Abstract In human immunodeficiency virus 1 (HIV-1)-in-fected patients, a hypoperfusion is seen by SPECT analy-ses in different brain regions but a specific pattern for thepredominance of a specific brain region has not been found.The vessels of the cerebral cortex of the frontal, temporal,parietal, and occipital lobes of acquired immunodeficiencysyndrome (AIDS) brains and control brains were analyzedby immunohistochemistry and lectin histochemistry. Im-munohistochemistry was performed for collagen IV, laminin(basal lamina), and factor VIII (endothelial cell) and lectinhistochemistry [Ricinus communis agglutinin (RCA-I), Ulexeuropaeus agglutinin (UEA-I), wheatgerm agglutinin (WGA)and soybean agglutinin (SBA)] was used to study changesof glycoproteins in the endothelial cell membrane. Vesselswere counted in the gray and white matter, and their stain-ing intensity for the different antibodies and lectins wasrated using a three-point scale. Immunoreactivity for col-lagen IV was reduced in AIDS brains, which may be re-lated to thinning of the basal lamina of cerebral vessels, ashas previously been shown by electron microscopy. Lectinhistochemistry with SBA, UEA-I and WGA indicated lossof glycoproteins in the membrane of endothelial cells. Thedata from the present study show morphological changesof the endothelial cells and of the basal lamina in the brainof individuals with AIDS, and might represent the mor-phological sequelae of a disturbed blood-brain barrier, ormay account for the hypoperfusion seen in SPECT analy-ses.

Key words Acquired immunodeficiency syndrome ·Blood-brain barrier · Cerebral vessels · Immunohistochemistry · Lectin histochemistry

Introduction

Infection of the brain by the human immunodeficiencyvirus 1 (HIV-1) leads to a variety of neurological and psy-chiatric signs and symptoms [14, 18, 22, 28]. The neuro-pathological changes encountered in HIV-1-infected brainsshould not be underestimated, since they are found in about95% of cases examined [3, 35]. They encompass HIV-1-specific changes and HIV-1-associated changes, as well aschanges due to secondary opportunistic infections and neo-plasias. HIV-1-specific and HIV-1-associated neuropatho-logical changes include HIV-1 encephalitis, HIV-1 leuko-encephalopathy, vacuolar myelopathy, vacuolar leukoence-phalopathy, and lymphocytic meningitis [3, 35]. Oppor-tunistic changes include infections with cytomegalovirus,papovavirus, herpes simplex virus, Toxoplasma gondii, As-pergillus, Candida albicans, and Cryptoccoccus neoformans[11, 15].

In HIV-1-infected patients, a focal or multifocal hy-poperfusion in different brain regions has been seen byHMPAO-SPECT analyses [17, 21, 26, 30]. The observedhypoperfusion was found in 69–100% of the patients ana-lyzed. A specific pattern for a predominance of a specificbrain region could not be determined, and possible mor-phological substrates for the hypoperfusion were not ad-vanced by the authors. Thus, it remains unclear if the hy-poperfusion is due to circulatory disturbances or to a re-duced demand from neurons.

Several authors have shown alterations of the corticalmicrovasculature in brains of HIV-1-infected patients in-dicating damage to the blood-brain barrier (BBB) [4, 29,34]. In HIV-leukoencephalopathy, changes in the cerebralvessels including mural thickening, increased cellularity,and enlargement and pleomorphism of endothelial cellshave been described [9, 27]. Leakage of serum proteinsthrough the BBB in the acquired immunodeficiency syn-drome (AIDS) was recently reported [20, 23]. In part I ofthis study, changes in cerebral vessels were reported at thelight and electron microscopic levels, as assessed by mor-phometry (Weis et al., submitted). The diameter, surface

Andreas Büttner · Parviz Mehraein · Serge Weis

Vascular changes in the cerebral cortex in HIV-1 infectionII. An immunohistochemical and lectinhistochemical investigation

Acta Neuropathol (1996) 92 :35–41 © Springer-Verlag 1996

Received: 17 February 1995 / Revised, accepted: 7 December 1995

REGULAR PAPER

A. Büttner · P. Mehraein · S. Weis (Y)Institute of Neuropathology, Ludwig-Maximilians University, Thalkirchnerstrasse 36, D-80337 Munich, GermanyFax: 5160 5396

area density, and the volume fraction of cerebral vesselswere significantly increased in AIDS brains. A thinningof the basal lamina was noted at the ultrastructural level.

