Lyme borreliosis (LB) is the most common vector-borne disease in the Northern Hemisphere. The infection is caused by spirochetes belonging to the Borrelia burgdorferi sensu lato complex, and it is transmitted to humans by ticks. LB is associated with several clinical manifestations, of which erythema migrans (EM) and neuroborreliosis (NB) are the most common inEurope. The course of the disease is usually benign, but can vary between individuals. The underlying pathogenic mechanisms are not fully understood, but the prognosis is probably determined by a complex interplay between the bacteria and the host’s immune response. Previous studies have indicated that a strong initial T helper (Th) 1-response followed by a Th2 response is beneficial for the clinical outcome in LB. The aims of this thesis were to follow the incidence of NB inJönköping County,Sweden, over time, to search for clinical and laboratory markers associated with the risk of developing long-lasting post-treatment symptoms, and to explore the role of the complement system as well as the relative balance between Th-associated cytokine/chemokine responses in LB. The number of NB cases, diagnosed by cerebrospinal fluid (CSF) analysis, increased from 5 to 10/100,000 inhabitants/year in Jönköping County during 2000-2005. Post-treatment symptoms persisting more than 6 months occurred in 13 %, and were associated with higher age, longer-lasting symptoms prior to treatment, higher levels of Borrelia-specific IgG in CSF, and reported symptoms of radiculitis. Facial palsy, headache and fever were frequent manifestations in children, whereas unspecific muscle and joint pain were the most commonly reported symptoms in older patients. Complement activation occurred both locally in the skin in EM and in CSF of NB patients. However, no activation could be detected in blood in NB patients. Elevated levels of C1q, C4 and C3a in CSF, along with correlation between C1q and C3a levels, suggest complement activation via the classical pathway locally in the central nervous system in NB. In vitro experiments with two clinical Borrelia isolates revealed that B. garinii LU59 induced higher complement activation in human plasma compared to B. afzelii K78 that recruited more of complement regulator factor H. To elucidate the role of complement in the phagocytosis process, experiments were performed using whole blood from healthy donors incubated with fluorescence-labelled spirochetes and different complement inhibitors. The results illustrated a central role of complement for phagocytosis of Borrelia spirochetes. We also studied the relative contribution of different Th-associated cytokines/chemokine responses in NB. The results support the notion that early NB is dominated by a Th1 response, eventually accompanied by a Th2 response. IL-17A was increased in CSF in half of the patients with confirmed NB, suggesting a hitherto unknown role of Th17 in NB. In conclusion, the risk of developing long-lasting post-treatment symptoms tend to increase mainly with age and duration of symptoms prior to treatment in NB. The complement system seems to play an important role in host defence to recognize and kill Borrelia spirochetes. However, complement activation in inappropriate sites or to an excessive degree may cause tissue damage, and therefore, the role of complement in relation to disease course needs to be studied further. Likewise, the role of Th17 in LB pathogenesis and host defence should be further evaluated in prospective studies.

Lymphocytic infiltration is a feature of basal breast cancer tumors. CXCL10 is a chemokine that was found expressed higher in basal tumor RNA compared to estrogen receptor positive tumor RNA. Both CXCL10 and its receptor CXCR3 were expressed in familial breast cancer tissues, a proportion of which have a basal phenotype. CXCL10 expression was associated with lymphocytic infiltration, and with CXCR3 expression. CXCL10 ligand and receptor were overexpressed individually or together in the human MCF7 cell line. Recombinant human CXCL10 was found to dose dependently decrease cell proliferation. CXCR3 and CXCL10-CXCR3 expressing cells had the potential for increased migration independent of CXCL10 concentration. Co-expression of both genes increased proMMP-2 levels, and conditioned media from one of these clones chemoattracted more of the CXCR3 clones, CXCL10-CXCR3 clones, and CD4+ T-lymphocytes. CXCL10 neutralization suggested that CXCL10 could play a role in this chemoattraction, though it is likely not the only factor involved.

Lymphocytic infiltration is a feature of basal breast cancer tumors. CXCL10 is a chemokine that was found expressed higher in basal tumor RNA compared to estrogen receptor positive tumor RNA. Both CXCL10 and its receptor CXCR3 were expressed in familial breast cancer tissues, a proportion of which have a basal phenotype. CXCL10 expression was associated with lymphocytic infiltration, and with CXCR3 expression. CXCL10 ligand and receptor were overexpressed individually or together in the human MCF7 cell line. Recombinant human CXCL10 was found to dose dependently decrease cell proliferation. CXCR3 and CXCL10-CXCR3 expressing cells had the potential for increased migration independent of CXCL10 concentration. Co-expression of both genes increased proMMP-2 levels, and conditioned media from one of these clones chemoattracted more of the CXCR3 clones, CXCL10-CXCR3 clones, and CD4+ T-lymphocytes. CXCL10 neutralization suggested that CXCL10 could play a role in this chemoattraction, though it is likely not the only factor involved.

