Comparative analysis of cell culture and prediction algorithms for phenotyping of genetically diverse HIV-1 strains from Cameroon

Article (PDF Available)inAIDS Research and Therapy 6(1):27 · November 2009with16 Reads
DOI: 10.1186/1742-6405-6-27 · Source: PubMed
Abstract
With the advent of entry inhibitors, monitoring of viral tropism in the clinical setting is important. Conventional methods are cell-based and lengthy, therefore V3 sequence based prediction algorithms are becoming increasingly attractive as monitoring tools. Here we report a comparative analysis of viral tropism of strains circulating in Cameroon where diverse and emerging variant strains are prevalent. Viruses were isolated from 17 HIV positive individuals from three cities in Cameroon. Ghost cell lines expressing either CCR5 or CXCR4 with CD4 or CD4 alone (NIH AIDS Reagent Program) were used to determine co-receptor usage. HIV replication was determined by measuring p24 antigen levels. Plasma viral load (VL) was determined using the Versant bDNA assay. Nucleotide sequencing was performed on the V3 region and sequences were edited, aligned and translated into amino acids as described in the algorithm. Bio-informatics tools based on the 11/25 and charge rule were used to predict co-receptor usage. The majority of patient isolates in our study were CRF02_AG or CRF02_AG containing recombinants. Tropism of these complex viruses based on the cell culture assay was determined to be R5 in 15/17 (88.2%) patients. However, two patient isolates were dual tropic R5X4 and had drug-specific mutations. Of these two patients, one was on antiretroviral treatment with a VL of 20,899 copies/ml and the other was drug-naïve with 141,198 copies/ml. Genotype based prediction was overall in good agreement with phenotype for R5 viruses, where 93% (14/15) of results were comparable, dual tropic viruses being reported as X4 viruses by prediction. Our results indicate that most HIV strains in Cameroon were R5 tropic and some harbored drug-resistant mutations. V3 sequence based prediction compared well with cell based assays for R5 strains and may be useful even in settings where highly diverse strains are prevalent.
BioMed Central
Page 1 of 4
(page number not for citation purposes)
AIDS Research and Therapy
Open Access
Short report
Comparative analysis of cell culture and prediction algorithms for
phenotyping of genetically diverse HIV-1 strains from Cameroon
Viswanath Ragupathy
1
, Jiangqin Zhao
1
, Xue Wang
1
, Owen Wood
1
,
Sherwin Lee
1
, Sherri Burda
2
, Phillipe Nyambi
2
and Indira Hewlett*
1
Address:
1
Laboratory of Molecular Virology, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, MD 20892,
USA and
2
Department of Pathology, NYU School of Medicine, 550 First Avenue, Medical Sciences Building, 5th Floor, New York, NY 10016 423,
USA
Email: Viswanath Ragupathy - viswanath.ragupathy@fda.hhs.gov; Jiangqin Zhao - Jiangqin.Zhao@fda.hhs.gov;
Xue Wang - xue.wang@fda.hhs.gov; Owen Wood - owen.wood@fda.hhs.gov; Sherwin Lee - sherwin.lee@fda.hhs.gov;
Sherri Burda - Sherri.Burda@nyumc.org; Phillipe Nyambi - Phillipe.Nyambi@nyumc.org; Indira Hewlett* - indira.hewlett@fda.hhs.gov
* Corresponding author
Abstract
Background: With the advent of entry inhibitors, monitoring of viral tropism in the clinical setting
is important. Conventional methods are cell-based and lengthy, therefore V3 sequence based
prediction algorithms are becoming increasingly attractive as monitoring tools. Here we report a
comparative analysis of viral tropism of strains circulating in Cameroon where diverse and
emerging variant strains are prevalent.
Methods: Viruses were isolated from 17 HIV positive individuals from three cities in Cameroon.
Ghost cell lines expressing either CCR5 or CXCR4 with CD4 or CD4 alone (NIH AIDS Reagent
Program) were used to determine co-receptor usage. HIV replication was determined by
measuring p24 antigen levels. Plasma viral load (VL) was determined using the Versant bDNA assay.
