HIV-1 infects the brain early in the course of infection and will, if untreated, cause neurologic disease such as HIV-associated dementia (HAD) in a substantial proportion of patients . In recent years also minor neurocognitive disorders in individuals under antiretroviral medication have been highlighted . Previous studies have revealed genotypic and phenotypic differences between virus derived from brain compartment and blood [3,4]. HIV target cells in the central nervous system (CNS) are mainly constituted by mononuclear phagocytes, for example perivascular macrophages and microglial cells . Several chemokine receptors, and related orphan receptors, have been shown to mediate HIV entry in CD4 expressing cell lines in vitro, but only CCR5 and CXCR4 are thought to be relevant as HIV-1 coreceptors in vivo . However, this dogma is largely based on studies that have assessed receptor use by blood-derived HIV-1 isolates. Also, the relatively few studies that have addressed coreceptor use by virus of CNS origin have mainly assessed CCR5 and CXCR4 tropism [4,7,8]. Several alternative coreceptors, in addition to CCR5 and CXCR4, are expressed in the brain, where they may mediate infection of target cells and/or contribute to the neuropathogenesis of AIDS . The Eotaxin receptor CCR3 was early identified as a HIV-1 coreceptor that together with CCR5 could mediate efficient infection of microglial cells by macrophage-tropic isolates  and more recent findings support that both receptors may be necessary for this process .
In our previous work we performed a phenotypic characterization of paired plasma and cerebrospinal fluid (CSF) isolates from 28 HIV-1-infected individuals and showed that discordant CCR5 and CXCR4 use between the two compartments was relatively common . In order to investigate if these differences also may be reflected in the use of alternative coreceptors that are expressed within the CNS, we conducted the current study. For this purpose we chose the NP-2 cell line, which is of astroglial cell origin and previously has been shown to lack endogenous expression of minor coreceptors . NP-2.CD4 cells were stably transfected with a panel of coreceptors with documented expression in the brain. Transfected cell-lines were used for the characterization of alternative coreceptor use by paired plasma and CSF isolates and the results were related to virological, immunological, and clinical parameters.
Materials and methods
Patients and virus isolates
The 28 HIV-1-infected individuals included in the study (Table 1) have been previously described . Briefly, the cohort was retrospectively selected to include individuals with varying levels of CD4+ T-cell counts (range 27–820, median 190 cells/μl), plasma viral load (range 1452–682000, median 52000 copies/ml), and CSF viral load (range 600 to >750000, median 66000 copies/ml). Seven patients had HAD, and none had received antiretroviral medication for at least 9 months prior to virus isolation. Virus isolation from plasma and CSF and virus propagation has previously been described [4,13]. High CSF viral load (CSFhigh) and low CSF viral load (CSFlow) were defined as at least or less than 40000 copies/ml, respectively. This cutoff was based on early previous studies and reviews suggesting the initiation of antiretroviral therapy at plasma viral load more than 30000–50000 copies/ml [14–16]. A subtype B isolate 25 (B117), which previously has been reported to display broad coreceptor use , was used as positive control in the infection assays.
HIV-1 subtype determination
RNA from plasma virus isolates was extracted and reverse transcribed as described [18,19]. The HIV-1 env gp120 V1-V3 region was amplified, cloned, and sequenced (six colonies were routinely picked). Sequences were aligned and manually edited in MEGA4 . A neighbor-joining tree was reconstructed to control for patient-specific clustering and to exclude the possibility of contamination. For subtype determination, one representative sequence from each patient was aligned with a reference sequence data set of all major subtypes, subsubtypes and circulating recombinant forms (downloaded from Los Alamos Sequence Database) in MEGA4 (LASDB; http://www.hiv.lanl.gov/). Finally, a maximum-likelihood phylogenetic tree was reconstructed in Garli 0.951 and bootstrapped 1000 times .
