Genetics and Evolution of Infectious Diseases (Elsevier Insights)
Bivariate models were defined as follows:. For all remaining traits, sex and tank were included as fixed effects, and MW and MA were fitted as covariates. Similarly, the effect of LogBL was included for ADGi, ST and BS that did not include LogBL as response variable; u i and e i are vectors of random animal genetic and residual effects, respectively, c 2 is the vector of random environmental effect associated with common rearing of full-sib families for HW prior to tagging; X i and Z i are the design matrices for the corresponding fixed and random effects for both traits, and W 2 is the design matrix for HW.
The random effects associated to each animal and residual effects, in conjunction with common environment effect for HW, were assumed to be normally distributed according to:. Common environment effect was evaluated for each trait using a single-trait likelihood ratio test [ 32 ]. Thus, C 0 represents a 1 x 1 scalar of common environment effect for HW.
Given that HW was recorded on a different population of individuals to the challenge population, environmental covariance was set to zero in the R 0 matrix. To assess the influence of different recording times for body weight and bacterial load between survivors and non-survivors on co- variance estimates, bivariate analyses for logBL, ADG0, ADGi, and HW were repeated for subsets of data containing either survivors or non-survivors only.
Similarly, bivariate analyses for ST were also performed exclusively for non-survivors. The following formula was used to estimate heritability values for the different traits:. Typical clinical signs and pathological lesions associated with a P. These signs included inappetence, lethargy and pale gills [ 4 ]. Mortality began on day 10 post IP injection.
Dead individuals showed a swollen kidney, splenomegaly and yellowish liver tone typical symptoms of SRS infection; [ 4 ].
Evolution in action: climate change, biodiversity dynamics and emerging infectious disease
During the 50 days of challenge, the three replicate tanks reached a cumulative mortality of Figure 1A shows the observed mortality from all challenged families, ranging from 5. The proportion of survivors and non-survivors among these 33 extreme families are shown in Figure 1B. Observed mortalities for all the coho salmon families challenged with P. The 16 most resistant and 17 most susceptible families for which bacterial load was quantified are highlighted in light and dark grey, respectively Fig 1A.
Percentage of survivors and non-survivors for each of the 16 most resistant and 17 most susceptible families selected after experimental challenge against P. Table 1 shows summary statistics of the phenotypic variation observed for the different measured traits. Prior to the challenge test, individuals gained on average 1. However, this average daily growth rate was reduced by almost half 0.
Some individuals continued to grow maximal gain of 6. Summary statistics for average daily gain prior P. Bacterial load quantification of fish from 33 families showed that 6. All these individuals survived the experimental challenge and belonged to the resistant families. The bacterial load measured on the log 10 scale in the other animals ranged from 0.
Harvest weight,obtained from 41, commercial fish with linked pedigree to the challenged population, had a mean of 6.
Survivors grew on average by 2. Furthermore, genetic correlations between the traits ADGi, and LogBL measured in survivors and non-survivors, respectively, were low 0. Interestingly, the same was true for growth prior to infection, where the genetic correlation between ADG0 in survivors and non-survivors was 0. Summary of the variables used for each mixed models and the significance effect. The latter suggests that differences in growth during infection occur due to other factors than differences in bacterial load.
Estimated heritabilities and genetic correlations obtained by including both survivors and non-survivors in the models, and by analyzing both categories separately, are presented in Table 3. This estimate decreased considerably during the experimental challenge test e. Higher heritability estimates were obtained for HW of related individuals 0. The survival traits ST and BS were found to be moderately heritable 0. The estimate of heritability for LogBL was low and not significantly different from zero 0. Genetic parameters and estimated heritabilities SE , genetic and phenotypic correlations below and above diagonal, respectively for average daily gain prior infection ADG0 , during infection ADGi , harvest weight HW , bacterial load logBL , day of death ST and binary survival BS for pooled, survivors and non-survivors animals.
Infection, Genetics and Evolution
Growth prior to infection and ST also showed a significant and positive genetic correlation 0. However, genetic correlations between ADG0 and BS were not found significantly different from zero 0.
- Researchers uncover novel virus type that may shed light on viral evolution.
- Marcos Perez-Losada | Computational Biology Institute | The George Washington University.
- Hypertension in an Anthropological and Evolutionary Paradigm | Hypertension.
- About the book.
- Portable sequencer in the fight against infectious disease | Journal of Human Genetics!
Binary survival was strongly and favorably correlated with ST 0. Similarly, the statistically significant undesirable positive correlation between HW and LogBL estimated with the pooled dataset was no longer observed when the analysis was stratified into survivors and non-survivors Table 3. Lastly, genetic correlations between ADGi and LogBL tended to be favorable, but were not found statistically significant in either of the datasets. In order to include P. The current dataset comprises animals that were exposed to P.
However, these authors only focused on the genetic association between harvest weight and day of death as a measure of resistance.
Improving viral genomic data analyses to gain new insights into HIV transmission and evolution
The current work provides novel insights into the genetic co variation between growth and P. By defining growth as average daily gain prior and during an infection with P. Moreover, estimates of the genetic correlation of all these traits with harvest weight was also determined using a large cohort from a coho salmon breeding population.
We found a moderate significant genetic variation for early growth rate 0. Similar heritability values have been reported for growth rate in others salmonid species, ranging from 0. When growth rate was measured during infection with P. A similar drop in heritability for average daily gain during infection, compared growth rate prior infection have been observed in pigs [ 37 ] and chickens [ 38 ]. In our study, this drop in heritability could be explained by a relatively stronger increase in the phenotypic variance with some fish losing rather than gaining weight due to infection , than in the genetic variance Table 3.
