THE DYNAMICS OF INTRA- AND INTER-ORGAN COMMUNICATION DURING AN ACUTE KIDNEY INFECTION

(1)

From the Department of Neuroscience Karolinska Institutet, Stockholm, Sweden

THE DYNAMICS OF INTRA- AND INTER- ORGAN COMMUNICATION DURING AN

ACUTE KIDNEY INFECTION

Svava Steiner

Stockholm 2022

(2)

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet.

Printed by Universitetsservice US-AB, 2022

Cover illustration: “Rodin’s Rodent” by Jari Radros, after Auguste Rodin “The Thinker”

(3)

THE DYNAMICS OF INTRA- AND INTER-ORGAN COMMUNICATION DURING AN ACUTE KIDNEY INFECTION

THESIS FOR DOCTORAL DEGREE (Ph.D.)

By

Svava Steiner

The thesis will be defended in public at Karolinska Institutet, lecture hall Eva & Georg Klein, Biomedicum floor 3, Solnavägen 9, February 18^th 2022 at 1:00 pm.

Principal Supervisor:

Professor Agneta Richter-Dahlfors Karolinska Institutet

AIMES

Department of Neuroscience Co-supervisors:

Professor Annelie Brauner Karolinska Institutet

Department of Microbiology, Tumor and Cell Biology

Dr. Keira Melican Karolinska Institutet AIMES

Department of Neuroscience

Opponent:

Professor David J Klumpp

Feinberg School of Medicine, Northwestern University

Department of Urology and Department of Microbiology-Immunology

Examination Board:

Adjunct Professor Jon Lampa Karolinska Institutet

Department of Medicine, Solna Professor Teresa Frisan Umeå University

Department of Molecular Biology Associate Professor Mikael Sellin Uppsala University

Department of Medical Biochemistry and Microbiology

(4)

(5)

“It is in moments of illness that we are compelled to recognize that we live not alone but chained to a creature of a different kingdom, whole worlds apart, who has no knowledge of us and by whom it is impossible to make ourselves understood: our body.”

Marcel Proust, 1920, The Guermantes Way

To my grandparents, who have inspired and still inspire me with their

creativity, curiosity, hard work and kindness (some in their own

special way).

(6)

(7)

POPULAR SCIENCE SUMMARY OF THE THESIS

Bacterial infections are once again becoming an increasing threat as we face increasing trends of antibiotic resistance. Infections in the urinary tract are among the most common bacterial infections. They can cause life threatening conditions as well as long-term complications.

Early detection and correct treatment of a urinary tract infection is of utmost importance to prevent such deleterious events. Rapid diagnostics, effective treatment and prevention of short-term and long-term complications requires that we understand how bacteria and the infected organism – the host – interact. The work in this thesis contributes to the mapping of such interactions, and the understanding of how they trigger reactions within the infected organ (intra-organ communication) as well as in distant organs (inter-organ communication).

In Papers I and II we focus on how the nervous system can sense bacterial toxins and signal for both local and distant responses in the body through nerve reflexes. We show that two different bacterial toxins, α-hemolysin and LPS, can be sensed by the nervous system and give rise to nerve responses. Neural sensing of α-hemolysin occurs when sensory nerves come in either direct contact with the toxin, or indirectly through products released from infected kidney cells. This sensing results in activation of a reflex that signals via nerves to the spleen. The spleen, in turn, responds by releasing a signaling molecule (IFNγ) that modulates the immune response. This reflex plays a role in controlling the earliest host responses to kidney infection. At a later timepoint another reflex appears to be activated by LPS. Both immune and reflex responses in sensory nerves can be triggered by LPS. The reflex results in the release of the signaling molecule CGRP in the infected kidney. This signaling molecule can in turn modulate the immune responses. Together, our results indicate that the nervous system has multiple roles in helping the body sensing and combatting a bacterial kidney infection.

