- Email: [email protected]
Distribution of HIV genomic DNA in brains of AIDS patients
Clinical
ELSEVIER
Distribution
and Diagnostic Virology 3 (1995) 61--72
of HIV genomic DNA in brains of AIDS patients
Larry E. Bockstahler a,*, Thomas Werner b, Herbert Festl ‘, Serge Weis I’, Karl Max Einhaeupl e, Volker Erfle ’ and Ruth Brack-Werner ’ ’ Molecular Biology Branch, DLS, OST, CDRH, Food and Drug Administration, Rockville, MD, USA Siiugetiergenetik, GSF-Forschungszentrum fiir Umweit und Gesundheit, Neuherberg, Germany b Institut fiir ’Institut fir Molekulare Virologie, GSF-Forschungszentrum fiir Umwelt und Gesundheit, Neuherberg. Germany d Institute of Neuropathology, Ludwig-Maximilians-University, Munich. Germany e Department of Neurology, Munich. Germany Received
21 October
1993; revision received
14 April 1994; accepted
19 April 1994
Abstract Background: Data concerning the distribution of HIV in the brains of AIDS patients at different stages of viral infection might contribute towards: ( 1) understanding the route(s) of HIV entry into the brain and virus dissemination within the brain and (2) establishing a possible correlation between the extent of CNS damage and the distribution of virus in AIDS brains. Objective: To determine the distribution of HIV- 1 genomic DNA within the brains of three deceased AIDS patients by polymerase chain reaction (PCR). Study design: The brains of three deceased AIDS patients were examined. Two brains had limited neuropathologic findings (brains I and II), and one brain (brain III) showed primary HIV-specific neuropathologic damage. Tissues were taken from different locations within each brain, and high molecular weight DNA isolated from the tissues was assessed for HIV-l genomic DNA by PCR. Results: HIV-l genomic DNA was found in all three brains, but the amount was low: order of magnitude of 1 viral genome per 1,000 cells. Multiple PCR analyses of DNA from brain I showed that the viral genomic DNA in this brain was non-uniformly distributed; only samples taken from the brainstem were clearly positive for HIV-l. HIV-l genomic DNA in brain II was found in portions of the lower and upper hemispheres. All but one of the brain III samples were clearly positive for HIV-I, and they had been taken from locations spread throughout this brain. Conclusions: Our results suggest that in early or latent stages of HIV-infection of the brain, viral genomic DNA is localized at restricted regions. At later stages this DNA is distributed
* Corresponding Abbreviations:
author. Fax: + 1 (301) 594-6775. PCR, polymerase chain reaction;
CNS, central
092%0197/95/$9.50 0 1995 Elsevier Science B.V. All rights reserved SSDI 0928-0197(94)00023-N
nervous
system
62
L.E. Bockstahler et al./Clinical and Diagnostic Virology 3 (1995) 61-72
more uniformly throughout the brain. Our data are compatible infection events followed by viral spreading within brain tissues.
with the concept of rare
Key words HIV-l; HIV-l proviral DNA; brain; polymerase chain reaction; HIV-l-associated cognitive/motor complex; AIDS dementia complex
1. Introduction
HIV infection of the central nervous system (CNS) is believed to occur early during systemic infection, progressing from an asymptomatic stage to full-blown disease (Chiodi and Fenyo, 1991; Cheng-Mayer and Levy, 1990; Epstein and Gendelman, 1993). AIDS-related neurological disease includes a plethora of clinical syndromes collectively termed HIV- 1-associated cognitive/motor complex (previously known as AIDS dementia complex) as well as CNS lesions and HIV leukoencephalopathy detectable by histopathology (Budka et al., 1991). CNS dysfunction is rare in the clinically asymptomatic stage of AIDS, whereas it is detected in the majority of AIDS patients prior to death (Price et al., 1991). Although the neuropathology of HIV infection has been defined at the tissue level (see, e.g., Budka, 1991), relatively little is known about exact mechanisms involved with the induction and development of HIV-specific CNS disease (Epstein and Gendelman, 1993). Studies concerning the association of HIV-l infection to CNS disease have been carried out mainly with brain specimens relating to stages of CNS infection in which patients exhibit severe neurological symptoms and visible signs of HIV-encephalitis (Budka, 1991). However, examination of AIDS brain specimens showing limited neuropathologic findings, which reflect earlier stages of CNS infection, is also important. Such examination is valuable in order to address questions concerning viral persistence and dissemination in the CNS and the relationship between the extent of neuropathologic findings and presence of the virus. The objective of this investigation was to detect HIV genomic DNA and determine its distribution in the brains of three deceased AIDS patients. Two of the brains selected had mild neurologic impairment, and the third showed extensive primary HIV-specific damage. The information obtained might contribute towards understanding the route(s) of HIV entry into the brain and virus dissemination within the brain.
2. Materials and methods 2.1. Patients HIV-l seropositivity of patient I, an adult male, was diagnosed 19 months before death and repeatedly confirmed. Neuropathological examination of the autopsied brain revealed no HIV-l specific neuropathologic changes (e.g., HIV encephalitis, HIV leukoencephalopathy). No signs of opportunistic infectious pathogens
L. E. Bockstahler et al./Clinicaland Diagnostic Virology 3 (1995) 61-72
63
(e.g., toxoplasma, cytomegalovirus) were observed. Malignant B-cell lymphoma was diagnosed in the right thalamus and extended into several neighboring tissues. Patient II, an adult male, was diagnosed as HIV-l seropositive five years before death, and the diagnosis was repeatedly confirmed. Neither HIV-l-specific lesions nor signs of opportunistic infections were observed upon gross anatomical examination of the autopsied brain. Malignant B-cell lymphoma (1 cm diameter) was diagnosed in the left parietal lobe about 4 cm distant from the hypothalamus. HIV-l seropositivity of patient III, an adult male, was diagnosed two years before death and repeatedly confirmed. Gross anatomical examination of the autopsied brain revealed typical neurological abnormalities associated with HIV-l infection: non-specific microglial nodules, vacuolar myelopathy, and HIV- 1 encephalitis. No signs of opportunistic infectious pathogens were visible in this brain. 2.2. DNA isolation Brain tissue samples (ca. 1 cm3) were taken from different locations of brains I, II and III after death and frozen at -70°C. High molecular weight DNA was isolated from the frozen tissues by standard procedures (Bowtell, 1987; Strauss et al., 1988). DNA samples from all 3 brains were precipitated with ethanol; genomic DNA was then spooled on a glass rod, dissolved in Tris-EDTA and stored at 4°C. DNA concentrations were determined by UV spectrophotometric analysis, and each DNA sample was examined by agarose gel electrophoresis and shown to contain high molecular weight DNA (data not shown). A DNA product of the expected size, as determined by agarose gel electrophoresis, could be PCR-amplified from 1 ug of each DNA sample using primers specific for the /3-actin gene (data not shown). 2.3. Polymerase chain reaction DNA samples (brain I) were assayed for the presence of HIV-l genomic DNA by amplifying a small DNA segment (loo-250 bp) from each of 3 different highly conserved regions (gag, LTR and pol) of the viral genome. DNA oligonucleotide primer pairs used for PCR included SK38/39 (gag), SK29/30 (LTR) (Ou et al., 1988) and HP4149N/4392C (pol) (Bootman and Kitchin, 1992). Oligonucleotide hybridization probes that were homologous to the amplified viral genomic DNA segments and used to confirm the PCR amplified DNA products as HIV-l specific by Southern blot hybridization analysis included SK19 (gag), SK31 (LTR) (Ou et al., 1988) and HP4316N (pol) (Bootman and Kitchin, 1992). The gag primer pair was found to detect HIV-l proviral DNA most frequently in experiments with brain I DNA samples and was therefore used exclusively for analysis of brain II and III DNA samples. PCR was performed as previously described (Bockstahler et al., 1992); the number of amplification cycles ranged from 35 to 40. PCR products were fractionated by agarose gel electrophoresis, transferred to nylon membranes and confirmed as HIV-l
64
L. E. Bockstahler et al.IClinical and Diagnostic Virology 3 (1995) 61-72
specific by hybridization with the homologous DNA oligonucleotide probes ( 5’[32P]end labelled) by procedures previously described (Bockstahler et al., 1992). At least 3 controls were used in each of the assay experiments performed: an HIV-l-positive control with genomic DNA’s from a chronically HIV-l infected cell line and two negative controls: DNA from uninfected human cells and a reagent (no DNA) control. Special care was taken to avoid contamination and amplicon ‘carry-over’. Steps taken included use of positive displacement pipettes or filtered pipette tips, frequent changing of laboratory clothing and gloves, aliquoting of reagents, physical isolation of PCR preparations and products and other recommendations of Kwok and Higuchi (1989). Throughout the study all DNA samples were coded, and the viral genome distribution in each brain was not determined until completion of PCR measurements.
