Immunoglobulins as Biomarkers for Gastrointestinal Nematodes Resistance in Small Ruminants: A systematic review

The rise of anthelmintic resistance worldwide has led to the development of alternative control strategies for gastrointestinal nematodes (GIN) infections, which are one of the main constraints on the health of grazing small ruminants. Presently, breeding schemes rely mainly on fecal egg count (FEC) measurements on infected animals which are time-consuming and requires expertise in parasitology. Identifying and understanding the role of immunoglobulins in the mechanisms of resistance could provide a more efficient and sustainable method of identifying nematode-resistant animals for selection. In this study we review the findings on immunoglobulin response to GIN in the literature published to date (june 2019) and discuss the potential to use immunoglobulins as biomarkers. The literature review revealed 41 studies which measured at least one immunoglobulin: 35 focused on lamb immune response (18 used non-naïve lambs) and 7 on yearlings. In this review we propose a conceptual model summarizing the role of immunoglobulins in resistance to GIN. We highlight the need for more carefully designed and documented studies to allow comparisons across different populations on the immunoglobulin response to GIN infection.


Results and Discussion
evaluation of the risk of bias. The results of the qualitative evaluation of the risk of bias are presented on Fig. 1. Ethical statement for the use of animal was specified in 26% of the studies. Indeed, this statement was not compulsory before the early 2000s, depending on the country. An adequate allocation sequence generation was specified in 74% of the studies, 88% used random housing and 91% reported the use of similar groups at baseline. However, allocation concealment was not performed in 60% of the studies. Random outcome assessment was specified in 84% of the studies but the outcome assessment was not blind in 56% of the studies. Incomplete outcome was adequately addressed in 91% of the studies and other sources of bias were reported in 98% of the studies. There was no indication of bias across studies reporting correlation coefficients based on the funnel plot ( Fig. 2) analysis and the results of the Egger's test (I 2 = 10.1%, P = 0.304).
immune response to Gin infection. Both cellular and humoral responses are actively involved in immune response against nematode infection. The main effectors of this immune response are T-and B-lymphocytes, plasma cells, mast cells, eosinophils, globule leukocytes, soluble cytokines and various immunoglobulin isotypes 25 . Incoming nematode larvae from GIN infection trigger local inflammation and mast cell degranulation, which damages the gastrointestinal mucosa 26 . Dendritic cells, macrophages and the other antigen-presenting cells, capture the nematode antigens within the intestinal mucosa and migrate to the regional lymph nodes to present these antigens to naïves T cells 24 . After T-cell differentiation, the secretion of type 1 T helper (Th1) or type 2 T helper (Th2)-associated cytokines induces the migration to the site of infection of activated effector cells such as eosinophils and mast cells 24,27 . The type of helper CD4 + T lymphocytes that develop following an infection with nematodes is critical for the ability of the host animal to overcome an infection 28 . The Th1 cells produce interferon-gamma (IFNγ), interleukin (IL-2) and tumor necrosis factor-beta (TNFβ) for the activation of macrophages and the initiation of the cell-mediated immunity and the phagocyte-dependent protective responses 11,29 .  The Th1 cells develop mainly following infections by intracellular parasites (viruses and some bacteria). When GIN antigens penetrate the gastrointestinal tissues of the host, macrophages and other cells which have receptors for nematode cell surface molecules are activated and induce a specific but mostly ineffective immune response 29 .
The phagocyte-independent protective responses characterized by antibodies production, eosinophils activation and inhibition of several macrophage functions, are activated by the production of Th2 cytokines (IL-4, IL-5, IL-10, and IL-13) 11,29 . The Th1 response inhibits the Th2 response through IL-10 30 , which makes Th1 and Th2 responses antagonistic to each other. Results from studies in sheep showed that CD4 + lymphocytes increase during the experimental infection of both susceptible and resistant sheep with GIN 31,32 . But compared with resistant sheep, susceptible produce more interferon gamma (IFNγ), fewer parasite-specific serum antibodies, blood and abomasal eosinophils 33 . The role of the two major types of T helper cells distinguishes resistant from susceptible sheep.