In papers III and IV our focus is on how bacterial infections can trigger blood clotting, and whether this is of benefit or harm for the host. In Paper III we found that LPS and α-hemolysin can trigger coagulation in different ways. While the structure of LPS alters the kinetics of blood clotting, α-hemolysin appear to contribute to activation of blood clotting. This activation occurs when kidney cells infected with bacteria that produce α-hemolysin release a signaling molecule (CD147) that in turn activate the first stages of blood clotting. Blood clotting can hinder bacteria from spreading further through the blood stream, and administration of a drug inhibiting blood clotting results in blood stream infections in an experimental model of kidney infection. This may imply that patients prescribed similar drugs (antithrombotics) might be at higher risk of developing blood stream infections and should be monitored extra carefully. In Paper IV we therefore investigated if patients on antithrombotic treatment are at higher risk of developing blood stream infections from a kidney infection. We did not find any such increased risk. Thus, there is no need to change the current clinical recommendations regarding antithrombotic treatments during kidney infection.

(8)

ABSTRACT

This thesis focuses on the roles of neuro-immune communication and infection-mediated coagulation during bacterial kidney infections. We demonstrate dynamic intra- and inter- organ host responses, where the outcome of an infection depends on both bacterial and host factors. Our full understanding of these interactions will enable us to improve diagnostics and develop new treatments of bacterial infections to aid us as we face the increasing threat of antibiotic resistance.

Urinary tract infections (UTIs) are very common and Uropathogenic Escherichia coli (UPEC) is the most common causative pathogen. If bacteria reach the kidneys, they can cause a more severe form of UTI – pyelonephritis. Untreated pyelonephritis can lead to acute complications, such as bacteremia or urosepsis, as well as chronic complications due to kidney injury, which might result in chronic renal scarring. UPEC kidney infection triggers a number of host responses that are both local (intra-organ) and systemic (inter-organ), but many of these responses are still poorly understood. Using both in vitro and in vivo techniques we have studied the initial stages of host-pathogen interaction and different aspects of intra- as well as inter-organ signaling during UPEC kidney infection. Stemming from our experimental findings we also conducted an epidemiological study investigating infection- medicated coagulation in patients with acute pyelonephritis.

Paper I and II focus on how the nervous system can sense a UPEC kidney infection and give rise to neuro-immune reflexes. By inducing in vivo UPEC kidney infections with high spatio- temporal precision, in Paper I we were able to zoom in on extremely early events and show that the spleen is activated within 4 h. This rapid inter-organ communication triggered by infection is enabled by intact nerve signaling to the spleen, and dependent on bacterial expression of the toxin α-hemolysin. Sensory nerves can be activated by ether direct contact with α-hemolysin, or indirect contact with the toxin as renal epithelial cells infected with strains expressing α-hemolysin release ATP that in turn activates neural responses. The splenic activation results in systemic Interferon-γ (IFNγ) release, and is found to modulate the inflammatory signaling at the infection site in the kidney.

In Paper II we continue to show that the bacterial endotoxin lipopolysaccharide (LPS) has a differential role as an activator of both inflammatory and neural signaling during UPEC kidney infection. In vivo, UPEC infection triggers a release of the neurotransmitter calcitonin gene-related peptide (CGRP) in the infected kidney tissue, independent of the presence of α- hemolysin. Together with in vitro data from cell culture models we suggest that LPS triggers this local neuro-immune reflex. Collectively Papers I and II demonstrate that nerves can sense a UPEC kidney infection and initiate both distant and local neuro-immune reflexes that can modulate the immune responses at the site of infection.

In parallel to activating the nervous system, UPEC kidney infections also trigger activation of the coagulation system. This infection-mediated coagulation is the focus of Papers III and IV. We have previously shown that coagulation occurs in peritubular capillaries as a rapid

(9)

response to in vivo UPEC kidney infection. In Paper III, we investigate potential signaling mechanisms responsible for this response. Again, we show that two UPEC virulence factors may be involved in the signaling mechanisms behind the initiation of infection-mediated coagulation. The kinetics of the activation of coagulation in vivo is altered by the acylation state of the lipid A of UPEC LPS. Further, CD147 is released by renal proximal epithelial cells infected with UPEC expressing α-hemolysin, and was found to be a potential host factor involved in infection-mediated coagulation. We found that CD147 can activate tissue factor on endothelial cells, which indicates that it may initiate the first steps of the coagulation cascade.