The anatomical substrate of the BBB are the non-fen-estrated capillaries surrounded by the foot processes of as-trocytes [10]. Capillaries are defined as vessels consistingof endothelial cells, basal lamina, and pericytes [33]; theendothelial cells are sealed together by continuous tightjunctions. Changes in capillaries have been reported in partI of this study (Weis et al., submitted). The basal lamina iscomposed of extracellular matrix proteins, the major com-ponents of which are collagen IV, laminin, heparan sulfatepreoteoglycan, fibronectin, nidogen, and entactin [7, 16].

Numerous glycoproteins can be demonstrated by lectinhistochemistry on the surface of the endothelial cell mem-brane [6]. These glycoproteins play an important role inintercellular recognition, cell adhesion, cell differentiation,signal transfer, receptor function, and transport processes[8]. Under pathological conditions, i.e. hypertension orbrain tumors, an increased endothelial permeability is as-sociated with altered surface charges due to changes ofglycoproteins [19]. An altered pattern of glycoproteins inthe endothelial cell membrane of capillaries has been seenwithin and at the borders of brain tumor tissue [8]. Thisfinding was associated with a defective BBB.

The manner in which HIV-1 invades the central nervoussystem is still unclear. Entry via infected lymphocytes andmonocytes is the most probable mechanism [1–3, 5, 12,36]; however, infection of endothelial cells with consecu-tive disturbance in the permeability of the BBB was alsoadvanced as a possible mechanism [2, 5, 12, 24, 36].

Taking into account the alterations described above,the aim of the present study was to investigate, by meansof immunohistochemistry and lectin histochemistry, changesin the endothelial cells and the basal lamina of capillariesin the cerebral cortex of HIV-1-infected patients. Histo-chemistry was performed using different antibodies directedagainst endothelial proteins (factor VIII) and componentsof the basal lamina (collagen type IV, laminin), as well aslectins directed against various glycoproteins of the en-dothelial cell membrane.

Materials and methods

In the present study, postmortem brain specimens from 11 AIDSpatients and 10 age-matched controls were used. All patients weretreated at the University Hospital of Ludwig-Maximilians Univer-sity, Munich. The AIDS patients were in the later stages accordingto the Walter-Reed classification (WR 5 and WR 6). The age of theAIDS patients ranged from 29 to 59 years (mean age 38.5 years);the age of the control cases ranged from 25 to 59 years (mean age43.7 years). Autopsy was performed at the same institution and thebrain tissue of all patients was processed in our laboratory. In allcases, the postmortem interval did not exceed 24 h; there was nosignificant difference between the control and the AIDS group.The brain weights in the control group ranged from 1150 g to 1490 g(mean 1335 g) and in the AIDS group from 1240 g to 1610 g (mean1379 g).

The brains had been fixed for 1 week in a 4% PBS-bufferedformalin solution. The following sections were removed for thisstudy: (1) the frontal cortex at the level of the genu of the corpus

callosum, including the superior and medial frontal gyrus; (2) theparietal cortex 1 cm posterior to the postcentral gyrus, includingthe superior and inferior parietal lobulus; (3) the temporal cortex atthe level of the lateral geniculate body, including the hippocampus,the parahippocampal gyrus, the inferior and medial temporal gyrusand the lateral occipito-temporal gyrus; and (4) the occipital cortexat the level of the calcarine fissure, including the primary and sec-ondary visual cortex. The specimens were washed in running wa-ter, dehydrated, and embedded in paraffin. Routine histological ex-amination included the following staining methods: hematoxylinand eosin, cresyl-violet (Nissl stain), and Luxol-Fast-Blue. In ad-dition, immunohistochemical stains were used for the demonstra-tion and localization of HIV-1 antigens (gp41, p24).

