Experimental population: Animals (n=457) of an experimental F2 population based on reciprocal crossbreeding of Duroc and Berlin Miniature Pig (DUMI)  were immunized with Mycoplasma hyopneumoniae (Mh), Aujeszky disease (ADV), and PRRS (PRRSV) vaccines at 6, 14 and 16 weeks of age, respectively. Blood samples were taken immediately prior to immunisation (day 0) and at day 4 and 10 after Mh and ADV vaccination, but only at day 10 days after PRRSV vaccination. Hemolytic complement activities in the classical (CH50) and alternative pathways (AH50), C3c (C3c) and haptoglobin (HP) acute phase protein levels were measured . The DUMI animals were genotyped at 88 loci (72 microsatellites, 16 biallelic markers) covering the porcine autosomes. Linkage analysis was performed using CRI MAP . The QTL analysis was done using QTLexpress  under the line cross and the half sib model and after adjustment of phenotypic data for systematic effects by mixed models with and without repeated measures statement. Association and linkage of SNPs detected in the cDNA sequence of C6, C7, C8A, C8B, and C9 were tested by means of analysis of variance using repeated measures mixed models and family based association test (FBAT) testing the null hypothesis of no association in the presence of linkage (option `eÂ´, computing the test statistic using the empirical variance ).
Commercial herds: Animals (n=311) of commercial herds of the breeds German Landrace, Large White and their crossbreeds (F1) and Piétrain × F1 were kept and performance tested at the Pig performance test station Jürgenstorf and slaughtered at the experimental abattoir of the Leibniz Institute of Farm Animals Biology (FBN). Liver tissue was immediately sampled after exsanguination for RNA isolation using Tri-Reagent (Sigma-Aldrich, Taufkirchen, Germany) and NucleoSpin RNA II kit (Macherey-Nagel, Düren, Germany) including DNase treatment. Lung lesions were recorded during veterinary inspection and relative abundance of transcripts of C6, C7, C8A, C8B, C8G, and C9 was determined in liver tissue. Therefore, real time PCR was performed using the LightCycler 480 system (Roche, Mannheim, Germany) with cDNA synthesized from 1µg of total RNA using random and oligo d(T) 13VN primers, Superscript III reverse transcriptase (Invitrogen, Karlsruhe, Germany), and the LightCycler 480 SYBR Green I Master kit (Roche). Standard curves were generated by amplifying serial dilutions of specific PCR products. Normalization of variation in RT-PCR efficiency and initial RNA input was performed using the RPL32 and ALB genes as internal standards by dividing the calculated mRNA copy numbers by a mean normalization factor derived from the expression of the reference genes (relative abundance). Alternatively, ΔCT was calculated by subtracting the mean ct-values of the reference genes from the ct-values obtained for single TCC.
Association of SNPs detected in the promoters or cDNAs of C6, C7, C8A, and C9 with relative transcript abundance were analyzed by mixed model analysis of variance (SAS, Proc mixed) taking into account the fixed effects of breed, farm where the piglets were born and raised until seven weeks of age when they were transported to the performance test station, genotype of the respective gene and the random effect of father. The relationship between the respective genotypes and the prevalence of lung lesions was analyzed by case control analysis using JMP genomics. The relationship between the transcript abundance and lung lesions was evaluated by analysis of variance taking breed and lung lesions (affected, non-affected) as fixed effects on the dependent variable of relative transcript abundance.