The infection-mediated coagulation has earlier been found to result in local ischemia in the kidney, but it has also been found to protect the host from progressing from a localized infection in the kidney to systemic spread through the blood stream. The protective role of coagulation seen during UPEC kidney infection in rodents suggests that patients prescribed antithrombotic treatments could be at higher risk of developing bacteremia or urosepsis. On the other hand, infection-mediated coagulation could contribute to ischemia induced kidney injury. Since an increasing number of patients are prescribed antithrombotic treatments, and UTIs are so common, mapping potential harms or benefits of antithrombotic treatment during kidney infections is clinically important. In Paper IV we conducted a retrospective cohort study to investigate the association between antithrombotic treatment and bacteremia or acute kidney injury in patients with acute pyelonephritis. We did not find any association between antithrombotic treatment and increased risk of bacteremia during acute pyelonephritis.

Rather, low-molecular-weight heparin (LMWH) at prophylactic doses was associated with both a lower risk of bacteremia and a lower risk of acute kidney injury, compared to no antithrombotic treatment. Thus, our results suggest that it is safe to continue antithrombotic treatment during acute pyelonephritis, with regards to bacteremia and acute kidney injury.

(10)

LIST OF SCIENTIFIC PAPERS

I. Steiner SE, Choong FX, Antypas H, Morado-Urbina CE, Schulz A, Bersellini Farinotti A, Bas DB, Svensson CI, Richter-Dahlfors A*, Melican K*. UPEC kidney infection triggers neuro-immune communication leading to modulation of local renal inflammation by splenic IFNγ.

PLoS Pathog 17, e1009553, doi:10.1371/journal.ppat.1009553 (2021).

* These authors contributed equally to this work.

II. Choong FX*, Steiner SE*, Zhang T, Bas DB, Svensson CI, Melican K, Richter-Dahlfors A. Nerve responses to UPEC LPS during kidney infection.

Manuscript.

* These authors contributed equally to this work.

III. Schulz A, Chuquimia OD, Antypas H, Steiner SE, Sandoval RM, Tanner GA, Molitoris BA, Richter-Dahlfors A, Melican K. Protective vascular coagulation in response to bacterial infection of the kidney is regulated by bacterial lipid A and host CD147.

Pathog Dis 76, doi:10.1093/femspd/fty087 (2018).

IV. Steiner SE, Edgren G, Melican K, Richter-Dahlfors A, Brauner A.

Effect of anticoagulant and platelet inhibition on the risk of bacteremia among patients with acute pyelonephritis: a retrospective cohort study.

Submitted manuscript under review in BMC Infect Dis.

SCIENTIFIC PAPERS NOT INCLUDED IN THIS THESIS

I. Sedin J, Giraud A, Steiner SE, Ahl D, Persson AEG, Melican K, Richter- Dahlfors A, Phillipson M. High Resolution Intravital Imaging of the Renal Immune Response to Injury and Infection in Mice.

Front Immunol 10, 2744, doi:10.3389/fimmu.2019.02744 (2019).

II. Choong FX, Bäck M, Steiner SE, Melican K, Nilsson KP, Edlund U, Richter- Dahlfors A. Nondestructive, real-time determination and visualization of cellulose, hemicellulose and lignin by luminescent oligothiophenes.

Sci Rep 6, 35578, doi:10.1038/srep35578 (2016).