The acquired immune response after infection with Haemonchus contortus was compared in Barbados Blackbelly sheep, which are generally defined as a resistant breed, and Columbia sheep, a breed classified as susceptible 34 . Sheep of the resistant breed developed and sustained a Th2 response through increasing and maintaining IgG and blood eosinophil levels. Meanwhile sheep of the susceptible breed showed changes in the response starting with an initial increase in IgG and blood eosinophils (Th2) but a later reduction in both, which suggests a switch to a Th1 response 34 . An earlier study suggested that in the relative absence of Th1 type secretions (i.e. cytokines), the Th2 cells secrete cytokines that promote mastocytosis, eosinophilia and the production of IgE and IgG1 35 . Gulf coast native (resistant) lambs showed a significantly higher expression of IL-4 mRNA (Th2) on day 10 post exposure to the nematode compared to Suffolk lambs (susceptible). On the other hand, the expression of IFN-γ mRNA and IL-10 (Th1 and regulatory T) on days 7, 10 and 14 post exposure was higher in Suffolk lambs (susceptible) compared with native lambs 25 .
This confirms that if T helper cells of the Th2 type gain ascendancy after GIN infection, then a protective immune response ensues. In contrast, if an inappropriate Th1 type response predominates, effective resistance is unlikely to develop. The Th1 type response for GIN infection is most likely associated with susceptibility, while a Th2 type response is associated with resistant phenotypes in sheep 25,33,34 . As Th2-associated cytokines target plasma cells to produce nematode-specific antibodies and generate protective immune responses 24 , we focus here on the immunoglobulin response against GIN.
Immunoglobulin response in sheep (IgA, IgE, IgG and IgM). The association between different immunoglobulin isotypes, including IgA, IgE, IgG and IgM and GIN resistance has been widely studied in sheep (Table 1). Studies which measured at least one immunoglobulin parameter during GIN infection (Table 1) differed in sheep breed used, type of breed (resistant or susceptible), age of animals in the experiment, immunological status (naïve or non-naïve), infected parasite genus and infection type (natural, artificial with single dose or artificial with trickle doses) which makes the comparison between them rather complex.
When comparing publications which measured different immunoglobulins levels against different larva stages during GIN infection in sheep it was found that the majority of the studies examined the presence of L3 antigen-specific immunoglobulins (Fig. 3). The third stage larva (L3) represents the stage with the first contact of the gastrointestinal nematodes with the host immune system. It can also be seen from the figure that IgA was the most commonly investigated immunoglobulin isoform in sheep. total antibody response. Results from Douch et al. 36 and Gauly et al. 37 suggested that resistant sheep have higher total antibody levels and that the antibody level could be used in selection for resistance. Romney rams, selected based on their serum antibody levels at an age of 6 months to reduce FEC, underwent a natural parasite challenge, and it was predicted that the genetic gain was 51 to 67% of the genetic gain achieved when FEC was directly used as a selection trait 36 . The total antibody level in both 4 and 5 months old Rhön (resistant) sheep were significantly higher compared to Merinoland (susceptible) lambs following experimental infection with H. contortus 37 .
The correlation between Trichostrongylus colubriformis-L3 or -adult total antibody and FEC (−0.62 and −0.55) or nematode burden (−0.56 and −0.63) was high in Romney progeny selected for low and high FEC following a natural GIN challenge 38 . igA response. It has been suggested that local immune effectors expressed in the abomasal mucosa, particularly IgA, play an important role in immunity acquired both naturally and experimentally 39 . In sheep, secretory and plasma IgA derive predominantly from the gastrointestinal tract 24 . The correlation between gastric mucus IgA and peripheral IgA is positive and highly significant, ranging from 0.618 to 0.779 40,41 . Several studies reported an increase of the IgA response after GIN infection, higher levels were recorded for resistant breeds 16,39,42 .
