Immunology Letters 66 (1999) 111–116

Subpopulations of peripheral blood dendritic cells show differentialsusceptibility to infection with a lymphotropic strain of HIV-1

Steven Patterson *, Stephen P. Robinson, Nicholas R. English, Stella C. Knight

Antigen Presentation Research Group, Imperial College School of Medicine at Northwick Park, Institute for Medical Research, Harrow,Middlesex, HA1 3UJ, UK

Received 19 October 1998; accepted 18 November 1998

Abstract

Blood dendritic cells (DC) express CD4 and are susceptible to HIV infection. By electron microscopy two morphologicallydistinct types of DC were identified in peripheral blood. Only one of these two types was susceptible to infection with alymphotropic strain of HIV-1. By FACS two populations could be defined based on the expression of CD11c. The morphologyof cultured FACS-purified CD11c negative DC was similar to that DC population shown to be susceptible to infection with thelymphotropic strain of HIV. Furthermore after several hours in culture CXCR4, the co-receptor for lymphotropic strains ofHIV-1, was expressed at a significantly higher level on the CD11c negative DC than on the CD11c positive cells. This studysuggests that there are subpopulations of DC that show differences in susceptibility to infection with some strains of HIV-1.© 1999 Elsevier Science B.V. All rights reserved.

Keywords: Dendritic cells; Infection; AIDS; HIV; Pathogenesis

1. Introduction

Dendritic cells (DC) have an unparalleled ability topresent antigen to naive T-cells and play a crucial rolein the initiation and maintenance of cell mediated im-mune responses in viral infections. High level expres-sion of MHC, adhesion proteins, the co-stimulatorymolecules CD40, CD80 and CD86, and the ability tocluster T-cells are thought to contribute to the uniquestimulatory properties of DC. Dendritic cells arederived from CD34+ progenitors in the bone marrowunder the influence of cytokines including GM-CSFand TNFa. They are distributed through the body viathe blood, where they constitute less than 1% of themononuclear cell population, and are found in smallnumbers in all tissues with the possible exception of thebrain. At strategically important sites, such as mucosal

tissue where dendritic Langerhans’ cells are situated,they intercept invading pathogens. After acquiring for-eign antigen they travel to the draining lymph node viathe afferent lymphatics and in the T-dependent areascluster and present antigen to T-cells [1,2].

Dendritic cells express CD4 and are susceptible toHIV-1 infection in vitro and in vivo [3–9] and clustersof DC and T-cells are thought to be important sites ofHIV replication [10,11]. In addition DC function isimpaired in HIV infection and this is believed to con-tribute to the development of immune deficiency [8,12].In an earlier electron microscopy study we investigatedthe infection of semi purified preparations of blood DCby the lymphotropic IIIB strain of HIV-1 [13]. Twodistinct morphological types of DC, termed type 1 and2, were identified but only one was susceptible toproductive virus infection. It was proposed that the twocell types represented different stages of a commondifferentiation pathway. New data is now presentedsuggesting that these different morphological types of

* Corresponding author:. Tel.: +44-181-8693420; fax: +44-181-8693532; e-mail: s.patterson@ic.ac.uk.

0165-2478/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 5 -2478 (98 )00166 -7

S. Patterson et al. / Immunology Letters 66 (1999) 111–116112

DC may represent two distinct DC lineages with differ-ential susceptibility to infection with lymphotropicstrains of HIV-1.

2. Materials and methods

2.1. Flow cytometry

Freshly isolated PBMC were simultaneously labelledwith PerCP-conjugated anti-MHC class II DR and acocktail of PE-conjugated antibodies directed againstCD3, CD14, CD16, CD19 and CD34 (purchased fromBecton–Dickinson). Dendritic cells were identified byhigh level expression of DR and absence of labellingwith the PE-conjugated lineage specific antibodies. Ex-pression of CD4 and CD11c on DC was detected withFITC-conjugated antibodies (Becton–Dickinson) andthe chemokine receptor, CXCR4, labelled with a bi-otinylated antibody (R&D) followed by a streptavidin-FITC conjugated reagent. Cells were also labelled withappropriate isotype control antibodies in eachexperiment.

2.2. Infection of dendritic cells

The PBMC were separated from herparinized bloodby centrifugation over Ficoll. The cells were then cul-tured overnight in 1640 RPMI medium, Dutch modifi-cation, containing 100 IU/ml penicillin, 100 mg/mlstreptomycin and 10% heat-inactivated foetal calfserum. A DC-enriched fraction, containing 8–15% DC,was obtained by centrifuging the non-adherent cellfraction over a 13.7% metrizamide (Nygaard) gradientfor 10 min at 600×g. These preparations (5×105−106 cells) were infected with 104 TCID50 units of theIIIB strain of HIV-1 and cultured for 5 days beforeexamination by electron microscopy as previously de-scribed [6].