(11)

LIST OF ABBREVIATIONS

α7nAChR α7 nicotinic acetylcholine receptor

ATC Anatomic Therapeutic Chemical

ATP Adenosine triphosphate

BMI Body mass index

CD-14/26/147 Cluster of differentiation-14/26/147

CFU Colony forming units

CGRP Calcitonin gene-related peptide

CI Confidence interval

CNF-1 Cytotoxic necrotizing factor 1 CXCL1 C-X-C Motif Chemokine Ligand 1 CXCR2 C-X-C Motif Chemokine Receptor 2 DAMP Damage-associated molecular pattern DIC Disseminated intravascular coagulation

DRG Dorsal root ganglia

ELISA Enzyme-linked immunosorbent assays

GFP Green fluorescent protein

Hla α-hemolysin of S. aureus

HlyA α-hemolysin of UPEC

IL-1β/6/8/10/12 Interleukin-1β/6/8/10/12

IFNγ Interferon-γ

LBP LPS-binding protein

LMWH Low-molecular-weight heparin

LPS Lipopolysaccharide

MD-2 Myeloid differentiation factor 2 NET Neutrophil extracellular trap

NK1R Neurokinin-1 receptor

OR Odds ratio

PAI-1 Plasminogen activation inhibitor 1 PAMP Pathogen-associated molecular pattern

PMB Polymyxin B

PRR Pattern recognition receptor Sat Secreted autotransporter toxin

SLS Streptolysin S

TLR Toll-like receptor

TNFα Tumor necrosis factor-α

TRP Transient receptor potential UPEC Uropathogenic Escherichia coli

UTI Urinary tract infection

(13)

1 INTRODUCTION

Humans have combatted bacterial infections for thousands of years. The study of bacterial pathogens goes back to the 1^st century AD, when a roman scholar described an association between “minute creatures that cannot be seen by the eyes” and disease¹. It was not until the 17^th century that bacteria were observed under the microscope for the first time², and the connection between bacteria and disease was not suggested until 200 years later by Koch³. Throughout the centuries, the study of bacteria and their interaction with a host has expanded, and it has become evident that many systems interact, within the complexity of the living mammal, to maintain homeostasis in the body. Thus, traditionally separated fields of study, like microbiology, coagulation and neuroscience have become integrated and new fields such as ‘Tissue Microbiology’, ‘Neuroimmunity’ and ‘Immunothrombosis’ have recently emerged.

Over the recent 30 years ‘Cellular Microbiology’, the nexus of microbiology and cell biology⁴, and ‘Infection Biology’, the nexus of microbiology and immunology⁵, have emerged. These two fields have enabled the study of host-pathogen interactions through both in vitro and in vivo studies. Traditional in vivo models of infection do not possess the same resolution with regards to time (when host-pathogen interactions occur) and space (where host-pathogen interactions occur) as in vitro studies. However, they offer the advantage that they allow signaling between all involved cell types, including possible distant signaling effects from other organs. Live models can thus be advantageous for studying the whole- animal responses to infection. In vivo models of urinary tract infections (UTIs) are traditionally models of ascending UTI, where infection is induced through a transurethral instillation of bacteria into the bladder. While this in many aspects mimics the natural progression of UTI, it is not possible to control where the bacteria ascend to in the kidney, or to study the earliest time-points of a kidney infection.

By using a unique renal infection model, where Uropathogenic Escherichia coli (UPEC) are microinfused into single proximal tubules of exposed kidneys in anesthetized rats, our lab has developed a spatio-temporally controlled rodent kidney infection model⁶. In combination with multiphoton based intravital microscopy our lab has been able to visualize the pathophysiological changes during pyelonephritis in real-time, within the complex living host^6-8. This tissue-based microbiology has been presented as the new field ‘Tissue Microbiology’⁹.

Studies on the interaction between the nervous system and the immune system goes back to the very first description of the inflammatory response, stipulated by Celsus (25 BC-50 AD) as the four signs of rubor (redness), calore (heat), tumor (swelling), and dolore (pain)¹⁰. In the second half of the 19^th century Goltz and Stricker found the first evidence of neurogenic inflammation, as they found that electrical stimulation of sensory nerves resulted in skin vasodilation and contributed to one of these cardinal signs – swelling (tumor)¹¹. In the recent centuries, researchers have demonstrated that the peripheral nervous system can modulate inflammation by signaling to the vasculature and immune system¹². More recently a number

(14)

of neural reflexes regulating immune responses have been described¹³, and the field

‘Neuroimmunity’ has emerged¹⁴.

