A Milk Protein Lactoferrin Enhances Human T Cell Leukemia Virus Type I and Suppresses HIV-1 Infection¹

Human T cell leukemia virus type I (HTLV-I)⁴ is a causative agent of a fatal hemological malignancy known as adult T cell leukemia/lymphoma and a complex neurological disorder designated as HTLV-I-associated myelopathy/tropical spastic paraparesis (1, 2, 3). HTLV-I infection has been found to cluster in certain geographic areas including southwestern Japan, the Caribbean, and central Africa (1, 2, 3). Within such endemic areas, HTLV-I has been transmitted vertically from mother to nursing infant via breast milk, and horizontally by sexual contacts. Milk-borne infection of HTLV-I was supported by the presence of infected cells in milk from carrier mothers (4) and by the experimental transmission of the virus to animals by oral administration of the carrier’s breast milk (5, 6, 7). An intervention in Nagasaki, an endemic area in southwestern Japan by refrain from breast-feeding blocked >80% of mother-to-infant transmission of HTLV-I (8). These studies clearly indicate that breast milk has been an important vehicle of HTLV-I for the maintenance in human populations, and also implicate that components in breast milk may help establish transmission of the virus.

In contrast to HTLV-I, it is generally considered that HIV-1, another human retrovirus, has been recently introduced to Homo sapiens. Milk-borne infection contributes considerably to vertical transmission of this deadly virus (9); however, whether breast milk components influence HIV-1 infection remains largely unknown.

To address these issues, we started to investigate whether breast milk components can influence replication of the two retroviruses. In this study we demonstrate that a milk protein lactoferrin (Lf) can facilitate replication of HTLV-I by up-regulating viral expression, whereas it can inhibit replication of HIV-1 by interfering with viral fusion/entry.

Recombinant human (h) Lf, provided by Agennix (Houston, TX), was derived from Aspergillus awamori and had a purity of 99.4%. Where indicated, hLf was heat-denatured by incubating at 56°C for 1 h. Bovine (b) Lf, provided by Morinaga Milk Industry (Kanagawa, Japan), was derived from bovine milk. In experiments aimed to compare effects of iron-saturated hLf and apolactoferrin, we used human milk-derived hLf reagents that had been purchased from Sigma (St. Louis, MO). PMA, ionomycin, AZT, and mitomycin C (MMC) were also purchased from Sigma.

PBMCs were isolated from HTLV-I-seropositive or -seronegative healthy individuals (Nagasaki Red Cross Blood Center, Nagasaki, Japan), as described previously (10). Cord blood mononuclear cells (CBMCs) were provided by the Pre-and Post-Natal Care Unit, Nagasaki University School of Medicine Hospital with consent by donors.

For HTLV-I infection studies, unfractionated PBMCs obtained from HTLV-I carriers were propagated with RPMI 1640 supplemented with 10% heat-inactivated FBS in the presence or absence of Lf and/or AZT. Also, PBMCs obtained from HTLV-I-seronegative healthy donors were infected with purified HTLV-I virions (Advanced Biotechnologies, Columbia, MD) in the presence or absence of Lf. For HTLV-I transmission studies, CBMCs were cocultured with MMC-treated PBMCs obtained from HTLV-I carriers in the presence or absence of Lf. Replication of HTLV-I was monitored by HTLV-I p19 Ag detection ELISA as per manufacturer’s instruction (Cellular Products, Buffalo, NY).

For HIV-1 infection studies, PBMCs obtained from healthy donors seronegative for both HIV-1 and HTLV-I were infected with HIV-1 AD8 at an approximate multiplicity of infection of 0.05, or PBMCs obtained from an asymptomatic HIV-1 carrier were incubated in the presence or absence of Lf. Replication of HIV-1 was monitored by reverse transcriptase (RT) assays as described previously (10, 11).

