RESEARCH ARTICLE

T cell derived HIV-1 is present in the CSF in the

face of suppressive antiretroviral therapy

Gila LustigID 1, Sandile CeleID

2,3, Farina KarimID 2,3, Anne Derache2, Abigail Ngoepe2,

Khadija Khan2,3, Bernadett I. Gosnell4, Mahomed-Yunus S. MoosaID 4, Ntombi Ntshuba2‡,

Suzaan Marais5, Prakash M. Jeena6, Katya GovenderID 2, John Adamson2,

Henrik Kløverpris2,7,8, Ravindra K. GuptaID 2,9, Rohen HarrichandparsadID

10, Vinod

B. Patel5, Alex SigalID 2,3,11*

1 Centre for the AIDS Programme of Research in South Africa, Durban, South Africa, 2 Africa Health

Research Institute, Durban, South Africa, 3 School of Laboratory Medicine and Medical Sciences, University

of KwaZulu-Natal, Durban, South Africa, 4 Department of Infectious Diseases, University of KwaZulu-Natal,

Durban, South Africa, 5 Department of Neurology, University of KwaZulu-Natal, Durban, South Africa,

6 Discipline of Pediatrics and Child Health, University of KwaZulu-Natal, Durban, South Africa, 7 Division of

Infection and Immunity, University College London, London, United Kingdom, 8 Department of Immunology

and Microbiology, University of Copenhagen, Copenhagen, Denmark, 9 Department of Medicine, University

of Cambridge, Cambridge, United Kingdom, 10 Department of Neurosurgery, University of KwaZulu-Natal,

Durban, South Africa, 11 Max Planck Institute for Infection Biology, Berlin, Germany

‡ Unavailable.

* [email protected]

Abstract

HIV cerebrospinal fluid (CSF) escape, where HIV is suppressed in blood but detectable in

CSF, occurs when HIV persists in the CNS despite antiretroviral therapy (ART). To deter-

mine the virus producing cell type and whether lowered CSF ART levels are responsible for

CSF escape, we collected blood and CSF from 156 neurosymptomatic participants from

Durban, South Africa. We observed that 28% of participants with an undetectable HIV blood

viral load showed CSF escape. We detected host cell surface markers on the HIV envelope

to determine the cellular source of HIV in participants on the first line regimen of efavirenz,

emtricitabine, and tenofovir. We confirmed CD26 as a marker which could differentiate

between T cells and macrophages and microglia, and quantified CD26 levels on the virion

surface, comparing the result to virus from in vitro infected T cells or macrophages. The

measured CD26 level was consistent with the presence of T cell produced virus. We found

no significant differences in ART concentrations between CSF escape and fully suppressed

individuals in CSF or blood, and did not observe a clear association with drug resistance

mutations in CSF virus which would allow HIV to replicate. Hence, CSF HIV in the face of

ART may at least partly originate in CD4+ T cell populations.

Author summary

The brain may be a site where HIV persists despite antiretroviral therapy (ART). Persis-

tence can manifest as cerebrospinal fluid (CSF) escape, where HIV is detectable in the

CSF but not the blood in some individuals. The reasons for CSF escape are incompletely

PLOS PATHOGENS

PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1009871 September 23, 2021 1 / 21

a1111111111

a1111111111

a1111111111

a1111111111

a1111111111

OPEN ACCESS

Citation: Lustig G, Cele S, Karim F, Derache A,

Ngoepe A, Khan K, et al. (2021) T cell derived HIV-

1 is present in the CSF in the face of suppressive

antiretroviral therapy. PLoS Pathog 17(9):

e1009871. https://doi.org/10.1371/journal.

ppat.1009871

Editor: Ronald Swanstrom, University of North

Carolina at Chapel Hill, UNITED STATES

Received: August 13, 2020

Accepted: August 6, 2021

Published: September 23, 2021

Copyright: © 2021 Lustig et al. This is an open

access article distributed under the terms of the

Creative Commons Attribution License, which

permits unrestricted use, distribution, and

reproduction in any medium, provided the original

author and source are credited.

Data Availability Statement: All relevant data are

within the manuscript and its Supporting

information files.