Inflammation and coagulation are also intertwined systems, and they share an evolutionary origin. Mammalian blood coagulation shares several features with clotting of hemolymph in invertebrates¹⁵. The hemocyte in invertebrates mediates both hemostasis and immune responses, and resembles both neutrophils and platelets of mammals¹⁶. Further, the hemolymph clotting is a component of the host defense against pathogens in invertebrates¹⁵. The hemolymph clotting can physically prevent dissemination of pathogens, and it contributes to the production of antimicrobial peptides¹⁵. Mammals also respond to pathogen invasion through activation of platelets and coagulation, and there are several interactions between the immune and coagulation systems during infections¹⁵. The study of these interactions has been termed ‘Immunothrombosis’¹⁵.

The fields of Infection Biology, Neuroimmunity and Immunothrombosis border on each other, and in recent years it has become more evident that all the body’s different systems are closely linked to combat infections. This has come further into focus during the COVID-19 pandemic, with multiple articles covering host-pathogen interactions and how they are linked to both Immunothrombosis¹⁷, and neural reflexes modulating immune responses¹⁸. The goal of this thesis has been to further elucidate the nexus between these fields, and to study host- pathogen interactions in the presence of all interplaying factors and with multiple systems in focus. Our full understanding of these interactions will enable us to develop better diagnostic and therapeutic strategies to combat life-threatening bacterial infections.

(15)

2 LITERATURE REVIEW

2.1 URINARY TRACT INFECTIONS

Urinary tract infections (UTIs) are among the most common infections in humans^19-21. Several bacterial species can infect the human urinary tract. Uropathogenic Escherichia coli (UPEC) is the cause of community-acquired UTI in up to approximately 80% of the cases, and causes the majority of hospital acquired UTIs^21-23. The severity of the UTI depends on both pathogen virulence and the host immune response²⁴.

The absolute majority of UTIs progress in an ascending manner, beginning when UPEC inoculating the periurethral area traverses the urethra and enters the bladder^25-27. Bacterial colonization of the bladder can cause asymptomatic bacteriuria, but when accompanied by symptoms it is called cystitis. Cystitis is typically characterized by painful and frequent urination, urgency of urination and suprapubic tenderness²⁸.

From the bladder, bacteria can ascend via the ureters to the kidneys, progressing to an upper UTI known as pyelonephritis^19,29. Pyelonephritis is often associated with additional symptoms such as fever, nausea, vomiting, and flank pain³⁰. Untreated pyelonephritis can lead to major chronic complications, including renal scarring and subsequent renal failure^29,31,32. Infants and children with UTIs have been found to be especially vulnerable.

Acute pyelonephritis causes acute kidney injury in many pediatric patients³³, and 15-60% of children with UTI have been reported to develop permanent renal scarring³⁴.

If pyelonephritis is left untreated, in the acute phase, there is a risk that invading bacteria can gain access to the bloodstream by breaching tissue barriers, and cause bacteremia¹⁹. It has been shown that 25-50% of patients with UPEC kidney infection also present with bacteremia³⁵. Bacteremia can result in sepsis, a serious and sometimes fatal complication^19,36, where the pathogen induces an overactivated systemic inflammatory response that can lead to multiple organ failure^24,37. Urosepsis is the second most common form of sepsis, and accounts for approximately 30% of sepsis cases in the USA and Europe³⁸. Mortality rates of patients with severe sepsis are 22-76%³⁸, and no improving trend has been seen over the last 40 years^38,39. This is mainly because the standard clinical approach to sepsis is based on early diagnosis and patient supportive measures⁴⁰.

