• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Oct 15, 2002; 99(21): 13795–13800.
Published online Oct 8, 2002. doi:  10.1073/pnas.202357499
PMCID: PMC129777
From the Cover
Medical Sciences

HIV-1 infection and AIDS dementia are influenced by a mutant MCP-1 allele linked to increased monocyte infiltration of tissues and MCP-1 levels

Abstract

Studies in humans and in experimental models of HIV-1 infection indicate an important role for monocyte chemoattractant protein-1 (MCP-1; also known as CC chemokine ligand 2), a potent chemoattractant and activator of mononuclear phagocytes (MP) in the pathogenesis of HIV-associated dementia (HAD). We determined the influence of genetic variation in MCP-1 on HIV-1 pathogenesis in large cohorts of HIV-1-infected adults and children. In adults, homozygosity for the MCP-1 –2578G allele was associated with a 50% reduction in the risk of acquiring HIV-1. However, once HIV-1 infection was established, this same MCP-1 genotype was associated with accelerated disease progression and a 4.5-fold increased risk of HAD. We examined the molecular and cellular basis for these genotype–phenotype associations and found that the mutant MCP-1 –2578G allele conferred greater transcriptional activity via differential DNA–protein interactions, enhanced protein production in vitro, increased serum MCP-1 levels, as well as MP infiltration into tissues. Thus, MCP-1 expression had a two-edged role in HIV-1 infection: it afforded partial protection from viral infection, but during infection, its proinflammatory properties and ability to up-regulate HIV-1 replication collectively may contribute to accelerated disease progression and increased risk of dementia. Our findings suggest that MCP-1 antagonists may be useful in HIV-1 infection, especially for HAD, and that HIV+ individuals possessing the MCP-1 –2578G allele may benefit from early initiation of antiretroviral drugs that effectively cross the blood–brain barrier. In a broader context, the MCP-1 –2578G allele may serve as a genetic determinant of outcome of other disease states in which MP-mediated tissue injury is central to disease pathogenesis.

Keywords: chemokine‖genotype‖leukocyte

The cardinal biological property of monocyte chemoattractant protein-1 (MCP-1; also known as CC chemokine ligand 2), a CC chemokine, is its ability to attract and activate with high-potency mononuclear phagocytes (MPs) such as monocytes and macrophages (1). Because of this remarkable target specificity, MCP-1 is thought to play an important role in a variety of diseases characterized by MP infiltration, including atherosclerosis and multiple sclerosis (1). Furthermore, despite the extensive redundancy in the chemokine system (2), studies in genetically modified mice place MCP-1 in a central position in MP trafficking and activation (1, 3).

This tight link between MP recruitment/activation and MCP-1 expression could potentially also influence HIV-1 pathogenesis. In addition to CD4+ T lymphocytes, MPs are targets for HIV-1 infection, and productively infected MPs may serve as “Trojan horses” for the dissemination of virus to different organs, including the lymph nodes and brain (48). Additionally, an MP-dependent mechanism of viral amplification at local sites of initial infection appears to be a prerequisite for efficient dissemination of infection (9). These observations, and the finding that MCP-1 expression and HIV-1 replication may be coregulated (1016), suggest that MCP-1 could play a central role in the early MP-mediated events necessary for the establishment and spread of infection. Furthermore, MCP-1 can inhibit HIV-1 infection in vitro (17), although it remains unclear whether these effects are direct, i.e., blockade of HIV-1 entry via its receptor CC chemokine receptor (CCR)2, or indirect, i.e., by inducing receptor-mediated heterologous desensitization of the major HIV-1 coreceptors CCR5 and/or CXC chemokine receptor (CXCR)4 (2, 18).

MP-mediated inflammatory processes are also major precipitating events in AIDS-defining illnesses such as HIV-1-associated dementia (HAD) (6, 11, 1316, 1922). Indeed, HIV-1 is thought to enter the nervous system via peripheral blood-derived MPs within hours of infection and remains present throughout the infection (6, 2023). Brain MPs (microglia) are not only the main target cells sustaining HIV replication but are also the major cellular determinants of HIV encephalitis, a pathological condition associated with HAD (16, 2022). Notably, the abundance of MPs is a better correlate of HAD than the extent of viral infection in the brain (16, 2022, 24). How and why so many MPs enter the brain in HAD is unclear. However, not only is there substantial evidence supporting a central contribution of MCP-1 in MP recruitment to the brain (1, 25), but there is also direct experimental evidence implicating MCP-1 as a critical factor in the neuropathogenesis of HAD. A striking example comes from studies in an animal model of HIV-1 (26). Simian immunodeficiency virus-infected macaques that developed moderate or severe encephalitis had elevated levels of MCP-1 in the cerebrospinal fluid (CSF) that preceded and predicted encephalitis (26). Substantiating these findings, there is selective accumulation of MCP-1 in the CSF of AIDS patients with HAD, and levels correlate with degree of dementia (10, 11, 1316, 1922, 2629).