Funding: This study was supported by National

Institute of Mental Health award R21MH104220

(AS) and National Institute of Allergy and Infectious

Diseases award R01AI138546 (AS). Salary support

was provided to GL and AS from National Institute

of Mental Health award R21MH104220 and

National Institute of Allergy and Infectious Diseases

award R01AI138546. The funders had no role in

understood. Evidence from individuals mostly on second line protease inhibitor-based

ART indicates that detectable HIV in this compartment may have acquired drug resis-

tance. In this work we investigated HIV in blood and CSF of 156 participants from Dur-

ban, South Africa. We observed a very high prevalence of CSF escape of 28%. We aimed

to find the cell type responsible for producing HIV in CSF escape and whether replication

occurred because of lower CSF drug levels or because the virus has developed resistance

to therapy. We found that at least some of the CSF HIV was produced by T cells, and that

drug resistance was not always present. This suggests that at least part of the CSF HIV res-

ervoir may be generated by either an infection mode not requiring drug resistance for

viral replication, or by latently infected CD4+ T cells trafficking to and releasing HIV in

the CSF without extensive viral replication taking place.

Introduction

HIV persistence in the face of ART necessitates lifelong adherence to treatment. The CNS may

serve as one reservoir for HIV persistence [1]. HIV infection in the CNS in the absence of sup-

pressive ART may lead to HIV-associated neurocognitive disorders (HAND). Yet, even in the

presence of ART mediated suppression, sub-clinical cognitive impairment is common [2–7]

and there is widespread immune activation and inflammation in the CNS [8–10]. Consistent

with a role for the CNS as an HIV reservoir, a subset of individuals show CSF escape, where

HIV is detectable in the CSF while being successfully suppressed below the level of detection

in the blood [11–16].

Potential reasons for a CNS reservoir include reduced drug levels. Drug levels of efavirenz

(EFV), emtricitabine (FTC), and tenofovir (TFV) in the CSF are reduced approximately

200-fold, 2-fold, and 20-fold, respectively in the CSF relative to blood [17–19]. Since the

majority of individuals do not show detectable virus in the CSF, these lowered ART levels

seem to be sufficient to suppress viremia. However, it is unclear if ART levels in the CSF are

lower in individuals with neurosymptomatic CSF escape in South Africa, accounting for the

lack of effective suppression in this compartment and possibly evolution of drug resistance

mutations. Mutations could include the M184V or M184I resistance mutation to FTC, a drug

which would provide selective pressure since it has good penetration to the CNS [14, 20–24].

There is evidence for compartmentalized HIV infection in the CNS, indicating that CNS

specific cell subtypes may be involved [25–27]. HIV infected, long lived CNS resident host

cells such as microglia and perivascular macrophages may be responsible for the HIV reservoir

in the CNS [11, 27–31]. HIV infection may not be appreciably cytotoxic in these cells [32] and

these cells are resistant to cytotoxic T lymphocyte killing [33], allowing long-term infected cell

persistence without new cycles of re-infection.

T cells are also present in the CNS. CSF contains trafficking T cells, mostly CD4+ memory

cells, which enter across the choroid plexus [34]. T cell-tropic HIV is present in the CSF in

some individuals [27, 28, 35] and CSF HIV was found to have fast decay kinetics upon ART

initiation, consistent with the short half-lives of infected T cells [36].

Here we aimed to determine the cellular source of HIV in the brain and whether lower

ART levels relative to fully suppressed individuals account for CSF escape in individuals from

Durban, South Africa. This is the first time in our knowledge where CSF escape in Sub-Saha-

ran Africa has been investigated in a relatively large number of participants on ART [37].

Since accessing HIV infected cells from the CNS is challenging, we chose a method which

could determine the cell-of-origin of cell-free HIV sampled from the CSF. Upon viral budding,

PLOS PATHOGENS T cell derived HIV-1 in the CSF in the face of suppressive antiretroviral therapy

PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1009871 September 23, 2021 2 / 21

study design, data collection and analysis, decision

to publish, or preparation of the manuscript.

