Inhibition of HIV infectivity by a natural human isolate of Lactobacillus jensenii engineered to express functional two-domain CD4

Abstract

The predominant mode of HIV transmission worldwide is via heterosexual contact, with the cervico-vaginal mucosa being the main portal of entry in women. The cervico-vaginal mucosa is naturally colonized with commensal bacteria, primarily lactobacilli. To address the urgent need for female-controlled approaches to block the heterosexual transmission of HIV, we have engineered natural human vaginal isolates of Lactobacillus jensenii to secrete two-domain CD4 (2D CD4) proteins. The secreted 2D CD4 recognized a conformation-dependent anti-CD4 antibody and bound HIV type 1 (HIV-1) gp120, suggesting that the expressed proteins adopted a native conformation. Single-cycle infection assays using HIV-1HxB2 carrying a luciferase reporter gene demonstrated that Lactobacillus-derived 2D CD4 inhibited HIV-1 entry into target cells in a dose-dependent manner. Importantly, coincubation of the engineered bacteria with recombinant HIV-1HxB2 reporter virus led to a significant decrease in virus infectivity of HeLa cells expressing CD4-CXCR4-CCR5. Engineered lactobacilli also caused a modest, but statistically significant, decrease in infectivity of a primary isolate, HIV-1JR-FL. This represents an important first step toward the development of engineered commensal bacteria within the vaginal microflora to inhibit heterosexual transmission of HIV.

Fig. 1.

Schematic of the pOSEL144 expression vector. The repA and erm genes are derived from a naturally occurring L. reuteri plasmid pLEM7 (14) and the ColE1 replicon is from pBluescript. An expression cassette comprising the P₂₃ promoter from Lactococcus lactis (15), the CbsA_ss from L. crispatus, and the 2D CD4 coding region was inserted into SacI and XbaI sites of pOSEL144. The resulting plasmid was designated as pOSEL651.

Fig. 2.

Time-dependent release of 2D CD4 by L. jensenii Xna harboring the 2D CD4 expression plasmid (Xna-651) relative to the control (Xna-144). Samples were collected at OD₆₀₀ 0.8 or 3.2. Proteins were collected by precipitation with TCA from culture supernatants from corresponding bacterial cells and subjected to electrophoretic separation under reducing SDS/PAGE for Western blot analysis with polyclonal anti-CD4 antibodies, T4-4.

Fig. 3.

Secretion of biologically active 2D CD4 by engineered L. jensenii Xna. L. jensenii Xna-144 or Xna-651 bacteria were grown in MRS broth (37°C and 5% CO₂). Culture supernatants were used for CD4 ELISA (A) and gp120 binding assay (B). 2D CD4 concentrations were determined from standard curves generated with refolded 2D CD4 expressed in E. coli.

Fig. 4.

L. jensenii-derived 2D CD4 inhibits infectivity of Env-pseudotyped HIV-1_HxB2. The inhibitory effect of the 2D CD4 protein produced by Xna-651 after partial purification (A) and conditioned medium 199 containing 2D CD4 (B) on Env-pseudotyped HIV-1_HxB2 infectivity of HeLa cells expressing CD4-CXCR4 was determined by viral entry assay (21). The results (mean ± SD) from triplicate determinations in a single experiment were presented and confirmed in three independent experiments. The luciferase activity from virus-infected cells without addition of 2D CD4 samples was defined as 100%. This corresponded to relative light unit values of 276 ± 21 in A and 350 ± 19 in B.

Fig. 5.

Coincubation of L. jensenii Xna-651 bacteria with HeLa-CD4 cells and recombinant viruses inhibits HIV-1 infectivity. L. jensenii Xna-144 or Xna-651 bacteria in medium 199 were coincubated with Env-pseudotyped HIV-1_HxB2 or HIV-1_JR-FL viruses and HeLa-CD4-CCR5 for 1 h at 37°C in the viral entry assay (21). The results (mean ± SD) from four experiments are presented. The luciferase activity from virus-infected cells in the presence of Xna-144 was defined as 100%. The statistical significance of the data (*, P < 0.01) was determined by Student's t test.