Nonsterile immunity to cryptosporidiosis in infants is associated with mucosal IgA against the sporozoite and protection from malnutrition

Abstract

We conducted a longitudinal study of cryptosporidiosis from birth to three years of age in an urban slum of Dhaka Bangladesh. Fecal DNA was extracted from monthly surveillance samples and diarrheal stool samples collected from 392 infants from birth to three years. A pan-Cryptosporidium qPCR assay was used to identify sub-clinical and symptomatic cryptosporidiosis. Anthropometric measurements were collected quarterly to assess child nutritional status. 31% (121/392) of children experienced a single and 57% (222/392) multiple infections with Cryptosporidium. Repeat infections had a lower burden of parasites in the stool (Cq slope = -1.85; p<0.0001) and were more likely to be sub-clinical (Chi square test for trend; p = 0.01). Repeat infections were associated with the development of growth faltering (Pearson correlation = -0.18; p = 0.0004). High levels of fecal IgA antibodies against the Cryptosporidium Cp23 sporozoite protein at one year of life were associated with a delay in reinfection and amelioration of growth faltering through three years of life (HAZ IgA high responders -1.323 ± 0.932 versus HAZ -1.731 ± 0.984 p = 0.0001). We concluded that nonsterile immunity to cryptosporidiosis in young children was associated with high levels of mucosal IgA anti-Cp23 and protection from diarrhea and growth faltering. Trial Registration: NCT02764918.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. COHORT diagram.

Study subjects, collected samples and new infection numbers. Abbreviations: RT-qPCR Real-Time quantitative polymerase chain reaction; DS: Diarrheal Stool samples; FU: Follow-up; MS: Monthly Stool Samples, Cq: Cycle Quantification.

Fig 2. Parasite burden was lower in recurrent infections.

A) Correlation between parasite burden and the number of Cryptosporidium infections. Each symbol represents the first detectable sample of an individual infection. Y-axis represents the quantitative cycle (Cq) of the diagnostic pan-Cryptosporidium PCR assay. X-axis shows the number of Cryptosporidium infections that had occurred in this child. The line represents the slope (-1.85 ± 0.21) and Y-intercept (26.95 ± 0.44) estimated from the GEE model with the exchangeable correlation structure (p<0.0001). B) Comparison of single infections (black symbol) with those that are part of a series (gray symbol). Bar graph (indicating data mean ± standard deviation) with individual data points. Each symbol on the box plot represents the first positive sample of an individual infection. X-axis refers to Infection number and, if the first infection, whether a second Cryptosporidium infection took place in the 3 years of life. Y-axis represents the quantitative cycle (Cq) of the diagnostic pan-Cryptosporidium PCR assay. Horizontal bars represent the result of a non-parametric Kruskal-Wallis test **** indicates p<0.0001.

Fig 3. Parasite burden in older children.

The amount of parasite in stool was determined as a function of the number of Cryptosporidium infections in a child by linear regression. The analysis was restricted to children between 2.25 and 2.75 years of age (n=140). Each symbol represents the first detectable sample of an individual infection. Y-axis, quantitative cycle of the diagnostic pan-Cryptosporidium PCR assay (Cq). X-axis, the number Cryptosporidium infections that had occurred in each child. Slope: -1.65, R squared value: 0.01125, Significance p<0.0001.

Fig 4. Cryptosporidiosis frequency was associated with growth faltering distinct from the impact of birth nutritional status.

A) Relationship between length for age z score (LAZ) at birth (Y-axis) and the total number of Cryptosporidium spp. infections (X-axis). The slope was not significantly different from one. B) Relationship between height for age z score (HAZ) at 3 years (Y-axis) and the total number of Cryptosporidium spp. infections (X-axis). This was initially analyzed by use of linear regression Slope: -0.152 ± 0.0429, R squared value: 0.0313, Significance p = 0.0004 and the relationship between the two data sets was then confirmed by the Pearsons correlation coefficient as described in the text. Children were defined to be at risk for growth faltering with a LAZ or HAZ score <-1 and malnourished at LAZ or HAZ score <-2. Orange box: birth LAZ or 3 year HAZ score -1 to -2; red box: birth LAZ or 3 year HAZ or score < -2.

Fig 5. Correlates of cryptosporidiosis-associated growth-faltering.

A) Comparison of three year-HAZ with birth LAZ. (Slope: -0.294 ± 0.05; R squared value: 0.08; Significance p

Fig 6. High anti-Cp23 IgA levels were associated with a reduction in cryptosporidiosis-associated growth-faltering.

The HAZ values of children in the upper and lower 50th percentile for fecal IgA anti-Cp23 year were analyzed. HAZ are shown for children in both year one and year three of life. Mean ± standard deviation with individual data points. Horizontal bars represent the result of a non-parametric Kruskal-Wallis test ***p