Plasmid pU3R-luc, which contains HTLV-I long terminal repeat (LTR) U3R region followed by the luciferase gene, was a gift of K.-T. Jeang (National Institute of Allergy and Infectious Diseases, Bethesda, MD). Plasmids pHIV-1 LTR-luc and pCMV-luc contain HIV-1 LTR or human CMV major immediate-early promoter (MIEP), respectively, followed by the luciferase gene. Transfections of PBMCs or CBMCs and luciferase assays were performed as described previously (12).

Recombinant vaccinia viruses (rVVs) were propagated as described previously (13, 14). rVV expressing HTLV-I Env (vWR-Env17) was a gift of H. Shida (Hokkaido University, Hokkaido, Japan) (15). For fusion assays, BSC-1 cells were infected with vCB21R (a rVV encoding the lacZ gene under the control of T7 promoter) as well as a rVV expressing the indicated viral Env at a multiplicity of infection of 10. Primary CD4⁺ T cells obtained from healthy volunteers were infected with vTF7-3 (a rVV expressing T7 RNA polymerase) at a multiplicity of infection of 10 each. After a 10-h culture at 37°C, both fusion effectors (BSC-1 cells) and fusion targets (CD4⁺ T cells) were mixed and incubated at 37°C for 6 h in the presence of cytosine arabinoside, and the mixed cell culture lysates were subjected to β-galactosidase assays. Where indicated, hLf (100 μg/ml) was added to CD4⁺ T cell cultures 1 h before and throughout cocultivation.

Single-round virus replication assays were performed using replication-incompetent luciferase reporter viruses that had been pseudotyped by Env derived from either T-tropic (X4) HIV-1 HXB2, M-tropic (R5) HIV-1 JR-FL, or amphotropic murine leukemia virus (AMLV), as described previously (13). In brief, 5 × 10⁵ cells were mock-treated or treated with hLf (100 μg/ml) for 1 h at 37°C and then infected with the indicated virus stock (50,000 cpm RT activity). Three hours later, the infected cells were resuspended in RPMI 1640/10% FBS with or without hLf (100 μg/ml), and incubated for 3 days. Luciferase assays were performed using commercially available reagents (Promega, Madison, WI).

Cell surface expression of CD69, CD25, CCR5, or CXC chemokine receptor 4 (CXCR4) was determined by staining cells with monoclonal anti-CD69 Ab FITC-conjugate, anti-CD25 Ab FITC-conjugate, anti-CCR5 Ab 2D7 PE-conjugate, or anti-CXCR4 Ab 12G5 PE-conjugate (PharMingen, San Diego, CA), respectively, and by analyzing in FACScan (Becton Dickinson Immunocytometry Systems, Mountain View, CA), as described previously (13).

Effect of Lf on HTLV-I infection was first evaluated in PBMCs derived from asymptomatic carriers. In preliminary experiments for selected donors, induction of HTLV-I replication that was determined by p19 Ag levels in cell-free culture supernatants was detected as soon as 3 days after stimulation with Lf, and peak p19 Ag titers were obtained by day 7 in most cases (data not shown). Therefore, p19 Ag titers were determined on day 7 in the subsequent experiments. HTLV-I replication was enhanced up to 5- to 10-fold by stimulation with hLf or bLf in a dose-dependent manner (Table I). Similarly enhanced HTLV-I replication was obtained in hLf-stimulated PBMCs that were infected with cell-free HTLV-I virions (Table I). Heat denaturation of hLf abrogated its effects on HTLV-I replication (Table II). Apolactoferrin also induced HTLV-I replication, although it appeared to have less pronounced effects (3- to 5-fold induction) as compared with iron-saturated Lf (5- to 10-fold induction) (data not shown). Because several different sources of Lf (recombinant, human milk, or bovine milk) had similar effects on HTLV-I infection as above, the ability of Lf to induce viral replication appears to be specific. Treatment with AZT (2 μM) resulted in a markedly decreased expression of p19 Ag (Table III). Because AZT has no antiretroviral effect on PBMCs already infected with HTLV-I (16), profound suppression of p19 expression by AZT would indicate that induction of HTLV-I expression in the presence of Lf required expansion of HTLV-I infection. Addition of Lf to the cell culture medium did not appear to influence cell viability or expression of CD69 or CD25, as determined by trypan blue exclusion and flow cytometric analysis, respectively (data not shown), suggesting that Lf effect does not rely on cell activation/proliferation. Thus, these results suggest that Lf may induce HTLV-I replication in carriers’ PBMCs.