The level of IgA against the CarLA antigen (a carbohydrate larval surface antigen expressed on the L3 of all trichostrongylid nematode species) has been suggested to be a suitable means to measure the level of resistance to GIN 43,44 . Different studies found that CarLA is a target antigen for host IgA which binds to the larval surface antigen and prevents larvae from establishing at their preferred sites in the intestinal epithelial folds [43][44][45] . In this context, a different L3-specific surface antigen (CarLA) was detected from Trichostrongylus colubriformis, Haemonchus contortus and Ostertagia circumcincta with similar molecular weight (35-kDa), and from Cooperia curticei and Nematodirus spathiger with a different molecular weight (22-kDa on blots of L3 extracts) 43 . IgA in saliva had a negative genetic correlation with FEC (r = −0.5) and animals with high levels of anti-CarLA IgA have typically 20-30% lower FEC than animals with low or undetectable titers 45 . A simple way to use these results for the selection of animals resistant to parasite infection could be to measure anti-CarLA IgA in saliva.
The faecal egg output in St. Croix hair-type sheep (resistant breed) can rapidly reduce in response to H. contortus artificial infection following a 45-day breeding season, which was accompanied with higher levels of Continued www.nature.com/scientificreports www.nature.com/scientificreports/ circulating antigen-specific antibody IgA compared to a susceptible composite line of wool-type sheep (50% Dorset, 25% Finnsheep and 25% Rambouillet breeding) 16 . The increase in anti-T. circumcincta IgA antibody and eosinophil concentrations were associated with an increase in the frequency of early L4 42 . Also Ellis et al. 39 found a correlation (r = 0.534, P = 0.007) between L3 antigen-specific IgA levels in efferent gastric lymph and the percentage of inhibited L4s. In addition, a negative correlation (r = −0.565, P = 0.005) between total T. circumcincta burden measured at necropsy and L3 antigen-specific IgA levels in efferent gastric lymph was reported..
A negative association was reported between IgA activity against L4 and both egg counts and worm length when studying resistance to T. circumcincta in Scottish Blackface and Churra sheep 40,46 . Gastric mucus IgA against L4 somatic antigen was highly and negatively correlated (r = −0.71, P < 0.01) with the number of eggs per female in utero and also with the length of adult females (r = −0.552, P < 0.01). Results for IgA against somatic antigen from the adult stage were similar to those with activity against L4, but correlations were somewhat weaker 40 . Negative genetic correlations were found between IgA and FEC, worm length, worm fecundity and worm burden (r = −0.78, −0.53 and −0.62, −0.36, respectively) in Scottish Blackface lambs exposed to natural mixed infection 47 . It was suggested that parasite development such as worm growth and fecundity in sheep can be regulated via an IgA response, possibly in conjunction with eosinophils 11,41 .
In this paper, we have collected data from previous studies that measured correlations between FEC and blood IgA levels against different larval stages. We also calculated correlations using raw data for FEC and blood IgA level from these studies before re-analyzing all the data. Figure 4 shows the correlations between FEC and blood IgA levels against different larval stages from different studies. The overall correlation between IgA in blood and FEC was negative (r = −0.36, 95% CI = −0.46, −0.26). Correlations between FEC and blood IgA activity against L3 or adult were −0.39 (95% CI = −0.51, −0.28) and −0.47 (95% CI = −0.85, −0.09), respectively. Only one study measured the correlation between FEC and blood IgA activity against L4, L5 or ESP. Although IgA is produced locally at the mucosal surfaces and serum IgA is derived from the gastrointestinal tract 24 , few studies measured mucosal IgA, possibly because of the difficulties in sampling mucus. Serum IgA is easier to measure and  www.nature.com/scientificreports www.nature.com/scientificreports/ highly correlated with mucosal IgA 40,41 . We were not able to re-analyze mucosal IgA. Although CarLA saliva IgA antibody test is currently being marketed (CARLA ® SALIVA TEST) as a powerful new tool for measuring parasite immunity in sheep (https://www.agresearch.co.nz/doing-business/products-and-services/carla-saliva-test/), no correlation was found between saliva IgA and serum or mucosal IgA 40 . However, antibodies in saliva may be binding directly to ingested L3. ige response. The degranulation of mucosal mast cells is induced by the cross-linking of IgE on their surface.