We therefore hypothesized that interindividual differences in MCP-1 production and, by extension, differences in efficiency of MP recruitment and activation are important contributing factors not only for the highly variable risk of developing HAD (6, 2022) but also for HIV-1 infection and rate of disease progression. A powerful approach to understanding the relationship between MCP-1 expression and HIV pathogenesis in vivo is to define the association between polymorphisms that may affect MCP-1 gene expression and risk of transmission and HAD as well as clinical progression rate. To this end, we genotyped 1,115 HIV-infected adults and 1,035 control blood donors as well as 592 children exposed perinatally to HIV-1 for two MCP-1 promoter single nucleotide polymorphisms (SNP), which are known to modulate both gene and protein expression (30). We also elucidated the molecular and cellular mechanisms by which MCP-1 promoter SNPs may influence their effects on HIV-1 pathogenesis.

Methods

Adult Cohorts.

Control European-, African-, and Hispanic-American normal blood donors were genotyped. The characteristics of the European- and African-American control groups were described previously (31). The Hispanic-American controls were from normal blood donors in San Antonio, TX. A total of 1,115 HIV+ adult patients were evaluated, including 515 seroconverting individuals. The demographic makeup of this cohort was 55% European-American, 36% African-American, 6% Hispanic-American, and 3% “other.” The presence of HIV-1-associated dementia was established by using the AIDS 1987 and 1993 revised criteria (32, 33), including neuroimaging studies and neuropsychological testing. The proportion of patients with HAD in this cohort was similar to that described previously in other U.S.-based cohorts (≈6%) (34). Additional epidemiological features of the HIV-1-infected cohort are published as supporting information on the PNAS web site, www.pnas.org, and as described (31, 35).

Cohort of Children Exposed Perinatally to HIV-1.

The cohort of children is as described (36) and in supporting information on the PNAS web site. This cohort includes 592 children perinatally exposed to HIV-1 between 1986 and 1998. A total of 322 HIV+ and 270 HIV− children born to HIV-1+ mothers were followed up prospectively. HIV-1 infection status, AIDS definition, and stage of immune suppression were established according to the 1994 criteria of the Center for Disease Control and Prevention classification for children (37).

SNPs, PCR, and Restriction Enzyme Digestion Conditions.

The two SNPs in the MCP-1 promoter region are as described by Rovin et al. (30), and their positions relative to the adenine in ATG are –2578 and –2136; Rovin et al. designated these two positions as –2518 and –2076, respectively. Detailed methods for genotyping these two sites are published as supporting information on the PNAS web site and as described (30).

Serum MCP-1 Levels.

The ELISA Quantikine kit (R&D Systems) was used to determine the serum MCP-1 levels in 77 European-American men. Each subject provided written informed consent under a protocol approved by the Institutional Review Board for Human Subjects of Ohio State University.

Renal Biopsy and Evaluation of Leukocyte Recruitment into the Kidney.

Archived renal biopsies of 21 patients (14 European-American, 7 African-American) with systemic lupus erythematosus (SLE) and severe SLE nephritis (World Health Organization class IV) were stained for infiltrating leukocytes (methods are published as supporting information on the PNAS web site). The diagnosis and class of SLE nephritis were determined by a pathologist not involved in the current study.

Statistical Analyses.