We next investigated whether Lf can accelerate transmission of HTLV-I to CBMCs. Uninfected CBMCs were cocultured with MMC-treated carriers’ PBMCs at a ratio of 10:1 in the presence or absence of Lf. It has been shown that MMC-treated cells can no longer support new cycles of retroviral infection (17). As shown in Table IV, p19 level in the coculture supernatants was increased by treatment with Lf. Cultures of MMC-treated carrier PBMC alone produced a much reduced amount of p19 Ag even in the presence of Lf, suggesting that expansion to susceptible CBMCs is required for the induction of HTLV-I replication. Similar results were obtained for experiments of HTLV-I transmission to adult PBMCs (data not shown). These results imply that Lf can expand viral infection to other cells including neonatal or adult lymphocytes.

Lf has been shown to display antiviral activity against a diverse array of viruses including HIV-1, another human retrovirus (18, 19, 20, 21). As demonstrated by those studies, Lf treatment inhibited HIV-1 replication in a dose-dependent manner (Table V). Thus, it appears that a milk protein Lf has opposite effects on the two human retroviruses that are transmitted via breast milk.

A variety of immunomodulatory activities have been attributed to Lf (22), indicating that sequestration of iron is not the only function of this protein. Recently, Lf has been shown to regulate transcription (23); therefore, Lf may exert its function, at least in part, through transcriptional regulation.

To explore the possibility that Lf can transactivate HTLV-I LTR promoter, transient expression assays were performed. PBMCs or CBMCs obtained from HTLV-I-uninfected individuals were transfected with pU3R-luc, and cell lysates were prepared for luciferase assays 2 days after transfection. Where indicated, transfected cells were stimulated with hLf for 12 h before harvest. As shown in Fig. 1, hLf could transactivate HTLV-I LTR but had little effect on activity of HIV-1 LTR or human cytomegalovirus (HCMV) MIEP, whereas stimulation with PMA/ionomycin transactivated all these promoters (Fig. 1; data not shown). Thus, our results strongly suggest that Lf can induce HTLV-I replication by up-regulating HTLV-I LTR promoter activity.

To explore the mechanism whereby HIV-1 infection is inhibited by Lf, we next investigated effects of Lf on viral fusion/entry, another critical step during viral replicative cycle. First, Env glycoproteins derived from HIV-1 or HTLV-I were expressed on the cell surfaces using rVV, and fusogenicity of these cells with CD4⁺ T cells was determined. As shown in Fig. 2, Lf markedly down-regulated fusogenicity with HIV-1 Env, irrespective of their coreceptor usage. Flow cytometric analyses also demonstrated that expression of CCR5 or CXCR4 was not significantly modified by stimulation with Lf (data not shown). Lf also down-regulated fusogenicity with HTLV-I Env, although to a lesser degree.

HIV-1 Env-mediated fusion/entry was further investigated by another system. CD4⁺ T cells were infected with a replication-incompetent, luciferase-reporter molecular clone NL4–3luc-R⁻E⁻ that had been pseudotyped by Env from X4-HIV-1, R5-HIV-1, or AMLV, and viral expression was assessed by luciferase activity. As shown in Fig. 3, pretreatment, but not post-treatment, of the cells with hLf markedly down-regulated HIV-1 expression. This effect was not dependent on coreceptor usage and was also observed in AMLV Env-pseudotyped virus. Taken together with previous studies suggesting that Lf is capable of inhibiting a number of viral infections at early stage(s) of replicative cycle (18, 19, 21), our results imply that Lf may interfere with cellular entry of a number of enveloped viruses in a nonspecific fashion.

