Effect of Broccoli Sprouts and Live Attenuated Influenza Virus on Peripheral Blood Natural Killer Cells: A Randomized, Double-Blind Study

Abstract

Enhancing antiviral host defense responses through nutritional supplementation would be an attractive strategy in the fight against influenza. Using inoculation with live attenuated influenza virus (LAIV) as an infection model, we have recently shown that ingestion of sulforaphane-containing broccoli sprout homogenates (BSH) reduces markers of viral load in the nose. To investigate the systemic effects of short-term BSH supplementation in the context of LAIV-inoculation, we examined peripheral blood immune cell populations in non-smoking subjects from this study, with a particular focus on NK cells. We carried out a randomized, double-blinded, placebo-controlled study measuring the effects of BSH (N = 13) or placebo (alfalfa sprout homogenate, ASH; N = 16) on peripheral blood mononuclear cell responses to a standard nasal vaccine dose of LAIV in healthy volunteers. Blood was drawn prior to (day-1) and post (day2, day21) LAIV inoculation and analyzed for neutrophils, monocytes, macrophages, T cells, NKT cells, and NK cells. In addition, NK cells were enriched, stimulated, and assessed for surface markers, intracellular markers, and cytotoxic potential by flow cytometry. Overall, LAIV significantly reduced NKT (day2 and day21) and T cell (day2) populations. LAIV decreased NK cell CD56 and CD158b expression, while significantly increasing CD16 expression and cytotoxic potential (on day2). BSH supplementation further increased LAIV-induced granzyme B production (day2) in NK cells compared to ASH and in the BSH group granzyme B levels appeared to be negatively associated with influenza RNA levels in nasal lavage fluid cells. We conclude that nasal influenza infection may induce complex changes in peripheral blood NK cell activation, and that BSH increases virus-induced peripheral blood NK cell granzyme B production, an effect that may be important for enhanced antiviral defense responses.

Trial registration: ClinicalTrials.gov NCT01269723.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. CONSORT 2010 Flow diagram for recruitment and randomization of subjects.

ASH = alfalfa sprout homogenate, BSH = broccoli sprout homogenate, LAIV = life attenuated influenza virus, NK cell = natural killer cell.

Fig 2. Overview of sample collection and processing.

(A) Study design and sample collection. Details of the complete study have been published previously [1]. (B) Blood samples were stained for total leukocyte populations or used for NK cell enrichment. NK cells were analyzed for surface marker expression or cytokine production either naive or stimulated with PMA and ionomycin (Iono). Half of the peripheral blood mononuclear cells (PBMCs) were frozen and used later for the cytotoxicity assay.

Fig 3. LAIV effect on cytotoxicity potential of systemic NK cells (regardless of treatment).

Following NK cell enrichment, NK cells were incubated with K562 target cells for 4hrs and the cell mixture was stained for viability. N = 26 (day-1 and 2), N = 22 (day21). Data are shown as whiskers with 10–90 percentiles. *significantly different (p = 0.015), tested with paired t test.

Fig 4. BSH effect on intracellular markers of systemic NK cells.

Following NK cell enrichment, NK cells were stimulated with PMA/Ionomycin and blocked with Brefeldin A for 4hrs. The differences of day2 or day21 and day-1 are shown. Data are presented as mean±std.dev. N = 9–14. Data are shown as whiskers with 10–90 percentiles. *significantly different (p = 0.049), tested with two sample t test.

Fig 5. Correlation between nasal influenza B virus load and granzyme B in systemic NK cells.

Viral load in nasal lavage fluid cells collected after inoculation with LAIV was detected via RT-PCR to Flu B RNA and expressed as log transformed area under the curve (details see [1]). The correlation with granzyme B levels in systemic NK cells was tested with the Pearson correlation for both groups.