Br Heart3r 1994;71:45-50

Is thickening of the basal lamina in the saphenousvein a hallmark of smoking?

D J Higman, J T Powell, R M Greenhalgh, A Coady, J Moss

AbstractObjective-To investigate whether smok-ing causes ultrastructural changes in theintima ofthe proximal saphenous vein.Design-Proximal saphenous veins fromheavy smokers and non-smokers wereexamined with scanning and transmis-sion electron microscopy to determinechanges in surface ultrastructure, in theintercellular junction, and in the thick-ness of the basal lamina. hnmunogoldlabelling was used to identify specificcomponents of the endothelial basallamina.Material-Vein specimens were obtainedfrom patients undergoing varicose veinsurgery (12 patients) or distal bypasssurgery (eight patients).Main results-The only ultrastructuralchange that discrilinated between spec-imens was thickening of the endothelialbasal lamina. All specimens with a thick-ened basal lamina were from heavysmokers. Inmunogold labelling studiesshowed that the thickened basal laminacontained specific accumulations offibronectin but not heparan sulphateproteoglycans, type IV collagen, orlamiliin.Conclusions-Two ultrastructural char-acteristics are associated with smoking:thickening of the endothelial basallamina and a specific accumulation offibronectin in the thickened basal lam-ina. Such abnormalities in the saphenousveins from smokers may contribute tothe poorer performance of these veins asbypass conduits.

(Br Heart3' 1994;71:45-50)

Department ofSurgeryD J HigmanJ T PowellR M GreenhalghDepartment ofHistopathology,Charing Cross andWestminster MedicalSchool, LondonA CoadyJ MossCorrespondence to:D J Higman, Department ofSurgery, Charing Cross andWestminster MedicalSchool, Fulham PalaceRoad, London W6 8RF.

Accepted for publication9 August 1993

The saphenous vein is widely used for bothdistal and coronary artery bypass surgery.'2Several studies have shown that continuedcigarette smoking is associated with a signifi-cant increase in the risk of occlusion of thebypass graft.35 The mechanisms of occlusioninclude intimal hyperplasia, graft thrombosis,and the development of focal stenoses.6 In thevasculature of the human umbilical cordnoticeable ultrastructural changes have beenassociated with maternal smoking.7 Thesechanges include "blebbing" of the endothe-lium, opening of endothelial intercellularjunctions, subendothelial oedema, andwidening of the basal lamina. Uterine arteriesfrom smokers also showed discontinuity

between endothelial cells on scanningelectron microscopy.8

Here we have investigated, with electronmicroscopy and immunogold labelling ofcomponents of the basal lamina, whetherthere were ultrastructural changes in thesaphenous veins of smokers, that might pre-dispose these veins to occlusion when used asarterial bypass grafts.

Patients and methodsThis study was approved by the local ethicscommittee and all patients gave informedconsent. Segments of proximal long saphe-nous vein were obtained with minimal traumafrom two groups of patients, those having lig-ation and excision of the long saphenous veinfor the correction of varicose veins (n = 12)and those having arterial bypass surgery forcritical ischaemia of the lower limb (n = 8).Even in patients with varicose veins the proxi-mal saphenous vein was minimally dilated,with no disease evident on light microscopy.Smokers were defined as current smokerswith at least a 10 pack-year history of smok-ing. In the varicose vein group, non-smokerswere patients who had never smoked and inthe vein bypass group, the non-smokers werethose patients who had not smoked for atleast five years. Serum samples were obtainedat the time of surgery for the estimation ofcotinine, the principal stable metabolite ofnicotine: cotinine concentrations of > 200nmolI were considered indicative of currentsmoking. Patients with diabetes, on antihy-pertensive treatment, or those who had a his-tory of venous thrombosis or venous surgerywere excluded from this study. Duplex scanconfirmed a competent long saphenous sys-tem in all patients undergoing bypass surgery.

After rinsing the vein specimens in phos-phate buffered saline (pH 7 2) at 37°C, speci-mens were processed for electron microscopyand immunogold studies as described later. Asegment of the specimen was also fixed in10% buffered formalin for light microscopy.