Establishment of NP-2.CD4 cell-lines expressing HIV coreceptors found in the central nervous system
NP-2.CD4 and NP-2.CD4.APJ cell-lines were kindly provided by Professor Hiroo Hoshino (Gunma University School of Medicine, Japan). NP-2.CD4 cells were stably transfected with sequence-verified c-DNA from either of the following receptors: CCR5, CXCR4, CCR3, CXCR6, GPR1, ChemR23, RDC1, and BLT1 as previously described . The CNS expression of included receptors has previously been documented [9,23–25]. Receptor expression was verified with flow cytometry (CCR5, CXCR4, CCR3, CXCR6, ChemR23 and BLT1) or mRNA expression (APJ, GPR1 and RDC1). Cells were maintained in transfection medium prior to infection experiments as previously described .
HIV-1 infections of coreceptor-transfected NP-2.CD4 cell-lines were performed as previously described [4,22]. In brief, 2 days before infection, 4 × 103 cells/well were seeded into 48-well plates using medium without antibiotics. At the time of infection, medium was removed and virus, corresponding to 30 ng p24 antigen/ml, was added to duplicate wells in 130 μl medium. Two hours after infection, medium was added to a total volume of 300 μl/well. After overnight incubation, cells were washed three times and 1 ml fresh medium was added to each well. Seven days after infection, medium was sampled from each well for the detection of viral p24 antigen and analyzed according to the manufacturer's description (Vironostika; Biomerieux, Boxtel, The Netherlands). Receptor use was defined as positive when p24 content in the supernatant reached 100 pg/ml (hereafter depicted as + after the indicated coreceptor). In a semiquantitative grading of receptor use we applied the following system: 100–1000 pg/ml, which is low-grade use (+), 1000–10000 pg/ml, which is moderate use (++), more than 10000 pg/ml, which is high-grade use (+++) .
Dual R3R5 tropism characterizes cerebrospinal fluid isolates from HIV-1-infected individuals with high cerebrospinal fluid viral load and low CD4+ T-cell count
Paired plasma and CSF isolates from 28 HIV-1-infected individuals were evaluated for their ability to utilize a panel of HIV coreceptors with documented expression in the CNS. The subtypes of the isolates are depicted in Table 1. All isolates were found to utilize CCR5 and/or CXCR4, as previously shown . As depicted in Fig. 1, 11 out of 14 CSF R5 virus isolates from individuals with CSFhigh were of dual R3R5 phenotype, whereas exclusively monotropic R5 CSF isolates were found within the CSFlow group (seven of seven) (P = 0.001, Two-tailed Fisher's exact test). We also noted that two out of three CCR3− R5 isolates in the CSFhigh group were of subtype C origin and able to use CXCR6. The level of CCR3 use was similar to CCR5 use in some dual tropic CSF isolates (Table 1) and was significantly more pronounced in CSF isolates than in corresponding plasma isolates (P = 0.035, Wilcoxon signed rank test). The median CSF viral load in individuals with dual R3R5 CSF isolates was 119000 versus 19500 copies/ml in individuals with CCR3− R5 CSF isolates (P = 0.001, Mann–Whitney rank sum test). In contrast, plasma viral load did not differ significantly between individuals harboring dual R3R5 plasma virus (median 52000 copies/ml) and those with monotropic R5 plasma isolates (median 23000 copies/ml). However, in patients with R5 HIV-1 infection, the presence of CCR3+ isolates in plasma and/or CSF was associated with lower CD4+ T-cell counts (P = 0.03 and P = 0.001, respectively, Mann–Whitney rank sum test). Among individuals with R5 HIV-1 infection, there was also an inverse correlation between CD4+ T-cell counts and CSF viral load (P = 0.003, Spearman's rank correlation test), which was lost when including individuals with X4/R5X4 viruses. Accordingly, no correlation was found between CCR3-use and CSF viral load in patients who harbored either X4 or R5X4 HIV-1 (Fig. 1). Dual R3R5 tropism and CSFhigh were not specifically confined to individuals with HAD, but were found also in neuroasymptomatic individuals with low CD4+ T-cell counts (Table 1). In summary, dualtropic R3R5 CSF isolates are commonly exhibited by individuals with high CSF viral load and coincide with low CD4+ T-cell counts. In the few individuals with X4/R5X4 virus phenotypes, no clear associations between CCR3 use, low CD4+ T-cell counts, and CSFhigh were observed.