Deparaffinized 5-µm-thick sections were washed in PBS andincubated in 10% H2O2 to inhibit endogenous peroxidase. Aftertreatment with nonspecific rabbit serum to block background stain-ing, sections were incubated with the monoclonal antibodies anti-collagen IV (dilution 1 :200), anti-laminin (dilution 1 :20) and anti-factor VIII (dilution 1 :100) (all antibodies purchased from DAKO,Hamburg, Germany) for 1 h at room temperature. Sections werethen treated with a rabbit anti-mouse antibody (dilution 1 :400) anda streptavidin/peroxidase complex (dilution 1 :1000) (from DAKO).Peroxidase was developed using a H2O2/diaminobenzidine chromo-gen solution. The sections were counterstained with hematoxylin.Sections treated with anti-collagen IV antibody were predigestedwith protease (Sigma, Deisenhofen, Germany). Immunohistochemi-cal controls included omission of the first antibody.

Lectin histochemistry was performed with Ricinus communisagglutinin (RCA-I; dilution 1 :100), Ulex europaeus agglutinin(UEA-I; dilution 1 :200), wheatgerm agglutinin (WGA; dilution 1 :200), and soybean agglutinin (SBA; dilution 1 :200. All lectinswere purchased from Medac Hamburg, Germany).

Vessels in the gray and white matter were counted at 400 ×magnification. The vessels in the gray matter were analyzed fol-lowing “systematic row sampling,” whereas the vessels in the whitematter were analyzed following “random systematic sampling.”The details of both sampling strategies have been described in de-tail elsewhere [31]. Briefly, systematic row sampling is done in thefollowing way: the first measuring field (MF) is positioned at thepial surface. After measurement of the vessels in the MF is com-pleted, the MF is moved downwards and positioned adjacent to thefirst one. MFs are positioned in this way until the white matter isreached. The result of positioning the MFs and measuring vesselswithin each MF is called a vertical “row.” Then, the MF is posi-tioned again at the pial surface laterally adjacent to the first MFand a second row is evaluated. Random systematic sampling isused for investigation of the white matter and subcortical gray mat-ter. In random systematic sampling the first MF is positioned ran-domly in the white matter; the subsequent MFs are positionedequidistant to each other.

The intensity of the immunostaining for every antibody andlectin was evaluated using a three-point rating scale. Staining reac-tivity was denoted as 1 for mild immunoreactivity, 2 for moderateimmunoreactivity, and 3 for strong immunoreactivity. After a cer-tain period of time, the same material was re-examined by thesame investigator using the three-point rating scale. Identical re-sults were obtained in 98.3% of cases.

The density of vessels with a defined staining intensity was cal-culated according to the following formula:

Vessel density = number of vessels/ (number × area of measur-ing fields)

The profile area of each measuring field was 0.15 mm2. Finally,the vessel density was calculated as the number of vessels per squaremillimeter.

During the staining procedures, the tissues from each case and eachregion were processed together with its matched control. In orderto assess the variation in the staining qualities, the day of process-ing and the number of each batch of histological sections were en-tered as variables in the statistical analyses. There were no signifi-cant differences that could be attributed to the staining procedure.

The data were processed with the statistical program SPSS(Statistical Package for the Social Sciences) on an IBM-compati-

36

ble computer (HIGHSCREEN PC 486). The one-way analysis ofvariance (ANOVA) and the nonparametric Mann-Whitney U-testand the Kruskal-Wallis H-test were used.

Results

The histopathological examination revealed no neuropatho-logical changes in the control brains, and in the AIDSbrains neither HIV-1 specific neuropathological changesnor changes due to opportunistic infectious agents werefound. Immunohistochemistry for gp41 and p24 was neg-ative in all of the AIDS brains.

Basal lamina

Table 1 and Fig.1 show the staining intensity of corticalvessels for collagen IV in AIDS brains as compared tocontrol brains. In AIDS brains, immunohistochemistry forcollagen type IV revealed a reduced number of vesselsshowing strong immunoreactivity, whereas the number ofvessels with mild immunoreactivity was significantly in-

creased as compared to controls. The number of vesselswith moderate immunostaining was not significantlychanged. This pattern was observed in all four cortical regions, and refers to the gray matter as well as the whitematter, with minor changes.