Kawasaki disease (KD) is the most common cause of paediatric acquired heart disease, which may be attributed to the combined effects of infection, immunological response, and genetic susceptibility. The most severe complication in KD is acute coronary artery lesions (CALs), including myocardial infarction and coronary artery aneurysms. Mounting evidence indicates that adhesion molecules and chemokines play an important role in inflammation and cardiovascular disease on basis of pathogenesis. Thus, this study aimed to investigate the association of seven single nucleotide polymorphisms (SNPs) of adhesion molecules and chemokines (P-selectin 290G>A, PSGL-1 62G>A, MCP-1 -2518A>G, SDF-1 -801G>A, PECAM-1 L125V, PECAM-1 S563N and PECAM-1 R670G) with the risk of KD, sequelae of CALs and initial intravenous immunoglobulin (IVIG) treatment failure. A total of 301 KD children (185 without acute and chronic CALs, 81 with acute but without chronic CALs, and 33 with acute and chronic CALs) and 246 sex-matched healthy controls were recruited in the case-control study. In addition, 166 cases from the above KD children and 332 parents were recruited to carry out case-parent trio study. We found that PECAM-1 3 SNPs polymorphisms were not associated with above several risks, except for CALs in chronic stage. As compared with non-Leu-Ser-Arg haplotype, Leu-Ser-Arg haplotype was associated with a significant increased risk for CALs in the chronic stage (AOR 2.50, 95% CI 1.05-6.00, P=0.039). Analyses based on the diplotypes of PECAM-1 also showed that Leu-Ser-Arg allele had a significant increased risk of CALs in chronic stage in dominant manner (AOR 2.98, 95% CI 1.15-7.72, P=0.024). In addition, carriers of Leu-Ser-Arg allele had significant increased counts of platelet (¡Ñ1000/Cumm) (672.6¡Ó207.6 versus 563.1¡Ó196.8; P=0.027) within 10 days of diagnosis of KD. Moreover, we also found a significant correlation between each SNP and polymorphonuclear neutrophil counts by genotype analysis. As for other genes, there were no markedly different outcomes regardless of the risk of KD, sequelae of CALs or initial IVIG treatment failure. In conclusion, the haplotype Leu-Ser-Arg of PECAM-1 is a genetic marker of susceptibility to sequelae of chronic CALs for KD patients. However, the role of PECAM-1 SNPs in CALs formation in the chronic stage in KD patients still needs further evaluation.

Zugl.: Giessen, Univ., Diss., 2007

Human colonic epithelial cell lines (T84, Caco2 and HT-29) were used to address the question of whether intestinal epithelial cells can detect and respond to activated complement via the anaphylatoxin receptors, considering the gut is host to large numbers of bacteria. All cell lines possess C3aR, C5aR and C5L2. Confocal microscopy confirmed that cells express apical C5aR and C5L2. C3a and C5a up-regulated CXCL8 and CXCL10 mRNA but not secreted protein levels within 48 hours. Protein levels were not increased using simultaneous treatment with subthreshold concentrations of LPS or TNF plus anaphylatoxin. C3a and C5a also increased the permeability of polarized monolayers. Anaphylatoxins also promoted the proliferation of T84 and HT-29. Inhibition of ERK signaling abolished these effects of anaphylatoxins. Our findings that multiple human cell lines possess functional anaphylatoxin receptors indicates that the colonic epithelium likely responds to the activation of complement in the lumen with an inflammatory outcome.