Milk-borne transmission of HTLV-I accounts for a majority of cases of vertical transmission, and infection during the infantile period is usually a prerequisite for the development of leukemia during late adulthood (1, 2, 3). Because HTLV-I appears to have coexisted with Homo sapiens for quite a long time by efficacious transmission via breast milk and semen, we hypothesized that HTLV-I may take advantage of constituent(s) in these body fluids. Lf is a mammalian iron-binding glycoprotein present in a variety of body fluids including breast milk, tears, semen, and plasma, and has been attributed to antimicrobial effects and immunomodulatory activities (22). In this study we have demonstrated that Lf is capable of enhancing HTLV-I infection.

Induction of HTLV-I replication by Lf is mediated, at least in part, by transcriptional activation of HTLV-I LTR promoter. Lf induction of HTLV-I LTR activity is specific, because Lf had little effect on the HIV-1 LTR or HCMV MIEP activity. Recently, it has been shown that Lf is transferred to the nucleus, directly binds to DNA, and is capable of activating transcription (23); however, no Lf-inducible promoter has actually been identified yet. Therefore, this may be the first study to identify a promoter that is transactivated by Lf. The precise mechanism whereby Lf transactivates HTLV-I LTR remains undetermined; it may directly transactivate HTLV-I LTR promoter, or may induce expression of host factor(s) that in turn transactivate the promoter. Interestingly, Lf induction of HTLV-I LTR promoter activity is more prominent in CBMCs than in PBMCs (Fig. 1). Although sample numbers are still small, these results may indicate that immature newborn cells may contain larger amounts of host factor(s) that are required for Lf-mediated effect than do adult cells. Such possibilities are currently under investigation.

Lf has been shown to have antiviral activity. Recently, Harmsen et al. showed that Lf inhibits HIV-1 and HCMV infections (18), probably at the level of viral entry (absorption or penetration). By performing two assays (rVV-based fusion assays and single-round viral replication assays), we have confirmed that Lf could inhibit HIV-1 entry into cell. We have also shown that Lf had similar effects on HTLV-I Env-mediated cell-cell fusion, although to a lesser degree. Thus, Lf appears to have the opposite effects on HTLV-I infection, induction of viral expression, and suppression of viral entry. Therefore, we hypothesize that the ability of Lf to enhance viral infection at later step(s) may overcome its negative effect on earlier step(s) in case of HTLV-I; however, infection with other viruses (e.g., HIV-1, CMV) may be simply suppressed by its inhibitory effect on viral entry. Further investigation is needed to clarify this issue.

Lf is a major milk protein, and the amounts of Lf are 5∼7 mg/ml in colostrum, 1∼3 mg/ml in mature milk, 1.5∼2.2 mg/ml in tears, and 0.5∼1 mg/ml in seminal fluid (24, 25). Therefore, we believe that the concentration used in this study (up to 100 μg/ml) is physiologically relevant even after dilution in the gastrointestinal and genital tracts. Although the precise portal of entry for HTLV-I or HIV-1 across mucosal surfaces in infants is unknown (9), it is not unreasonable to hypothesize that Lf actually exerts its effects on viral transmission or infection in vivo. Finally, it is important to rule out the possibility that other components or contaminants rather than Lf contain the relevant activity. However, based upon our observations that several different sources (recombinant, human milk, or bovine milk) of Lf had similar effects, it is reasonable to assume that Lf did actually exhibit activities demonstrated in this study.

In summary, this study suggested that a milk protein Lf enhances replication and transmission of HTLV-I and suppresses infection with HIV-1, the two milk-borne pathogens, and may have implications for host-pathogen interactions during transmission via milk or other biological secretions.

We thank K.-T. Jeang for providing pU3R-luc; Agennix and Morinaga Milk Industry for Lf products; Nagasaki Red Cross Blood Center and Pre-and Post-Natal Care Unit (Nagasaki University Hospital) for blood samples; H. Shida for vWR-Env17; E. Berger for other rVV constructs; N. Landau and J. Sodroski for plasmids used in single-round virus replication assays; T. Theodore for HIV-1 AD8; S. Katamine and H. Ishimaru for critical review; and K. Iijima and M. Yokoyama for technical assistance and graphic work.