Drastic fitness loss in human immunodeficiency virus type 1 upon serial bottleneck events

Abstract

Muller's ratchet predicts fitness losses in small populations of asexual organisms because of the irreversible accumulation of deleterious mutations and genetic drift. This effect should be enhanced if population bottlenecks intervene and fixation of mutations is not compensated by recombination. To study whether Muller's ratchet could operate in a retrovirus, 10 biological clones were derived from a human immunodeficiency virus type 1 (HIV-1) field isolate by MT-4 plaque assay. Each clone was subjected to 15 plaque-to-plaque passages. Surprisingly, genetic deterioration of viral clones was very drastic, and only 4 of the 10 initial clones were able to produce viable progeny after the serial plaque transfers. Two of the initial clones stopped forming plaques at passage 7, two others stopped at passage 13, and only four of the remaining six clones yielded infectious virus. Of these four, three displayed important fitness losses. Thus, despite virions carrying two copies of genomic RNA and the system displaying frequent recombination, HIV-1 manifested a drastic fitness loss as a result of an accentuation of Muller's ratchet effect.

FIG. 1

HTA as a tool for quantification of mixed viral populations. (A) Quantification of two viruses present in different proportions. Mixed infections with clones J1 and A1 were carried out with decreasing (from 100 to 0%) and increasing (from 0 to 100%) proportions of the two viruses. Proviral DNA obtained from the infections was subjected to HTA with, as a labeled probe, cDNA of the V1-V2 region of the env gene, as described in Materials and Methods. (B) The same HTA analysis with DNAs of all initial clones, A1 to J1, using cDNA of clone J1 as a probe. All clones formed heteroduplexes (HT), whereas J1 formed a homoduplex (HO). The H1 population was probably a mixture of two clones. (C) Stability of the HTA pattern of each initial clone after five passages in cell culture. HTA patterns of amplified proviral DNAs from the first (lanes 1) and fifth (lanes 5) passages are shown.

FIG. 2

Schematic representation of the derivation of HIV-1 clones and of serial plaque-to-plaque transfers. Procedures for isolation of virus from individual plaques of the HIV-1 isolate s61 on MT-4 cells are detailed in Materials and Methods. Viruses from randomly chosen individual plaques were diluted and plated again (circles and arrows). Filled boxes indicate the viral populations from which infectious virus could not be rescued upon subsequent plating or infection in liquid medium in different cell lines (as described in Materials and Methods). The 15 intended serial plaque transfers could be completed only for clones B1, D1, F1, I1, J1, and K1, although infectious virus could be rescued only from populations D15, F15, I15, and K15.

FIG. 3

Fitness vector determination for HIV-1 clone D1. The fitness value was obtained from competition passages between clones D1 and J1 as described in Materials and Methods. A total of five competition passages were carried out with, as starting viruses, mixtures of D1 and J1 at ratios of 1:9, 1:1, and 9:1. (A) Result of the HTA assay. HT (heteroduplex) represents the proportion of D1 virus; HO (homoduplex) represents the proportion of J1. (B) From the 1:1 initial mixture, the proportions of D1 and J1 DNAs (measured and quantified by densitometry of the autoradiogram) (Rn) were obtained for the five competition passages. This proportion was compared to the ratio of the two viruses in the initial mixtures (Ro). The value in each passage (solid squares) was used to derive the fitness vector for D1 (24). Details of all procedures involved are given in Materials and Methods.

FIG. 4

Fitness vectors and corresponding fitness values of the initial clones. (A) Fitness vectors of all of the initial clones, determined as indicated in Materials and Methods and in the legend to Fig. 3. In all lineages, the 1:1 competition cultures were used in the calculation of the vectors, except for clones I1 and H1. In these viruses the vectors were drawn with 1:9 competition cultures because of the very high fitness value of clone I1, which completely suppressed the J1 population, and because of the presence of two populations in clone H1 (Fig. 1B). Vectors were drawn with data from at least three competing passages. (B) Slopes of fitness vectors (fitness values) for the initial HIV-1 clones. Clone J1 was used as a reference, and consequently its slope is zero.

FIG. 5

Fitness alterations produced in the viral populations after 15 serial plaque-to-plaque transfers. (A) HTA patterns obtained from the competitions at the 1:9, 1:1, and 9:1 proportions with cDNAs from the initial clone K1 and the final clone K15, using J1 cDNA as a probe. The arrow points to a new variant genome with a different HTA mobility arising at passage 5 of the competition series carried out at a 1:9 ratio of K1 and J1 in competition culture. (B) Comparison of the vectors in lineages for D1 to D15, F1 to F15, I1 to I15, and K1 to K15. The continuous lines represent initial fitness vectors, and the dashed lines represent final fitness vectors. For I1 to I15, a different scale has been used due to the wider range of values plotted. All of the vectors corresponding to final populations (transfers 15) are below the corresponding values for the initial clones, except for clone F1 (see text). Procedures are described in Materials and Methods, and fitness values are summarized in Table 1.