ELECTRON MICROSCOPYA 2-3 cm segment of saphenous vein wascannulated with a modified polythene con-nector and distended with 3% glutaraldehydein 0 1 M cacodylate buffer (pH 7 2) from areservoir at a pressure of 40 cm of water; thiswas the minimum pressure required torestore the vein segment to the approximatediameter in vivo. The distended vessel wasfixed overnight by immersion in 3% buffered

45

Higman, Powell, Greenhalgh, Coady, Moss

glutaraldehyde. For scanning electron micro-scopy, the vessel was divided longitudinally,thoroughly washed with 0-1 M cacodylatebuffer (pH 7 2) for two hours, postfixed in 1% osmium tetroxide for one hour,dehydrated with graded ethanols (30-100%)for three hours and critical point dried.Specimens were then mounted on aluminiumstubs and coated with a layer of gold about50 nm thick. For transmission electronmicroscopy, specimens were post fixed in 2%aqueous osmium tetroxide, dehydratedthrough ascending grades of ethanol, andembedded in Spurr's resin. Ultrathin sectionswere stained with aqueous uranyl acetate andReynold's lead citrate.

EXAMINATION OF LIGHT AND ELECTRONMICROSCOPY SPECIMENSAll sections and micrographs were examinedblind,-that is, with no prior knowledge of thepatients concerned or their smoking habits.

IMMUNOGOLD STUDIESThe processing of samples was based on thetechnique of Woodrow et a19 Specimens wereplaced in 4% paraformaldehyde in 01 Mphosphate buffer (pH 7 2), for four hours at4°C, dissected into blocks 4 mm x 1 mm x1 mm, and stored in 1% paraformaldehydefor three days. The fixed samples were rinsedbriefly in phosphate buffer at 4°C. All furtherprocessing was carried out at about - 25°C.The tissue was rapidly dehydrated throughascending grades of ethanol over 30 minutes,infiltrated with 1:1 then 1:2 ethanol: LowicrylK4M mixtures (Taab Laboratories UK) for30 minutes, and left for one hour and thenovernight in fresh Lowicryl K4M. Embeddingwas carried out in brimming, sealed gelatincapsules of Lowicryl, with polymerisationunder ultraviolet light for 20 hours at -30°Cin a UVF-35 cold cabinet (Agar Scientific,Stanstead, UK), then by two days at roomtemperature.

Gold coloured sections were collected onuncoated nickel grids. These were labelledwith gold by immersion in drops of the fol-lowing solutions at room temperature: 1%ovalbumin (grade V, Sigma, UK) in modifiedphosphate buffered saline (MPBS) for 25minutes, primary antibody in MPBS for twohours, washes in five drops of MPBS over 15minutes, goat anti-rabbit IgG conjugated to10 nm gold (Biocell Research Laboratories,Cardiff, UK) in MPBS for 30 minutes,

Table Origin ofsaphenous vein samples and the presence of thickening of the basallamina

CotinineMedian age range Basal lamina

Number (male) (range, yr) (nmol/l) thickening

Non-smokers:Varicose vein correction 6 (3) 50 (40-65) 35-161 0/6Bypass surgery 4 (2) 72 (65-78) 38-82 0/4

Smokers:Varicose vein correction 6 (1) 40 (34-60) 207-1541 6/6Bypass surgery 4 (1) 71 (60-82) 958-1489 4/4

The immunogold labelling studies were carried out on a subset of samples obtained at surgeryfor correction of varicose veins. These veins were taken from four non-smokers aged 40-61(two men) and four smokers aged 39-54 (one man).

washes in five drops of MPBS over 10 min-utes, and a final wash in several drops of dis-tilled water. The sections were stained for 30minutes in 1% aqueous uranyl acetate and inReynolds' lead citrate for five minutes.The MPBS contained phosphate buffered

saline (pH 7 2) with 0-05% Tween 20, 0-1%bovine serum albumin (Sigma, UK) and 0-02M sodium azide. In control sections the pri-mary antiserum was omitted or replaced withnormal rabbit serum.

IMMUNOLOGICAL REAGENTSRabbit antihuman plasma fibronectin (A245)was obtained from DAKO, Buckinghamshire,UK, rabit antibovine heparan sulphate pro-teoglycans was a gift from Professor RGSpiro, Harvard Medical School, Boston,USA,10 rabbit antihuman placental type IVcollagen (10411) was obtained from InstitutPasteur de Lyon, France, and rabbit anti-human placental laminin (AB949) fromChemicon Intemational, London.