HIV-1 cerebrospinal fluid isolates of subtype C origin commonly exhibit CXCR6 use
CXCR6 use was displayed by four out of five CSF isolates of subtype C origin, as compared to one out of 23 nonsubtype C isolates (Two-tailed Fisher's exact test, P = 0.001) (Fig. 1 and Table 1). CXCR6 use was highly efficient and equal to CCR5 use in one CSF isolate (patient 11), whereas CSF isolates of patients three, nine and 10 had low-grade CXCR6 use and efficient CCR5 use. The CSF isolate of patient 11, also exhibited efficient CCR3 use and was derived from an individual with extremely high CSF viral load (>750.000 copies/ml), while the plasma viral load was low (1450 copies/ml). This patient had a chronic infection and suffered no signs of neurocognitive impairment. It is worth noting that CXCR6 use coincided with CSFhigh in subtype C-infected individuals despite relatively preserved CD4+ T-cell counts (Fig. 1 and Table 1). Two of the corresponding subtype C plasma isolates displayed low-grade CXCR6 use (patient nine and 11). Low-grade use of mainly GPR1 but also, RDC-1, APJ, and ChemR23 was exhibited by some CSF and/or plasma isolates (Table 1). BLT1 use was only exhibited by the subtype B control isolate, but not by any of the isolates included in the study (data not shown). Taken together, CXCR6 use was displayed by CSF isolates of subtype C origin, which coincided with high CSF viral load despite CD4+ T-cell counts well above the level of WHO recommendations for antiretroviral treatment .
To our knowledge, this is the first study to have assessed the use of coreceptors other than CCR5 and CXCR4 in CSF-derived HIV-1 isolates. Although CSF is not identical to brain tissue, it is a more readily sampled site that provides an important ‘window’ into HIV CNS infection . In this study we show that CCR3 use characterizes CSF-derived R5 HIV-1 isolates in individuals with high CSF viral load, and that this coincides with low CD4+ T-cell counts. CCR3 was early identified as a coreceptor able to promote infection of microglial cells by selected HIV-1 isolates , which has support in a more recent study . Using a gene knockout strategy, Agrawal et al.  showed that CCR3 and CCR5 colocalize with CD4 molecules at the cell surface and that both coreceptors are required to convey an efficient infection of microglial cells and macrophages by R3R5 HIV-1 isolates. Given that these are the major target cells for HIV-1 in the brain, it is reasonable to speculate that the acquisition of CCR3 use, specifically and more pronounced in CSF isolates, reflects a viral adaption to replication in mononuclear phagocytes within the CNS. This could theoretically explain the correlation between dual R3R5 tropism in CSF-derived isolates and elevated CSF viral load found in this study. Consequently, a more efficient CCR3 use in CSF isolates than in plasma isolates may reflect differences in available target cells between the two compartments. Furthermore, increased macrophage tropism is exhibited by HIV-1 during late-stage disease , which in our study may be mirrored by the frequent appearance of R3R5 isolates in individuals with low CD4+ T-cell counts.
Unexpectedly, CXCR6 use was exhibited by four out of five CSF-derived isolates from individuals with subtype C HIV-1 infection. Although efficient CXCR6 use previously has been shown for the HIV simian counterpart SIV and HIV-2 [29,30], mainly weak and infrequent receptor use has been exhibited by HIV-1 plasma isolates [31,32]. Worth noting, CXCR6+ CSF isolates in subtype C-infected individuals coincided with CSFhigh despite relatively preserved CD4+ T-cell counts. In fact, all other individuals in the cohort with CSFhigh had AIDS, as defined by CD4+ T-cell counts of 200 cells/μl or less and/or ADD. Clearly, further studies are warranted to verify if CXCR6 use is more common in HIV-1 CSF isolates of subtype C origin and also if this may correlate with elevated CSF viral load. The latter is specifically important because high viral load may correlate with the development of neurological morbidity such as HAD . Furthermore, subtype C HIV-1 represents approximately 50% of the global epidemic. Studies addressing the influence of subtype on HIV-1-related neurological complications are few, but have shown that HAD poses a major health concern, also in areas of the world where subtype C HIV-1 infection dominate . Our study cohort was retrospectively selected to include individuals with varying clinical, virological, and immunological parameters, which may have affected the results. Nevertheless, despite the limited number of individuals included in the cohort, the results presented merit future validation.