3. Results 3.1. Brain I Initially, the DNA isolated from brain I (limited neuropathologic findings) was assayed for the presence of HIV-l specific DNA by Southern blot analysis using a 32P-labelled viral genomic DNA probe (pBHlO-R3). Although suitable positive and negative controls were utilized, HIV-l genomic DNA could not be detected in any of the samples taken from 11 different locations of this brain (data not shown). This result suggested that if any HIV-l genomic DNA were present in the samples, it was relatively low in amount. Each sample was then assayed by the highly sensitive PCR technique using primers designed to amplify a loo-250 bp DNA segment from each of 3 different highly conserved regions of the viral genome. Results of an assay Expt. for sequences from the gag region of the HIV-l genome are shown in Fig. 1 (Expt. 1 in Table 1). Five of the 11 brain DNA samples were HIV-l genomic DNA positive, including No. 7, which was weakly positive as determined from the original autoradiogram. Six samples were HIV-l-negative, and DNA from lung (not further investigated in this study) was HIV-l-positive. In this experiment control DNA samples were either HIV-l-positive, or HIV-l-negative as expected, Lane 13b of Fig. 1 and separate sensitivity experiments (data not shown) indicated that the sensitivity of our assay was about 15 molecules of HIV-l proviral DNA per ug total DNA using the positive control genomic DNA from chronically HIV- 1-infected glial cell line TH4-7-5 [single HIV-l provirus integration site (BrackWerner et al., 1992)]. Compiled results of 6 PCR experiments (Table 1) show that 3 of the brain DNA samples were clearly HIV-l-positive ( 1, 4 and lo), whereas 3 were HIV-l-negative (5,8 and 9) in at least 5 of 6 experiments. The remaining samples were indeterminate: 2, 3, 6, 7, and 11. Such PCR data scatter occurs with DNA samples that contain extremely low numbers of target HIV-l genomic DNA molecules (Bootman and Kitchin, 1992). Sample 2 was the only indeterminate sample positive for all three
L. E. Bockstahler
et al./Clinical
C
12
3
----------_-_
and Diagnostic
Virology 3 (1995) 61-72
5
6
7
8
9
13b13c
14
14
15
15
4
1011
65
C
114 bp-
B
C 12 13a -------------
16
16
C
114 bp-
Fig. 1. Detection of HIV-I in AIDS brain I DNA samples by polymerase chain reaction. (A,B) Representative autoradiograms showing Southern blot analysis of PCR-amplified DNA (40 cycles) using primer pair SK38/39 derived from gag region of HIV-l genome. Amplified DNA (expected 114 bp) was fractionated by agarose gel electrophoresis, transferred to nylon membranes and detected by hybridization with 32P-labelled DNA probe SK19. (A) Amplified DNA. Lanes l-11, from tissues taken from 11 different locations within brain I of AIDS patient I; C, positive control from chronically HIV-l infected T-lymphoma cells. (B) Amplified DNA. Lane 12, from lung tissue of patient; lane 13 (a-c), positive control from chronically HIV-l-infected glial cells [TH4-7-5 (Brack-Werner et al., 1992); # molecules of HIV-l proviral DNA per ug total DNA: a = 150, b = 15, c = 1.51; lane ‘-‘, blank; lane 14, negative control from uninfected T-lymphoma cells; lane 15, negative control from uninfected human skin fibroblast cells; lane 16, reagent control (no DNA); C, same as in (A).
of the viral genomic regions tested; this suggests that it may in fact be positive. Thus, the distribution of viral genomic DNA in this brain was non-uniform. The quantities of HIV-l genomic DNA in the HIV-l-positive samples from brain I (as well as from brains II and III) were low: order of magnitude of 1 viral genome per total of 1,000 cells as estimated by parallel PCR analysis of genomic DNA from chronically HIV- 1-infected glial cell line TH4-7-5 [single HIV- 1 provirus integration site (Brack-Werner et al., 1992)]. Fig. 2A shows the approximate locations of the 11 tissue samples in a schematic diagram of a human brain. The HIV-l-positive samples (1, 4, 10 and possibly positive 2) originated from tissues located in the midbrain, medulla oblongata and pons regions. HIV-l-negative samples 5, 8 and 9 were located in the frontal lobe (white matter), parieto-occipital lobe (white matter) and basal ganglia. The indeterminate samples were taken mainly from the hindbrain [ 3, 6, 11: cerebellum, parietooccipital lobe (gray matter) and temporal lobe].
66
L.E. Bockstahler
et a/./Clinical
and Diagnostic
Virology 3 (1995) 61-72
TABLE 1 Detection of HIV-l genomic DNA in AIDS patient brain I (limited neuropathologic polymerase chain reactionasb Expt. no.