Tables 2 and 3 show the staining intensity of corticalvessels for laminin in AIDS brains as compared to controlbrains. Immunohistochemistry for laminin revealed a re-duced number of vessels showing mild immunoreactivity,whereas the number of vessels with strong immunoreac-tivity was significantly increased in AIDS brains as com-

37

Fig. 1 Numerical density (n/mm2) of microvessels in HIV-1 infec-tion in cortical gray matter of the temporal and parietal lobes, ac-cording to the staining intensity for collagen IV (T temporal lobe,P parietal lobe, 1 mild immunoreactivity, 2 moderate immunore-activity, 3 strong immunoreactivity)

Table 1 Numerical density (n/mm2) of vessels in the gray matter,according to the staining intensity for collagen type IV (1 mild im-munoreactivity, 2 moderate immunoreactivity, 3 strong immunore-activity)

Region Rating Controls AIDSscore Mean (SE) Mean (SE) P

Frontal 1 13.60 (3.84) 40.44 (9.33) 0.022 54.61 (8.90) 60.22 (7.20) 0.573 55.07 (12.75) 24.91 (5.90) 0.02Total 123.11 (11.16) 125.56 (5.15) 0.53

Occipital 1 16.56 (5.65) 60.49 (10.42) 0.0032 50.57 (5.62) 50.56 (7.65) 0.943 71.06 (18.65) 16.79 (4.82) 0.06Total 138.19 (14.93) 127.88 (5.39) 0.40

Table 3 Numerical density (n/mm2) of vessels in the white mat-ter, according to the staining intensity for laminin

Region Rating Controls AIDSscorea Mean (SE) Mean (SE) P

Frontal 1 17.47 (4.69) 3.76 (1.34) 0.0092 21.60 (3.80) 11.88 (2.12) 0.053 24.53 (5.77) 55.88 (3.28) 0.002Total 63.60 (3.02) 71.52 (4.44) 0.31

Temporal 1 20.00 (5.12) 7.03 (2.27) 0.042 17.87 (2.80) 12.24 (2.60) 0.143 26.67 (6.26) 44.61 (5.65) 0.03Total 64.53 (4.97) 63.88 (3.05) 0.72

Parietal 1 23.60 (5.58) 4.61 (1.49) 0.0072 18.80 (1.96) 9.46 (2.35) 0.013 16.80 (4.60) 46.18 (4.72) 0.001Total 59.20 (2.19) 60.24 (2.43) 0.70

Occipital 1 16.80 (4.37) 1.94 (0.58) 0.00042 19.74 (3.30) 6.42 (1.50) 0.0073 26.13 (6.80) 62.18 (4.34) 0.002Total 62.67 (3.68) 70.67 (4.73) 0.36

a See Table 1 for explanations

Table 2 Numerical density (n/mm2) of vessels in the gray matter,according to the staining intensity for laminin

Region Rating Controls AIDSscorea Mean (SE) Mean (SE) P

Frontal 1 74.90 (9.45) 23.58 (6.66) 0.0012 45.32 (9.25) 43.06 (4.07) 0.733 18.35 (6.56) 100.00 (13.99) 0.0004Total 138.57 (8.67) 166.63 (7.88) 0.04

Temporal 1 64.96 (8.20) 26.21 (4.22) 0.0012 26.52 (6.65) 43.23 (4.51) 0.913 18.92 (9.46) 77.17 (13.09) 0.002Total 110.39 (8.73) 146.31 (8.15) 0.01

Parietal 1 68.22 (12.52) 18.51 (5.72) 0.0052 40.89 (11.51) 41.52 (6.21) 0.703 29.56 (11.80) 110.16 (11.36) 0.0006Total 138.67 (12.09) 170.19 (6.60) 0.05

Occipital 1 79.94 (9.11) 13.27 (3.44) 0.00012 41.72 (12.20) 36.40 (3.99) 0.753 17.95 (7.34) 136.83 (9.67) 0.0001Total 139.94 (11.95) 185.13 (6.43) 0.006

a see Table 1 for explanations

pared to controls. The number of vessels with moderateimmunostaining did not differ significantly between thegroups. The numerical density of vessels in the gray mat-ter was increased in AIDS brains, whereas the numericaldensity of vessels in the white matter did not differ be-tween the groups. In general, this pattern could be ob-served in all four cortical regions and is valid for the graymatter as well as for the white matter, with minor changes.

