Identification of potential modifier loci To determine the contributions of B6 and 129 alleles to the mixed genetic background, and to identify potential modifier regions, tail DNA was analyzed by PCR for markers of polymorphisms between these two mouse strains. Mice congenic on the B6 background, which were derived from mice originally on a B6/129 mixed background, were also analyzed to confirm they are congenic B6. Twelve CF mice on the B6 background averaged 99.5% B6 alleles. One of the twelve mice had alleles from both strains at chr.9, 40 cM. All twelve mice had both B6 and 129 markers on chromosome 6, 1 cM which is probably due to the targeted Cftr gene which is at chr.6, 3.1 cM [36]. Three CF mice on the mixed background were initially analyzed and were found to be 95% B6, 5% 129. A second group of 8 CF mice was analyzed and the markers used were refined to focus on the chromosomal regions showing variations from the B6 strain. The differences found from the two analyses are combined in Table 3. The differences in the mixed strain CF mice were at chr.1, 92 cM (9 of 11 mice had both B6 and 129 alleles); chr.9, 9 cM (all 11 mice had two 129 alleles); and chr.10, 65 cM (10 mice had both B6 and 129 alleles). Table 3 Genome scanning analysis of CF mice on the mixed background. Chromosome cM B6/B6 129/129 B6/129 Number of Mice 1 92 2 0 9 11 1 102 3 2 6 11 2 7 2 0 1 3 9 9 0 11 0 11 9 20 8 1 1 11* 9 33 5 0 3 8 9 40 1 1 1 3 10 40 6 0 2 8 10 65 0 1 10 11 11 35 5 0 3 8 11 43 6 0 5 11 12 19 6 0 2 8 The table shows the distribution of B6 and 129 alleles in mice from the two genetic backgrounds. An initial analysis was performed on 3 samples and a second analysis on 8 more samples. Only the markers that were not uniformly B6 alleles are shown. Some of the markers were refined based on the initial analysis, so not all markers were used for all 11 mice. (*) Eleven samples were analyzed but one sample was ambiguous. Because the spacing of markers used was about 12 cM, genes within 75% of this interval on either side of the markers were looked at for potential relevance to the milder CF intestinal phenotype. None of the known chloride channels that might substitute for the missing CFTR are in the regions of the three chromosomes associated with the milder phenotype. There are several potassium channel genes in the identified regions that potentially could affect electrolyte and fluid transport: Kcnj9 (chr.1, 94.2 cM), Kcnj10 (chr.1, 93.5 cM), Kcnj5 (chr.9, 11 cM), and Kcnc2 (chr.10, 62 cM). All gene names are from the Mouse Genome Informatics website . Inflammation is a hallmark of CF, and whether there is an inherent defect in CF that predisposes to excessive inflammation is controversial. Several genes involved in inflammation and the immune system are located in the regions of the markers identified: TNF superfamily members Tnfsf4, 6, and 8 (chr.1, 84.9–85 cM) which are involved in T cell activation [37,38]; three selectin genes (Sele, Sell, Selp, chr.1, 86.6 cM) which are involved in immune cell infiltration into inflamed tissues [39]; several members of immune cell surface proteins of the Slam family (slamf1, 2, 5, 6, and 9; chr.1, 89.5–93.3 cM) [40]; the chemokine gene Xcl1 (chr.1, 87 cM) which is expressed by mast cells and recruits lymphocytes [41]; several immunoglobulin Fc receptor genes (Fcrl3, Fcgr2b, and Fcgr3 at chr.1, 92.3 cM; Fcer1g at chr.1, 93.3 cM; Fcer1a at chr.1, 94.2 cM); the flagellin receptor Tlr5 (chr.1, 98 cM); Mmp3 (chr.9, 1 cM) which recruits CD4+ lymphocytes [42]; Mmp7 (chr.9, 1 cM) which activates Paneth cell-derived cryptdins (α-defensins) [43]; Icam1 (chr.9, 7 cM) which is involved in lymphocyte infiltration into inflamed tissues [44]; Kitl (chr.10, 57 cM) which is also known as stem cell factor, and is crucial for mast cell differentiation [45]; Im5 (chr.10, 65 cM) which is involved in antibody-responsiveness [46]; Lyzs (chr.10, 66 cM) which is a Paneth cell product that digests cell walls of bacteria [47]; Ifng (chr.10, 67 cM) which is an important inflammatory signal in CF as well as other conditions [48]; Il22 (chr.10, 67 cM), a member of the anti-inflammatory IL-10 interleukin family [49]; and the Stat2 and 6 genes (chr.10, 70 cM) which are important components of intracellular signaling pathways [50]. Also near the identified markers are a number of QTL associated with body weight: Cfbw1, CF mouse body weight at chr.1, 85 cM; Obq9, obesity 9 at chr.1, 88 cM; Bw8q1, body weight 8 at chr.1, 100 cM; Lbm6, lean body mass 6 at chr.9, 7.7 cM; Bwtq4, body weight 4 at chr.9, 8 cM; Bgeq8, body growth early 8 at chr.10, 57 cM; and Pbwg5, postnatal body weight growth 5 at chr.10, 68 cM. Clearly, there are numerous genes in the three regions identified in this study. Because the CF mouse intestinal phenotype is characterized by an innate type immune response, with increases in mast cells and neutrophils, the genes that affect these cells are of special interest. The Kitl gene is crucial for differentiation of mast cells, and CF mice on the mixed background have much fewer mature mast cells than on the B6 background as revealed by less expression of Mcpt2. Similarly, for neutrophils the selectins and Icam1 are of interest, as these proteins are required for extravasation of neutrophils from the circulation into the inflamed tissue. Altered immune responses may also relate to excessive mucus accumulation in the CF intestine. It is unclear why mucus accumulates to high levels in CF tissues. In part it may be due to reduced fluid secretion and a more acidic environment in the lumina of affected organs. However, there is also evidence for hypersecretion of mucus in CF [12], and it is likely that effector molecules released by mast cells and neutrophils (histamine, proteases, prostaglandins) have an important role in stimulating mucus secretion.