Id 2 3 4 5 6
Primer pair
gag gag gag PO1 PO1 LTR
Overall result:’
findings) by
Brain sample no.’ 1
2
3
4
5
6
7
8
9
10
11
+e
_
_
+
_
+
+
_
_
+
_
++ ++ + ++ +
+ + +-+
+ +
++ + ++ -
+ _
+ + ++
+
-
-
++ + ++ ++
_
+
**+
_ --
-+-+
-+
+--+
++ nd *
a Compilation of results from 6 separate PCR experiments. b 23 out of 24 negative controls (l/2 normal human cell DNA and l/2 no DNA) were HIV-l negative; 30 out of 30 positive controls (DNA from HIV-infected human cells) were HIV-l-positive (Expt. 1-6; control data not shown). ’ DNA extracted from tissues taken from 11 different locations within AIDS patient’s brain. d Autoradiogram for Expt. 1 shown in Fig. 1. e Labels, + +, + , - denote high-level, low-level and negative, respectively, for relative intensities of bands observed in autoradiograms; nd, not done. ‘Overall result defined as HIV-l-positive (+) or -negative (-) if at least 5 of all 6 experiments were positive or negative, respectively; f denotes indeterminate result.
3.2. Brain II Tissues were taken from 8 different locations within brains II and III (without knowledge of brain I PCR results). DNA samples isolated from the tissues were assayed for the presence of HIV-l genomic DNA using the gag primers and probes. Compiled results of 6 PCR experiments on brain II (limited neuropathologic findings) DNA samples are shown in Table 2. Five of the brain DNA samples were HIV-l-positive (3, 4, 5, 7 and 12) in at least 5 of 6 experiments, whereas 3 were indeterminate (9, 11 and 13). None of the 8 samples gave an overall result of HIV- 1-negative. The approximate locations of the 8 tissue samples in a brain schematic diagram are shown in Fig. 2B. The HIV-l-positive samples 3, 4 and 12 were located in the lower hemisphere of the brain (cerebellum, medulla oblongata and temporal lobe), while samples 5 and 7 were located in the upper hemisphere (frontal lobe, white and gray matter). Thus, the HIV-l-positive samples of brain II were not localized in one region of the brain. 3.3. Brain III Table 3 shows a compilation of results of 6 PCR experiments performed on brain III (extensive neuropathologic findings) DNA samples. This DNA was isolated from
L.E. Bockstahler et al./CIinical and Diagnostic Virology 3 (199.5) 61-72
67
Fig. 2. Distribution of HIV-l genomic DNA in AIDS brains I, II and III. Three-dimensional view of brain designates approximate location of brain tissue samples/DNA’s used for polymerase chain reaction study. The brain is shown with the front left quarter section (planes xz and yz) cut out as indicated in the inset. The gray area on the left side of the figure represents uncut brain surface from the right hemisphere, which would be visible in a view approximately from the lower left corner of the inset, slightly below the plane of the inset, and looking towards the center of the brain. The xz plane indicates the lateral distribution of the samples, which are not necessarily located strictly on the surface shown, Samples are depicted in the cut sections of the brains for clarity even if they originated from the symmetric location of the other half of the brain. Major blood supply in the brain indicated by the round, dark grey region between samples 5 and 9. Sample numbers correspond to those used in the tables. Source of tissues: 1, midbrain; 2, pons (inside); 3, cerebellum; 4, medulla oblongata; 5, frontal lobe (white); 6, parieto-occipital lobe (gray); 7, frontal lobe (gray); 8, parieto-occipital lobe (white); 9, basal ganglia; 10, pons (outside); 11, temporal lobe; 12, temporal lobe; 13, ammon’s horn. (A) Brain 1, limited neuropathologic findings. (B) Brain II, limited neuropathologic findings. (C) Brain III, extensive neuropathologic findings.
8 different brain tissue samples taken from locations close to those of samples taken from brain II. Seven of the 8 DNA samples were HIV-l-positive in at least 5 of 6 experiments. One sample was indeterminate (12), and none gave an overall result of HIV- 1-negative.
68
L. E. Bockstahler
et al.IClinical and Diagnostic
Virology 3 (1995) 61-72
Fig. 2. (continued)
Fig. 2C shows the approximate locations of the 7 HIV-l-positive samples and indeterminate sample 12. The HIV-l-positive samples are distributed throughout brain III.
4. Discussion The main findings of this study are: (1) HIV-l genomic DNA was detected in all three brains investigated; it was found, as expected, in the tissues of brain III, which had extensive CNS damage, but it was also present in brains I and II even though they contained limited neuropathologic findings (lymphoma, but no primary HIV-l specific neuropathologic changes). (2) The number of copies of viral genomes found was low in all three brains (order of magnitude of 0.1% infected cells or less). (3) Some quantitative differences appeared to be present, e.g., brain III contained more clearly positive HIV-l samples than brains I and II. (4) The HIV-l-positive samples of brain I were concentrated in the brainstem, although lower amounts of viral
L. E. Bockstahler
et al./Clinical
and Diagnostic
Virology 3 (I 995) 61-72
69
Fig. 2. (continued)
genomic DNA (indeterminate samples) were observed in other regions of this brain. (5) Brain II was also HIV-l-positive for the brainstem area and other regions in the lower and upper hemispheres, and (6) in brain III, HIV-l DNA appeared to be more uniformly distributed throughout the brain both qualitatively and quantitatively as judged from the number of PCR experiments that yielded positive results. The number of clearly HIV-l-positive samples in brain III was greater than that found in brain I, but was not significantly greater than that found in brain II (Fig. 2). Also, the amount of viral genomic DNA in brain III positive samples in terms of per cent infected cells was not markedly greater than that found in brain I and II positive samples (data not shown). Therefore, no obvious correlation was found between the extent of damage observed by neuropathologic examination and the distribution of viral genomic DNA in these brains. Another reason for performing this study was to obtain data that might help illuminate the route(s) of HIV entry into the brain and dissemination of virus within the brain. We believe that brain II (Fig. 2B) may have been at a slightly later stage of CNS infection than brain I (Fig. 2A), because it contained more clearly positive
70
L. E. Bockstahler et al. JClinical and Diagnostic Virology 3 (1995) 61-72
Table 2 Detection of HIV-l genomic DNA in AIDS patient brain II (limited neuropathologic polymerase chain reactionasb Expt. no.
Primer pair
gag gag gag gag gag gag Overall result:’
findings) by
Brain sample no.’ 3”
4
5
7
9
11
12
13
+d ++ ++ + ++ ++
++ ++ + + ++
+ + + + ++
++ ++ ++ + ++ ++
+ _
+ + _
+ +
++ _ _
+ + + _ ++ +
+ + + + _
+
+
+
+
+
*
+
*
a Compilation of results from 6 separate PCR experiments. b 10 out of 12 negative controls (l/2 normal human cell DNA and l/2 no DNA) were HIV-l-negative; 6 out of 6 positive controls (DNA from HIV-infected human cells) were HIV-l-positive (Expts. 1-6; control data not shown). ’ DNA extracted from tissues taken from 8 different locations within AIDS patient’s brain. d Labels + f, f, - denote high-level, low-level and negative, respectively, for relative intensities of bands observed in autoradiograms. e Brain sample numbers represent approximately identical sampling locations in brains I, II and II. f Overall result defined as HIV-l-positive (+) if at least 5 of all 6 experiments were positive; k denotes indeterminate result.