The association of MCP-1 alleles, genotypes, haplotypes, and haplotype pairs with the risk of acquiring HIV was assessed by using the χ2 test, Fisher's exact test, and unconditional logistic regression. We estimated the crude odds ratios (ORs) and adjusted (for race/continent-of-origin) ORs by using the Mantel–Haenszel test. For significance testing, we made appropriate corrections to the α value by using the Bonferroni procedure. Similar statistical analyses were used to evaluate altered risk of developing an AIDS-defining illness. To study the association of genetic variation in MCP-1 and the clinical course, we chose two outcomes: time to AIDS [1987 criterion for adults (32) and 1994 criterion for children (37)] and time to death. Only HIV+ adults who had a followup time of at least 6 months were included in the analysis. Between-group analyses were done by using the log–rank test. Survival curves for both outcomes were prepared by using the Kaplan–Meier (KM) method. The relative hazards were estimated by using Cox proportional hazards models. Statistical analysis for the serum MCP-1 levels and renal leukocyte recruitment were done by using the median test and the Mann–Whitney test, respectively.

Electrophoretic Mobility-Shift Assay (EMSA).

Nuclear extracts were prepared from K562 cells and tumor necrosis factor-α-stimulated (20 ng/ml for 4 h) MG-63 osteoblast cells. Methods for nuclear protein extraction, EMSA, and supershift analysis were as described previously (38). The oligomers used in EMSAs spanned the region encompassing –2578A/G and –2136A/T (sequences available from S.K.A., ude.ascshtu@sajuha). Densitometry analysis of the EMSA gels was performed with the nih image software package (ftp://rsbweb.nih.gov).

Results

MCP-1 SNPs, Genotypes, Haplotypes, and Haplotype Pairs.

The ancestral state of the two MCP-1 promoter SNPs (−2578A, −2136A) was determined by genotyping 58 unrelated chimpanzees and by sequencing the MCP-1 promoter region from two chimpanzees (Fig. (Fig.11a). The frequency of these SNPs in control blood donors and adult HIV-infected subjects did not deviate from the Hardy–Weinberg equilibrium (Table 4, which is published as supporting information on the PNAS web site). The presence of MCP-1 –2578G or –2136T was mutually exclusive (Table 5, which is published as supporting information on the PNAS web site). For example, −2578G homozygosity is always linked to −2136A homozygosity, whereas −2136T homozygosity is linked to −2578A homozygosity (Table 5). Thus, the SNPs at –2578 and –2136 defined three MCP-1 haplotypes, designated as AA (the ancestral haplotype), GA, and AT, and six haplotype pairs (Fig. (Fig.11 b and c; Table 5). The frequencies of the MCP-1 alleles, genotypes, haplotypes, and haplotype pairs varied significantly among different populations (Fig. (Fig.11 b and c; Table 4), with African-Americans having the highest prevalence of the MCP-1 ancestral haplotype pair AA/AA among the four populations studied. The MCP-1 allele, genotype, haplotype, and haplotype pair frequencies in Argentinean children resembled more closely those found in European- and Hispanic-Americans than in African-Americans (Fig. (Fig.11 and Table 4).

Figure 1
MCP-1 SNPs, haplotypes and haplotype pairs. (a) Schematic illustration of the genomic organization and SNPs (*) in human MCP-1. Open boxes represent the three exons. The sequence at the position corresponding to the human SNPs in chimpanzees (AA is the ...

Genetic Variation in MCP-1 and Risk of Acquiring HIV-1.

We first determined the association between MCP-1 SNPs, genotypes, haplotypes, and haplotype pairs and the risk of acquiring HIV-1 infection in adults. All statistical comparisons were to the ancestral state for the alleles (−2578A and −2136A), haplotype (AA), and haplotype pair (AA/AA), and corrections were made for multiple comparisons. Table Table11 shows that even after controlling for the possible confounding effect of continent-of-origin (race), possession of the MCP-1 −2578G allele, the –2578G/G genotype, the MCP-1 GA haplotype, or the GA/GA haplotype pair was associated with a significantly lower risk of acquiring HIV-1 (P < 0.003). In contrast, an effect associated with the –2136T allele or −2136T-containing haplotypes or haplotype pairs was not observed in U.S. adults (Table (Table1).1). The frequency of the MCP-1 –2578G and –2136T alleles was similar in infected and uninfected Argentinean children (Table 4), suggesting that in the setting of mother-to-child transmission, these alleles may not influence the risk of acquiring HIV-1.

Table 1
Association of genetic variation in MCP-1 with the risk of acquiring HIV-1

We next examined whether the association between the MCP-1 –2578G allele and a lower risk of acquiring HIV-1 was more prominent in specific U.S. populations (Table (Table2).2). In all three populations, possession of the MCP-1 −2578G allele, the –2578G/G genotype, the GA haplotype, and the GA/GA haplotype pair was associated with a lower risk of acquiring HIV-1 (OR <1; Table Table2).2). This effect was most prominent in European- and Hispanic-Americans and least prominent in African-Americans (Table (Table2).2).