Endothelial cells

In AIDS brains, immunostaining for factor VIII revealeda significantly reduced number of vessels showing strongimmunoreactivity, whereas the number of vessels with mildimmunoreactivity was increased as compared to controlbrains. The number of vessels with moderate immuno-staining did not differ significantly between the groups.The numerical density of vessels in the gray matter wasincreased in AIDS brains, whereas the numerical densityof vessels in the white matter did not differ between thegroups. In general, this pattern could be observed in allfour cortical regions, in the gray matter as well as in thewhite matter, with minor changes.

Table 4 and Fig.2 show the results of the staining in-tensity of cortical vessels for WGA in AIDS brains as com-pared to control brains. In AIDS brains, immunohisto-chemistry for WGA revealed a reduced number of vesselsshowing strong immunoreactivity, whereas the number ofvessels with mild immunoreactivity was significantly in-creased as compared to controls. The number of vesselswith moderate immunostaining and the numerical densityof vessels were not significantly changed. This patternwas observed in all four cortical regions and holds true for

the gray matter as well as for the white matter, with minorchanges.

The results of the staining intensity of cortical vesselsfor UEA-I and SBA in AIDS brains as compared to con-

38

Table 4 Numerical density (n/mm2) of vessels in the gray matter,according to the staining intensity for wheatgerm agglutinin (WGA)

Region Rating Control AIDSscorea Mean (SE) Mean (SE) P

Frontal 1 20.56 (4.66) 63.24 (8.54) 0.0012 46.78 (4.54) 46.86 (4.38) 0.833 68.56 (14.89) 24.87 (6.38) 0.009Total 135.89 (12.79) 134.97 (5.50) 0.78

Temporal 1 16.46 (3.46) 61.74 (6.87) 0.00032 35.43 (2.18) 21.36 (5.64) 0.063 56.57 (12.27) 13.50 (5.36) 0.004Total 108.46 (10.71) 96.60 (6.12) 0.53

Parietal 1 22.17 (4.62) 65.34 (9.69) 0.0022 49.56 (4.94) 48.03 (5.38) 0.943 61.83 (13.44) 20.94 (3.91) 0.02Total 133.56 (13.09) 134.31 (7.14) 0.94

Occipital 1 17.16 (4.44) 66.69 (9.34) 0.00092 44.02 (8.62) 47.53 (8.21) 0.623 86.28 (16.39) 34.22 (10.11) 0.02Total 147.45 (15.23) 148.45 (12.16) 0.83

a See Table 1 for explanations

Fig.3 Immunohistochemistry for collagen type IV, demonstratingthe staining intensities of the basement membrane ranging frommild ( ), moderate (x) to strong (c) immunoreactivity (tempo-ral cortex, AIDS patient, magnification × 120)

Fig.2 Numerical density (n/mm2) of microvessels in HIV-1 infec-tion in the cortical white matter, according to the staining intensityfor wheat germ agglutinin (WGA) (F frontal lobe, O occipitallobe; 1: for other abbreviations see Fig.1)

trol brains are not shown in table form. Briefly, lectin his-tochemistry for UEA-I and SBA showed a reduced num-ber of vessels with strong immunoreactivity, whereas thenumber of vessels with mild immunoreactivity was in-creased in AIDS brains as compared to controls. In gen-eral, this pattern was observed in all four cortical (frontal,temporal, parietal, occipital) regions, in both the gray andwhite matter, with minor changes.

For all four cortical regions including gray and whitematter, no significant differences were found in the histo-chemical staining intensity for RCA-I between the AIDSgroup and the control group. Microglial cells were also la-belled by RCA-I; they could easily be differentiated fromendothelial cells, but were not included in this investiga-tion.