HIV-l genomic DNA samples. If the results for brains I, II and III (Fig. 2) are compared, there is at least a tendency for increasing amounts of HIV-l genomic DNA to appear in areas close to the brainstem and upper hemisphere of the brain, and finally in the center of the brain, while the brainstem region appears to remain almost unchanged with respect to amount of viral genomic DNA. The amount of viral genomic DNA in the HIV-l-positive samples of all three brains was of the same order of magnitude (0.1% infected cells) regardless of neuropathologic findings. Our data could be interpreted in terms of virus entry into the brain and localization of virus at restricted regions followed by a cell-to-cell spread of viral infection over wider areas within the brain, rather than local build-up of infection. In other words, our findings are compatible with relatively rare, primary infection events at restricted areas followed by spreading of virus from these entry points, in contrast to multiple, simultaneous viral infection of different brain regions. Invasion of the CNS by HIV-l may have occurred by several mutually nonexclusive pathways including transport of viral infected macrophages/monocytes across the blood-brain barrier (Cheng-Mayer and Levy, 1990) and direct viral infection of choroid plexus (Harouse et al., 1989). The viral genomic DNA distribution pattern found in brain I (Fig. 2A), showing localization of HIV-l in the brainstem, suggests the possibility in this brain of neural spread of virus, a pathway known to be utilized by other neuropathogenic viruses such as poliovirus, herpesvirus and rabies virus (Tyler and Fields, 1990). If HIV-l had entered this brain via virus-
L. E. Bockstahler
et al.IClinical and Diagnostic
71
Virology 3 (1995) 61-72
Table 3 Detection of HIV-l genomic DNA in AIDS patient brain III (extensive neuropathologic polymerase chain reactio&’ Expt. no.
Primer pair
gag gag gag gag gag gag Overall result!
findings) by
Brain sample no.’ 3”
4
5
1
9
11
12
13
fd _ + ++ + +
++ ++ + + + _
++ ++ + + + +
+ + + + + ++
+ + + + +
+ ++ + ++ ++ ++
++ + + +
-__ __ __ _-
+
+
+
+
+
+
*
i-
*- +
’ Compilation of results from 6 separate PCR experiments. b 10 out of 12 negative controls (l/2 normal human cell DNA and l/2 no DNA) were HIV-l-negative; 6 out of 6 positive controls (DNA from HIV-infected human cells) were HIV-l-positive (Expts. L-6; control data not shown). ’ DNA extracted from tissues taken from 8 different locations within AIDS patient’s brain. d Labels ++, +, - denote high-level, low-level and negative, respectively, for relative intensities of bands observed in autoradiograms. ’ Brain sample numbers represent approximately identical sampling locations in brains I, II and III. ’Overall result defined as HIV-l positive (+) if at least 5 of all 6 experiments were positive; + denotes indeterminate result.
infected macrophages that had been transported there from peripheral blood, we would have expected samples close to the region of major blood supply (see round dark grey region between samples 5 and 9 in Fig. 2A) to have been HIV-l-positive rather than HIV- 1-negative or indeterminate.
Acknowledgements
This work was supported by Grant BGA III-002%89/FVP 2 from the Bundesgesundheitsamt and was performed in part during a U.S. Public Health Service Foreign Work/Study Program assignment in Neuherberg by one of us (LEB). The authors are grateful to T. Goerblich (GSF, Neuherberg, Germany) for advice and assistance with DNA isolations. We thank A. Kleinschmidt, A. Luz, W. Schmahl (GSF), L. Vitkovic (NIH, Rockville, Maryland, USA), A. Buchbinder (VAMC, New York, New York), J. Bootman (NIBSC, South Mimms, Hertfordshire, UK) and FDA colleagues for helpful discussions.
References Bockstahler, L.E., Carney, P.G., Bushar, G. and Sagripanti, J.L. (1992) Detection of Junin virus by the polymerase chain reaction. J. Virol. Methods 39, 231-235.
12
L. E. Bockstahler et al.IClinical and Diagnostic Virology 3 (1995) 61-72
Bootman, J.S. and Kitchin, P.A. (1992) An international collaborative study to assess a set of reference reagents for HIV-l PCR. J. Virol. Methods 37, 23-42. Bowtell, D.D.L. (1987) Rapid isolation of eukaryotic DNA. Anal. Biochem. 162, 463-465. Brack-Werner, R., Kleinschmidt, A., Ludvigsen, A., Mellert, W., Neumann, M., Herrmann, R., et al. (1992) Infection of human brain cells by HIV-l: restricted virus production in chronically infected glial cell lines. AIDS 6, 2733285. Budka, H. (1991) Neuropathology of human immunodeficiency virus infection. Brain Pathol. 1,163-175. Budka, H., Wiley, CA., Kleihues, P, Artigas, J., Asbury, A.K., Cho, E.-S., et al. ( 1991) HIV-associated disease of the nervous system: review of nomenclature and proposal for neuropathology-based terminology. Brain Pathol. 1, 1433152. Cheng-Mayer, C. and Levy, J.L. (1990) Human immunodeficiency virus infection of the CNS: characterization of ‘neurotropic’ strains. Current Topics Microbial. Immunol. 160, 1455156. Chiodi, F. and Fenyo, E.M. (1991) Neurotropism of human immunodeficiency virus. Brain Pathol. 1,185-191. Epstein, L.G. and Gendelman, H.E. (1993) Human immunodeficiency virus type 1 infection of the nervous system: pathogenetic mechanisms. Ann. Neurol. 33, 4299436. Harouse, J.M., Wroblewska, Z., Laughlin, M.A., Hickey, W.F., Schonwetter, B.S. and GonzalezScarano, F. (1989) Human choroid plexus cells can be latently infected with human immunodeficiency virus. Annals Neurol. 25, 406-411. Kwok, S. and Higuchi, R. (1989) Avoiding false positives with PCR. Nature 339, 237-238. Ou, C-Y., Kwok, S., Mitchell, S.W., Mack, D.H., Sninsky, J.J., Krebs, J.W. et al. (1988) DNA amplification for direct detection of HIV-l in DNA of peripheral blood mononuclear cells. Science 239, 2955297. Price, R.W., Sidtis, J.J. and Brew, B.J. (1991) AIDS dementia complex and HIV-l infection: a view from the clinic. Brain Pathol. 1, 155-162. Strauss, P. G., Schmidt, J., Pedersen, L. and Erfle, V. (1988) Amplification of endogenous proviral MuLV sequences in radiation-induced osteosarcomas. Int. J. Cancer 41, 616-621. Tyler, K.L. and Fields, B.N. (1990) Pathogenesis of viral infections. In: B.N Fields and D.M. Knipe (Eds.), Fields Virology, 2nd edn., Raven Press, New York, pp. 191-239.
