Overexpression of a Mycobacterium ulcerans Ag85B-EsxH Fusion Protein in Recombinant BCG Improves Experimental Buruli Ulcer Vaccine Efficacy

Abstract

Buruli ulcer (BU) vaccine design faces similar challenges to those observed during development of prophylactic tuberculosis treatments. Multiple BU vaccine candidates, based upon Mycobacterium bovis BCG, altered Mycobacterium ulcerans (MU) cells, recombinant MU DNA, or MU protein prime-boosts, have shown promise by conferring transient protection to mice against the pathology of MU challenge. Recently, we have shown that a recombinant BCG vaccine expressing MU-Ag85A (BCG MU-Ag85A) displayed the highest level of protection to date, by significantly extending the survival time of MU challenged mice compared to BCG vaccination alone. Here we describe the generation, immunogenicity testing, and evaluation of protection conferred by a recombinant BCG strain which overexpresses a fusion of two alternative MU antigens, Ag85B and the MU ortholog of tuberculosis TB10.4, EsxH. Vaccination with BCG MU-Ag85B-EsxH induces proliferation of Ag85 specific CD4+ T cells in greater numbers than BCG or BCG MU-Ag85A and produces IFNγ+ splenocytes responsive to whole MU and recombinant antigens. In addition, anti-Ag85A and Ag85B IgG humoral responses are significantly enhanced after administration of the fusion vaccine compared to BCG or BCG MU-Ag85A. Finally, mice challenged with MU following a single subcutaneous vaccination with BCG MU-Ag85B-EsxH display significantly less bacterial burden at 6 and 12 weeks post-infection, reduced histopathological tissue damage, and significantly longer survival times compared to vaccination with either BCG or BCG MU-Ag85A. These results further support the potential of BCG as a foundation for BU vaccine design, whereby discovery and recombinant expression of novel immunogenic antigens could lead to greater anti-MU efficacy using this highly safe and ubiquitous vaccine.

Fig 1. Heterologous expression of the MU-Ag85B-EsxH fusion protein by M. bovis BCG.

A. The 6698 base pair (bp) plasmid map for pSL402 is shown. Constitutive expression of MU-Ag85B-EsxH (green, purple) C-terminally fused to an HA epitope tag (yellow) is driven by the promoter region from hsp60 (black). Plasmid replication is regulated in in mycobacteria using oriM (red) and in E. coli shuttle vectors using oriE (blue). Plasmid retention is selected by a hygromycin resistance cassette (brown). B. Three vaccine accession lot vials of BCG transformed with pSL402 were chosen at random for Western analysis of MU-Ag85B-EsxH expression using anti-HA antibody (lanes 1–3). The expected band size was 46 kDa. The negative control was a lysate from BCG transformed with an empty expression vector (lane 4) and the positive control was a lysate from M. smegmatis transformed with pSL402 (lane 5).

Fig 2. BCG MU-Ag85B-EsxH vaccination induces proliferation of antigen-specific CD4⁺ T cells.

C57BL/6 mice remained unprimed (dotted black, open circle) or were intravenously primed (A) or subcutaneously primed (B) with 10⁷ BCG transformed with empty-vector (pHA, square), the MU-Ag85A pSL401 expression vector (triangle), or with the MU-Ag85B-EsxH pSL402 vector (circle). At weekly time points, retro-orbital blood was collected for T lymphocyte isolation. Flow cytometric analysis was performed to quantify levels of CD4⁺ T cells bound to MHCII-Ag85 tetramer. Asterisks and number signs indicate statistical analysis by the student’s t-test (n = 5 for each group) comparing the empty-vector BCG group to BCG MU-Ag85B-EsxH and BCG MU-Ag85A groups, respectively. Error bars represent standard deviation. *p<0.05, **p<0.002, #p<0.05, ##p<0.001.

Fig 3. BCG MU-Ag85B-EsxH induces replication of antigen-specific CD4⁺ effector and central memory T cells.

At 4 weeks post-prime with empty-vector BCG (gray), BCG MU-Ag85A (black), or BCG MU-Ag85B-EsxH (striped), mice were bled retro-orbitally and peripheral leukocytes were stained CD4 and the memory markers CD62L and CD44. Percentages of total CD4⁺ T cells (A) and CD62L^loCD44^lo naïve CD4⁺ T cells (B) were quantified, as were antigen-specific effector memory CD4⁺ T cells (CD62L^loCD44^hi) and central memory CD4⁺ T cells (CD62L^hiCD44^hi) which specifically bound Ag85-MHCII-tetramer (C, D). Asterisks indicate statistical analysis by the student’s t-test (n = 4 for each group). Error bars represent standard deviation. *p<0.05.