In conclusion, increased CSF viral load in late-stage HIV-1 infection correlates with the emergence of dual R3R5 tropic viruses in CSF. In addition, our results indicate that subtype-specific differences in HIV coreceptor use may exist and remain unnoticed when analyzing plasma isolates.
This project was supported by grants from The Swedish Research Council (grant numbers K2008–58P-20930–04–1, K2010-56X-14588-08-3 and K2008–58X-20931–01–1), the Crafoord Foundation, The Clas Groschinsky Foundation, Lund University/Region Skane and the Sahlgrenska Academy at the University of Gothenburg (ALFGBG-11067). The technical assistance of Ulrika Mårtensson, Elsbieta Vincic and Jesper Bristulf is greatly appreciated.
The funding sources had no involvement in in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Conflicts of interest
The authors have no actual or potential conflict of interest including any financial, personal or other relationships with other people or organizations within 3 years of beginning the submitted work that could inappropriately influence, or be perceived to influence, the presented study.
1. Price RW, Brew B, Sidtis J, Rosenblum M, Scheck AC, Cleary P. The brain in AIDS: central nervous system HIV-1 infection and AIDS dementia complex
2. Mothobi NZ, Brew BJ. Neurocognitive dysfunction in the highly active antiretroviral therapy era
. Curr Opin Infect Dis
3. Pillai SK, Pond SL, Liu Y, Good BM, Strain MC, Ellis RJ, et al. Genetic attributes of cerebrospinal fluid-derived HIV-1 env
4. Karlsson U, Antonsson L, Repits J, Medstrand P, Owman C, Kidd-Ljunggren K, et al. Mode of coreceptor use by R5 HIV type 1 correlates with disease stage: a study of paired plasma and cerebrospinal fluid isolates
. AIDS Res Hum Retroviruses
5. Gendelman HE, Meltzer MS. Mononuclear phagocytes and the human immunodeficiency virus
. Curr Opin Immunol
6. Berger EA, Murphy PM, Farber JM. Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease
. Annu Rev Immunol
7. Spudich SS, Huang W, Nilsson AC, Petropoulos CJ, Liegler TJ, Whitcomb JM, et al. HIV-1 chemokine coreceptor utilization in paired cerebrospinal fluid and plasma samples: a survey of subjects with viremia
. J Infect Dis
8. Soulie C, Tubiana R, Simon A, Lambert-Niclot S, Malet I, Canestri A, et al. Presence of HIV-1 R5 viruses in cerebrospinal fluid even in patients harboring R5X4/X4 viruses in plasma
. J Acquir Immune Defic Syndr
9. Gabuzda D, Wang J. Chemokine receptors and virus entry in the central nervous system
. J Neurovirol
10. He J, Chen Y, Farzan M, Choe H, Ohagen A, Gartner S, et al. CCR3 and CCR5 are co-receptors for HIV-1 infection of microglia
11. Agrawal L, Maxwell CR, Peters PJ, Clapham PR, Liu SM, Mackay CR, et al. Complexity in human immunodeficiency virus type 1 (HIV-1) co-receptor usage: roles of CCR3 and CCR5 in HIV-1 infection of monocyte-derived macrophages and brain microglia
. J Gen Virol
12. Soda Y, Shimizu N, Jinno A, Liu HY, Kanbe K, Kitamura T, et al. Establishment of a new system for determination of coreceptor usages of HIV based on the human glioma NP-2 cell line
. Biochem Biophys Res Commun
13. Andersson LM, Svennerholm B, Hagberg L, Gisslen M. Higher HIV-1 RNA cutoff level required in cerebrospinal fluid than in blood to predict positive HIV-1 isolation
. J Med Virol
14. British HIV Association guidelines for antiretroviral treatment of HIV seropositive individuals. BHIVA Guidelines Co-ordinating Committee