ELSEVIER
Distribution
and Diagnostic Virology 3 (1995) 61--72
of HIV genomic DNA in brains of AIDS patients
Larry E. Bockstahler a,*, Thomas Werner b, Herbert Festl ‘, Serge Weis I’, Karl Max Einhaeupl e, Volker Erfle ’ and Ruth Brack-Werner ’ ’ Molecular Biology Branch, DLS, OST, CDRH, Food and Drug Administration, Rockville, MD, USA Siiugetiergenetik, GSF-Forschungszentrum fiir Umweit und Gesundheit, Neuherberg, Germany b Institut fiir ’Institut fir Molekulare Virologie, GSF-Forschungszentrum fiir Umwelt und Gesundheit, Neuherberg. Germany d Institute of Neuropathology, Ludwig-Maximilians-University, Munich. Germany e Department of Neurology, Munich. Germany Received
21 October
1993; revision received
14 April 1994; accepted
19 April 1994
Abstract Background: Data concerning the distribution of HIV in the brains of AIDS patients at different stages of viral infection might contribute towards: ( 1) understanding the route(s) of HIV entry into the brain and virus dissemination within the brain and (2) establishing a possible correlation between the extent of CNS damage and the distribution of virus in AIDS brains. Objective: To determine the distribution of HIV- 1 genomic DNA within the brains of three deceased AIDS patients by polymerase chain reaction (PCR). Study design: The brains of three deceased AIDS patients were examined. Two brains had limited neuropathologic findings (brains I and II), and one brain (brain III) showed primary HIV-specific neuropathologic damage. Tissues were taken from different locations within each brain, and high molecular weight DNA isolated from the tissues was assessed for HIV-l genomic DNA by PCR. Results: HIV-l genomic DNA was found in all three brains, but the amount was low: order of magnitude of 1 viral genome per 1,000 cells. Multiple PCR analyses of DNA from brain I showed that the viral genomic DNA in this brain was non-uniformly distributed; only samples taken from the brainstem were clearly positive for HIV-l. HIV-l genomic DNA in brain II was found in portions of the lower and upper hemispheres. All but one of the brain III samples were clearly positive for HIV-I, and they had been taken from locations spread throughout this brain. Conclusions: Our results suggest that in early or latent stages of HIV-infection of the brain, viral genomic DNA is localized at restricted regions. At later stages this DNA is distributed
* Corresponding Abbreviations:
author. Fax: + 1 (301) 594-6775. PCR, polymerase chain reaction;
CNS, central
092%0197/95/$9.50 0 1995 Elsevier Science B.V. All rights reserved SSDI 0928-0197(94)00023-N
nervous
system
62
L.E. Bockstahler et al./Clinical and Diagnostic Virology 3 (1995) 61-72
more uniformly throughout the brain. Our data are compatible infection events followed by viral spreading within brain tissues.
with the concept of rare
Key words HIV-l; HIV-l proviral DNA; brain; polymerase chain reaction; HIV-l-associated cognitive/motor complex; AIDS dementia complex
1. Introduction
HIV infection of the central nervous system (CNS) is believed to occur early during systemic infection, progressing from an asymptomatic stage to full-blown disease (Chiodi and Fenyo, 1991; Cheng-Mayer and Levy, 1990; Epstein and Gendelman, 1993). AIDS-related neurological disease includes a plethora of clinical syndromes collectively termed HIV- 1-associated cognitive/motor complex (previously known as AIDS dementia complex) as well as CNS lesions and HIV leukoencephalopathy detectable by histopathology (Budka et al., 1991). CNS dysfunction is rare in the clinically asymptomatic stage of AIDS, whereas it is detected in the majority of AIDS patients prior to death (Price et al., 1991). Although the neuropathology of HIV infection has been defined at the tissue level (see, e.g., Budka, 1991), relatively little is known about exact mechanisms involved with the induction and development of HIV-specific CNS disease (Epstein and Gendelman, 1993). Studies concerning the association of HIV-l infection to CNS disease have been carried out mainly with brain specimens relating to stages of CNS infection in which patients exhibit severe neurological symptoms and visible signs of HIV-encephalitis (Budka, 1991). However, examination of AIDS brain specimens showing limited neuropathologic findings, which reflect earlier stages of CNS infection, is also important. Such examination is valuable in order to address questions concerning viral persistence and dissemination in the CNS and the relationship between the extent of neuropathologic findings and presence of the virus. The objective of this investigation was to detect HIV genomic DNA and determine its distribution in the brains of three deceased AIDS patients. Two of the brains selected had mild neurologic impairment, and the third showed extensive primary HIV-specific damage. The information obtained might contribute towards understanding the route(s) of HIV entry into the brain and virus dissemination within the brain.
2. Materials and methods 2.1. Patients HIV-l seropositivity of patient I, an adult male, was diagnosed 19 months before death and repeatedly confirmed. Neuropathological examination of the autopsied brain revealed no HIV-l specific neuropathologic changes (e.g., HIV encephalitis, HIV leukoencephalopathy). No signs of opportunistic infectious pathogens
L. E. Bockstahler et al./Clinicaland Diagnostic Virology 3 (1995) 61-72
63
(e.g., toxoplasma, cytomegalovirus) were observed. Malignant B-cell lymphoma was diagnosed in the right thalamus and extended into several neighboring tissues. Patient II, an adult male, was diagnosed as HIV-l seropositive five years before death, and the diagnosis was repeatedly confirmed. Neither HIV-l-specific lesions nor signs of opportunistic infections were observed upon gross anatomical examination of the autopsied brain. Malignant B-cell lymphoma (1 cm diameter) was diagnosed in the left parietal lobe about 4 cm distant from the hypothalamus. HIV-l seropositivity of patient III, an adult male, was diagnosed two years before death and repeatedly confirmed. Gross anatomical examination of the autopsied brain revealed typical neurological abnormalities associated with HIV-l infection: non-specific microglial nodules, vacuolar myelopathy, and HIV- 1 encephalitis. No signs of opportunistic infectious pathogens were visible in this brain. 2.2. DNA isolation Brain tissue samples (ca. 1 cm3) were taken from different locations of brains I, II and III after death and frozen at -70°C. High molecular weight DNA was isolated from the frozen tissues by standard procedures (Bowtell, 1987; Strauss et al., 1988). DNA samples from all 3 brains were precipitated with ethanol; genomic DNA was then spooled on a glass rod, dissolved in Tris-EDTA and stored at 4°C. DNA concentrations were determined by UV spectrophotometric analysis, and each DNA sample was examined by agarose gel electrophoresis and shown to contain high molecular weight DNA (data not shown). A DNA product of the expected size, as determined by agarose gel electrophoresis, could be PCR-amplified from 1 ug of each DNA sample using primers specific for the /3-actin gene (data not shown). 2.3. Polymerase chain reaction DNA samples (brain I) were assayed for the presence of HIV-l genomic DNA by amplifying a small DNA segment (loo-250 bp) from each of 3 different highly conserved regions (gag, LTR and pol) of the viral genome. DNA oligonucleotide primer pairs used for PCR included SK38/39 (gag), SK29/30 (LTR) (Ou et al., 1988) and HP4149N/4392C (pol) (Bootman and Kitchin, 1992). Oligonucleotide hybridization probes that were homologous to the amplified viral genomic DNA segments and used to confirm the PCR amplified DNA products as HIV-l specific by Southern blot hybridization analysis included SK19 (gag), SK31 (LTR) (Ou et al., 1988) and HP4316N (pol) (Bootman and Kitchin, 1992). The gag primer pair was found to detect HIV-l proviral DNA most frequently in experiments with brain I DNA samples and was therefore used exclusively for analysis of brain II and III DNA samples. PCR was performed as previously described (Bockstahler et al., 1992); the number of amplification cycles ranged from 35 to 40. PCR products were fractionated by agarose gel electrophoresis, transferred to nylon membranes and confirmed as HIV-l
64
L. E. Bockstahler et al.IClinical and Diagnostic Virology 3 (1995) 61-72
specific by hybridization with the homologous DNA oligonucleotide probes ( 5’[32P]end labelled) by procedures previously described (Bockstahler et al., 1992). At least 3 controls were used in each of the assay experiments performed: an HIV-l-positive control with genomic DNA’s from a chronically HIV-l infected cell line and two negative controls: DNA from uninfected human cells and a reagent (no DNA) control. Special care was taken to avoid contamination and amplicon ‘carry-over’. Steps taken included use of positive displacement pipettes or filtered pipette tips, frequent changing of laboratory clothing and gloves, aliquoting of reagents, physical isolation of PCR preparations and products and other recommendations of Kwok and Higuchi (1989). Throughout the study all DNA samples were coded, and the viral genome distribution in each brain was not determined until completion of PCR measurements.
