Molecular Evidence for Transmission of Human T-Lymphotropic Virus Type II Infection by a Human Bite

Abstract

Investigation of a human T-lymphotropic virus type II (HTLV-II) infection in a female Australian blood donor identified a human bite as the likely mode of transmission, confirmed by nucleotide sequencing of the proviral tax/rex from both donor and contact. We believe this to be the first report of the transmission of an HTLV by a human bite.

Human T-lymphotropic virus type II (HTLV-II) was first isolated in 1982 (8) and is closely related to HTLV-I (13). HTLV-II transmission occurs through the transfer of infected lymphocytes via breast-feeding, blood transfusion, sexual contact, and injecting drug (ID) use (7). Infection is common in ID users (IDU) in the United States and Europe (7) but not in Australia (5). HTLV-II occurs in two major molecular subtypes, subtype A (HTLV-IIa), principally occurring in IDU and some selected Amerindian tribes, and subtype B (HTLV-IIb), associated with endemic HTLV-II infection of a number of Amerindian tribes and also with IDU in the United States and Europe (7). Additional molecular variants have also been reported (6, 14).

The first Australian case of HTLV-II, detected during blood donor screening (3), was shown to be subtype B by nucleotide sequence analysis (4). Originally, no risk factor for HTLV-II infection could be determined. Subsequent clinical and laboratory investigation have suggested that the blood donor (subject D) acquired HTLV-II through a severe bite from her HTLV-II-infected son (subject S), who had probably acquired the infection through ID use in North America in 1984.

Subject D has been previously described (3, 4). Genomic DNA from a peripheral blood mononuclear cell lysate prepared for the previous study and collected in 1993 was used for PCR in this investigation. Serum samples from donations made by subject D prior to the commencement of routine donor testing for anti-HTLV-I were retrieved from −30°C storage. Serum and anti-coagulated whole blood were collected from subject S in July 1995. Genomic DNA was extracted from thawed whole blood with an extraction kit (QIA-Amp; Qiagen, Hilden, Germany). Subject D and a partner were interviewed regarding risk factors for HTLV infection. Information regarding the ID use of subject S was also obtained from subject D and the partner (subject S was not available).

Anti-HTLV-I testing was performed with a p21E-enhanced HTLV-I viral lysate enzyme immunoassay (EIA) (Cambridge Biotech, Worcester, Mass.) and an HTLV-I viral lysate EIA (Abbott Diagnostics, Chicago Ill.). These assays detected anti-HTLV-II on the basis of serological cross-reactivity with HTLV-I. HTLV-II antibodies were confirmed by Western blotting (HTLV blot versions 2.3 and 2.4, Genelabs Diagnostics, Singapore) with an HTLV-II type-specific recombinant envelope protein in addition to HTLV-I viral lysate and recombinant envelope proteins.

Nested PCR amplification of a 746-bp fragment of the proviral tax/rex gene was conducted with outer primers TR101 and HFL104 (15) and inner primers SK43 (9) and HFL104 (15). PCR on subjects D and S was carried out in separate experiments, on different days, and in different buildings to prevent cross-contamination, and all amplifications included “no DNA” controls, which gave appropriate results. The PCR products were precipitated and directly sequenced on an ABI 373A automated sequencer (model 1993; Perkin-Elmer, Foster City, Calif.) with the PCR primers HFL108 (15) and WB109 (5′GGAGCCTTCCTCACCAA 3′).

Derived sequences were aligned with reference sequences by using CLUSTALW software. Phylogenetic analysis was performed with TREECON software. Nucleotide database searches were performed by BLAST software.

A follow-up interview with subject D and her partner revealed no obvious risk factors for HTLV-II infection, and the partner was nonreactive by anti-HTLV-I EIA and Western blotting and PCR negative for HTLV-I and -II (data not shown). Subject D reported that she had suffered a severe bite from subject S (then 26 years old) during an epileptic fit in April 1992. The bite to the right index finger occurred during an attempt to secure the tongue of subject S and resulted in extensive bleeding. Blood was visible in the mouth of subject S prior to the bite. The interview determined that subject S had apparently engaged in ID use during a visit to the United States and Mexico in 1984.