Discussion

In the present study, a reduced immunoreactivity for col-lagen IV of the basal lamina of cortical vessels was found

in AIDS brains as compared to age-matched controls. Aspecific pattern for their predominance in a specific brainregion could not be found. The alterations were heteroge-neously distributed among the frontal, temporal, parietal,and occipital brain regions. There is a parallel betweenour morphological observations and those from HMPAO-SPECT analyses. HMPAO-SPECT is used for perfusionstudies of the cerebral cortex whereas (S)-N-[(1-ethyl-2-pyr-rolidinyl)]methyl-2-hydroxy-3-iodo-6-methoxybenzamide([123I]) (IBZM) perfusion studies are especially useful forinvestigation of the basal ganglia. As a matter of fact, thein vivo studies with HMPAO-SPECT showed the sameheterogeneous distribution of the uptake defects. To ourknowledge, no studies with IBZM in HIV-1 infected pa-tients have been performed so far. Furthermore, the find-ings of the present study correlate well with the quantita-tive electron microscopical data of area 7 and 11 presentedin part I of this investigation (Weis et al., submitted), whichshowed a significantly reduced thickness of the basal laminaof cortical vessels in AIDS brains.

The immunoreactivity for laminin was significantly in-creased in cortical vessels of AIDS brains. This finding isin accordance with data reported by Taruscio et al. [29],who found a marked increase in reactivity for laminin incapillary walls in AIDS cases and suggested that thesemodifications may play a role in the pathogenesis of neu-rological dysfunction. They did not find significant differ-ences in immunoreactivity for collagen IV between AIDSand control brains. These authors did not use a morpho-metric approach and did not count the number of vesselswith a defined staining intensity, as was done in this in-vestigation. They interpreted the increased laminin reac-tivity as being related to the increase in the volume frac-tion, as preliminarily reported by our group [34]. Takinginto account the thinning of the basal lamina with loss ofcollagen IV, the increase of laminin reactivity can also beinterpreted as being only a relative increase. Thus, theamount of laminin may not be changed between the AIDSgroup and controls. However, this hypothesis has to beconfirmed by quantitative immunocytochemistry at theelectron microscopic level. Such studies might, however,be limited due to the poor tissue preservation for ultra-structural analyses.

Immunohistochemistry for factor VIII and lectin histo-chemistry for SBA, UEA-I, and WGA revealed a loss ofimmunoreactivity in the endothelial cell membrane inAIDS brains. Lectin histochemistry with RCA-I revealedno significant difference between the AIDS and controlgroup. Since numerous glycoproteins are present on theendothelial cell membrane, the data show that some gly-coproteins are changed in AIDS brains. The loss of glyco-proteins in the membrane of endothelial cells in HIV-1-in-fected individuals might be interpreted with regard to thealterations of endothelial cells seen at the light and elec-tron microscopic level [32, (Weis et al., submitted)]. Theloss of glycoproteins in the endothelial cell membrane canbe interpreted as a morphological substrate of the hypo-perfusion seen in SPECT analyses. However, it is not yetestablished if the observed alterations are responsible for

39

Fig. 4 Immunohistochemistry for factor VIII, demonstrating thestaining intensities of the endothelial cells ranging from mild ( ),moderate (x) to strong (c) immunoreactivity (temporal cortex,AIDS patient, magnification × 250)

the hypoperfusion seen in SPECT analyses, or for a dis-turbance of the BBB, or for both.

In the present study, the vessels of the grey and whitematter of the cerebral cortex were investigated. The ves-sels included capilleries, arterioles, and larger arteries, es-pecially of the white matter. Thus, one could observe inthe same histological specimen heterogenous staining forbasement membrane antigens, ranging from mild to strong.These results can be explained by the fact that the thick-ness of the basement membrane varies even within thegroup of capillaries, as was shown in part I of our study(Weis et al., submitted).

The changes in the basal lamina and the endothelialcell membrane might be due to a direct or indirect toxiceffect of HIV-1. However, the indirect toxic effects are themost likely to occur. Indeed, in this sample, HIV-1 antigenhas never been demonstrated by immunohistochemistry inendothelial cells [32, 35]. We therefore believe that an in-direct effect of HIV-1 infection might be the cause of theabove-mentioned changes. As described by other authors,tumor necrosis factor alpha, interleukin 1 (IL-1) or IL-6secreted by HIV-1-infected macrophages or multinucle-ated giant cells might be possible candidates [3, 13, 25].