3. Results 3.1. Brain I Initially, the DNA isolated from brain I (limited neuropathologic findings) was assayed for the presence of HIV-l specific DNA by Southern blot analysis using a 32P-labelled viral genomic DNA probe (pBHlO-R3). Although suitable positive and negative controls were utilized, HIV-l genomic DNA could not be detected in any of the samples taken from 11 different locations of this brain (data not shown). This result suggested that if any HIV-l genomic DNA were present in the samples, it was relatively low in amount. Each sample was then assayed by the highly sensitive PCR technique using primers designed to amplify a loo-250 bp DNA segment from each of 3 different highly conserved regions of the viral genome. Results of an assay Expt. for sequences from the gag region of the HIV-l genome are shown in Fig. 1 (Expt. 1 in Table 1). Five of the 11 brain DNA samples were HIV-l genomic DNA positive, including No. 7, which was weakly positive as determined from the original autoradiogram. Six samples were HIV-l-negative, and DNA from lung (not further investigated in this study) was HIV-l-positive. In this experiment control DNA samples were either HIV-l-positive, or HIV-l-negative as expected, Lane 13b of Fig. 1 and separate sensitivity experiments (data not shown) indicated that the sensitivity of our assay was about 15 molecules of HIV-l proviral DNA per ug total DNA using the positive control genomic DNA from chronically HIV- 1-infected glial cell line TH4-7-5 [single HIV-l provirus integration site (BrackWerner et al., 1992)]. Compiled results of 6 PCR experiments (Table 1) show that 3 of the brain DNA samples were clearly HIV-l-positive ( 1, 4 and lo), whereas 3 were HIV-l-negative (5,8 and 9) in at least 5 of 6 experiments. The remaining samples were indeterminate: 2, 3, 6, 7, and 11. Such PCR data scatter occurs with DNA samples that contain extremely low numbers of target HIV-l genomic DNA molecules (Bootman and Kitchin, 1992). Sample 2 was the only indeterminate sample positive for all three
L. E. Bockstahler
et al./Clinical
C
12
3
----------_-_
and Diagnostic
Virology 3 (1995) 61-72
5
6
7
8
9
13b13c
14
14
15
15
4
1011
65
C
114 bp-
B
C 12 13a -------------
16
16
C
114 bp-
Fig. 1. Detection of HIV-I in AIDS brain I DNA samples by polymerase chain reaction. (A,B) Representative autoradiograms showing Southern blot analysis of PCR-amplified DNA (40 cycles) using primer pair SK38/39 derived from gag region of HIV-l genome. Amplified DNA (expected 114 bp) was fractionated by agarose gel electrophoresis, transferred to nylon membranes and detected by hybridization with 32P-labelled DNA probe SK19. (A) Amplified DNA. Lanes l-11, from tissues taken from 11 different locations within brain I of AIDS patient I; C, positive control from chronically HIV-l infected T-lymphoma cells. (B) Amplified DNA. Lane 12, from lung tissue of patient; lane 13 (a-c), positive control from chronically HIV-l-infected glial cells [TH4-7-5 (Brack-Werner et al., 1992); # molecules of HIV-l proviral DNA per ug total DNA: a = 150, b = 15, c = 1.51; lane ‘-‘, blank; lane 14, negative control from uninfected T-lymphoma cells; lane 15, negative control from uninfected human skin fibroblast cells; lane 16, reagent control (no DNA); C, same as in (A).
of the viral genomic regions tested; this suggests that it may in fact be positive. Thus, the distribution of viral genomic DNA in this brain was non-uniform. The quantities of HIV-l genomic DNA in the HIV-l-positive samples from brain I (as well as from brains II and III) were low: order of magnitude of 1 viral genome per total of 1,000 cells as estimated by parallel PCR analysis of genomic DNA from chronically HIV- 1-infected glial cell line TH4-7-5 [single HIV- 1 provirus integration site (Brack-Werner et al., 1992)]. Fig. 2A shows the approximate locations of the 11 tissue samples in a schematic diagram of a human brain. The HIV-l-positive samples (1, 4, 10 and possibly positive 2) originated from tissues located in the midbrain, medulla oblongata and pons regions. HIV-l-negative samples 5, 8 and 9 were located in the frontal lobe (white matter), parieto-occipital lobe (white matter) and basal ganglia. The indeterminate samples were taken mainly from the hindbrain [ 3, 6, 11: cerebellum, parietooccipital lobe (gray matter) and temporal lobe].
66
L.E. Bockstahler
et a/./Clinical
and Diagnostic
Virology 3 (1995) 61-72
TABLE 1 Detection of HIV-l genomic DNA in AIDS patient brain I (limited neuropathologic polymerase chain reactionasb Expt. no.