Following reinfection, a negative correlation between the concentrations of globule leucocytes (intraepithelial mast cells) and the T. circumcincta worm burden have been shown in sheep 48 . Indeed, the IgE antibody response mainly directed against L3 antigens, would be more prominent in previously infected sheep 49 .
In Romney sheep selected for almost two decades for high or low FEC following T. colubriformis infection, sheep from the low FEC line had higher total IgE (97 to 103%) and T. colubriformis-specific IgE (59 to 98%) compared to sheep from the high FEC line 50,51 . Similarly, after 7 weeks of grazing on a contaminated pasture, low FEC (based on the accumulated weekly measure) lambs (Greyface × Suffolk) had significantly higher systemic levels of IgE anti-HMWTc (a major high molecular weight complex allergen from T. circumcincta L3) than high FEC lambs 52 . In addition, after 13 and 18 weeks on pasture, serum IgE anti-HMWTc assays demonstrated an even greater difference between low and high FEC lambs. We conclude from these studies that high levels of IgE are associated with low levels of FEC in sheep. Pettit et al. 53 showed that during the first and second year on pasture a higher concentration of blood circulating IgE-bearing cells was associated with a lower FEC of Scottish Blackface lambs.
Other studies comparing resistant and susceptible sheep breeds confirm the role of IgE as an important effector of the immune response to nematode infection in sheep 13,33 . Sayers et al. 13 reported that the most notable difference of mucosal antibody isotype when comparing Texel (GIN resistant breed) and Suffolk (GIN susceptible breed) is IgE, after natural GIN infection. Mucosal IgE in Texel was significantly negative correlated with FEC (r = −0.48) and abomasum worm burden (r = −0.47). Also, the total IgE content in 5-6 month old Gulf Coast Native lambs increased significantly and was highest 7-14 days after artificial infections with H. contortus compared to the level in Suffolk lambs, which was confirmed in a natural infection experiment with the highest level of IgE at 14-42 days post infection 33 . igG response. A review of studies investigating IgG and FEC after parasite infection in different sheep breeds 12,34,38,54 suggested that IgG (especially IgG1) increases following infection and is associated with increased resistance to some GIN infections in sheep. Romney lambs had elevated levels of antibodies to T. colubriformis L3 excretory/secretory antigens which consisted predominantly of IgG1, reaching peak levels between days 42 and 77 post infection 54 . Bisset et al. (1996) found that the T. colubriformis-specific IgG1 response against both L3 and adult secretory/excretory antigens, was significantly higher in Romney lambs bred for low FEC compared to those selected for high FEC. IgG1 was negatively correlated with FEC (r = −0.60 and −0.48) and strongyle burden (r = −0.53 and −0.57) for both L3 and adult worm secretory/excretory antigens 38 .
Barbados Blackbelly lambs, which are resistant to H. contortus, had higher amounts of IgG anti-L3 between weeks 9 and 15 after infection. On the other hand, in the 9th week, sheep from the less resistant Columbia breed had lower IgG levels compared to the uninfected group 34 . An increase in IgG anti-L3 and IgG anti-adult H. contortus was also observed in Santa Ines ewes (a resistant breed) and its cross with Dorper, Ile de France, Suffolk www.nature.com/scientificreports www.nature.com/scientificreports/ and Texel, as a consequence of exposure to GIN larvae on pasture 55 . Santa Ines lambs had, after artificial infection with T. colubriformis, significantly higher specific serum levels of IgG anti-L3 and IgG anti-adult than the uninfected control group from week 4 (P < 0.05) to 13 (P < 0.01) post-infection 56 . A significant negative correlation was reported between H. contortus burden and IgG against L3 (r = −0.72), IgA against L5 (r = −0.61) and mast cells (r = −0.73) in Santa Ines male lambs, but not in Ile de France, which may be the reason that Santa Ines sheep had a lower FEC and worm burden than Ile de France sheep 12 . In a serial infection trial with Santa Ines crossbred lambs, H. placei infection induced high levels of IgG anti-L3 and IgG anti-adult compared with a control group, while animals serially infected with H. contortus induced high levels of IgG anti-adult but not IgG anti-L3 compared with the uninfected control group 57 . igM response. IgM has an important role in the immune response as it represents the first class of antibodies produced following the initial exposure to a foreign antigen 58 . However, it is not normally present in the gastrointestinal mucus 59 . Only a few studies have investigated the IgM response in sheep during GIN infection and the findings from two studies did not suggest a major effect of IgM in sheep 33,38 . In selected Romney ram progeny, bred for genetic divergence in FEC, the correlations between IgM and both FEC or GIN burden were weak following an extended period of exposure to naturally contaminated pasture 38 . IgM did not differ between Gulf Coast Native (a resistant breed) and Suffolk lambs (a susceptible breed) in both artificial infections with H. contortus and natural infection experiments 25 .