In conclusion, changes in the basal lamina and the en-dothelial cells were demonstrated by immunohistochem-istry and lectin histochemistry. As shown in part I of thisinvestigation (Weis et al., submitted) these changes couldonly be detected by applying quantitative methods, in ad-dition to the histological techniques. At various levels,vascular changes were noted, which differ from those re-ported by other groups. Until now, the functional corre-lates of these changes cannot be clearly shown. Future in-vestigations have to be directed towards the elucidation ofthis problem. The combination of the data from animalmodels of BBB damage and from functional in vivo stud-ies will help in clarifying the above-mentioned vascularchanges seen HIV-1 infection.

Acknowledgements The authors are grateful to Mrs. A. Henn forexcellent technical assistance, and thank Ms. I. C. Llenos, M.D. forcorrecting the manuscript. This study was supported by the Ger-man Federal Ministry for Research and Technology (FKZ BGAIII-002-89/FVP).

References

1.Achim CL, Schrier RD, Wiley CA (1991) Immunopathogene-sis of HIV encephalitis. Brain Pathol 1 :177–184

2.Budka H (1989) Human immunodeficiency virus (HIV)-in-duced disease of the central nervous system: pathology and im-plications for pathogenesis. Acta Neuropathol 77 :225–236

3.Budka H (1991) Neuropathology of human immunodeficiencyvirus infection. Brain Pathol 1 :163–175

4.Cenacchi G, Giangaspero F, Martinelli M (1991) Immunohis-tochemical localization of laminin and type IV collagen in HIVencephalitis (abstract). Proceedings of the VII InternationalConference on AIDS, Florence, M.A. 1274

5.Dal Canto MC (1989) AIDS and the nervous system. Currentstatus and future perspectives. Hum Pathol 20 :410–418

6.Damjanov I (1987) Lectin cytochemistry and histochemistry.Lab Invest 57 :5–20

7.D’Ardenne AJ (1989) Use of basement membrane markers intumour diagnosis. J Clin Pathol 42 :449–457

8. Debbage PL, Gabius H-J, Bise K, Marguth F (1988) Cellularglycoconjugates and their potential endogenous receptor in thecerebral microvasculature in man: a glycohistochemical study.Eur J Cell Biol 46 :425–434

9. DeGirolami U, Smith TW, Hénin D, Hauw J-J (1990) Neuro-pathology of the acquired immunodeficiency syndrome. ArchPathol Lab Med 114 :643–655

10.Goldstein GW (1988) Endothelial cell-astrocyte interactions. Acellular model of the blood-brain barrier. Ann NY Acad Sci529 :31–39

11.Gray F, Gherard R, Scaravilli F (1988) The neuropathology ofthe acquired immune deficiency syndrome (AIDS). A review.Brain 111 :245–266

12.Ho DD, Pomerantz RJ, Kaplan JC (1987) Pathogenesis of in-fection with human immunodeficiency virus. N Engl J Med317 :278–286

13.Ho DD, Bredesen DE, Vinters HV, Daar ES (1989) The ac-quired immunodeficiency syndrome (AIDS) dementia complex.Ann Intern Med 111 :400–410

14.Levy RM, Bredesen DE, Rosenblum ML (1985) Neurologicalmanifestations of the acquired immunodeficiency syndrome(AIDS): experience at UCSF and review of the literature. J Neurosurg 62 :475–495

15.Levy RM, Bredesen DE, Rosenblum ML (1988) Opportunisticcentral nervous system pathology in patients with AIDS. AnnNeurol 23 [Suppl] : 7–12

16.Martinez-Hernadez A, Amenta PS (1983) The basement mem-brane in pathology. Lab Invest 48 :656–677

17.Masdeu JC, Yudd A, Van Heertum RL, Grundman M, Hriso E,O’Connell RA, Luck D, Camli U, King LN (1991) Single-pho-ton emission computed tomography in human immunodefi-ciency virus encephalopathy: a preliminary report. J Nucl Med32 :1471–1475

18.Naber D, Perro C, Schielke E, Goebel FD, Hippius H (1992)Neuropsychological deficits and other psychiatric symptoms inHIV-1 infected patients. In: Weis S, Hippius H (eds) HIV-1 in-fection of the central nervous system. Clinical, pathological,and molecular aspects. Hogrefe and Huber, Seattle, pp 67–85