Id 2 3 4 5 6
Primer pair
gag gag gag PO1 PO1 LTR
Overall result:’
findings) by
Brain sample no.’ 1
2
3
4
5
6
7
8
9
10
11
+e
_
_
+
_
+
+
_
_
+
_
++ ++ + ++ +
+ + +-+
+ +
++ + ++ -
+ _
+ + ++
+
-
-
++ + ++ ++
_
+
**+
_ --
-+-+
-+
+--+
++ nd *
a Compilation of results from 6 separate PCR experiments. b 23 out of 24 negative controls (l/2 normal human cell DNA and l/2 no DNA) were HIV-l negative; 30 out of 30 positive controls (DNA from HIV-infected human cells) were HIV-l-positive (Expt. 1-6; control data not shown). ’ DNA extracted from tissues taken from 11 different locations within AIDS patient’s brain. d Autoradiogram for Expt. 1 shown in Fig. 1. e Labels, + +, + , - denote high-level, low-level and negative, respectively, for relative intensities of bands observed in autoradiograms; nd, not done. ‘Overall result defined as HIV-l-positive (+) or -negative (-) if at least 5 of all 6 experiments were positive or negative, respectively; f denotes indeterminate result.
3.2. Brain II Tissues were taken from 8 different locations within brains II and III (without knowledge of brain I PCR results). DNA samples isolated from the tissues were assayed for the presence of HIV-l genomic DNA using the gag primers and probes. Compiled results of 6 PCR experiments on brain II (limited neuropathologic findings) DNA samples are shown in Table 2. Five of the brain DNA samples were HIV-l-positive (3, 4, 5, 7 and 12) in at least 5 of 6 experiments, whereas 3 were indeterminate (9, 11 and 13). None of the 8 samples gave an overall result of HIV- 1-negative. The approximate locations of the 8 tissue samples in a brain schematic diagram are shown in Fig. 2B. The HIV-l-positive samples 3, 4 and 12 were located in the lower hemisphere of the brain (cerebellum, medulla oblongata and temporal lobe), while samples 5 and 7 were located in the upper hemisphere (frontal lobe, white and gray matter). Thus, the HIV-l-positive samples of brain II were not localized in one region of the brain. 3.3. Brain III Table 3 shows a compilation of results of 6 PCR experiments performed on brain III (extensive neuropathologic findings) DNA samples. This DNA was isolated from
L.E. Bockstahler et al./CIinical and Diagnostic Virology 3 (199.5) 61-72
67
Fig. 2. Distribution of HIV-l genomic DNA in AIDS brains I, II and III. Three-dimensional view of brain designates approximate location of brain tissue samples/DNA’s used for polymerase chain reaction study. The brain is shown with the front left quarter section (planes xz and yz) cut out as indicated in the inset. The gray area on the left side of the figure represents uncut brain surface from the right hemisphere, which would be visible in a view approximately from the lower left corner of the inset, slightly below the plane of the inset, and looking towards the center of the brain. The xz plane indicates the lateral distribution of the samples, which are not necessarily located strictly on the surface shown, Samples are depicted in the cut sections of the brains for clarity even if they originated from the symmetric location of the other half of the brain. Major blood supply in the brain indicated by the round, dark grey region between samples 5 and 9. Sample numbers correspond to those used in the tables. Source of tissues: 1, midbrain; 2, pons (inside); 3, cerebellum; 4, medulla oblongata; 5, frontal lobe (white); 6, parieto-occipital lobe (gray); 7, frontal lobe (gray); 8, parieto-occipital lobe (white); 9, basal ganglia; 10, pons (outside); 11, temporal lobe; 12, temporal lobe; 13, ammon’s horn. (A) Brain 1, limited neuropathologic findings. (B) Brain II, limited neuropathologic findings. (C) Brain III, extensive neuropathologic findings.
8 different brain tissue samples taken from locations close to those of samples taken from brain II. Seven of the 8 DNA samples were HIV-l-positive in at least 5 of 6 experiments. One sample was indeterminate (12), and none gave an overall result of HIV- 1-negative.
68
L. E. Bockstahler
et al.IClinical and Diagnostic
Virology 3 (1995) 61-72
Fig. 2. (continued)
Fig. 2C shows the approximate locations of the 7 HIV-l-positive samples and indeterminate sample 12. The HIV-l-positive samples are distributed throughout brain III.
4. Discussion The main findings of this study are: (1) HIV-l genomic DNA was detected in all three brains investigated; it was found, as expected, in the tissues of brain III, which had extensive CNS damage, but it was also present in brains I and II even though they contained limited neuropathologic findings (lymphoma, but no primary HIV-l specific neuropathologic changes). (2) The number of copies of viral genomes found was low in all three brains (order of magnitude of 0.1% infected cells or less). (3) Some quantitative differences appeared to be present, e.g., brain III contained more clearly positive HIV-l samples than brains I and II. (4) The HIV-l-positive samples of brain I were concentrated in the brainstem, although lower amounts of viral
L. E. Bockstahler
et al./Clinical
and Diagnostic
Virology 3 (I 995) 61-72
69
Fig. 2. (continued)
genomic DNA (indeterminate samples) were observed in other regions of this brain. (5) Brain II was also HIV-l-positive for the brainstem area and other regions in the lower and upper hemispheres, and (6) in brain III, HIV-l DNA appeared to be more uniformly distributed throughout the brain both qualitatively and quantitatively as judged from the number of PCR experiments that yielded positive results. The number of clearly HIV-l-positive samples in brain III was greater than that found in brain I, but was not significantly greater than that found in brain II (Fig. 2). Also, the amount of viral genomic DNA in brain III positive samples in terms of per cent infected cells was not markedly greater than that found in brain I and II positive samples (data not shown). Therefore, no obvious correlation was found between the extent of damage observed by neuropathologic examination and the distribution of viral genomic DNA in these brains. Another reason for performing this study was to obtain data that might help illuminate the route(s) of HIV entry into the brain and dissemination of virus within the brain. We believe that brain II (Fig. 2B) may have been at a slightly later stage of CNS infection than brain I (Fig. 2A), because it contained more clearly positive
70
L. E. Bockstahler et al. JClinical and Diagnostic Virology 3 (1995) 61-72
Table 2 Detection of HIV-l genomic DNA in AIDS patient brain II (limited neuropathologic polymerase chain reactionasb Expt. no.
Primer pair
gag gag gag gag gag gag Overall result:’
findings) by
Brain sample no.’ 3”
4
5
7
9
11
12
13
+d ++ ++ + ++ ++
++ ++ + + ++
+ + + + ++
++ ++ ++ + ++ ++
+ _
+ + _
+ +
++ _ _
+ + + _ ++ +
+ + + + _
+
+
+
+
+
*
+
*
a Compilation of results from 6 separate PCR experiments. b 10 out of 12 negative controls (l/2 normal human cell DNA and l/2 no DNA) were HIV-l-negative; 6 out of 6 positive controls (DNA from HIV-infected human cells) were HIV-l-positive (Expts. 1-6; control data not shown). ’ DNA extracted from tissues taken from 8 different locations within AIDS patient’s brain. d Labels + f, f, - denote high-level, low-level and negative, respectively, for relative intensities of bands observed in autoradiograms. e Brain sample numbers represent approximately identical sampling locations in brains I, II and II. f Overall result defined as HIV-l-positive (+) if at least 5 of all 6 experiments were positive; k denotes indeterminate result.