Nucleotide sequencing was performed on the V3 region and sequences were edited, aligned and
translated into amino acids as described in the algorithm. Bio-informatics tools based on the 11/25
and charge rule were used to predict co-receptor usage.
Results: The majority of patient isolates in our study were CRF02_AG or CRF02_AG containing
recombinants. Tropism of these complex viruses based on the cell culture assay was determined
to be R5 in 15/17 (88.2%) patients. However, two patient isolates were dual tropic R5X4 and had
drug-specific mutations. Of these two patients, one was on antiretroviral treatment with a VL of
20,899 copies/ml and the other was drug-naïve with 141,198 copies/ml. Genotype based prediction
was overall in good agreement with phenotype for R5 viruses, where 93% (14/15) of results were
comparable, dual tropic viruses being reported as X4 viruses by prediction.
Conclusion: Our results indicate that most HIV strains in Cameroon were R5 tropic and some
harbored drug-resistant mutations. V3 sequence based prediction compared well with cell based
assays for R5 strains and may be useful even in settings where highly diverse strains are prevalent.
Published: 25 November 2009
AIDS Research and Therapy 2009, 6:27 doi:10.1186/1742-6405-6-27
Received: 28 July 2009
Accepted: 25 November 2009
This article is available from: http://www.aidsrestherapy.com/content/6/1/27
© 2009 Ragupathy et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0
),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
AIDS Research and Therapy 2009, 6:27 http://www.aidsrestherapy.com/content/6/1/27
Page 2 of 4
(page number not for citation purposes)
Findings
Human Immunodeficiency virus type 1 (HIV-1) enters the
cell by a multistep process that involves CD4 binding and
the use of co-receptors CCR5 or CXCR4. Co-receptor
usage in many cases correlates with disease pathogenesis
and progression [1,2]. Furthermore, changes in viral tro-
pism occurs in many HIV positive individuals over time,
indicated by a shift in co-receptor use from CCR5 to
CXCR4 which has been shown to generally correlate with
increased disease progression [3]. Some viruses are capa-
ble of using both co-receptors and are termed dual tropic
or R5X4 viruses. In the era of antiretroviral therapeutics,
co-receptor antagonists are now in use for treatment of
HIV infected individuals [4], and it therefore becomes
necessary to identify strains circulating in a given popula-
tion or region on the basis of their tropism. This should be
helpful to clinicians by providing additional information
for better management of disease.
Currently there are two methods in practice for co-recep-
tor determination a) bio informatics tools based on V3
sequence to predict co-receptor use and b) transfected cell
culture based methods. The latter method is widely used
in many clinical settings but is labor intensive and time
consuming. Prediction of co-receptor usage based on V3
sequence data on plasma viral RNA may be a useful alter-
native tool to assist clinicians in situations where virus
culture based phenotyping methods that rely on isolation
of peripheral blood mononuclear cells (PBMC) from
patient specimens may not be practical while also being
labor-intensive and time consuming.
The present investigation was aimed at characterizing
genetically diverse HIV-1 strains circulating in Cameroon
in terms of co-receptor usage and comparing cell culture
based methods with V3 sequence based prediction algo-
rithms for virus phenotyping and co-receptor usage of
complex, emerging HIV strains.
Virus isolates (n = 17) were obtained from patients attend-
ing clinics in three cities in Cameroon - Bamenda, Limbe
and Buea. Demographic information was collected in the
Performa and analyzed. Viruses were propagated in PBMC
derived from buffy coats and cell free viruses stored in liq-
uid nitrogen for subsequent analysis. Ghost cell lines
(Human osteosarcoma cells) expressing CCR5, CXCR4
with CD4 or CD4 alone (received from NIH AIDS Reagent
Program) were used to determine co-receptor use. Briefly,
cells were seeded at a concentration of 10e5 cells/well in a
24 well plate. After 24 hours, cells were infected with 5 ng
of p24 antigen of different HIV strains, incubated at 37°C
with 5% CO2 for 2 hours, washed thoroughly and cul-
tured in MEM media with10% FBS and antibiotics.