Fig 4. Expression of MU antigens increases responsiveness of T_h1 cells to M. ulcerans and purified components.

C57BL/6 mice remained unprimed (white) or were intravenously primed with 10⁷ BCG transformed with empty-vector (gray), the MU-Ag85A pSL401 expression vector (black), or with the MU-Ag85B-EsxH pSL402 vector (striped). At 8 weeks post-prime, mice were euthanized and splenocytes were isolated for stimulation with MU-Ag85A peptide, MU-Ag85B peptide, whole heat killed MU1615 (HKMU), or MU whole cell lysate. ELISPOT spot forming units (SFU) were used to quantify the number of IFNγ-producing cells after 24 hours of stimulation. Asterisks indicate statistical analysis by the student’s t-test (n = 5 for each group). Error bars represent standard deviation. *p<0.05, **p<0.01.

Fig 5. Ag85-specific humoral responses are induced by BCG MU-Ag85B-EsxH.

C57BL/6 mice remained unprimed (dotted, open circle) or were intravenously primed with 10⁷ BCG transformed with empty-vector (square), the MU-Ag85A pSL401 expression vector (triangle), or with the MU-Ag85B-EsxH pSL402 vector (circle). At 2, 4, and 6 weeks post-vaccination peripheral submandibular blood was collected for serum isolation and IgG ELISA against recombinant M. tuberculosis Ag85A (rAg85A, A), recombinant M. tuberculosis Ag85B (rAg85B, B), or purified M. tuberculosis Ag85 complex proteins (C). Asterisks and number signs indicate statistical analysis by the student’s t-test (n = 5 for each group) comparing the empty-vector BCG group to BCG MU-Ag85B-EsxH or BCG MU-Ag85A groups, respectively. Error bars represent standard deviation. *p<0.05, #p<0.05.

Fig 6. BCG MU-Ag85B-EsxH vaccination is significantly more protective than BCG or BCG MU-Ag85A against MU challenge and bacterial burden.

A. Naïve C57BL/6 mice (black) or mice subcutaneously primed with 10⁷ empty-vector BCG (blue), BCG MU-Ag85A (red), or BCG MU-Ag85B-EsxH (green) were challenged with 10⁵ MU1615 via the footpad 8 weeks post-vaccination. At various time points, point post-challenge, the area of footpad swelling was measured. Mice were euthanized if the height of footpad swelling reached 4.5 mm. Survival curves represent time to euthanasia. Asterisk indicates statistical analysis of empty-vector BCG versus BCG MU-Ag85A (red) or BCG MU-Ag85B-EsxH (green) and BCG MU-Ag85A versus BCG MU-Ag85B-EsxH (black) by the Mantel-Cox test (n = 10 for each group). *p<0.05, **p<0.01, ***p<0.001 B. C57BL/6 mice were subcutaneously vaccinated as described in A (unprimed; white, empty-vector BCG; gray, BCG MU-Ag85A; black, BCG MU-Ag85B-EsxH; striped) and intradermally challenged with 10⁵ MU1615. At 6 and 12 weeks post-challenge, mice were euthanized, footpad homogenates were prepared for auramine-rhodamine staining, and acid-fast bacilli were quantified under 1000x magnification. Asterisks indicate statistical analysis by the student’s t-test (n = 16 images per group). Error bars represent standard deviation. *p<0.05, ***p<0.001 C-E. Tissue sections from 12 week MU-infected mouse footpads were H&E stained (row C) or Ziehl-Neelsen (ZN) stained (rows D, E) and visualized by light microscopy. Footpads from unprimed mice consistently displayed loss of epidermis (white arrows) associated with necrosis (N) and inflammatory infiltrate (*). Black arrows indicate fully intact epidermis for the BCG MU-Ag85A and BCG MU-Ag85B-EsxH groups. Edema (E) and thinning epidermis remained prominent in empty-vector BCG -primed mice but was rarely presented in either of the recombinant BCG-primed groups. Red arrows denote masses of pink ZN⁺ extracellular bacilli observed in unprimed and empty-vector-primed sections (rows D, E) and the yellow arrow highlights and granulomatous lesion in an unprimed footpad. Scale bars in rows C and D represent 100 μm (40X magnification) and in row E represent 50 μm (100x magnification).