2.3. Dendritic cell purification

The PBMC were isolated from buffy coat cell prepa-rations on Ficoll density gradients. The PBMC werethen centrifuged over a 50% percoll gradient. The cellsat the top of the gradient, containing the DC fraction,were harvested and stored overnight at 4°C. Cells werethen labelled with antibodies specific for CD3, CD14and CD20 and labelled cells depleted with immunomag-netic beads (Dynal). The remaining cells were antibodylabelled with PE-conjugated anti CD3, CD14, CD16,CD19, CD34, cy-chrome-conjugated MHC class IIanti-DR and FITC-conjugated anti CD11c. The DC(i.e. cells expressing DR but not labelled with the PEreagents) were purified by flow cytometry using a Bec-ton–Dickinson FACS Vantage and separated into

CD11c positive and negative fractions. The purified DCwere cultured in GM-CSF (100 ng/ml) and TNFa (50m/ml) for 7 days before examination by electronmicroscopy.

3. Results

3.1. Flow cytometry

Dendritic cells can be clearly identified by flow cy-tometry based on high expression of MHC class IIexpression and the absence of labelling for markers thatidentify T, B, NK, monocytic and progenitor cells (Fig.1a). These cells can be divided into CD11c negative andCD11c positive populations (Fig. 1b). It is possible toseparate the two populations on labelling with the PEantibody cocktail and anti DR reagent since the CD11cpositive population expresses slightly higher levels ofMHC class II DR and are very weakly labelled by thePE CD14 antibody (Fig. 2). This observation enabledus to look for other receptors on the two populations ofDC using three colour analysis. In freshly isolatedPBMC CD4 was expressed on both populations of DCbut at a slightly higher level on the CD11c negativepopulation (Fig. 3a,b). CD4 levels were unchanged onDC in PBMC cultured for 3 h but after overnightincubation expression was reduced on the CD11c posi-tive cells and maintained on the CD11c negative frac-tion (Fig. 3c,d,e,f). The chemokine receptor CXCR4was not detectable on either type of DC in freshlyisolated PBMC but was detectable on both DC popula-tions after incubation for 3 h at 37°C (Fig. 4a,b,c,d).This chemokine receptor was more highly expressed onthe CD11c negative population whereas on the CD11cpositive cells the labelling was only marginally abovethat of the isotype control. Overnight culture did notresult in an increase in CXCR4 expression on theCD11c positive cells but was maintained on the CD11cnegative DC (Fig. 4e,f).

Fig. 1. Identification of two populations of blood DC. PBMC werelabelled with PE-conjugated anti-CD3, CD14, CD16, CD19 andCD34, with PerCP conjugated anti DR and with FITC conjugatedCD11c. In (a) DC are defined by the absence of labelling with PE(FL2) and strong labelling with PerCP. When DC are analysed forCD11c expression (b) two populations are observed, FL1 representsFITC labelling.

S. Patterson et al. / Immunology Letters 66 (1999) 111–116 113

Fig. 2. The two DC populations can be separated by two colourFACS. (a) A population of DC was acquired by gating on the PEantibody cocktail negative (FL2) DR (FL3) positive cells. Gates R5and R6 in (a) define the CD11c negative (b) and CD11c positive (c)DC, respectively. Histograms (b) and (c) show labelling of the cells inthese regions with FITC-conjugated anti CD11c.

Fig. 4. Expression of CXCR4 on CD11c negative and positive DC infresh PBMC and after culture. CD11c negative cells are shown in (a)fresh, (c) after 3 h and (e) after overnight culture. The CD11c positivecells are shown in (b) fresh, (d) after 3 h and (f) after overnightculture.

3.2. Infection of DC with the lymphotropic IIIB strainof HIV-I

In a previous investigation we identified two distinctmorphological types of DC in semi-purified prepara-tions. By immunogold labelling they were shown to

express MHC class II but did not label for CD3, CD14,CD16 or CD19. The first cell, termed a type 1 DC, hadnumerous short processes, a nucleus containing a sig-nificant amount of electron dense heterochromatin ad-jacent to the nuclear membrane and occasionalcytoplasmic vacuoles (Fig. 5a). The second cell, termeda type 2 DC, showed a much smoother cell boundarywith fewer processes, a nucleus containing less hete-rochromatin than the type 1 DC and typically areas ofcytoplasm devoid of organelles except for occasionalstrands of endoplasmic reticulum (Fig. 5b,d). In DC-en-riched cell preparations infected for 5 days with theIIIB strain of HIV-1, virus replication was never ob-served in the type 1 DC but was seen in type 2 DC.Virus was observed on the surface of these cells andalso budding through the cell membrane indicating thatthey were productively infected and had not merelyadsorbed virus onto their surface (Fig. 5c,e). Electronmicroscopy of FACS-purified CD11c positive and neg-ative DC cultured for 7 days revealed cells that mor-phologically resembled the type 1 and type 2 DCrespectively (Fig. 6).