Appropriate controls included uninfected cells, and cells
treated with co-receptor antagonists TAK 779 (9.14 μmol/
ml) and AMD 3100 (100 ng/ml) to block CCR5 and
CXCR4. Culture supernatants were harvested at days 4
and 8 and HIV replication was determined by measuring
p24 antigen levels using the Perkin Elmer kit (Cat No:
NEK050B).
Viral RNA was isolated using the QIAGEN (Cat No:
52906) viral RNA extraction kit The V3 region was ampli-
fied by nested PCR and sequenced using primers ED31 5'-
CCTCAGCCATTACACAGGCCTGTCCAAAG and ES8 5'-
CACTTCTCCAATTGTCCCTCA and sequences were
edited, aligned in Clustal program and translated into
amino acids as described in the algorithm. Plasma VL was
determined using the Versant HIV RNA 3.0 Assay (bDNA;
Siemens, IL) for 15 of the 17 samples studied.
In our analysis, we used 11/25 and charge rule bio-infor-
matics tool to predict co-receptor usage [5]. The positively
charged amino acids at 11/25 and net charge >5 predict
CXCR4 in the V3 loop aligned against a consensus
sequence although other positions may also be impor-
tant.
The 17 viruses studied were obtained from nine males and
eight females whose ages ranged from 37-54 Yrs for men
and 20-60 Yrs for women. Because 9/17 (53%) of the
patients were on anti-retroviral therapy with Triamune
(3TC/d4T/NVP) or a combination of lamivudine(3TC)/
Stavudine(d4T) or lamivudine(3TC)/Nevirapine(NVP),
we examined the pol sequences of their viral isolates for
evidence of drug specific resistance mutations. Genotyp-
ing revealed that 5/9 (55.5%) had drug specific muta-
tions. Of these 5 patients, 2 had drug resistance for all
classes of RT antiretroviral drugs NRTI (A62V, K65R, T69I,
V75I, F77L, F116Y, Q151M, M184V/I) and NNRTI (V90I,
V108I, Y181C, Y188L, M184I) and 3 had one or more
drug specific mutations, K103N, Y181C and G190A in the
RT region. HIV disease progression is generally associated
with high VL and X4 phenotype, therefore we determined
plasma VL for both treatment-experienced and drug naïve
patients. Those who received ARV therapy had a VL range
of <75-28987 copies/ml while drug naive patients ranged
from <75-141,198 copies/ml.
Since multiple, diverse strains are responsible for the HIV
epidemic in Cameroon, we analyzed viral sequences of
the strains to identify genetic subtype. Phylogenetic anal-
ysis of partial sequences of gp41, p17 and pol region
revealed that 53% belonged to CRF02_AG, 6% subtype F2
and 41% were Unique Recombinant Forms (URFs) (Table
1). It is interesting to note that majority of viruses were
recombinants of CRF02_AG with gene segments of other
HIV CRFs and subtypes suggesting that newly emerging
HIV strains in Cameroon may be recombinants of the pre-
dominant CRF strain with other lesser CRF variants con-
AIDS Research and Therapy 2009, 6:27 http://www.aidsrestherapy.com/content/6/1/27
Page 3 of 4
(page number not for citation purposes)
tributing to the evolving diversity of HIV in this region.
The tropism of these complex viruses based on cell culture
experiments was determined to be R5 for 15/17 (88.2%)
strains (Table 1). A possible explanation for the predomi-
nance of R5 tropism observed with the CRF02_AG viruses
we studied may be attributed to the subtype A of the env
gene segment in these strains and it has been reported ear-
lier [6] that subtype A strains generally show CCR5 tro-
pism. In our study virus isolates from two patients
(06CMLPH02MG; 60CMBDSH05) were found to be dual
tropic R5X4 and their genotypes were assigned as URFs.
Thus, most representative viruses circulating in the
regions we studied were classified as R5 viruses. As entry
inhibitors for CCR5 and CXCR4 co-receptors are currently
being used for HIV treatment, fast and reliable methods
for determination of viral tropism will be of value to clini-
cians. Many previous studies have shown that V3
sequence based prediction algorithms can be compara-
tively rapid and reliable for population studies [7-10].