19.Nag S (1984) Cerebral endothelial surface charge in hyperten-sion. Acta Neuropathol (Berl) 63 :276–281

20.Petito CK, Cash KS (1992) Blood-brain barrier abnormalitiesin the acquired immunodeficiency syndrome: immunohisto-chemical localization of serum proteins in postmortem brain.Ann Neurol 32 :658–666

21.Pohl P, Vogl G, Fill H, Rössler H, Zangerle R, Gerstenbrand F(1988) Single photon emission computed tomography in AIDSdementia complex. J Nucl Med 29 :1382–1386

22.Price RW, Brew B, Sidtis J, Rosenblum M, Scheck AC, ClearyP (1988) The brain in AIDS: central nervous system HIV-1 in-fection and AIDS dementia complex. Science 239 :586–592

23.Rhodes RH (1991) Evidence of serum-protein leakage acrossthe blood-brain barrier in the acquired immunodeficiency syn-drome. J Neuropathol Exp Neurol 50 :171–183

24.Rhodes RH, Ward JM (1991) AIDS meningoencephalomyelitis.Pathogenesis and changing neuropathologic findings. PatholAnnu 26 :247–275

25.Rosenblum MK (1990) Infection of the central nervous systemby the human immunodeficiency virus type 1. Morphology andrelation to syndromes of progressive encephalopathy and myelo-pathy in patients with AIDS. Pathol Annu 25 :117–169

26.Schielke E, Tatsch K, Pfister HW, Trenkwalder C, LeinsingerG, Kirsch CM, Matuschke A, Einhäupl KM (1989) Reducedcerebral blood flow in early stages of human immunodeficiencyvirus infections. Arch Neurol 47 :1342–1345

27.Smith TW, DeGirolami U, Hénin D, Bolgert F, Hauw J-J(1990) Human immunodeficiency virus (HIV) leucoencephalo-pathy and the microcirculation. J Neuropathol Exp Neurol 49 :357–370

28.Snider WD, Simpson DM, Nielsen S, Gold JWM, Metroka CE,Posner JB (1983) Neurological complications of acquired im-mune deficiency syndrome: analysis of 50 patients. Ann Neu-rol 14 :403–418

40

29. Taruscio D, Malchiodi Albedi F, Bagnato R, Bagnato R,Pauluzzi S, Francisci D, Cavaliere A, Donelli G (1991) In-creased reactivity of laminin in the basement membranes ofcapillary walls in AIDS brain cortex. Acta Neuropathol 81 :552–556

30. Tran Dinh YR, Mamo H, Cervoni J, Caulin C, Saimot AC(1990) Disturbances in the cerebral perfusion of human im-mune deficiency virus-1 seropositive asymptomatic subjects: aquantitative tomography study of 18 cases. J Nucl Med 31 :1601–1607

31. Weis S (1991) Morphometry in the neurosciences. In: WengerE, Dimitov L (eds) Image processing and computer graphics.Theory and applications. Oldenburg, Munich, pp 306–326

32. Weis S (1992) Morphometric aspects of the brain in HIV-1 in-fection. In: Weis S, Hippius H (eds) HIV-1 infection of thecentral nervous system. Clinical, pathological, and molecularaspects. Hogrefe and Huber, Seattle, pp 199–224

33.Weis S, Haug H (1989) Capillaries in the human cerebral cor-tex: a quantitative electronmicroscopical study. Acta Stereol 8 :139–144

34.Weis S, Haug H, Budka H (1988) Stereological investigation ofthe microvasculature of cerebral cortex in AIDS-demented pa-tients. Clin Neuropathol 7 :221

35.Weis S, Bise K, Llenos IC, Mehraein P (1992) Neuropatholog-ical features of the brain in HIV-1 infection. In: Weis S, Hip-pius H (eds) HIF-1 infection of the central nervous system.Clinical, pathological, and molecular aspects. Hogrefe and Hu-ber, Seattle, pp 159–190

36.Wiley CA, Schrier RD, Nelson JA, Lampert PW, OldstoneMBA (1986) Cellular localization of human immunodeficiencyvirus infection within the brains of acquired immune deficiencysyndrome patients. Proc Natl Acad Sci USA 83 :7089–7093

41