4. Discussion

A number of studies have recognised different sub-

Fig. 3. Expression of CD4 on CD11c negative and positive DC infresh PBMC and after culture. CD11c negative cells shown in (a)fresh, (c) after 3 h and (e) after overnight culture. CD11c positive cellsare shown in (b) fresh, (d) after 3 h and (f) after overnight culture.

S. Patterson et al. / Immunology Letters 66 (1999) 111–116114

Fig. 5. (a) Electron micrograph of a type 1 DC immunogold labelled with anti MHC class II DQ. Bar represents 1 m. (b) and (d), Electronmicrographs of type 2 DC infected for 5 days with HIV-1 IIIB. Bar represents 1 m. (c) and (e), show higher magnifications of the areas indicatedin (b) and (d), respectively. The virus can be observed on the cell surface and budding through the cell membrane. Bar represents 300 nm.

populations of DC in peripheral blood but the func-tional and developmental relationship between thesecell types is not clear [3,13,14]. Based on morphologywe have previously identified two major types of bloodDC, termed ‘types 1’ and ‘2’. It was suggested thatthey represented different stages of a common devel-opmental pathway, the type 1 DC was proposed to bethe precursor of type 2 cells. Similarly, O’Doherty etal. [3] identified two types of blood DC based on theexpression of the b2-integrin, CD11c, and suggestedthat the CD11c negative cells were immature cells mi-grating from the bone marrow to the tissues whereasthe CD11c positive cells are mature DC travellingfrom the tissues to the secondary lymphoid organs.The data presented here suggest that the type 1 and 2DC described earlier represent the CD11c positive andnegative DC, respectively. Recent work in our labora-tory also indicates that these cells are not different

stages of a common developmental pathway but repre-sent two distinct lineages of DC which maintain differ-ential expression of CD11c after maturing for 7 daysin culture (S. Robinson et al., submitted for publica-tion).

Dendritic cells are postulated to play an importantrole in the pathogenesis of HIV infection [15]. Theyexpress CD4 and are susceptible to HIV infection invitro and in vivo. Studies in the SIV macaque animalmodel suggest that dendritic Langerhans’ cells in themucosal tissue of the genital tract are amongst the firstcells to become infected during sexual transmission ofthe virus [16]. An inability of lymphotropic virus vari-ants of HIV-1 to infect tissue Langerhans’ cells hasbeen proposed to explain the finding that host infec-tion is usually initiated by macrophage tropic variants[17].

S. Patterson et al. / Immunology Letters 66 (1999) 111–116 115

Fig. 6. Electron micrographs of FACS purified CD11c positive (a) and negative (b) DC cultured for 7 days in GM-CSF and TNFa. Bar represents1 m.

The present flow cytometry and electron microscopystudies have identified two distinct populations of DC.A correlation was observed between the morphology ofthe CD11c positive DC and the type 1 DC which werenot susceptible to productive infection with alymphotropic strain of HIV-1. The type 2 DC, in whichproductive infection with a lymphotropic virus wasobserved, were similar by electron microscopy to theCD11c negative DC. These cells may therefore be aseparate lineage from those cells that develop into tissueLangerhans’ cells.

It is now well established that although lymphotropicand macrophage tropic strains of HIV-1 both utiliseCD4 for cellular entry, they have different secondaryreceptor requirements. Macrophage tropic variants aredependent on b-chemokine receptors, particularlyCCR5, whereas lymphotropic virus uses the chemokinereceptor CXCR4 [18–20]. By FACS both populationsof DC were shown to express CD4 although this wasdown regulated in the CD11c positive DC afterovernight culture. Neither DC population in freshlyisolated PBMC expressed CXCR4 but after a few hoursin culture this receptor was expressed on the CD11cnegative population. By contrast labelling for CXCR4in the CD11c positive population was barely abovebackground levels. These findings may explain the dif-ferential susceptibility of the two DC populations toinfection with HIV-1 IIIB. One study has identifiedthree populations of cells with DC morphology inblood [14]. Two of these cell types were suggested torepresent immature and mature cells of the same devel-opmental pathway. The third cell type was the only onefound to be susceptible to HIV infection but sinceCD11c was not used as a marker and electron mi-

croscopy was not performed it is difficult to correlatethis data with the present findings [14]. RegardingCCR5 expression, only one commercial antibodyreagent was available to us and this did not givesatisfactory labelling of a cell line (PM1) expressingCCR5 or of the DC preparations.

