Phylogenetic Analysis of Murine Leukemia Virus Sequences from Longitudinally Sampled Chronic Fatigue Syndrome Patients Suggests PCR Contamination Rather than Viral Evolution

Aris Katzourakis; Stéphane Hué; Paul Kellam; Greg J. Towers

doi:10.1128/JVI.00827-11

DOI: 10.1128/JVI.00827-11

Article Figures & Data

Figures

Tables
Additional Files

Open in new tab
Download powerpoint
Fig. 1.
Maximum likelihood phylogeny of XMRV, xenotropic MLV (MLV-X), polytropic MLV (PMLV; Pmv), and modified polytropic MLV (MPMLV; Mpmv) gag gene sequences (710 nt). The initial chronic fatigue syndrome patient-derived XMRV sequences from Lo et al. (17) are indicated in blue. The eight sequences taken from the first time point are represented by a single branch, colored green. XMRV sequences sampled 15 years later from the same patients (GenBank accession numbers HQ601957 to HQ601962) are colored in red, while other sequences that do not have corresponding second-time-point sequences from Lo et al. are colored blue. The tree is rooted with Moloney MLV (GenBank accession number AF033811). Bootstrap scores of >50% are indicated on the corresponding branches. The scale bar represents the number of nucleotide substitutions per site.

Table 1.

Comparison of maximum likelihood phylogenetic tree with hypotheses consistent with within-patient viral evolution^{^a}

Constraint	Log likelihood	Decrease in likelihood	Standard deviation	P value		Expected likelihood wt
Constraint	Log likelihood	Decrease in likelihood	Standard deviation	SH	AU	Expected likelihood wt
Best tree	−3,658.76	0.00	NA	NA	NA	0.99
All 5 MLV 2010	−3,721.22	62.46	14.92	<0.001	<0.0001	<0.0001
MLV001 2010	−3,689.93	31.17	9.62	0.001	<0.0001	<0.0001
MLV002 2010	−3,696.66	37.90	10.80	0.002	<0.0001	<0.0001
MLV003 2010	−3,679.73	20.96	9.31	0.022	0.002	0.0086
MLV005 2010	−3,687.62	28.86	10.15	0.004	0.0003	0.0004
MLV006 2010	−3,701.63	42.87	12.90	0.002	<0.0001	<0.0001

↵a The most stringent constraint involved all 5 MLV sequences from the second time point, while each of the 5-s time point sequences was also constrained individually to cluster with the CFS type 1 sequence from the first time point. The trees were compared using pairwise Shimodaira-Hasegawa (SH) tests, the approximately unbiased (AU) test, and expected likelihood weights. All of the constrained trees were significantly worse than the maximum likelihood tree. NA, not applicable.

Table 2.

Comparison of patient-derived MLV sequences and known MLV sequences within the mouse genome^{^a}

Sequence	GenBank accession no.	Sampling yr(s)	Closest relative in the C57BL/6J mouse genome (July 2007 assembly)
Sequence	GenBank accession no.	Sampling yr(s)	Chr.	Strand	Span (nt)	Start (nt)	End (nt)	% identity	Changes^{^b}	Gaps (length in nt)
BD22	HM630560	Mid-1990s	10	−	697	8269377	8270073	99.30	5/697	0
			10	−	696	50145707	50146402	99.30	5/696	0
			X	+	696	15052167	15052862	99.30	5/696	0
BD26	HM630561	Mid-1990s	6	−	339	73242571	73242909	98.30	6/339	0
			2	−	339	15949590	15949928	98.30	6/339	0
			10	−	339	4627251	4627589	98.30	6/339	0
BD28	HM630557	Mid-1990s	9	+	331	62288048	62288378	98.10	6/318	1 (21)
			5	+	331	23722164	23722494	98.10	6/318	1 (21)
			4	+	331	133716363	133716693	98.10	6/318	1 (21)
CFS type 1	HM630562	Mid-1990s	8	+	697	125689652	125690348	99.30	5/697	0
			11	+	697	102946013	102946709	99.30	5/697	0
			6	−	696	73242413	73243108	98.90	5/696	0
CFS type 2	HM630558	Mid-1990s	8	+	698	125689652	125690349	99.00	7/698	1 (1)
			11	+	698	102946013	102946710	99.00	7/698	1 (1)
			6	−	697	73242412	73243108	98.60	7/697	1 (1)
CFS type 3	HM630559	Mid-1990s	2	−	697	15949431	15950127	99.90	1/697	0
			13	+	697	21905315	21906011	99.80	2/697	0
			6	−	696	73242413	73243108	99.30	1/696	1 (1)
MLV001	HQ601957	2010	12	+	340	19250254	19250593	99.70	1/276	2 (63; 1)
			9	+	339	62288048	62288386	99.00	2/276	1 (63)
			5	+	339	23722164	23722502	99.00	3/276	1 (63)
MLV002	HQ601958	2010	4	−	339	107826090	107826428	99.80	1/339	0
			2	−	339	57074273	57074611	99.80	1/339	0
			15	−	339	76395902	76396240	99.80	1/339	0
MLV003	HQ601959	2010	6	−	339	73242571	73242909	99.20	3/339	0
			4	−	339	107826090	107826428	99.20	3/339	0
			2	−	339	57074273	57074611	99.20	3/339	0
MLV004	HQ601960	2010	6	−	339	73242571	73242909	98.60	5/339	0
			2	−	339	15949590	15949928	98.60	5/339	0
			10	−	339	4627251	4627589	98.60	5/339	0
MLV005	HQ601961	2010	4	−	339	107826090	107826428	99.50	2/339	0
			2	−	339	57074273	57074611	99.50	2/339	0
			15	−	339	76395902	76396240	99.50	2/339	0
MLV006	HQ601962	2010	4	−	339	107826090	107826428	100.00	0/339	0
			2	−	339	57074273	57074611	100.00	0/339	0
			15	−	339	76395902	76396240	100.00	0/339	0

↵a Due to the short lengths of the patient-derived sequences, several mouse sequences are equally similar. Thus, three sequences are shown. The span column refers to the total length of the best match in the murine chromosome (Chr.).
↵b Number of nucleotides different/total number of nucleotides.

Additional Files

Figures
Tables

Supplemental material

Files in this Data Supplement:

Supplemental file 1 - Table S1 (Previously published nonecotropic endogenous MLV gag sequences.)
Zipped MS Word document, 17K.