Serum samples from blood donations by subject D made prior to the introduction of donor anti-HTLV-I screening were tested by EIA and Western blotting. A sample collected on 4 February 1992 was nonreactive by both EIAs and negative by Western blotting (Fig. 1, strip 31). A sample collected on 14 December 1992 was reactive by both EIA assays (Cambridge EIA sample-to-cutoff ratio [s/co], 4.0; Abbott EIA s/co, >2.8) and HTLV-II seropositive by Western blotting (Fig. 1, strip 32).

FIG. 1.

HTLV-I and -II Western blot (Genelabs HTLV blot version 2.4) analysis of serum samples from the female blood donor. Strip 30 is the sample collected at the anti-HTLV-seropositive donation in 1993, strip 31 is the archived sample from 4 February 1992, and strip 32 is the archived sample from 14 December 1992. Strips 33, 34, and 35 are the negative and HTLV-I- and HTLV-II-positive controls, respectively, from the kit; strip 36 is an in-house, low-level anti-HTLV-I control.

Serum from subject S, collected in July 1995, was reactive by the Cambridge EIA (s/co, 5.2) and HTLV-II seropositive by Western blotting (data not shown).

PCR amplified a 746-bp fragment from each subject (data not shown). The resulting DNA sequences from subject D (SBB1452) and subject S (H3293) were identical, and they differed by only a single base change (A→C at nucleotide 7910) from the prototype HTLV-IIb isolate NRA (L20734) (10). This substitution is predicted to produce a conservative M217L substitution in the tax protein (Fig. 2B). A BLAST search of nonredundant GenBank and EMBL databases identified 21 HTLV-II tax/rex sequences containing the region of interest, and none demonstrated the (A→C) change at nucleotide 7910. The close relationship of SBB1452 and H3293 to the HTLV-IIb reference sequences NRA and G12 is demonstrated by phylogenetic analysis (Fig. 2A).

FIG. 2.

(A) Phylogenetic analysis demonstrating clustering of SBB1452 and H3293 with HTLV-IIb reference sequences in a neighbor-joining tree based on a Kimura distance matrix with 100 bootstrap resamplings. Bootstrap values over 70 are shown at the relevant node. The tree is outgroup rooted on PP1664. The reference sequences are as follows: HTLV-I strains CH (GenBank accession no. M69044), ATK (J02029), HS35 (D00294), MOMS (X83118), EL (M67514), and Mel5 (L02534); HTLV-II strains CG (M63881), Mo (M10060), KAY1 (U32874), NRA (L20734), and G12 (L11456); and simian T-lymphotropic virus strains PtM3 (M11373), PHSu1 (X83120), and PP1664 (Z46344). (B) Peptide alignment of part of the tax reading frame of HTLV-IIb reference strain NRA, HTLV-I reference strain ATK (accession no. J02029), and SBB1452. Codon 217 is highlighted.

We conclude that HTLV-II was transmitted from subject S to subject D by a bite. This contention is supported by both the seroconversion of subject D over a period contemporaneous with the bite and the observation that the rarity of HTLV-II infection in Australia, combined with the shared single-nucleotide change in the highly conserved tax/rex gene, makes it extremely unlikely that these family members acquired their infections independently. The demonstrated seroconversion of subject D precludes mother-to-child transmission, such as breast-feeding.

It is probable that blood in the mouth of subject D prior to the bite provided the source of the infected lymphocytes, although HTLV-I proviral sequences can be detected in lymphocytes from mouthwash samples from HTLV-I infected individuals without oral lesions (1). In all probability, subject S acquired the HTLV-IIb infection in 1984 North America, where HTLV-II has been endemic in IDU since the 1970s (2) and HTLV-IIb occurs in IDU and blood donors (7, 11).

The peptide substitution seen in the tax protein is unlikely to be of functional significance, as leucine is found at codon 217 in strains of HTLV-I (Fig. 2B) and this region of the protein is not critical for tax activation of the CREB-ATF or NF-κB-Rel pathways (12).

We believe this to be the first reported transmission of a human retrovirus, other than human immunodeficiency virus (HIV), by a bite. Interestingly, this case is strikingly similar to a transmission of HIV-1 which occurred after an attempt to prevent airway obstruction of an infected individual during a seizure (16). Biting has also been implicated in the transmission of HIV-1 between siblings (17). While rare, transmission of human retroviruses by biting is worthy of investigation in appropriate situations where other modes of infection cannot be elucidated.