Table 2
Association of genetic variation in MCP-1 with the risk of acquiring HIV-1 in different U.S. populations

Genetic Variation in MCP-1 and Clinical Outcome in HIV-1+ Individuals.

We next determined the association between MCP-1 SNPs, haplotypes, and haplotype pairs and disease progression in HIV-1+ adults. In European-Americans, homozygosity, but not heterozygosity for the MCP-1 −2578G allele, i.e., the GA/GA haplotype pair, was associated with accelerated progression to AIDS and death (Fig. (Fig.22 a and b). Homozygosity for –2578G was also associated with accelerated progression to AIDS in infected children (Fig. (Fig.22c).

Figure 2
Disease-modifying effects of MCP-1 alleles in a cohort of infected European- and African-American adults or in a cohort of Argentinean children with perinatally acquired HIV-1. (a and b) KM curves of the development of AIDS or death in the entire European-American ...

European-Americans who had at least one ancestral MCP-1 haplotype (AA) had a slower disease course than those who lacked this haplotype (Fig. (Fig.22 d and e). Of the three haplotype pairs that lacked the AA haplotype (Fig. (Fig.11c; GA/GA, GA/AT, AT/AT), only GA/GA was associated with more rapid progression to AIDS and death (Fig. (Fig.22 f and g). Most African-Americans had at least one ancestral MCP-1 AA haplotype (Fig. (Fig.11c). For this reason, the disease-modifying effects of AA- vs. each of the non-AA-containing haplotype pairs (e.g., GA/GA) could not be determined. In African-Americans, compared with AA/AA, the MCP-1 haplotype pair AA/GA was associated with rapid progression to AIDS (Fig. (Fig.22h), and AA/AT was associated with a trend toward a more rapid progression to death (Fig. (Fig.22i). The number of Hispanic-Americans with the different MCP-1 haplotype pairs who had clinical events was too few for statistical analysis.

Genetic Variation in MCP-1 and Risk of HAD.

In light of the hypothesis outlined in the introduction and the finding that possession of GA/GA was associated with an accelerated disease course in European-Americans, we determined whether this haplotype pair specifically increased the susceptibility to MP-mediated diseases such as HAD in this population. Compared with AA/AA, GA/GA was associated with a 4.7-fold higher likelihood of developing HAD (P < 0.015; Fig. Fig.22j). This association remained significant even after adjusting for the year in which HAD developed [P = 0.048, OR = 3.69, confidence interval (CI) = 1.01–12.45], suggesting that the administration of new classes of antiretroviral agents did not mitigate the increased susceptibility of individuals with GA/GA to HAD. The effects of GA/GA were specific for HAD and infection with Mycobacterium avium complex (MAC) and were not seen for AIDS-defining illnesses such as Pneumocystis carinii pneumonia (PCP), cytomegalovirus infection or Kaposi's sarcoma (Fig. (Fig.22k). Furthermore, compared with the ancestral haplotype pair AA/AA, possession of GA/GA was associated with accelerated progression to HAD and MAC infection, but not to other common AIDS-defining illnesses such as PCP (Fig. (Fig.22l and data not shown). There were too few HIV+ European-Americans with tuberculosis to determine the association between the MCP-1 GA/GA genotype and risk of developing this infection. The limited number of individuals with the GA/GA haplotype pairs in HIV-1-infected African- and Hispanic-Americans precluded an analysis of the association between this genotype and risk of HAD in these two populations.

Mechanisms Underlying the Associated Effects of the MCP-1 −2578G Allele.

Mechanistically, the differential effects of the MCP-1 –2578G allele versus the –2578A allele in European-Americans fits a model wherein increased levels of MCP-1 (e.g., in individuals with GA/GA) is protective at time of virus exposure, although detrimental after infection is established. Consistent with this model, we have previously demonstrated that peripheral blood mononuclear cells from individuals homozygous for MCP-1 –2578A produced lower amounts of MCP-1 than do cells from individuals homozygous or heterozygous for −2578G (30). Extending these in vitro findings, we found that there is an in vivo association between possession of the –2578G allele and MCP-1 serum levels. European-American men who possessed a –2578G allele had higher MCP-1 levels than those who lacked this allele (Table (Table3).3).