- A Succubus Tale: Volumes 1-3.
- Black Rabbit Summer: Roman (German Edition).
- SALUBRIOUS LIVING.
- Infection, Genetics and Evolution?
- Pour lamour de Bilouette (Les oeuvres de Robert Reus et leurs dossiers t. 7) (French Edition)!
- Sneaker Store Athletic Shoe Shop Start Up Sample Business Plan NEW!;
The results suggest that differences in growth under infection are primarily controlled by environmental rather than genetic factors, once individual differences in early growth or in disease resistance represented by log-transformed bacterial load included as fixed covariate are accounted for. Nevertheless, heritability estimates for growth under infection were still significantly different from zero, which is indicative for genetic variation in tolerance, in addition to resistance [ 25 , 31 ].
A moderate and positive genetic correlation was found between growth prior to and under infection. This favorable and significant genetic correlation was also estimated between growth prior to infection and harvest weight. The results indicate not only that fish with greater genetic growth potential at early stage in a pathogen-free environment in fresh water also tend to have greater growth potential during infection with P.
Significant additive genetic variation was estimated for ST and BS. These estimates are in agreement with previous estimates for the same and other types of pathogens for salmonid species for a detailed review see [ 9 ]. Furthermore, a moderate and favorable genetic correlation between early growth and ST was found. These results corroborate findings indicating that fish with faster growth prior to and during infection are more likely to survive after an experimental challenge with a bacterial agent [ 39 , 40 ]. Hence, together the results of this study suggest that selection for early growth is expected to have a positive effect on growth under P.
One of the novelties of the present studies is the inclusion of bacterial load as additional measures of host resistance to infection. Even though pathogen load is commonly used as measure of disease resistance in domestic livestock [ 41 — 43 ], it is rarely used in aquaculture for practical reasons [ 44 ]. Measurements of individual pathogen load not only provide novel insights into different genetic response mechanisms to infection, such as resistance and tolerance or endurance [ 31 , 45 , 46 ] and their impact on survival [ 21 ], but may also help to predict potential epidemiological effects of selection, as individuals with high pathogen load may be more infectious.
Medical microbiology - Wikipedia
In our study, the regression coefficient for logBL, when fitted into the statistical models for ST and BS, was significantly different from zero and negative, indicating that individuals with higher bacterial load were more likely to die and tended to die faster when infected with P. Although the sample size for bacterial load in our study was too small to obtain accurate genetic parameter estimates, we found significant genetic variation for bacterial load in surviving animals 0. Furthermore, a strong favorable genetic correlation was found between log-transformed bacterial load and binary survival, and genetic correlations between LogBL and ST or growth traits tended to be negative, suggesting that selection for growth or survival post P.
In our study, final body weight used to calculate ADGi, and BL were measured at time of death for non-survivors and at the end of the trial for survivors. This implies that the trait measurements may relate to different stages of infection in survivors and non-survivors. The low genetic and phenotypic correlations for these traits measured in survivors and non-survivors indicate that these traits should be considered as biologically different in both groups of individuals.
Indeed, survivors may have already fully or partially recovered from infection at the time of recording and may thus have had reduced bacterial load in contrast to non-survivors whose bacterial load may have peaked at the time of death. Similarly, in the case of ADGi, non-survivors fish may have died when body weight reached a minimum, whereas survivors may have experienced compensatory growth at the later stages of the experiment.
For these reasons, the analyses were carried out with data from survivors and non-survivors pooled with BS fitted as fixed effect to partly account for these differences in order to maximize statistical power, and for survivors and non-survivors separately to disentangle the effects of confounding with recording times. Furthermore, genetic parameter estimates may be slightly upward biased due to the fact that bacterial load was only measured in families from the extreme ends of the survival time breeding value distributions. Nevertheless, results of the genetic estimates within survivors and non-survivors analyzed separately were overall consistent with those obtained with pooled animals, although standard errors were higher.
In particular, growth prior to infection showed a generally favorable correlation with growth and survival during P. From a resource allocation theory point of view, a negative correlation between resistance and growth would be expected, given that these are two competing resource-demanding mechanisms [ 47 ]. Indeed, previous studies found a negative genetic correlation between body weight and resistance as day of death to SRS and viral haemorrhagic septicaemia VHS in salmonid species [ 12 , 15 , 48 ].
Instead, the estimated positive and favorable genetic correlations in pooled and non-survivors individuals, suggest that fish with higher genetic growth rate measured in freshwater at early stage are also genetically more resistant to P. Similar results have been obtained in Atlantic salmon and rainbow trout [ 39 , 49 ]. This trade-off was only observed due to the unfavorable genetic correlation between ST and HW when the former was measured in pooled animals.
However, the genetic correlation between ST and HW was not significantly different from zero when only non-survivors animals were used, suggesting a less robust estimation compared to ST and ADG0, which was positive and significantly different from zero when using only susceptible animals. Differences at the development of the immune system at early life stages, given by body size at time of infection may explain the lack of trade-off [ 39 ].
Furthermore, the role of insulin-like growth factor IGF could play a key role as has been associated with increased survival and detected in higher levels in faster growing fish [ 50 — 52 ]. Previously, an up-regulation of pro-inflammatory genes has been detected in Atlantic salmon families with early mortality following a P. Moreover, using genome-wide association studies, candidate genes related with pro-inflammatory response proximate to markers associated with P.
We propose that an exacerbated, ineffective inflammatory response may have led to tissue damage and the subsequent weight reduction in these individuals, with subsequent mortality.