In contrast, GIN challenge in Romney lambs gave rise to elevated levels of IgM anti-L3 after artificial infection with T. colubriformis L3 54 . A moderate increase of IgM serum antibody levels against both larval and adult antigens was found in Texel lambs in both primary infections and challenge infections with H. contortus 60 23 reported no consistent differences in IgA, IgE or IgG levels between Scottish Cashmere goats selected for low FEC (generations F2 to F9) and control lines (unselected) after artificial infection with T. circumcincta and during the grazing season. On the other hand, IgA anti-ESP, IgA anti-L3 and IgE anti-L3 were genetically correlated with FEC (0.84 ± 0.13, 0.72 ± 0.18 and −0.32 ± 0.08, respectively), while they did not find any phenotypic correlation between them 65 . A phenotypic correlation is the correlation between records of two traits on the same animal. Genetic correlation, traditionally calculated from pedigree data, is a measure of the genetic relationship between two traits 66 . A high genetic correlation between two traits is generally supported by genes that are usually co-inherited. The phenotypic correlation estimate the correlation between two traits and is depend both on additive genetic and of environmental effects 66 . Recently, it has been suggested that the lack of functionality of the immune response mediated by IgA and eosinophil against natural nematode infection in Boer goats would be due to a dysfunctional transmembrane domain of the high affinity IgA receptor 67 . Altogether these results indicate that the humoral response against GIN infection is less effective in goats than sheep and does not probably play a major role in resistance to nematode infections in goats.
Genomic studies on Gin resistance. Identifying genetic markers of resistance and/or susceptibility could improve the efficiency of breeding programs. The first approach used to search for genetic markers associated with host resistance was a study of associations between variants of genes related to the host immune response and phenotypic resistant traits 68 . The genetic markers that have been most frequently associated with nematode resistant are those from the major histocompatibility complex (MHC) region on Ovis aries chromosome 20 [69][70][71][72][73][74] . MHC genes are highly polymorphic 69,74 and play important roles in presenting processed antigens to host T lymphocytes, causing T cell activation. The second most frequently region identified in studies for resistance to GIN infection is the interferon ɣ (IFNG) gene on O. aries chromosome 3 75,76 . The majority of studies of quantitative trait loci (QTL) focused on the association/ linkage between genomic regions and FEC. Only few studies examined the association between genomic regions and immunoglobulins level during GIN infection. Table 2 shows the genomic regions that were reported to be linked or associated with immunoglobulin-mediated resistance to GIN infection. Similarly to phenotypic studies, the majority of genomic studies focussed on the IgA response during GIN infection 73,75,[77][78][79] , two studies considered the IgG response 80,81 and only one study examined the IgE response 80 . factors that impact animal response to Gin infection. A number of factors impact the immune response to parasite infection. We focus here on the factors of the animal itself, such as genetic differences, age of the animal and immunological experience, infection period, or type of infection (i.e. natural or experimental).

