First Identification and Molecular Characterization of Lymphotropic Polyomavirus in Peripheral Blood from Patients with Leukoencephalopathies

Lymphotropic polyomavirus (LPV) was first isolated in 1979 from a B-lymphoblastoid cell line of an African green monkey by zur Hausen and Gissman (10). This virus has some characteristics common to human polyomavirus, such as virion morphology, the presence of a closed circular double-stranded molecule of DNA, and in vitro transforming activity (7), but it is antigenically distinct from simian virus 40, BK virus, and JC virus (1). In cell cultures, LPV has a highly restricted host range for both human and monkey B lymphoblasts (2, 3), and seroepidemiological studies revealed that, in addition to sera from monkeys, many human sera had strong reactions in the presence of LPV antigens (8).

More recently, Lednicky and colleagues were able to detect a fragment of the LPV genome in the blood of an immunocompromised rhesus monkey (5), but neither LPV nor any other related virus has been isolated from human biological fluids or tissues.

In order to investigate the possible presence of LPV in human specimens, we examined DNA extracted from peripheral blood mononuclear cells (PBMCs) that were collected from Italian human immunodeficiency virus-positive (HIV⁺) patients who were affected with different forms of leukoencephalopathies or other neurological diseases (OND), as well as HIV⁺ subjects with no neurological disorders (NND). We also collected cells from healthy control (HC) subjects. All the individuals gave written consent for enrollment in the study.

In particular, the LPV genome was searched by means of nested PCR targeting the viral transcriptional control region (TCR) (5) in DNA from 10 patients with progressive multifocal leukoencephalopathy (PML), 10 with not determined JC virus-negative leukoencephalopathy (NDLE), 10 with OND, 10 with NND, and 10 HC subjects. Plasmid pLPV-K38, containing a full-length genome of strain LPV-K38 (6), was used as a positive-control template for the amplification assay.

The LPV control region was successfully amplified from three DNA extracts: one from PBMCs collected from a PML patient and two from PBMCs collected from two NDLE patients. However, the bands created by the inner PCR were different in size (520 bp and 270 bp). The amplified fragments, together with the positive control, were subjected to direct automated sequencing according to published procedures (4), and the obtained nucleotide sequences were aligned with the GenBank database. DNA isolation, PCR, and sequencing were repeated twice. Cross-referencing with the database showed the expected complete nucleotide identity (100%) of LPV (GenBank accession number K02562.1) to the LPV-K38 control, whereas lower nucleotide identities of the LPV genome to the amplified TCR from the PBMCs of the patients were shown.

In particular, the molecular organization of the LPV TCR amplified from the PML patient was rearranged, as well as the LPV-K38 control, but five nucleotide (nt) changes were detected between the two viral strains (nt 110, 222, 335, 394, and 408; numbering is relative to strain LPV-K38). Two archetypal LPV regulatory regions were identified in the PBMCs from the two NDLE patients. To rule out the hypothesis that a cross-contamination between the samples could have happened during the nested-PCR procedures, the two archetypal sequences were aligned and differences in six places were discovered between the two strains (nt 126, 129, 133, 201, 269, and 555; numbering is relative to strain LPV-K38). As shown in Fig. 1, the ori region and the enhancer 63-bp subgenomic element of the rearranged sequence found in the PML patient were duplicated in comparison to the archetypal sequence that was amplified from the two NDLE patients.

FIG. 1.

Schematic representation of the amplified TCR sequences. (A) Comparison between the rearranged TCR form amplified from the PML patient (pm1) and the one amplified from the plasmid of LPV-K38. (B) Comparison between the two TCR archetypal forms amplified from the two NDLE patients. The LPV PCR primer binding sites are indicated by LPR1 and LPR2. The arrows indicate the point mutation positions. The T-antigen recognition pentanucleotides are boxed with a solid line, while the enhancer 63-bp subgenomic elements are boxed with a dashed line. The ori region is underlined with a solid line.

In order to confirm these findings, another PCR targeting the VP1 region was performed. Specific primers LPV1 (nt 1765 to 1784) and LPV2 (nt 1966 to 1985) were used, and a 221-bp VP1 fragment from the PML patient and one of the NDLE patients was successfully amplified. A lack of material prevented us from performing the PCR with the PBMCs from the third patient (data not shown).

Even if the investigated human population was limited, it is noteworthy that the LPV genome was found exclusively in the peripheral blood of patients with the two different forms of leukoencephalopathies but not in the OND, NND, or HC cases. Moreover, two different molecular organizations of the viral TCR were detected, and based on what is known about human polyomaviruses (9), we can hypothesize that the archetypal organization is probably the form naturally transmitted (5), whereas rearranged forms, derived from partial duplication of the archetypal nucleotide sequence, could be generated as a consequence of highly active viral replication.

To the best of our knowledge, this is the first direct detection of the LPV genome in human peripheral blood, and this validates current seroepidemiological evidence suggesting the presence of LPV-like infections in human subjects. Due to the small number of studied subjects, we cannot draw a clear conclusion, but the data suggest at least the possible involvement of LPV in central nervous system leukoencephalopathies occurring in HIV⁺ individuals.

The isolation of LPV and/or whole-genome sequencing of the virus will be necessary to additionally support the significance of our results; however, the presence of the viral DNA in human blood cells is an important finding and warrants further investigations that may correlate LPV infection with specific human diseases.

Nucleotide sequence accession numbers.

The GenBank accession numbers for the two archetypal sequences and the rearranged sequences determined in this study are EU559168, EU559169, and EU559170, respectively.