These studies have presented evidence for two typesof blood DC which show differential susceptibility toinfection with lymphotropic strains of HIV-1. This mayhave relevance to disease pathogenesis but as yet we donot have evidence for any differences in function be-tween the CD11c negative and positive DC. However,we have data indicating some phenotypic similaritiesbetween Langerhans’ cells and cultured CD11c positiveDC (S. Robinson et al., submitted). The differentialsusceptibility of subpopulations of DC to infection withmacrophage and lymphotropic strains of HIV-1 mayalso be relevant to vaccine design since virus with anability to replicate in DC is thought to be responsiblefor initiating infection. In view of this observationfurther studies on the infection of subpopulations ofDC by different variants of HIV-1 are now underway.

Acknowledgements

SP, NRE and SCK were supported by the MedicalResearch Council. SPR was supported by the Associa-tion for International Cancer Research.

References

[1] J. Bancherau, R.M. Steinman, Nature 392 (1998) 245–252.[2] R.M. Steinman, Ann. Rev. Immunol. 9 (1991) 271–296.

S. Patterson et al. / Immunology Letters 66 (1999) 111–116116

[3] U. O’Doherty, R.M. Stemman, M. Peng, P.U. Cameron, S.Gezelter, I. Kopeloff, W.J. Swiggard, M. Pope, N. Bhardwaj, J.Exp. Med. 178 (1993) 1067–1078.

[4] J.J. Ferbas, J F. Toso, A.J. Logar, J.S. Navrath, C.R. Rinaldo,J. Immunol. 152 (1994) 4649–4662.

[5] S. Patterson, J. Gross, N. English, A. Stackpoole, P. Bedford,S.C. Knight, J. Gen. Virol. 76 (1995) 1155–1163.

[6] S. Patterson, S.C. Knight, J. Gen. Virol. 68 (1987) 1177–1181.[7] E. Langhoff, E.F. Terwilliger, H.J. Bos, K H. Kalland, M.C.

Pozansky, O.M. Bacon, W.A. Haseltine, Proc. Natl. Acad. Sci.USA 88 (1991) 7998–8002.

[8] S.E. Macatonia, R. Lau, S. Patterson, A.J. Pinching, S.C.Knight, Immunology 71 (1990) 38–45.

[9] S. Patterson, N.R. English, H. Longhurst, P. Balfe, M. Helbert,A.J. Pinching, S.C. Knight, J. Gen. Virol. 79 (1998) 247–257.

[10] M. Pope, M.G.H. Betjes, N. Romani, H. Hirmand, P.U.Cameron, L. Hoffman, S. Gezelter, G. Schuler, R.M. Stemman,Cell 78 (1994) 389–398.

[11] M. Pope, S. Gezelter, N. Gallo, L. Hoffman, R.M. Steinman, J.Exp. Med. 182 (1995) 2045–2056.

[12] S.C. Macatonia, S. Patterson, S.C. Knight, Immunology 67(1989) 285–289.

[13] S. Patterson, J. Gross, P. Bedford, S.C. Knight, Immunology 72(1991) 361–367.

[14] D. Weissman, Y. Li, J. Ananworanich, L. Zhou, J. Adelsberger,T.F. Tedder, M. Baseler, A.S. Fauci, Proc. Natl. Acad. Sci. USA92 (1995) 826–830.

[15] S.C. Knight, S. Patterson, Ann. Rev. Immunol. 15 (1997) 593–616.

[16] A.I. Spira, P.A. Marx, B.K. Patterson, J. Mahoney, R.A. Koup,S.M. Wolinsky, D.D. Ho, J. Exp. Med. 183 (1996) 215–225.

[17] M. Zaitseva, A. Blauvelt, S. Lee, C.K. Lapham, V. Klaus-Kov-tun, H. Mastowski, J. Manischewitz, H. Golding, Nature Med. 3(1997) 1369–1375.

[18] H. Deng, R Lui, W. Ellmeier, S. Choe, D. Unutmaz, M.Burkhart, P. Dimarzio, S. Marmon, R.E. Sutton, C.M. Hill,C.B. Davis, S.C. Peiper, S.C. Schall, D.R. Littman, N.R.Landau, Nature 381 (1996) 661–666.

[19] T.V. Dragic, V. Litwin, G.P. Allaway, S.R. Martin, Y. Huang,K.A. Nagashimu, C. Cayanan, P.J. Maddon, R.A. Koup, J.Moore, W.A. Paxton, Nature 381 (1996) 667–673.

[20] Y. Feng, C.C. Broder, P.E. Kennedy, E.A. Berger, Science 272(1996) 872–877.

..