HIV-l genomic DNA samples. If the results for brains I, II and III (Fig. 2) are compared, there is at least a tendency for increasing amounts of HIV-l genomic DNA to appear in areas close to the brainstem and upper hemisphere of the brain, and finally in the center of the brain, while the brainstem region appears to remain almost unchanged with respect to amount of viral genomic DNA. The amount of viral genomic DNA in the HIV-l-positive samples of all three brains was of the same order of magnitude (0.1% infected cells) regardless of neuropathologic findings. Our data could be interpreted in terms of virus entry into the brain and localization of virus at restricted regions followed by a cell-to-cell spread of viral infection over wider areas within the brain, rather than local build-up of infection. In other words, our findings are compatible with relatively rare, primary infection events at restricted areas followed by spreading of virus from these entry points, in contrast to multiple, simultaneous viral infection of different brain regions. Invasion of the CNS by HIV-l may have occurred by several mutually nonexclusive pathways including transport of viral infected macrophages/monocytes across the blood-brain barrier (Cheng-Mayer and Levy, 1990) and direct viral infection of choroid plexus (Harouse et al., 1989). The viral genomic DNA distribution pattern found in brain I (Fig. 2A), showing localization of HIV-l in the brainstem, suggests the possibility in this brain of neural spread of virus, a pathway known to be utilized by other neuropathogenic viruses such as poliovirus, herpesvirus and rabies virus (Tyler and Fields, 1990). If HIV-l had entered this brain via virus-
L. E. Bockstahler
et al.IClinical and Diagnostic
71
Virology 3 (1995) 61-72
Table 3 Detection of HIV-l genomic DNA in AIDS patient brain III (extensive neuropathologic polymerase chain reactio&’ Expt. no.
Primer pair
gag gag gag gag gag gag Overall result!
findings) by
Brain sample no.’ 3”
4
5
1
9
11
12
13
fd _ + ++ + +
++ ++ + + + _
++ ++ + + + +
+ + + + + ++
+ + + + +
+ ++ + ++ ++ ++
++ + + +
-__ __ __ _-
+
+
+
+
+
+
*
i-
*- +
’ Compilation of results from 6 separate PCR experiments. b 10 out of 12 negative controls (l/2 normal human cell DNA and l/2 no DNA) were HIV-l-negative; 6 out of 6 positive controls (DNA from HIV-infected human cells) were HIV-l-positive (Expts. L-6; control data not shown). ’ DNA extracted from tissues taken from 8 different locations within AIDS patient’s brain. d Labels ++, +, - denote high-level, low-level and negative, respectively, for relative intensities of bands observed in autoradiograms. ’ Brain sample numbers represent approximately identical sampling locations in brains I, II and III. ’Overall result defined as HIV-l positive (+) if at least 5 of all 6 experiments were positive; + denotes indeterminate result.
infected macrophages that had been transported there from peripheral blood, we would have expected samples close to the region of major blood supply (see round dark grey region between samples 5 and 9 in Fig. 2A) to have been HIV-l-positive rather than HIV- 1-negative or indeterminate.
Acknowledgements
This work was supported by Grant BGA III-002%89/FVP 2 from the Bundesgesundheitsamt and was performed in part during a U.S. Public Health Service Foreign Work/Study Program assignment in Neuherberg by one of us (LEB). The authors are grateful to T. Goerblich (GSF, Neuherberg, Germany) for advice and assistance with DNA isolations. We thank A. Kleinschmidt, A. Luz, W. Schmahl (GSF), L. Vitkovic (NIH, Rockville, Maryland, USA), A. Buchbinder (VAMC, New York, New York), J. Bootman (NIBSC, South Mimms, Hertfordshire, UK) and FDA colleagues for helpful discussions.
References Bockstahler, L.E., Carney, P.G., Bushar, G. and Sagripanti, J.L. (1992) Detection of Junin virus by the polymerase chain reaction. J. Virol. Methods 39, 231-235.
12
L. E. Bockstahler et al.IClinical and Diagnostic Virology 3 (1995) 61-72
Bootman, J.S. and Kitchin, P.A. (1992) An international collaborative study to assess a set of reference reagents for HIV-l PCR. J. Virol. Methods 37, 23-42. Bowtell, D.D.L. (1987) Rapid isolation of eukaryotic DNA. Anal. Biochem. 162, 463-465. Brack-Werner, R., Kleinschmidt, A., Ludvigsen, A., Mellert, W., Neumann, M., Herrmann, R., et al. (1992) Infection of human brain cells by HIV-l: restricted virus production in chronically infected glial cell lines. AIDS 6, 2733285. Budka, H. (1991) Neuropathology of human immunodeficiency virus infection. Brain Pathol. 1,163-175. Budka, H., Wiley, CA., Kleihues, P, Artigas, J., Asbury, A.K., Cho, E.-S., et al. ( 1991) HIV-associated disease of the nervous system: review of nomenclature and proposal for neuropathology-based terminology. Brain Pathol. 1, 1433152. Cheng-Mayer, C. and Levy, J.L. (1990) Human immunodeficiency virus infection of the CNS: characterization of ‘neurotropic’ strains. Current Topics Microbial. Immunol. 160, 1455156. Chiodi, F. and Fenyo, E.M. (1991) Neurotropism of human immunodeficiency virus. Brain Pathol. 1,185-191. Epstein, L.G. and Gendelman, H.E. (1993) Human immunodeficiency virus type 1 infection of the nervous system: pathogenetic mechanisms. Ann. Neurol. 33, 4299436. Harouse, J.M., Wroblewska, Z., Laughlin, M.A., Hickey, W.F., Schonwetter, B.S. and GonzalezScarano, F. (1989) Human choroid plexus cells can be latently infected with human immunodeficiency virus. Annals Neurol. 25, 406-411. Kwok, S. and Higuchi, R. (1989) Avoiding false positives with PCR. Nature 339, 237-238. Ou, C-Y., Kwok, S., Mitchell, S.W., Mack, D.H., Sninsky, J.J., Krebs, J.W. et al. (1988) DNA amplification for direct detection of HIV-l in DNA of peripheral blood mononuclear cells. Science 239, 2955297. Price, R.W., Sidtis, J.J. and Brew, B.J. (1991) AIDS dementia complex and HIV-l infection: a view from the clinic. Brain Pathol. 1, 155-162. Strauss, P. G., Schmidt, J., Pedersen, L. and Erfle, V. (1988) Amplification of endogenous proviral MuLV sequences in radiation-induced osteosarcomas. Int. J. Cancer 41, 616-621. Tyler, K.L. and Fields, B.N. (1990) Pathogenesis of viral infections. In: B.N Fields and D.M. Knipe (Eds.), Fields Virology, 2nd edn., Raven Press, New York, pp. 191-239.