15. Carpenter CC, Fischl MA, Hammer SM, Hirsch MS, Jacobsen DM, Katzenstein DA, et al. Antiretroviral therapy for HIV infection in 1996. Recommendations of an international panel. International AIDS Society-USA
16. Saag MS, Holodniy M, Kuritzkes DR, O’Brien WA, Coombs R, Poscher ME, et al. HIV viral load markers in clinical practice
. Nat Med
17. Bjorndal A, Deng H, Jansson M, Fiore JR, Colognesi C, Karlsson A, et al. Coreceptor usage of primary human immunodeficiency virus type 1 isolates varies according to biological phenotype
. J Virol
18. Esbjornsson J, Mansson F, Martinez-Arias W, Vincic E, Biague AJ, da Silva ZJ, et al. Frequent CXCR4 tropism of HIV-1 subtype A and CRF02_AG during late-stage disease: indication of an evolving epidemic in West Africa
19. Esbjornsson J, Mild M, Mansson F, Norrgren H, Medstrand P. HIV-1 molecular epidemiology in Guinea-Bissau, West Africa: origin, demography and migrations
. PLoS One
20. Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0
. Mol Biol Evol
21. Zwickl DJ. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion [dissertation]. The Univeristy of Texas at Austin; 2006.
22. Martensson UE, Fenyo EM, Olde B, Owman C. Characterization of the human chemerin receptor – ChemR23/CMKLR1 – as co-receptor for human and simian immunodeficiency virus infection, and identification of virus-binding receptor domains
23. Law NM, Rosenzweig SA. Characterization of the G-protein linked orphan receptor GPRN1/RDC1
. Biochem Biophys Res Commun
24. Croitoru-Lamoury J, Guillemin GJ, Boussin FD, Mognetti B, Gigout LI, Cheret A, et al. Expression of chemokines and their receptors in human and simian astrocytes: evidence for a central role of TNF alpha and IFN gamma in CXCR4 and CCR5 modulation
25. Noguchi K, Okubo M. Leukotrienes in nociceptive pathway and neuropathic/inflammatory pain
. Biol Pharm Bull
27. Price RW, Staprans S. Measuring the ‘viral load’ in cerebrospinal fluid in human immunodeficiency virus infection: window into brain infection?
. Ann Neurol
28. Gorry PR, Churchill M, Crowe SM, Cunningham AL, Gabuzda D. Pathogenesis of macrophage tropic HIV-1
. Curr HIV Res
29. Blaak H, Boers PH, Gruters RA, Schuitemaker H, van der Ende ME, Osterhaus AD. CCR5, GPR15, and CXCR6 are major coreceptors of human immunodeficiency virus type 2 variants isolated from individuals with and without plasma viremia
. J Virol
30. Riddick NE, Hermann EA, Loftin LM, Elliott ST, Wey WC, Cervasi B, et al. A novel CCR5 mutation common in sooty mangabeys reveals SIVsmm infection of CCR5-null natural hosts and efficient alternative coreceptor use in vivo
. PLoS Pathog
31. Tscherning-Casper C, Vodros D, Menu E, Aperia K, Fredriksson R, Dolcini G, et al. Coreceptor usage of HIV-1 isolates representing different genetic subtypes obtained from pregnant Cameroonian women. European Network for In Utero Transmission of HIV-1
. J Acquir Immune Defic Syndr
32. Edinger AL, Hoffman TL, Sharron M, Lee B, O’Dowd B, Doms RW. Use of GPR1, GPR15, and STRL33 as coreceptors by diverse human immunodeficiency virus type 1 and simian immunodeficiency virus envelope proteins
33. Robertson K, Fiscus S, Kapoor C, Robertson W, Schneider G, Shepard R, et al. CSF, plasma viral load and HIV associated dementia
. J Neurovirol
34. Joska JA, Westgarth-Taylor J, Myer L, Hoare J, Thomas KG, Combrinck M, et al. Characterization of HIV-Associated Neurocognitive Disorders among individuals starting antiretroviral therapy in South Africa
. AIDS Behav