Competing interests: The authors have declared

that no competing interests exist. Author Ntombi

Ntshuba was unable to confirm their authorship

contributions. On their behalf, the corresponding

author has reported their contributions to the best

of their knowledge.

the HIV envelope contains host surface markers [38, 39] since HIV uses the cellular plasma

membrane as its envelope. The host surface markers can be bound with antibodies and

detected using a variety of techniques [40–42] including electron microscopy [39], mass spec-

trometry [43], flow cytometry [44, 45], and immunomagnetic capture [46–50]. Many of these

studies report that HIV derived from macrophage lineage cells expresses CD36 [43–46, 48–

50], a scavenger receptor [51–56], on its envelope. HIV derived from T cells expresses CD26

[44, 46, 48–50], a dipeptidyl-peptidase involved in T cell activation [57]. We tested the ability

of CD26 and CD36 to differentiate between T cells and other cell types. We found that in sam-

ples obtained from the South African study participants, CD26 was T cell specific. CSF escape

HIV had CD26 on its surface, consistent with at least partial T cell origin of CSF escape virus.

We also observed that CSF ART concentrations in CSF escape were not significantly different

from those of participants with viral suppression and did not detect a clear association with

drug resistance mutations, indicating that detectable T cell origin HIV can persist in the CSF

despite suppressive ART.

Materials and methods

Ethical statement

CSF and matched blood were obtained from participants indicated for lumbar puncture

enrolled at Inkosi Albert Luthuli Central Hospital and King Edward VIII Hospital in Durban,

South Africa after written informed consent (University of KwaZulu-Natal Institutional

Review Board approval BE385/13). Discarded tissue from the field of neurosurgery was

obtained from two participants enrolled at Inkosi Albert Luthuli Central Hospital indicated

for neurosurgery after written informed consent (University of KwaZulu-Natal Institutional

Review Board approval BE315/18). Blood for PBMC, CD4+ and CD14+ cell isolation was

obtained from adult healthy volunteers after written informed consent (University of Kwa-

Zulu-Natal Institutional Review Board approval BE022/13 and BE083/18). Lymph nodes were

obtained from the field of surgery of participants undergoing surgery for diagnostic purposes

and/or complications of inflammatory lung disease. Informed consent was obtained from each

participant, and the study protocol approved by the University of KwaZulu-Natal Institutional

Review Board (approval BE024/09).

Statistical tests

Data is described with the non-parametric measures of median and interquartile range, and

significance determined using the non-parametric Mann-Whitney U test for pairwise compar-

isons, Fisher exact test for pairwise comparisons of frequencies, and the Kruskal-Wallis test

with multiple comparison correction by the Dunn Method for comparisons involved more

than two populations. All tests were performed using Graphpad Prism 8 software.

CSF sample collection and processing

Fresh CSF and matching blood samples were transported to the laboratory and processed

immediately. Two separate EDTA tubes of 4 mL blood were sent for testing in parallel: one for

a CD4/CD8 count at an accredited diagnostic laboratory (Ampath, Durban, South Africa) and

one for viral load at an accredited diagnostic laboratory (Molecular Diagnostic Services, Dur-

ban, South Africa, using the RealTime HIV-1 viral load test on an Abbott machine). CSF sam-

ples were spun for 10 min at 1000 g to remove debris. CSF supernatant was frozen in 1 mL

aliquots at -80˚C. One aliquot of 100 μl CSF was sent for viral load (Molecular Diagnostic

PLOS PATHOGENS T cell derived HIV-1 in the CSF in the face of suppressive antiretroviral therapy

PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1009871 September 23, 2021 3 / 21

Services) directly after the spin. Plasma was processed by spinning the whole blood for 10 min

at 1300 g, no brake. The top layer was removed and stored in 1 mL aliquots at -80˚C.