Rheumatoid arthritis is an autoimmune disease that can cause chronic inflammation of the joints and other areas of the body. Neutrophils contribute to the pathogenesis of arthritis, and are recruited to the site of inflammation by chemokines. CXCL8/IL-8 is a member of a sub-family of chemokines (ELR+CXC chemokines) that activate and attract neutrophils through the CXCR1 and CXCR2 receptors. Our lab developed a high affinity human CXCR1/CXCR2 antagonist, called human CXCL8(3-72)K11R/G31P (hG31P). This antagonist has been shown to be highly effective in blocking ELR+CXC chemokine-driven neutrophilic inflammation. In this study we looked at the therapeutic effect of blocking ELR+CXC chemokine receptors (CXCR1 and CXCR2) in an experimental model of arthritis. We induced type II collagen (CII)-induced arthritis (CIA) in mice and treated them with hG31P after the onset of disease. The parameters we looked at to assess disease severity were clinical scores (paws were graded on the severity of edema), clinical measurements (measuring inflammation by change in circumference of paw), serum levels of anti-CII antibodies, and inflammatory cytokines mRNA (IL-1β, TNF, KC, and MIP-2) and protein levels (IL-1β, IL-6, KC, and MIP-2) in paw tissue. Initially, when we analyzed all mice together, we were unable to see a change in clinical scores and measurement when CIA mice were treated with hG31P. All CIA mice did not develop arthritis simultaneously, but rather in a serendipitous fashion; therefore we subdivided our mice and analyzed data from mice that developed arthritis early versus those that developed it late. Treatment with hG31P in mice that developed arthritis early (within 5 weeks of initial CII injection) significantly reduced clinical scores (p=0.02) in one, but not both, of our experiments. When CIA mice were treated with hG31P we saw a significant reduction (p<0.05) in CII-specific IgG1 and MIP-2 protein levels in one of our experiments. Our results were variable and we did not see these changes in our other experiment. Treatment of CIA mice with G31P did not significantly affect inflammatory cytokine mRNA levels in the paws. During this study we found the production of anti-hG31P antibodies in our hG31P-treated mice. We used a Ca2+ influx assay to determine if these hG31P antibodies were neutralizing. When these antibodies were non-neutralizing we were able to see a significant reduction in the clinical scores (p=0.02) of our hG31P-treated CIA mice (that had developed early-onset arthritis) when compared to our saline-treated CIA mice. In the experiment in which we detected significant levels of neutralizing anti-hG31P antibodies, treatment with hG31P did not affect the clinical scores of our CIA mice. Although we cannot definitively say that hG31P has a therapeutic effect in CIA, we believe this line of research merits further investigation. Our research suggests to us that after some experimental refinement and reduction of the immune response mounted to hG31P, there could still be potential for hG31P to have a therapeutic effect in arthritis.

Respiratory infection remains a leading cause of morbidity and death in severely immunocompromised febrile neutropenic haematology patients, despite the introduction of numerous prophylactic strategies and advances in diagnosis and treatment. Prognosis is improved if an organism can be isolated and specific therapy commenced as soon as possible. Current practice in this population group is to commence empirical antibiotics and perform flexible bronchoscopy (FB) if temperature does not settle or after patients develop clinical or radiological features suggesting a respiratory source. This delay may result in a lower procedural diagnostic yield due to prior or prolonged anti-microbial treatment, and increased risk of respiratory compromise and procedural complications due to advanced respiratory infections. We hypothesised that proceeding to FB as early as possible after developing febrile neutropenia would improve treatment outcomes. With this randomised, prospective trial, we aim to further define the role of FB with reference to optimal timing of the procedure and its impact on diagnostic yield, future management and complication rate. We also aim to analyse the impact of proven infection on the cytokine profile of immunocompromised patients. Methods: Patients with acute leukaemia, allogeneic bone marrow transplantation or chronic lymphocytic leukaemia (CLL) being treated with Fludarabine/ Mabthera without an obvious non-respiratory source of infection were prospectively randomised into early bronchoscopy or conventional management groups at onset of febrile neutropenia. Bronchoalveolar lavage (BAL) fluid chemokine levels (IP-10, RANTES, MIG, IL-8, MCP-1) were measured using a human Chemokine cytometric bead array (CBA) kit. Results: Thirty-one episodes of febrile neutropenia in 29 patients were analysed; 17 conventional and 14 early. There was an increased yield in fungal growth in the early bronchoscopy group, which was not predicted by prior clinical or radiological changes. However, this had no impact on clinical management in the short-term due to the delayed growth. Overall diagnostic yield was not significantly different between the two groups. Procedural complication rate was negligible overall and there was no difference associated with either group. IP-10 and MIG were significantly lower in those patients who had a fungal pathogen isolated, compared with those study patients who did not (175.17 vs 1157.8, p=0.03, 30.33 vs 247.8, p=0.03 respectively). IP-10 levels were higher in the conventional than early group (1253.0 vs 261.14, p = 0.035) and the study population had higher MCP-1 (734 vs 2.83, p=0.006) and IL-8 levels (606.9 vs 14.25, p=0.00655) than normal controls. Those cases with fungal infection had higher mean MCP-1, RANTES and IL-8 levels than in normal controls (844.0 vs 2.83, p=0.007; 17.5 vs 2.1, p=0.03; 156.0 vs 14.25, p=0.004). Conclusions: Early bronchoscopy as a component of the septic screen in febrile neutropenic patients was feasible and safe. A significant difference in fungal yield was seen in the early bronchoscopy group compared to conventional methods, with a negligible complication rate, but this did not result in a change in immediate clinical management or outcomes. / Thesis (M.Clin.Sc.) - University of Adelaide, School of Medicine, 2009

Diss. (sammanfattning) Stockholm : Karolinska institutet, 2005. / Härtill 4 uppsatser.

Hannover, Univ., Diss., 2007