Table 3
Biological effects associated with MCP-1 −2578 genotypes

To extend these genotype–phenotype associations, we next sought genetic proof that possession of the MCP-1 –2578G allele is linked to an in vivo functional effect such as increased leukocyte recruitment. Recognizing the inherent limitations of obtaining brain biopsies to evaluate the relationship between MCP-1 genotypes and leukocyte recruitment to the brain in HAD, we determined this relationship in a non-HIV setting and in a more readily accessible organ, namely the kidney. The relationship of MCP-1 genotype to the number of infiltrating renal leukocytes during class IV SLE nephritis is shown in Table Table3.3. Patients having at least one MCP-1 −2578G allele had significantly more interstitial macrophages than those lacking this allele (P < 0.019; Table Table33 and Fig. Fig.3).3). Interstitial T cells, as well as macrophages and T cells in the glomerular compartment, were numerically higher in patients with a MCP-1 –2578G allele but did not reach statistical significance in this limited sample population (Table (Table3).3). There was no relationship between infiltrating neutrophils and MCP-1 genotype (data not shown).

Figure 3
MCP-1 genotype and renal leukocyte infiltration in SLE. (ad) Renal biopsy tissue from patients with class IV SLE nephritis was stained for macrophages (CD68+ cells) and T lymphocytes (CD3+ cells) by immunoperoxidase labeling (see ...

To elucidate further the molecular basis for the effects associated with the MCP-1 GA-haplotype, we tested the hypothesis that –2578G and –2136A, the nucleotide sequences that define the GA haplotype, mediate their effects by altering protein–DNA interactions. We now show that these effects of the GA haplotype on transcription may be linked in part to the differential avidity with which the nuclear factors IRF-1 and GATA-1 bind to −2578G and −2136A alleles, respectively (Fig. 4, which is published as supporting information on the PNAS web site). These findings are highly reminiscent of our recent observation that differential haplotype-specific transcriptional activity and altered transcription factor binding to polymorphic nucleotides may also underlie the HIV-1 disease- and transmission-modifying effects associated with CCR5 haplotypes (38).

Discussion

Our results are significant for several reasons. First, although CCR2, the only known high-affinity receptor for MCP-1, is generally considered as an HIV-1 coreceptor of minimal significance in vivo (2), we provide data that would imply an unexpected, albeit important, role for the MCP-1–CCR2 axis in the pathogenesis of HIV-1 infection. Second, we have elucidated the genotype–phenotype relationships of SNPs in the regulatory region of the MCP-1 gene at four levels (ref. 30 and this paper): gene expression, protein production, function (i.e., leukocyte trafficking), and biological effects (i.e., modulation of disease pathogenesis). We demonstrate that the MCP-1 –2578G allele is a genetic marker for: (i) greater MCP-1 transcriptional activity mediated by differential protein/DNA intereactions; (ii) increased MCP-1 protein production in vitro and in vivo [a similar association was recently reported (39)]; (iii) enhanced leukocyte trafficking to tissues; and (iv) differential HIV-1 transmission- and disease-modifying effects.

Third, the contrasting effects of the MCP-1 GA/GA haplotype pairs on HIV-1 transmission versus disease progression suggest that this genotype may be a two-edged genetic risk factor in HIV-1 infection. Our findings indicate that MCP-1 expression via direct (17) or indirect (2, 18) mechanisms appears to provide partial protection from viral infection in adults. However, once infection has been established, MCP-1 expression serves primarily as a mediator of inflammation and leukocyte recruitment rather than as an inhibitor of HIV-1 infection. Consistent with this notion, plasma and CSF levels of MCP-1, but not those of other MP-attracting CC chemokines such as macrophage inflammatory protein-1α or RANTES, correlate with plasma (19) and CSF viral load (11, 29), respectively. Of importance in this study, the disease-promoting effects of the MCP-1 –2578G allele were confirmed in two independent cohorts, each reflecting distinct facets of HIV-1 infection.