Antiretrovirals, viruses and cells

The following reagents were obtained through the AIDS Research and Reference Reagent Pro-

gram, National Institute of Allergy and Infectious Diseases, National Institutes of Health: the

antiretrovirals EFV, FTC, and TFV and the plasmid for the macrophage tropic pNL4–3(AD8)

HIV molecular clone. NL4–3 and NL4–3(AD8) HIV stocks were produced by transfection of

HEK293 cells with the molecular clone plasmids using TransIT-LT1 (Mirus) transfection

reagent. Supernatant containing released virus was harvested two days post-transfection and

filtered through a 0.45 micron filter (GVS) and stored in 0.5 mL aliqouts at -80˚C. The number

of HIV RNA genomes in viral stocks was determined using the RealTime HIV-1 viral load test

(Molecular Diagnostic Services, Durban, South Africa). RevCEM-E7 cells were generated as

previously described [58]. Cell culture medium was complete RPMI 1640 supplemented with

L-glutamine, sodium pyruvate, HEPES, non-essential amino acids (Lonza), and 10% heat-

inactivated FBS (Hyclone). PBMCs were isolated by density gradient centrifugation using His-

topaque 1077 (Sigma). CD4+ or CD14+ cells were positively selected using either CD4 or

CD14 Microbeads loaded onto MACS separation columns according to manufacturer’s

instructions (Miltenyi Biotec). CD4+ PBMCs were grown in the above cell media supple-

mented with 5 ng/mL IL-2 (Peprotech) and 10 μg/mL PHA (Sigma-Aldrich). Monocyte-

derived macrophages were grown in RPMI 1640 supplemented with 10% human serum

(Sigma) with added L-glutamine, sodium pyruvate, HEPES, and non-essential amino acids

(Lonza), and differentiated with 20 ng/mL M-CSF (Peprotech) for 10 days.

Surface staining and detection of CD26 and CD36 markers by flow

cytometry

Staining of MDM and PBMC T cells: MDM and CD4+ PBMCs were generated as described

above. PBMCs were washed once in PBS-/-. MDM were washed once in PBS-/- then incubated

in 5 mM EDTA in PBS-/- for 30 minutes on ice. Macrophages were collected by pipetting vig-

orously and the remaining attached cells were removed by gentle scraping. Cells were then

incubated with either CD3-APC and CD8-Bv500 (PBMC) or CD68-APC (MDM) and

CD26-FITC and CD36-PE (Biolegend) fluorescently labelled antibodies in staining buffer

(PBS-/- with 3%FCS) for 30 min on ice. The samples were then washed, resuspended in 400 μL

staining buffer and acquired on a FACSCalibur machine (BD Biosciences). Results were ana-

lyzed using FlowJo software. Staining of LN cells: LN from the field of indicated cardiothoracic

surgery were cellularized by gentle mechanical dissociation and cryopreserved. For staining,

LN cells were thawed, washed once in PBS-/-, then incubated with the following fluorescently

conjugated antibody mix for 30 min: CD45-HV500, CD19-BV785, CD3-PE-CF594,

CD4-AF700 HLA-DR-BV605 (all BD Biosciences), CD26-FITC, CD36-PE (Biolegend), and

the LIVE/DEAD Fixable Near-IR Dead Cell Stain (ThermoFisher Scientific). Cells were then

fixed and permeabilized using the BD Cytofix/Cytoperm Fixation/Permeabilization kit (BD

Biosciences) according to the manufacturer’s instructions. Cells were then stained with

CD68-APC antibodies (Biolegend). Cells were washed, fixed in 2% formaldehyde and acquired

on a BD ARIA Fusion flow cytometer (BD Biosciences).

Isolation and surface staining of human microglia

Meninges discarded tissue samples were transported immediately to the laboratory for pro-

cessing. The tissue was dissociated (Brain Dissociation kit, Miltenyi). The dissociated cells

PLOS PATHOGENS T cell derived HIV-1 in the CSF in the face of suppressive antiretroviral therapy

PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1009871 September 23, 2021 4 / 21

underwent myelin removal (Myelin removal kit, Miltenyi) and CD11b+ cells were further

purified (CD11b microglia Microbeads, Miltenyi). All kits were used according to the manu-

facturers instructions. The cells were then surface stained with fluorescently conjugated anti-

bodies for microglial markers: CD11b-APC, CD45-Bv605 and P2RY12-Bv421, CD26-FITC,

and CD36-PE (Biolegend). The cells were incubated with antibodies for 30 min on ice, washed,

resuspended in 500 μL staining buffer, and acquired on a FACS Fortessa (BD Biosciences).