Age of the animal and immunological experience. Our review on immune responses to parasite
infection in sheep suggests that the age of the animal influences the immune response to GIN infection. Even though the traits investigated were different, three studies have identified age-related differences in the immune response [82][83][84] . High eosinophil counts were significantly correlated with low FEC in naturally infected with T. circumcincta Scottish Blackface lambs which were at least 3 months of age, where the correlations were −0.33, −0.14 and −0.24 after 3, 4 and 5 months of age, respectively 83 . Also the correlations between IgE activity and FEC in Texel lambs following 4 weeks of natural mixed nematode infection were only significant at 5 and 6 months, Scientific RepoRtS | (2020) 10:7765 | https://doi.org/10.1038/s41598-020-64775-x www.nature.com/scientificreports www.nature.com/scientificreports/ while it was not significant at 7 months of age when natural exposure would be declining due to the onset of late autumn and winter 82 . We found that many studies discussed their results without taking the age structure of the chosen experimental cohort as an important factor into account. However, from the studies using age different cohorts, we conclude that differences in immune reactivity might be partly explained by age.
Immunological experience has a huge impact on the response to parasite infection and is partly confounded with animal age. Immune response was reported to be low and delayed in primary-infected lambs while it is higher and rapid in previously-infected animals 60,85,86 . The majority of the studies (85%, n = 35) focus on lamb immune response as model for nematode infection and half of these studies (n = 18) used non-naïve lambs. It has been shown that genetic variation in resistance to GIN is associated with the development of an acquired immune response, which explains why the pathophysiological impact of these parasitic infections is more important in growing lambs compared to mature sheep 87 . Also, in lambs genetic variation in FEC is not correlated with genetic variation in the total number of worms but rather to female worms length and consequently their fecundity 88 . In contrast to lambs, mature sheep may have the ability to limit both fecundity and worm numbers 87 . Hence the immune mechanisms could differ. infection period. Days post infection (d.p.i) represents also a very important factor that may affect the results obtained. The majority of the research measured host immune response within 0 to 6 weeks post infection, while other studies indicate that they did not find significant immune response until about 9 weeks post infection 34,37,42 .
IgG level at 60 d.p.i was significantly correlated with worm burden (r = 0.235 to 0.247) and total antibody level was significantly correlated with worm length (r = 0.316) and FEC (r = −0.148), whereas the correlations were  37 . A similar difference was also observed in Rhön sheep, the total antibody value was significantly correlated with worm burden (r = 0.372 to 0.378) at 60 d.p.i but not at 30 d.p.i 37 . Similarly, in Blackface lambs exposed continuously to infection of T. circumcincta, anti-T. circumcincta IgA levels were inversely correlated with FEC and increased with time from r = −0.17 (NS) at 14 d.p.i to r = −0.44 at 84 d.p.i (P < 0.001) 42 . Moreover, the difference of the anti-T. circumcincta IgA level between resistant and susceptible sheep diverged with time and was significant from 56 d.p.i. onwards 42 .
Comparing immune response to GIN in resistant and susceptible sheep breeds, sheep of both Barbados Blackbelly (resistant) and Columbia (susceptible) breeds showed a significant increase in blood eosinophils from 1 to 9 weeks post infection, these levels decreased suddenly thereafter in Columbia lambs 34 . A similar pattern was observed in the same study for IgG anti-L3 levels, Barbados Blackbelly lambs had a higher amount of IgG anti-L3 between weeks 9 and 15 after infection with a positive significant regression (0.79), while the regression in Columbian lambs was negative and not significant (−0.59) at week 9. These results suggest that at the beginning of the infection the immune response may not differ a lot between resistant and susceptible animals but differences become more apparent 9 weeks after infection.
An interlaced issue is the effect of age and infection period, significantly greater nematode-specific serum antibody activities were reported in Texel compared to Suffolk sheep for all isotypes (IgG1, IgG2, IgA and IgE) at 14 and 17 weeks of age with increasing divergence between the breeds as age increased 13 . Meanwhile these differences could be explained as the effect of infection period as in this experiment the lambs were exposed to natural infection after staying on pasture within 1 to 3 days of birth, so these differences in response could be due to age or infection period or a combination of both.  Table 3. These studies suggest a role for L3 antigen-specific IgA, IgE and IgG responses in resistance to GIN in sheep. Also the role of L4 antigen-specific IgA for the resistance to GIN in sheep appears more important than the role of IgA against L3.