Fourth, our findings provide strong genetic corroboration for the clinical and pathological findings in humans and in experimental non-human primate models of HIV that link MCP-1 expression to the neuropathogenesis of HAD (1, 10, 11, 1316, 1922, 2629). The current working model of HAD pathogenesis invokes HIV-1 gaining access to the central nervous system (CNS) via the bloodstream, either by direct infection of capillary endothelial cells or, more likely, by egress of infected peripheral MPs, which in turn transmit virus to perivascular brain MPs (6, 8, 14, 16, 2022). Over time, the parenchymal microglia are infected by virus produced within the CNS. The mechanisms that initiate the trafficking of peripheral MPs to the brain are unknown but are thought to involve up-regulation of chemokines such as MCP-1 and the expression of adhesion molecules on endothelial cells. Thus, we posit that homozygosity for the MCP-1 –2578G allele is a determinant of interindividual variation in MCP-1 expression and in turn serves as an important host genetic susceptibility factor in the risk of HAD.

Finally, although unique in its etiology, HAD is similar to other neurodegenerative disorders (40). These common links include inflammation, monocyte/macrophage recruitment, and glial activation. Indeed, it has been proposed that MP activation also plays a key role in the neuropathogenesis of other central nervous system disorders such as Alzheimer's disease and multiple sclerosis (40). Thus, in a broader context, our findings provide the impetus to understand the influence of genetic variation in MCP-1 in susceptibility to these neurological diseases as well as to conditions such as atherosclerosis, glomerulonephritis, and neoplasia, for which an important role of the MCP-1–CCR2 axis and MP activation has recently become evident (1, 4145). Validating this concept, recent studies have implicated the MCP-1 –2578G allele as a genetic susceptibility factor for coronary artery disease (46) and asthma (47). Furthermore, given that the prevalence of the MCP-1 –2578G allele varies significantly among populations (this study and refs. 39, 46, and 47), it may serve as an important genetic contributor to interindividual as well as interpopulation differences in MCP-1 production and MCP-1-dependent immune responses such as tissue leukocyte infiltration and disease pathologies (48, 49).

In summary, our findings implicate MCP-1 as an important host factor in several different facets of HIV-1 pathogenesis. Because MCP-1-mediated MP recruitment and activation can fuel HIV-1 pathogenesis, MCP-1 may represent an important target in the design of strategies that interfere with virus replication and spread. Furthermore, HAD is one of the most common causes of dementia worldwide among individuals aged 40 or less and is an independent risk factor for death due to AIDS (50). Thus, our findings offer the potential for assessing patients with a higher risk for developing HAD and for allowing for the early initiation of drugs that achieve high concentrations in the brain.

Supplementary Material

Supporting Information:

Acknowledgments

We thank the anonymous referees for their helpful suggestions, Dr. R. A. Clark for insightful discussions, R. Geevarghese for technical assistance, and A. S. Ahuja for forbearance. The Henry M. Jackson Foundation and the Military HIV Program, Walter Reed Army Institute of Research (WRAIR), contributed support for the Wilford Hall Medical Center patient cohort as part of the Tri-Service HIV Program. This work was supported in part by National Institutes of Health (NIH) Grant AI46326 to S.K.A. and by the Veterans Administration Research Center for AIDS and HIV-1 Infection in San Antonio. B.R. is supported in part by the S. L. E. Foundation, Incorporated, and by NIH Grant P01 DK55546. S.M. was supported in part by a research grant awarded by the American Heart Association. Support for the work conducted in Argentina was from Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) 0494. S.K.A. is a recipient of the Elizabeth Glaser Scientist Award and the Burroughs Wellcome Clinical Scientist Award in Translational Research.

Abbreviations

MCP-1
monocyte chemoattractant protein-1
MP
mononuclear phagocyte
CCR
CC chemokine receptor
HAD
HIV-1-associated dementia
SNP
single-nucleotide polymorphism
SLE
systemic lupus erythematosus
MAC
Mycobacterium avium complex
PCP
Pneumocystis carinii pneumonia
KM
Kaplan–Meier
OR
odds ratio
CI
confidence interval

Footnotes

This paper was submitted directly (Track II) to the PNAS office.