Results were analyzed using FlowJo software.

in vitro generation of virus from PBMC or MDM

To generate PBMC origin HIV, PBMCs isolated and activated as described above were

infected with 2 × 106 RNA copies/mL NL4–3(AD8) for 24 hours. Cells were then washed 4

times in growth medium to remove the input viral stock and incubated for 4 days (approxi-

mately two full virus cycles) for a total infection time of 5 days. Virus containing supernatant

was collected, centrifuged at 300 g for 5 minutes then filtered through a 0.45 micron syringe fil-

ter (GVS) to remove cells and cellular debris. The number of virus genomes in the virus stock

was determined using the RealTime HIV-1 viral load test (Molecular Diagnostic Services). To

generate monocyte-derived macrophage virus, CD14+ monocytes were isolated and differenti-

ated as described above. Cells were then infected with with 2 × 107 RNA copies/mL NL4–3

(AD8) for 24 hours. Cells were washed 6 times with RPMI to remove input virus, and growth

media was replaced. Half the volume of media was replaced every 3 days for 14 days for a total

infection time of 15 days. Virus containing supernatant was collected, centrifuged and filtered,

and viral load determined as for PBMC virus. We used 7 different donors blood donors: 6009,

6013, 6017, 6019, 6026, 6033 and 6049.

Host cell marker detection on virion surface

The following protocol was adapted from the μMACS Streptavidin Kit protocol (Miltenyi): 1

μg of biotinylated antibodies to CD26 or CD36 (Ancell) were added to 1 mL of virus, mixed

and incubated for 30 min at room temperature. Next, 30 μL of strepavidin MicroBeads (Milte-

nyi) were added per sample, mixed and incubated at room temperature for 10 min. The sam-

ples were then loaded onto a μColumn, washed three times and bound virus eluted. Clinical

virus samples were centrifuged for 13,000 g for 30 seconds to clear debris before addition of

antibodies. To avoid overloading columns, in vitro generated virus stocks from either PBMCs

or macrophages were diluted to approximately 104 RNA copies/mL in PBS before incubation

with antibodies. We used the 7 donors above. Virus was precipitated once for six of the donors,

whereas 6049 was done in duplicate. The number of virus genomes in elutions from μColumns

was determined using the RealTime HIV-1 viral load test (Molecular Diagnostic Services). Out

of 22 CSF escape samples, 11 had sufficient volume for the assay (2 mL). One sample which

showed neither detectable CD26 nor CD36 was excluded from the analysis due to possible deg-

radation of the virus.

Generation of YFP-NL4–3(AD8)

pNL4–3(AD8) was used as the source of the macrophage tropic HIV ENV which was excised

from pNL4–3(AD8) using BamHI and EcoRI restriction enzymes (NEB) and ligated using T4

ligase (Invitrogen) into the pNL4–3-YFP vector (gift from David Levy), replacing the NL4–3

X4 specific HIV envelope gene between the unique EcoRI-BamHI restriction sites.

PLOS PATHOGENS T cell derived HIV-1 in the CSF in the face of suppressive antiretroviral therapy

PLOS Pathogens | https://doi.org/10.1371/journal.ppat.1009871 September 23, 2021 5 / 21

Detection of ART concentrations in CSF and matched plasma by LC-MS/MS

Sample analysis was performed using an Agilent High Pressure Liquid Chromatography

(HPLC) system coupled to the AB Sciex 5500, triple quadrupole mass spectrometer equipped

with an electrospray ionization (ESI) TurboIonSpray source. The LC-MS/MS method was

developed and optimised for the quantitation of tenofovir, emtricitabine, efavirenz, lopinavir,

ritonavir, nevirapine, zidovudine, lamivudine, abacavir, atazanvir and dolutegravir in the same

sample. A protein precipitation extraction method using acetonitrile was used to process clini-

cal plasma and CSF samples. The procedure was performed using 50 μL of plasma or CSF. 50

μL of water and 50 μL of ISTD solution was added and the sample was briefly mixed. 150 μL of

acetonitrile was subsequently added to facilitate protein precipitation, vortex mixed and cen-

trifuged at 16000 g for 10 min at 4˚C. 170 μL of the clear supernatant was then transferred to a

clean micro-centrifuge tube and dried down using a SpeedVac dryer set at 40˚C. The dried

samples were then reconstituted in 100 μL of 0.02% sodium deoxycholate (Sigma) in Millipore