However, the ability of such tools to accurately predict co-
receptor usage of viruses in a population that harbors
genetically diverse HIV strains needs evaluation in order
to determine their appropriateness and suitability for
determination of viral tropism.
In our study, V3 sequence based prediction was found to
be in good agreement with phenotype, where 93% (14/
15) of results were comparable (Table 1). One patient iso-
late (07CMLPH128) was an R5 virus but genotype predic-
tion scored it as X4 tropic because of mutations in the V3
region and a net charge of 8. For these field isolates the
combined 11/25 and charge rule based prediction was
found to be most appropriate when compared with other
methods (data not shown). Similar observations were
reported in a study with the combined use of the 11/25
and net charge rules [5]. Furthermore, eight web based
prediction algorithms when used individually to deter-
mine viral tropism, had a low sensitivity and specificity
for non-B subtypes. However, for clade B viruses, only
PSSM
X4R5
and geno2pheno yielded good sensitivity and
specificity when compared with other algorithms [11].
Our study findings suggest that direct V3 sequencing may
provide an alternative to phenotypic assays for assessing
HIV-1 tropism. As reported earlier [12] dual tropic viruses
could be under estimated and we observed similar find-
ings in our data set. Our findings are in good agreement
with earlier observations for R5 viruses that prediction
methods based on V3 sequencing are comparable with the
phenotypic method suggesting their potential applicabil-
ity in clinical settings for the future.
Another observation in our limited study is that among
those who received antiretroviral therapy in our study
population, two individuals (ID = 06CMBDSH05;
06CMARC007) had resistance to all classes of RT inhibi-
tors with the first strain showing R5X4 and the second, R5
tropism. Viruses from both individuals were classified as
URFs having CRF02_AG with CRF11 for 06CMBDSH05
and a small fragment of subtype B in the gag region for
06CMARC007 (full length, unpublished data). A third
individual (ID 06CMLPH02MG) had never received anti-
viral therapy but his viral phenotype was R5X4 with VL
being 141,198 copies/ml suggesting potential advance-
ment of disease accompanied by X4 tropism and drug
resistance. Interestingly the virus was a URF composed of
the predominant strain in Cameroon, CRF02_AG with a
minor CRF37, suggesting that complex recombinants
emerging in this region could likely be dual tropic viruses.
Larger studies and data sets are needed from these regions
to further substantiate these findings and to understand
the complexity of the emerging diversity of HIV in this
region of high diversity.
Finally, although higher frequencies of drug resistance in
patients harboring X4 viruses have been reported recently
[13], in our study of the 5 individuals on therapy, we iden-
tified drug resistant mutations in four with R5 tropic and
one with X4 tropic viruses. Studies with larger sample
sizes are required to determine the impact of drug resist-
ance on tropism and whether emergence of new recom-
binants, drug resistance and viral tropism play a major
role in the spread and diversification of HIV strains in
Cameroon. In conclusion, our results, based on a limited
number of specimens and viruses isolated from PBMC
sampled for phenotypic and genotypic studies indicate
that even in a geographic region where highly complex
Table 1: Comparison of genotypic prediction vs phenotype
Sample ID Genotype Prediction Ghost cells Assay
06CMARC007 URF CCR5 CCR5
06CMARC009 CRF02 -- --
06CMARC036 CRF02 -- --
06CMARC058 CRF02 -- --
06CMLPH01OJ URF -- --
06CMLPH03VJ CRF02 -- --
06CMLPH11TT URF -- --
06CMLPH016SL CRF02 -- --
06CMLPH17HT CRF02 -- --
06CMLPH19CM URF -- --
06CMLPH20SL CRF02 -- --
06CMBDHS019 F2 -- --
06CMBDHS024 URF -- --
06CMBDHS064 CRF02 -- --
07CMLPH128 CRF02 CXCR4 CCR5
06CMLPH02MG URF CXCR4 CXCR4/CCR5
06CMBDSH05 URF CXCR4 CXCR4/CCR5
Columns indicate sample ID's and their corresponding HIV genotype
from partial sequences (gp41/p17/pol), its tropism by prediction and
cell culture assay.
Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral
AIDS Research and Therapy 2009, 6:27 http://www.aidsrestherapy.com/content/6/1/27
Page 4 of 4
(page number not for citation purposes)
viruses circulate, the V3 prediction algorithm compared
favorably with cell culture for R5 viruses that were pre-
dominant in this region. These findings suggest that V3
sequence based co-receptor prediction may potentially be
an alternate tool to cell culture assays for phenotypic char-
acterization of emerging, new and diverse HIV strains in a
population where genetic diversity is high and continuing
to evolve.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
VR have made substantial contributions to conception,
design, analysis and interpretation of data, JZ helped with
sequence data analysis, XW with tissue culture work, OW
& SL assisted with virus culture from stocks, SB & PN pro-
vided the viruses for the work, PN and IH have been
involved in drafting the manuscript or revising it critically
for publication, IH provided supervision for the project.
Acknowledgements
The authors wish to acknowledge Drs. Krishna Devadas, Ming Jie Zhang
and Hira Nakhasi for review of the manuscript. This work was supported
by an Inter Agency Agreement with the National, Heart, Lung and Blood
Institute, IAA-NHLBI, BY1-HB-5026-01. We wish to acknowledge The NIH
AIDS Research and Reference Reagent Program for providing us cells and
chemokine blockers. The findings and conclusions in this article have not
been formally disseminated by the Food and Drug Administration and
should not be construed to represent any agency determination or policy.
References
1. Dragic T, Litwin V, Allaway GP, Martin SR, Huang Y, Nagashima KA,
Cayanan C, Maddon PJ, Koup RA, Moore JP, Paxton WA: HIV-1
entry into CD4+ cells is mediated by the chemokine recep-
tor, CC-CXR-5. Nature 1996, 381:667-673.
2. Dittmar MT, McKnight A, Simmons G, Clapham PR, Weiss RA, Sim-
monds P: HIV-1 tropism and co-receptor use. Nature 1997,
385(6616):495-6.
3. Delobel P, Sandres-Sauné K, Cazabat M, Pasquier C, Marchou B, Mas-
sip P, Izopet J: R5 to X4 switch of the predominant HIV-1 pop-
ulation in cellular reservoirs during effective highly active
antiretroviral therapy. J Acquir Immune Defic Syndr 2005,
38(4):382-92.
4. Dau B, Novel Holodniy M: Targets for antiretroviral therapy:
clinical progress to date. Drugs 2009, 69(1):31-50.
5. Raymond S, Delobel P, Mavigner M, Cazabat M, Souyris C, Sandres-
Sauné K, Cuzin L, Marchou B, Massip P, Izopet J: Correlation
between genotypic predictions based on V3 sequences and
phenotypic determination of HIV-1 tropism. AIDS 2008,
22(14):F11-F16.
6. Tscherning-Casper C, Vödrös D, Menu E, Aperia K, Fredriksson R,
Dolcini G, Chaouat G, Barré-Sinoussi F, Albert J, Fenyö EM: Core-
ceptor 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 2000, 24(1):1-9.
7. Sierra S, Kaiser R, Thielen A, Lengauer T: Genotypic coreceptor
analysis. Eur J Med Res 2007, 12(9):453-62.
8. Lengauer T, Sander O, Sierra S, Thielen A, Kaiser R: Bioinformatics
prediction of HIV coreceptor usage. Nature Biotechnology 2007,
25(12):1407-10.
9. Boisvert S, Marchand M, Laviolette F, Corbeil J: HIV-1 coreceptor
usage prediction without multiple alignments: an application
of string kernels. Retrovirology 2008, 5:110.
10. de Mendoza C, Van Baelen K, Poveda E, Rondelez E, Zahonero N,
Stuyver L, Garrido C, Villacian J, Soriano V, Spanish HIV Serocon-
verter Study Group: Performance of a population-based HIV-1
tropism phenotypic assay and correlation with V3 genotypic
prediction tools in recent HIV-1 seroconverters. J Acquir
Immune Defic Syndr 2008, 48(3):241-4.