References

1. Gu L, Tseng S C, Rollins B J. Chem Immunol. 1999;72:7–29. [PubMed]
2. O'Brien S J, Moore J P. Immunol Rev. 2000;177:99–111. [PubMed]
3. Lu B, Rutledge B J, Gu L, Fiorillo J, Lukacs N W, Kunkel S L, North R, Gerard C, Rollins B J. J Exp Med. 1998;187:601–608. [PMC free article] [PubMed]
4. Ho D D, Rota T R, Hirsch M S. J Clin Invest. 1986;77:1712–1715. [PMC free article] [PubMed]
5. Gabuzda D H, Ho D D, de la Monte S M, Hirsch M S, Rota T R, Sobel R A. Ann Neurol. 1986;20:289–295. [PubMed]
6. Gartner S. Science. 2000;287:602–604. [PubMed]
7. Koenig S, Gendelman H E, Orenstein J M, Dal Canto M C, Pezeshkpour G H, Yungbluth M, Janotta F, Aksamit A, Martin M A, Fauci A S. Science. 1986;233:1089–1093. [PubMed]
8. Williams K C, Corey S, Westmoreland S V, Pauley D, Knight H, deBakker C, Alvarez X, Lackner A A. J Exp Med. 2001;193:905–915. [PMC free article] [PubMed]
9. Hirsch V M, Sharkey M E, Brown C R, Brichacek B, Goldstein S, Wakefield J, Byrum R, Elkins W R, Hahn B H, Lifson J D, et al. Nat Med. 1998;4:1401–1408. [PubMed]
10. Mengozzi M, De Filippi C, Transidico P, Biswas P, Cota M, Ghezzi S, Vicenzi E, Mantovani A, Sozzani S, Poli G. Blood. 1999;93:1851–1857. [PubMed]
11. Cinque P, Vago L, Mengozzi M, Torri V, Ceresa D, Vicenzi E, Transidico P, Vagani A, Sozzani S, Mantovani A, et al. AIDS. 1998;12:1327–1332. [PubMed]
12. Vicenzi E, Alfano M, Ghezzi S, Gatti A, Veglia F, Lazzarin A, Sozzani S, Mantovani A, Poli G. J Leukocyte Biol. 2000;68:405–412. [PubMed]
13. Weiss J M, Nath A, Major E O, Berman J W. J Immunol. 1999;163:2953–2959. [PubMed]
14. Conant K, Garzino-Demo A, Nath A, McArthur J C, Halliday W, Power C, Gallo R C, Major E O. Proc Natl Acad Sci USA. 1998;95:3117–3121. [PMC free article] [PubMed]
15. Wu D T, Woodman S E, Weiss J M, McManus C M, D'Aversa T G, Hesselgesser J, Major E O, Nath A, Berman J W. J Neurovirol. 2000;6, Suppl. 1:S82–S85. [PubMed]
16. Hou J, Major E O. Adv Exp Med Biol. 2001;493:29–34. [PubMed]
17. Frade J M R, Llorente M, Mellado M, Alcami J, Gutierrez-Ramos J C, Zaballos A, Real G, Martinez A C. J Clin Invest. 1997;100:497–502. [PMC free article] [PubMed]
18. Lee B, Doranz B J, Rana S, Yi Y, Mellado M, Frade J M, Martinez A C, O'Brien S J, Dean M, Collman R G, et al. J Virol. 1998;72:7450–7458. [PMC free article] [PubMed]
19. Weiss L, Si-Mohamed A, Giral P, Castiel P, Ledur A, Blondin C, Kazatchkine M D, Haeffner-Cavaillon N. J Infect Dis. 1997;176:1621–1624. [PubMed]
20. Zink W E, Zheng J, Persidsky Y, Poluektova L, Gendelman H E. FEMS Immunol Med Microbiol. 1999;26:233–241. [PubMed]
21. Gonzalez-Scarano F, Baltuch G. Annu Rev Neurosci. 1999;22:219–240. [PubMed]
22. Kaul M, Garden G A, Lipton S A. Nature. 2001;410:988–994. [PubMed]
23. Davis L E, Hjelle B L, Miller V E, Palmer D L, Llewellyn A L, Merlin T L, Young S A, Mills R G, Wachsman W, Wiley C A. Neurology. 1992;42:1736–1739. [PubMed]
24. Glass J D, Fedor H, Wesselingh S L, McArthur J C. Ann Neurol. 1995;38:755–762. [PubMed]
25. Bell M D, Taub D D, Perry V H. Neuroscience. 1996;74:283–292. [PubMed]
