-ised function and unique microscopic appearancesmay be linked and a few have demonstrated a relationbetween metabolism, structure and lymphocytediapedesis.2' On the basis of these studies, it isbecoming apparent that not only do structural andcytochemical characteristics distinguish this set ofblood vessels, with specialised function, from allothers but that they also represent an integral part ofthat function.

This correlation of function and structure raises thepossibility that, within the context of lymphocytemigration into any tissue, vessels exhibiting themicroscopic features of HEV may share theirspecialised function.

In this study a population of blood vessels,

exhibiting those features previously reported asspecific to HEV, has been identified in the lympho-cytic infiltrates associated with several completelyunrelated disease states and may therefore, representthe site at which circulating lymphocytes gain accessto the tissue.As it has not proved possible to obtain direct

evidence for these HEV-like vessels being the routeof lymphocyte migration into infiltrates in humans,evidence for this function is only circumstantial;nevertheless there are a number of observationswhich would support this possibility. Firstly, thevessels are identical in all histological respects tolymph node HEV; secondly, they have a moreintimate relation with lymphocytes than all othervessels; thirdly, vessels with the characteristics ofHEV are not found in normal tissue, nor indeedoutside areas of lymphocyte infiltration in diseasedtissues and, finally, serial sections show that infiltratescontaining significant numbers of lymphocytes(> 150/mm2) will always contain vessels of this sort.Thus there would appear to be a small, hitherto

undescribed, group of venules within areas oflymphocytic infiltration which may well be the site atwhich lymphocytes preferentially enter the tissue.These vessels may be recognised by their distinctivemicroscopic appearances, shared only by the HEV oflymphoid tissues.The HEV-like vessels cannot be the only site at

which circulating lymphocytes enter tissue aslymphocyte migration through healthy tissues, whichdo not contain these vessels, is well recognised. It ismore probable that they represent a local vascularresponse facilitating the rapid transfer of lymphocytesinto the diseased area. Whilst they suggest a possibleroute for lymphocyte migration from blood todiseased tissues, these observations apparently offerno direct information about its control; however,similarities between HEV and the vessels inlymphocytic infiltrates may indicate that themechanisms by which HEV are believed to influencelymphocyte diapedesis may also apply to the HEV-like vessels.

I am indebted to Dr CJP Jones for her assistance inproducing the electronmicrographs and to theManchester Central District Grants Committee fortheir support of this work.

References

'Edwards SJ, Rowlands GF, Lee MR. Reduction of lymphocytetransformation by a factor produced by gastrointestinal cancer.Lancet 1973;i:687-9.

2 Schoefl GI. The migration of lymphocytes across the vascularendothelium in lymphoid tissue. A re-examination. J Exp Med1972;136:568-8.

165



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

1663 Wenk EJ, Oric D, Reith E, Rhodin JAG. The ultrastructure of

mouse lymph node venules and the passage of lymphocytesacross their walls. J Ultrastruct Res 1974;47:214-41.

Gowans JL, Knight EJ. The route of recirculation of lymphocytesin the rat. Proc R Soc Lond [Bioll 1964;159:257-82.

Schumaker S von. Ueber Phagocytose und die Abfurwege derLeucocyten in den Lymphdrussen. Archiv fur mikroskopischeAnatomie und Entwicklungsmechanik 1899;54:311-29.

Smith C, Henon BK. Histological and histochemical study of highendothelium of post-capillary veins of the lymph node. Anat Rec1959;135:207-13.

7Smith JB, Mcintosh GH, Morris Bede. The migration of cellsthrough chronically inflamed tissues. J Pathol 1970;100:21-9.

Ball J, Jackson DS. Histological, chromatographic andspectrophotometric studies of toluidine blue. Stain Technol1953;28:33-40.

Holt SJ, Hobbiger EE, Pawan GL. Preservation of rat tissues forcytochemical staining purposes. J Biophys Biochem Cytol1960;7:383-6.

Freemont AJ, Davies SJ. Acid esterase activity in lymphocytesand other cells: a comparison of six alpha-naphthyl-basedsubstrates. Med Lab Sci 1982;39:405-7.

Underwood JCE. Lymphoreticular infiltrations in humantumours: prognostic and biological implications: a review. Br JCancer 1974;30:538-48.

2 Bloom HJG, Richardson WW, Field JR. Host resistance andsurvival in carcinoma of breast: a study of 104 cases of medullarycarcinoma of breast followed for 20 years. Br Med J 1970;i: 181-8.

Graham RC, Shannon SL. Peroxidase arthritis. I1, lymphoid cell-endothelial interactions during a developing immunologicinflammatory response. Am J Pathol 1972;69:7-24.

Freemont

Claesson MO, Jorgensen 0, Ropke C. Light and electronmicroscopic studies of the paracortical post-capillary highendothelial venules. Z Zellforsch Mikrosk Anat 1971;119:195-207.

Thome R. Endothelien als Phagocyten (aus den Lymphdrussenvon Macacus cyfomalgus). Arch Microsk Anat 1898;52:820-41.

Schulze W. Untersuchungen uber die Capillaren und post-capillaren Venen lymphatische Organe. Z Anat 1925 ;76:421-62.

7 Anderson AO, Anderson N. Studies on the structure andpermeability of microvasculature in normal rat lymph nodes.Am J Pathol 1975;80:387-418.

Ropke C, Jorgensen 0, Claesson MH. Histochemical studies ofhigh-endothelial venules of lymph nodes and Peyer's patches inthe mouse. Z Zellforsch Mikrosk Anat 1972;131:287-97.

Anderson ND, Anderson AO, Wyllie RG. Specialised structureand metabolic activities of high endothelial venules in ratlymphatic tissues. Immunology 1976;31:455-73.

2lFreemont AJ, Jones CJP. Light microscopic, histochemical andultrastructural studies of human lymph node paracorticalvenules. J Anat (in press).

21 Andrews P, Ford WL, Stoddart RW. Metabolic studies of high-walled endothelium of post capillary venules in rat lymph node.Blood Cells and Vessel Walls: functional interactions. CibaFound Symp 1980;71 new series:211-30.

Requests for reprints to: Dr AJ Freemont, Department ofPathology, University of Manchester Medical School,Oxford Road, Manchester M 13 9PT, England.



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

A route for diseased

Documents

Transcript of A route for diseased