TY - JOUR
T1 - Viral genetic variation accounts for a third of variability in HIV-1 set-point viral load in Europe
AU - BEEHIVE Collaboration
AU - Blanquart, Francois
AU - Wymant, Chris
AU - Cornelissen, Marion
AU - Gall, Astrid
AU - Bakker, Margreet
AU - Bezemer, Daniela
AU - Hall, Matthew
AU - Hillebregt, Mariska
AU - Ong, Swee Hoe
AU - Albert, Jan
AU - Bannert, Norbert
AU - Fellay, Jacques
AU - Fransen, Katrien
AU - Gourlay, Annabelle J.
AU - Grabowski, M. Kate
AU - Gunsenheimer-Bartmeyer, Barbara
AU - Guenthard, Huldrych F.
AU - Kivela, Pia
AU - Kouyos, Roger
AU - Laeyendecker, Oliver
AU - Liitsola, Kirsi
AU - Meyer, Laurence
AU - Porter, Kholoud
AU - Ristola, Matti
AU - van Sighem, Ard
AU - Vanham, Guido
AU - Berkhout, Ben
AU - Kellam, Paul
AU - Reiss, Peter
AU - Fraser, Christophe
N1 - PPU
PY - 2017
Y1 - 2017
N2 - HIV-1 set-point viral load-the approximately stable value of viraemia in the first years of chronic infection-is a strong predictor of clinical outcome and is highly variable across infected individuals. To better understand HIV-1 pathogenesis and the evolution of the viral population, we must quantify the heritability of set-point viral load, which is the fraction of variation in this phenotype attributable to viral genetic variation. However, current estimates of heritability vary widely, from 6% to 59%. Here we used a dataset of 2,028 seroconverters infected between 1985 and 2013 from 5 European countries (Belgium, Switzerland, France, the Netherlands and the United Kingdom) and estimated the heritability of set-point viral load at 31% (CI 15%-43%). Specifically, heritability was measured using models of character evolution describing how viral load evolves on the phylogeny of whole-genome viral sequences. In contrast to previous studies, (i) we measured viral loads using standardized assays on a sample collected in a strict time window of 6 to 24 months after infection, from which the viral genome was also sequenced; (ii) we compared 2 models of character evolution, the classical "Brownian motion" model and another model ("Ornstein-Uhlenbeck") that includes stabilising selection on viral load; (iii) we controlled for covariates, including age and sex, which may inflate estimates of heritability; and (iv) we developed a goodness of fit test based on the correlation of viral loads in cherries of the phylogenetic tree, showing that both models of character evolution fit the data well. An overall heritability of 31% (CI 15%-43%) is consistent with other studies based on regression of viral load in donor-recipient pairs. Thus, about a third of variation in HIV-1 virulence is attributable to viral genetic variation.
AB - HIV-1 set-point viral load-the approximately stable value of viraemia in the first years of chronic infection-is a strong predictor of clinical outcome and is highly variable across infected individuals. To better understand HIV-1 pathogenesis and the evolution of the viral population, we must quantify the heritability of set-point viral load, which is the fraction of variation in this phenotype attributable to viral genetic variation. However, current estimates of heritability vary widely, from 6% to 59%. Here we used a dataset of 2,028 seroconverters infected between 1985 and 2013 from 5 European countries (Belgium, Switzerland, France, the Netherlands and the United Kingdom) and estimated the heritability of set-point viral load at 31% (CI 15%-43%). Specifically, heritability was measured using models of character evolution describing how viral load evolves on the phylogeny of whole-genome viral sequences. In contrast to previous studies, (i) we measured viral loads using standardized assays on a sample collected in a strict time window of 6 to 24 months after infection, from which the viral genome was also sequenced; (ii) we compared 2 models of character evolution, the classical "Brownian motion" model and another model ("Ornstein-Uhlenbeck") that includes stabilising selection on viral load; (iii) we controlled for covariates, including age and sex, which may inflate estimates of heritability; and (iv) we developed a goodness of fit test based on the correlation of viral loads in cherries of the phylogenetic tree, showing that both models of character evolution fit the data well. An overall heritability of 31% (CI 15%-43%) is consistent with other studies based on regression of viral load in donor-recipient pairs. Thus, about a third of variation in HIV-1 virulence is attributable to viral genetic variation.
KW - IMMUNODEFICIENCY-VIRUS TYPE-1
KW - PROGNOSTIC MARKERS
KW - RNA LEVELS
KW - INFECTION
KW - AFRICA
KW - TRANSMISSION
KW - VIRULENCE
KW - PLASMA
KW - SEROCONVERSION
KW - HERITABILITY
U2 - 10.1371/journal.pbio.2001855
DO - 10.1371/journal.pbio.2001855
M3 - A1: Web of Science-article
VL - 15
JO - PloS Biology
JF - PloS Biology
SN - 1545-7885
IS - 6
M1 - 2001855
ER -