26. Zink M C, Coleman G D, Mankowski J L, Adams R J, Tarwater P M, Fox K, Clements J E. J Infect Dis. 2001;184:1015–1021. [PubMed]
27. Persidsky Y, Ghorpade A, Rasmussen J, Limoges J, Liu X J, Stins M, Fiala M, Way D, Kim K S, Witte M H, et al. Am J Pathol. 1999;155:1599–1611. [PMC free article] [PubMed]
28. Bernasconi S, Cinque P, Peri G, Sozzani S, Crociati A, Torri W, Vicenzi E, Vago L, Lazzarin A, Poli G, et al. J Infect Dis. 1996;174:1098–1101. [PubMed]
29. Kelder W, McArthur J C, Nance-Sproson T, McClernon D, Griffin D E. Ann Neurol. 1998;44:831–835. [PubMed]
30. Rovin B H, Lu L, Saxena R. Biochem Biophys Res Commun. 1999;259:344–348. [PubMed]
31. Gonzalez E, Bamshad M, Sato N, Mummidi S, Dhanda R, Catano G, Cabrera S, McBride M, Cao X H, Merrill G, et al. Proc Natl Acad Sci USA. 1999;96:12004–12009. [PMC free article] [PubMed]
32. Anonymous. Morbid Mortal Wkly Rep. 1987;36:1S–15S. [PubMed]
33. Anonymous. Morbid Mortal Wkly Rep. 1992;41:1–19.
34. Janssen R S, Nwanyanwu O C, Selik R M, Stehr-Green J K. Neurology. 1992;42:1472–1476. [PubMed]
35. Mummidi S, Ahuja S S, Gonzalez E, Anderson S A, Santiago E N, Stephan K T, Craig F E, O'Connell P, Tryon V, Clark R A, et al. Nat Med. 1998;4:786–793. [PubMed]
36. Mangano A, Gonzalez E, Dhanda R, Catano G, Bamshad M, Bock A, Duggirala R, Williams K, Mummidi S, Clark R A, et al. J Infect Dis. 2001;183:1574–1585. [PubMed]
37. Anonymous. Morbid Mortal Wkly Rep. 1994;43:1–10.
38. Mummidi S, Bamshad M, Ahuja S S, Gonzalez E, Feuillet P M, Begum K, Galvis M C, Kostecki V, Valente A J, Murthy K K, et al. J Biol Chem. 2000;275:18946–18961. [PubMed]
39. Jibiki T, Terai M, Shima M, Ogawa A, Hamada H, Kanazawa M, Yamamoto S, Oana S, Kohno Y. Arthritis Rheum. 2001;44:2211–2212. [PubMed]
40. Cotter R L, Burke W J, Thomas V S, Potter J F, Zheng J, Gendelman H E. J Leukocyte Biol. 1999;65:416–427. [PubMed]
41. Rovin B H. Curr Opin Nephrol Hypertens. 2000;9:225–232. [PubMed]
42. Peters W, Charo I F. Curr Opin Lipidol. 2001;12:175–180. [PubMed]
43. Huang D, Han Y, Rani M R, Glabinski A, Trebst C, Sorensen T, Tani M, Wang J, Chien P, O'Bryan S, et al. Immunol Rev. 2000;177:52–67. [PubMed]
44. Xia M Q, Hyman B T. J Neurovirol. 1999;5:32–41. [PubMed]
45. Saji H, Koike M, Yamori T, Saji S, Seiki M, Matsushima K, Toi M. Cancer. 2001;92:1085–1091. [PubMed]
46. Szalai C, Duba J, Prohaszka Z, Kalina A, Szabo T, Nagy B, Horvath L, Csaszar A. Atherosclerosis. 2001;158:233–239. [PubMed]
47. Szalai C, Kozma G T, Nagy A, Bojszko A, Krikovszky D, Szabo T, Falus A. J Allergy Clin Immunol. 2001;108:375–381. [PubMed]
48. Gosling J, Slaymaker S, Gu L, Tseng S, Zlot C H, Young S G, Rollins B J, Charo I F. J Clin Invest. 1999;103:773–778. [PMC free article] [PubMed]
49. Gu L, Okada Y, Clinton S K, Gerard C, Sukhova G K, Libby P, Rollins B J. Mol Cell. 1998;2:275–281. [PubMed]
50. Price R W, Yiannoutsos C T, Clifford D B, Zaborski L, Tselis A, Sidtis J J, Cohen B, Hall C D, Erice A, Henry K. AIDS. 1999;13:1677–1685. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...