Infection type (natural/artificial
Three major mechanisms of immunity to nematodes have been described in sheep, regulation of the establishment rate of infective larvae, suppressed nematode growth and thus fecundity, and the expulsion of adult worms; a combination of these mechanisms is possible 24 . Figure 5 shows the gastrointestinal nematode life cycle and summarizes the suggested role of different immunoglobulins in the three major mechanisms of immunity to nematodes. Reduced parasite establishment and survival is associated with IgE activity mainly against incoming third stage larvae (L3) in concert with mast cells as cross-linking of IgE on the mast cell surface leading to mast cell degranulation 82,89 with more prominent response in previously infected animals 49 . Reduced parasite growth and fecundity is correlated with increased local IgA activity against fourth stage larvae 59,89,90 . Increased number of inhibited larvae is correlated with IgG1 activity against the third stage larvae 54,60 beside IgA activity against the third and fourth stage larvae 89,90 . conclusions The selection of animals with a high immune response to GIN infection is a promising method of reducing the negative impact of these infections on grazing small ruminants. We have highlighted factors that differ across studies and affect the immune response to GIN infection. Indeed, it is essential that future studies take into account and mention the age of the animals, the infection experience and the type of infection (i.e. single vs trickle). One another important point is the need to normalize the measurements of immunoglobulin concentrations. The use of different units and/or optical density to measure immunoglobulin levels at different time post-infection produces results that are incomparable between studies. One way to overcome this issue is to use the change ratio between day 0 or an uninfected group and different time post-infection or to develop methods for the quantification of the immunoglobulins with standard curves. This effort for standardization should potentially allow to take advantage of the research results produced by implementing breeding programs for higher resistance to GIN infection.

Material and Methods
The study methodology followed the guideline of "Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement" 91 24,28,29,89,92 . Studies were closely evaluated and selected for inclusion if they measured at least one immunoglobulin type and the data were extractable. The literature review revealed 41 studies which measured at least one immunoglobulin: 35 focused on lamb immune response (18 used non-naïve lambs) and 7 on yearlings. Overall, the methodological quality of included studies was evaluated qualitatively with an 9-point scoring system based on SYRCLE's risk of bias tool 93 : (1) Statement on ethical approval or guidelines followed for using animals in the study clearly mentioned, (2) Adequate allocation sequence generation, (3) similar groups at baseline, (4) Allocation concealment, (5) Random housing, (6) Random outcome assessment, (7) Outcome assessment blinded and (8) incomplete outcome adequately addressed (Fig. 1).
Information extracted from each study included nematode genus, host (sheep or goat), infected breed, host classification (resistant or susceptible), host sex, host age (months), age at weaning (weeks), infection period (days post infection), sample tissue, infection type (artificial or natural), number of animals per group, immunological state (naïve or non-naïve), total number of infected larvae, larval details and infection method for artificial infection (single or trickle). Measured traits of interest included both parasitological and immunological parameters. If data were provided in graphical form, traits means were extracted using WebPlotDigitizer (version 3.8) 93 . These summary measures beside the study information previously mentioned were entered into an electronic spreadsheet in Microsoft Excel and a dataset was built. The Metafor package in R (version 3.5.1) was used to analyze the www.nature.com/scientificreports www.nature.com/scientificreports/ correlation coefficient between FEC and immunoglobulins. A REML model was used in which the effect size was calculated according to the number of animals used in each study with 95% confidence interval. The possibility of publication bias across studies reporting correlation coefficients was assessed by a visual analysis of a funnel plot of the standard error by the Fischer's Z transformed correlation coefficients and formally by using Egger weighted regression test (Metafor package in R version 3.5.1).