ResultsThe table shows the demographic details ofpatients. All the patients considered as non-smokers had cotinine concentrations < 200nmol/I and all the smokers had cotinine con-centrations in excess of this value. Lightmicroscopy examination of vein segments didnot show any gross intimal thickening ordamage in any of the specimens.

SCANNING ELECTRON MICROSCOPYCytoplasmic blebs were not seen in any of thespecimens examined. Microvilli were seen inspecimens from both smokers and non-smok-ers, with no apparent difference between thetwo groups. Intercellular boundaries wereconspicuous and cells were in close apposi-tion (fig 1). In some regions, separation ofendothelial cells had occurred, but this wasnot confined to specimens from smokers, andwas associated with splits in the cytoplasm,suggesting that the spaces between cells weredue to a preparation artifact.

TRANSMISSION ELECTRON MICROSCOPYWidening of the interendothelial cell junc-tions and subendothelial oedema were notnoted in any of the specimens examined.Intracellular organelles were generally wellpreserved and there was no dilatation ofendoplasmic reticulum or vacuolation of thecytoplasm. A distinct difference was noted,however, between specimens with respect tothe thickness of the endothelial basal lamina,and specimens could be divided into twogroups; those having normal basal lamina andthose showing areas of thickening. In the nor-mal group, the basal lamina appeared as anarrow continuous layer immediately beneaththe endothelial cells, and elastic fibres andbundles of microfibrils were seen beneath thebasal lamina (fig 2). In the other group, thebasal lamina was grossly thickened, up to

about 50 times normal, and appeared as awide amorphous or faintly fibrillar layer

46

Is thickening of the basal lamina in the saphenous vein a hallmark ofsmoking?

Figure 1 Scanningelectron micrograph ofsaphenous vein from anon-smoker. Bulges (*)denote the position of thenuclei. Cell boundaries areconspicuous (arrow) andsplits are seen in theadjacent cytoplasm (whitearrow). Originallyx 2600.

beneath the endothelial cells (fig 3). In someregions the thickening was continuous, run-ning beneath several cells, whereas in others,the gross thickening seemed to be focal, withadjacent areas of basal lamina being of nor-mal thickness (fig 3). The basal lamina wasalways distinct from the banded collagen andmicrofibrils. Areas of thickened basal laminawere seen in specimens from both the vari-cose group and the bypass group.When the serum cotinine concentrations of

the patients of these specimens were com-pared, it was found that specimens containingareas of thickened endothelial basal laminawere from patients with high cotinine concen-trations (smokers, table). Although non-thickened normal basal lamina was seen inspecimens from both smokers and non-

smokers, areas of thickening were not seen inspecimens from the non-smoking group.

IMMUNOGOLD STUDIESIn specimens from non-smokers, immuno-gold labelling for fibronectin was seen alongregions of the basal lamina of endothelial cellsand the external lamina of smooth musclecells (fig 4); similar regions were also seen inspecimens from smokers. Also in smokers,there were accumulations of numerous goldparticles over the finely fibrillar material ofthe thickened regions of basal lamina (fig 5).Fibronectin however was not associated withsubintimal bundles of microfibrils. The typi-cal amorphous thickening of basal laminaseen on routine section was not identified onthe Lowicryl sections, in which the thickened

Figure 2 Electronmicrograph ofsaphenousvein intima from a non-smoker showing thinendothelial basal lamina(arrow), microfibrils (m)and intimal smooth musclecell (s). Originallyx 21 000.

47

Higman, Powell, Greenhalgh, Coady, Moss

Figure 3 Electronmicrograph ofsaphenousvein intima from a smokershowingfocal thickening ofthe endothelial basallamina (bl), adjacentbasal lamina of normalthickness (arrow), coUlagenfibres (c) and the externallamina (ex) ofa smoothmuscle cell (s). Originallyx 21 000.

regions appeared as finely fibrillar material,similar morphologically to the subintimalbundles of microfibrils. Therefore it was diffi-cult to correlate the morphology of tissueembedded in Spurr's resin for routine elec-tron microscopy with the tissue embedded inLowicryl for immunogold labelling, whichhad low contrast.

