Engineered vaginal lactobacillus strain for mucosal delivery of the human immunodeficiency virus inhibitor cyanovirin-N

Abstract

Women are at significant risk of human immunodeficiency virus (HIV) infection, with the cervicovaginal mucosa serving as a major portal for virus entry. Female-initiated preventatives, including topical microbicides, are urgently needed to help curtail the HIV/AIDS pandemic. Here we report on the development of a novel, live microbicide that employs a natural vaginal strain of Lactobacillus jensenii engineered to deliver the potent HIV inhibitor cyanovirin-N (CV-N). To facilitate efficient expression of CV-N by this bacterium, the L. jensenii 1153 genome was sequenced, allowing identification of native regulatory elements and sites for the chromosomal integration of heterologous genes. A CV-N expression cassette was optimized and shown to produce high levels of structurally intact CV-N when expressed in L. jensenii. Lactobacillus-derived CV-N was capable of inhibiting CCR5-tropic HIV(BaL) infectivity in vitro with a 50% inhibitory concentration of 0.3 nM. The CV-N expression cassette was stably integrated as a single copy into the bacterial chromosome and resolved from extraneous plasmid DNA without adversely affecting the bacterial phenotype. This bacterial strain was capable of colonizing the vagina and producing full-length CV-N when administered intravaginally to mice during estrus phase. The CV-N-producing Lactobacillus was genetically stable when propagated in vitro and in vivo. This work represents a major step towards the development of an inexpensive yet durable protein-based microbicide to block the heterosexual transmission of HIV in women.

FIG. 1.

Proteolytic processing of L. jensenii-derived CV-N (P51G). (A) Schematic representation of CV-N (P51G) with the CbsA signal sequence (italicized). CV-N (P51G) possesses two disulfide bonds. (B) Proteolytic processing of Lactobacillus-derived CV-N (P51G). CV-N (P51G) is first synthesized as a CbsAss-bearing precursor protein. Following cleavage by a signal peptidase, the mature protein is released into the extracellular milieu. The predicted and observed cleavage sites are shown. (C) Production of proteolytically processed CV-N (P51G) by L. jensenii 1153. L. jensenii 1153 containing the empty expression vector (lane 1) and nonmodified CV-N (P51G) expression construct (lane 2) were cultured in Rogosa broth. Proteins from 300 μl of cell-free conditioned Rogosa media were separated by reducing SDS-PAGE, stained with Coomassie blue R-250, and compared to a CV-N (P51G) reference standard (NIH) (lane 3). Proteins in 5 μl of cell-free conditioned media were also probed with rabbit anti-CV-N polyclonal antibody (PAb). MALDI-TOF MS analysis revealed an average mass of 9,614 Da for the truncated CV-N (P51G).

FIG. 2.

Biological activity of L. jensenii-derived APVT-CV-N (P51G). The anti-HIV activity was evaluated in a CCR5-tropic HIV-1_BaL attachment inhibition assay. Cell viability was monitored on a replicate plate using the MTS dye reduction assay. The results (means ± standard deviations) from triplicate determinations in a single experiment are presented and were confirmed in three independent protein preparations.

FIG. 3.

Phenotypic assessment of CV-N expressing pox1 and pepO resolvant strains. Single colonies of the L. jensenii 1153-1666 (pox1::CV-N), L. jensenii 1153-1666 (pepO::CV-N), and parental strain (1153 wt) were inoculated in MRS and cultured at 37°C. Absorbance at 600 nm was performed at various time points to determine the rate of growth. The results (means ± standard deviations) from triplicate determinations in a single experiment are presented and were confirmed in three independent experiments.

FIG. 4.

Expression of full-length CV-N (P51G) by the stably integrated strain L. jensenii 1153-1666. L. jensenii 1153-1666 (Lj1153-1666) and the parental strain L. jensenii 1153 (Lj1153) were cultured in Rogosa broth to different bacterial densities. (A) Proteins from 300 μl of cell-free conditioned Rogosa media were separated by reducing SDS-PAGE and stained with Coomassie blue R-250. The CV-N proteins were quantified based on the band density on a Coomassie blue-stained SDS-PAGE gel in reference to a CV-N (P51G) reference standard (NIH) at different concentrations. (B) Proteins in 5 μl of cell-free conditioned media were also probed with rabbit anti-CV-N polyclonal antibody (PAb).

FIG. 5.

Expression of full-length CV-N (P51G) by L. jensenii strains cultured in pigtailed macaque CVL fluid. The following strains were used for the evaluation: L. jensenii 1153 (Lj1153, lanes 1, 4, and 7), L. jensenii 1153-1666 (lanes 2, 5, and 8), and L. jensenii 1153-2666 (lanes 3, 6, and 9). CVL fluid (400 μl) obtained from pigtailed macaques was inoculated with ∼5 × 10⁷ bacteria and incubated at 37°C and 5% CO₂ for 6 h (lanes 4 to 6) or 20 h (lanes 7 to 9). Proteins in cell-free supernatants (10 μl) were resolved by reducing SDS-PAGE and electroblotted onto a polyvinylidene difluoride membrane for immunodetection with anti-CV-N polyclonal antibodies.

FIG. 6.

Vaginal persistence of L. jensenii 1153-1666 and in vivo CV-N expression in the estrogenized CD1 mouse. (A) Brown and Hopps stain of Lactobacillus in the vagina 24 h postinoculation. Histology of the CD1 mouse vagina during estrus phase shows stratified squamous epithelium (SSE) with a cornified epithelial layer (C). Superficial columnar epithelial cells (CEC) slough into the vaginal lumen (L) and Lactobacillus, indicated by arrows, appear to associate with the sloughed columnar cells. (B) In vivo CV-N expression. Bacteria (∼5 × 10⁷ CFU) or the PBS vehicle were administrated intravaginally to CD1 mice. Vaginal washes were collected in approximately 50 μl of PBS at 48 h postinoculation. The vaginal washes with total of 8 × 10⁴ CFU bacteria were pooled from eight mice and trichloroacetic acid precipitated, and the expression of full-length CV-N was detected by Western blotting. Lane 1, CV-N (P51G) standard lacking APVT; lane 2, vaginal washes at 48 h postinoculation with L. jensenii 1153-1666; lane 3, vaginal washes at 48 h postinoculation with PBS control.