11. Garrido C, Roulet V, Chueca N, Poveda E, Aguilera A, Skrabal K,
Zahonero N, Carlos S, García F, Faudon JL, Soriano V, de Mendoza
C: Evaluation of eight different bioinformatics tools to pre-
dict viral tropism in different human immunodeficiency virus
type 1 subtypes. J Clin Microbiol 2008, 46(3):887-91.
12. Mefford ME, Gorry PR, Kunstman K, Wolinsky SM, Gabuzda D: Bio-
informatic Prediction Programs Underestimate the Fre-
quency of CXCR4 Usage by R5X4 HIV Type 1 in Brain and
Other Tissues. AIDS Research and Human Retroviruses 2008,
24(9):1215-1220.
13. Wagner TA, Frenkel LM: Potential limitation of CCR5 antago-
nists: drug resistance more often linked to CXCR4-utilizing
than to CCR5-utilizing HIV-1. AIDS 2008, 22(17):2393-5.
    • "Isolation of viral RNA from plasma specimens was performed by using Virus QIAamp Viral RNA Mini Kit (QIAGEN) according to manufacturers' instructions. cDNA was obtained by using Superscript III First-strand synthesis System for RT-PCR (Invitrogen 18080-051) in a 20 ml volume with in-house primers that have been previously published [26]. Amplification was performed by a nested polymerase chain reaction (nested-PCR) with ready-to-use Master Mix from Promega (Promega, Madison, WI, USA cat M7505). "
    [Show abstract] [Hide abstract] ABSTRACT: Background: The use of CCR5 antagonists involves determination of HIV-1 tropism prior to initiation of treatment. HIV-1 tropism can be assessed either by phenotypic or genotypic methods. Genotypic methods are extensively used for tropism prediction. However, their validation in predicting tropism of viral isolates belonging to group M non-B subtypes remains challenging. In Cameroon, the genetic diversity of HIV-1 strains is the broadest reported worldwide. To facilitate the integration of CCR5 antagonists into clinical practice in this region, there is a need to evaluate the performance of genotypic methods for predicting tropism of highly diverse group M HIV-1 strains.
    Full-text · Article · Nov 2014
    • "Therefore, necessary precautions are required to protect the eye from these opportunistic organisms. These microorganisms and their pathogenic effects might be different from country to country, particularly in the developing countries (Jabs et al., 1995; Ragupathy et al., 2009). Therefore, the timely identification of the microorganisms found in contact lenses of Saudi wearers is of paramount importance. "
    [Show abstract] [Hide abstract] ABSTRACT: The aim of this study was to investigate the prevalence and type of microbial contamination, associated with contact lenses and lens care accessories used by a group of contact lens wearers. Results show that a total of 178 strains were isolated, including, 100 Gram positive and 78 Gram negative bacteria. Staphylococcus epidermidis and Pseudomonas aeruginosa were the most common microorganisms isolated (25.281 and 13.483% respectively. 24 P. aeruginosa were isolated from lenses and eyes of contact lens and non-contact lens wearers. All isolates were susceptible to the tested aminoglycosides and fluoroquinolones. Aminoglycosides and fluoroquinolones (ciprofloxacin) were more efficient than β-lactams. 91.67% of the strains had intermediate-resistant to cefotaximee; 4.17% were resistant and 4.17% were sensitive to Cefotaximee. 95.83% of the strains were sensitive, while 4.17% were resistant to Imipenem. 20.83 and 12.5% of P. aeruginosa strains were resistant and sensitive to Ceftriaxone respectively, while 66.67% were moderately-resistant to ceftriaxone of the β-lactam class. All isolates were sensitive to the other tested β-lactam.
    Full-text · Article · Jan 2013 · PLoS ONE
  • [Show abstract] [Hide abstract] ABSTRACT: Induction motor is a multivariable, complex, and nonlinear controlled plant In the field-oriented control system, there may be a wide parameter variation problem. For the reason, this paper presents a speed controller based on fuzzy logic. The fuzzy speed controller adopts a fuzzy control rule table which can have a strong adjustability to the error. The algorithm for fuzzy speed controller can ensure to implement the control in real time. The results of simulation and experiment verify that the fuzzy speed controller is valid and feasible.
    Article · Jan 2004 · PLoS ONE
Show more