Heparan sulphate proteoglycans were

localised to the endothelial basal lamina andexternal lamina of the smooth muscle cells, apattern similar to that seen in the fibronectinstained sections from the non-smoking group,but there was no increase in the thickenedbasal lamina seen in sections from smokers.Immunogold labelling for type IV collagenand laminin was diffuse throughout thesubendothelium, and for type IV collagen was

Figure 4 Electronmicrographs oflabellingforfibronectin in saphenousvein from a non-smoker toshow immunogold paricleslocalised mainly to thebasal lamina of theendothelial cell (arrows)and external lamina of thesmooth muscle cell (shortarrows). Originallyx 44 000.

48

Is thickening of the basal lamina in the saphenous vein a hallmark ofsmoking?

Figure 5 Electronmicrograph of labelling forfibronectin in thesaphenous vein from asmoker to show localisationofgold particles toaccumulations of materialin the region of theendothelial basal lamina;microfibrils. Ornginallyx 42 000.

associated particularly with the subintimalbundles of microfibrils. No localisation ofgold particles was seen associated with thefinely fibrillar thickening of the basal laminapresent in smokers, and there was no obviousdifference in distribution between specimensfrom non-smokers and smokers.

DiscussionSmoking related ultrastructural changes havebeen reported in human umbilical and uter-ine arteries.78 On scanning electronmicroscopy, gaps between endothelial cellswere associated with smoking in both vesselsbut this finding has not been confirmed inour study of the saphenous vein. Separationof the cells seemed to be artifactually pro-duced by shrinkage of the specimen, whichcan occur during critical point drying." Theblebs described in human umbilical arteriesfrom smokers were not seen in the uterineartery, or in our study of the saphenous vein,and would not therefore be a consistentmarker of smoking.The most prominent feature associated

with smoking on transmission electron micro-scopy in umbilical arteries was focal thicken-ing of the basal lamina, and subendothelialoedema. Less noticeable differences weredilatation of the endothelial endoplasmicreticulum and widening of endothelial celljunctions. Thickened basal lamina has beenreported in conditions other than smoking. Itis well documented in diabetic microangiopa-thy,'2 13 and has also been reported as an ageand sex related change in muscle capillaries,being greatest in elderly men.'4 In animalmodels of endothelial injury, thickenedendothelial basal lamina has been found afterexposure to radiation or cytotoxic drugs.'5

In this study, focal thickening of theendothelial basal lamina enabled discrimina-tion between saphenous vein specimens fromsmokers and non-smokers, when examinedblind with no prior knowledge of the patient'ssmoking habits, irrespective of whether thevein was harvested at bypass surgery orsurgery for varicose veins. This morphologicalfinding is in agreement with the reportedstudies on human umbilical vessels. Diabetesand advanced age can be excluded as causesof this thickening- as diabetic patients werenot included in this study, and thickening ofthe basal lamina was not seen in the non-smokers of the older bypass group. By con-trast with the reports on umbilical vessels,subendothelial oedema, dilatation of endo-plasmic reticulum, and widening of inter-cellular junctions were not seen in thesaphenous vein of smokers.The endothelial basal lamina is a complex

mesh like structure and with immunohisto-chemical techniques the main protein compo-nents have been identified as fibronectin, typeIV collagen, heparan sulphate proteoglycans,and laminin.'6 17 Immunogold labelling ofsaphenous vein specimens from smokers sug-gest that there is a specific accumulation offibronectin in the thickened basal lamina.Fibronectin functions principally as an adhe-sion molecule in basal lamina, providinganchorage for endothelial cells. Recent exper-imental evidence has reported that the syn-thesis of fibronectin by endothelial cells maybe specifically increased by a rise in intra-cellular cyclic adenosine monophosphate.'8 Itis possible that this may be a mechanism bywhich circulating products of cigarette smokeinfluence the function of the endothelial cell,inducing the production of specific compo-nents of the basal lamina.

49

Higman, Powell, Greenhalgh, Coady, Moss

The venous wall is nurtured by the luminalblood supply; thickened endothelial basallamina may interfere with this process by act-ing as a barrier to the diffusion of macromole-cules across the intima. The thickened basallamina may also interfere with the normal sig-nalling and communication between endothe-lial cells and underlying mesenchymal cells.Such hindrance to cell to cell communicationby cytokines or endothelium derived relaxingfactor during the arterialisation of a vein graftcould contribute to the development of inti-mal hyperplasia and graft stenosis. Hence thethickened basal lamina in veins from smokerscould contribute to the increased rate of fail-ure of vein grafts in smokers. The thickenedbasal lamina may also have use as a patholog-ical marker of smoking, enabling the identifi-cation of the vein at risk of graft occlusion.Further work is required to determine moreprecisely the time scale over which thesechanges are reversible.Our results suggest that the endothelial

basal lamina is thickened in heavy smokersand that this thickening contains fibronectin.Smoking may therefore influence the functionof the endothelial cell and the structure of thevenous intima and contribute to an increasedrisk of graft occlusion.

We thank Mr Ian Shore for his help with the preparation ofspecimens for electron microscopy.

1 Kent KC, Whittemore AD, Mannick JA. Short-term andmidterm results of an all-autologous tissue policy forinfrainguinal reconstruction. J7 Vasc Surg. 1989;9:107-14.

2 Bourassa MG, Enjalbert M, Campeau L, Lesperance J.Progression of atherosclerosis in coronary artery andbypass grafts ten years later. Am J Cardiol 1984;53: 102C-7C.

3 Fitzgibbon GM, Leach AJ, Kafka HP. Atherosclerosis ofcoronary artery bypass grafts and smoking. Can MedAssoc J 1987;136:45-7.

4 Rutherford RB, Jones- DN, Bergqvist D. Factors affectingthe patency of infrainguinal bypass. J Vasc Surg 1988;8:236-46.

5 Wiseman S, Kenchington G, Dain R, et al. Influence ofsmoking and plasma factors on patency offemoropopliteal vein grafts. BMJ 1989;299:643-6.

6 Donaldson MC, Mannick JA, Whittemore AD. Causes ofprimary graft failure after in situ saphenous vein bypassgrafting. _J Vasc Surg 1992;15:113-20.

7 Asmussen I, Kjeldsen K. Intimal ultrastructure of humanumbilical arteries: observations on arteries from new-born children of smoking and non-smoking mothers. CirRes 1975;36:579-89.

8 Bylock A, Bondjers G, Jansson I, et al. Surface ultrastruc-ture of human arteries with special reference to theeffects of smoking. Acta Path Microbiol Scan 1979;87:201-9.

9 Woodrow DF, Shore I, Moss J, et al. Immunoelectronmicroscopic studies of immune complex deposits andbasement membrane components in IgA nephropathy.J Pathol 1989;157:47-57.

10 Mohan PS, Spiro RG. Macromolecular organisation ofbasement membranes. Characterisation and comparisonof glomerular basement membrane and lens capsulecomponents by immunochemical and lectin affinity pro-cedures. J Biol Chem 1986;261:4328-36.

11 Fonkalsrud EW, Sanchez M, Zerubavel R, et al. Arterialendothelial changes after ischaemia and perfusion. SurgGynecol Obstet 1976;142:715-21.

12 Bloom A. The nature of diabetes. J R Soc Med 1978;71:170-83.

13 Ashton N. Vascular basement membrane changes in dia-betic retinopathy. BrJ Ophthalmol 1974;58:344-51.

14 Kilo C, Vogler N, Williamson JR. Muscle capillarybasement membrane changes related to ageing and todiabetes mellitus. Diabetes 1972;21:881-99.

15 Pierce GB, Nakane PK. Basement membranes: synthesisand deposition in response to cellular injury. Lab Invest1969;21:27-37.

16 Martinez-Hernandez A, Amenta PS. The basement mem-brane in pathology. Lab Invest 1983;48:656-77.

17 Fujiwara S, Wiedemann H, Timpl R, et al. Structure andinteractions of heparan sulphate proteoglycans from amouse tumour basement membrane. Eur J Biochem1984;143: 145-57.

18 Cagliero E, Roth T, Roy S, et al. Expression of genesrelated to the extracellular matrix in human endothelialcells. J Biol Chem 199 1;266: 14244-50.

50