2.1.1 UPEC in the urinary tract

Bacteria must overcome several environmental challenges in the urinary tract to successfully establish an infection. Compared to commensal fecal strains of E. coli, UPEC express genes for virulence or fitness factors that enable colonization of and persistence within the urinary tract^41,42. These include adhesion molecules, factors that aid the bacteria in evading the host immune system, iron acquisition systems, and toxins^41,43 (Figure 1).

(16)

2.1.1.1 Virulence factors that enable UPEC attachment and establishment of an infection Bacterial adhesion to urinary tract epithelial cells is critical for the establishment of UTIs^41,44. To prevent its removal from the urinary tract by the natural flow of urine, UPEC strains have adhesive organelles called fimbriae located on their surface^41,45. The most well studied fimbriae of UPEC are Type 1 fimbriae and P fimbriae, traditionally associated with bladder infection and kidney infection respectively^41,45. In cystitis, the FimH adhesin of Type 1 fimbriae binds to mannose containing host glycoproteins uroplakin I and II on bladder epithelial cells^46-49. In the kidney, the tip adhesin PapG of P fimbriae binds to glycolipid structures on the surface of renal epithelial cells to establish a kidney infection⁵⁰. Both Type 1 and P fimbriae have, however, been found to act in synergy to facilitate adhesion and colonization (P fimbriae) as well as maintenance through inter-bacterial adhesion (Type 1 fimbriae) in the renal tubule during kidney infection⁸.

UPEC may also express virulence factors that aid in evading the immune system. For example the capsule, a cell-surface associated polysaccharide, is known to protect the bacteria from phagocytosis by immune cells and inhibit complement-mediated killing of the bacteria⁵¹. Some UPEC strains can produce biofilm once they have attached to a surface⁵². The biofilm provides increased adhesion of bacteria to biotic and abiotic surfaces, and increases their resistance to mechanical stress⁵². Further, studies have shown that biofilm formation protects the bacteria from innate immune responses, both indirectly by protecting the bacterial colony inside the biofilm, as well as directly by inhibiting actions of antimicrobial peptides and neutrophils^52-54. Biofilm formation also results in a decreased susceptibility to antibiotics⁵².

Figure 1. Important virulence and fitness factors of uropathogenic E.coli (UPEC). Virulence factors that UPEC can express include factors that aid in immune evasion (e.g. capsule, ability to form biofilm, the O- antigen of lipopolysaccharide [LPS]), toxins (e.g. LPS, α-hemolysin [HlyA], secreted autotransporter toxin [Sat], and cytotoxic necrotizing factor 1 [CNF1]),adhesins (e.g. curli, fimbriae, pili and afimbrial adhesins), iron acquisition systems, and flagella for motility. Created with BioRender.com.

(17)

To reach the kidneys, bacteria must ascend the ureters against the urine flow. It has been well described that UPEC expression of flagellum mediates motility that enables the bacteria to ascend to the upper urinary tract^55,56. However, the pathophysiological and molecular details for the bacterial ascension to the kidneys is not completely understood. One part of the explanation may be that UPEC interactions with the ureters can alter the ureteric contractility, and thus enable UPEC to more easily ascend to the kidneys⁵⁷.

2.1.1.2 UPEC toxins

Some UPEC strains can produce and release toxins that enable the bacteria to invade deeper in tissue and gain access to nutrients. Among the toxins reported to be released by UPEC strains are the exotoxins -hemolysin (HlyA), the secreted autotransporter toxin (Sat), and cytotoxic necrotizing factor 1 (CNF1), as well as the endotoxin lipopolysaccharide (LPS)⁵¹. In this thesis we have focused on LPS and HlyA.

LPS is a component of the outer membrane of the Gram-negative bacteria cell wall and has protective and mechanical supportive functions^58,59. LPS consists of an O-antigen that is attached to a core of oligosaccharides^58,59. The core of oligosaccharides, in turn, are anchored to the bacterial outer membrane by lipid A⁵⁸. The lipid A moiety has fatty acyl chains embedded in the outer leaflet of the bacterial outer membrane⁵⁸. The O-polysaccharide of LPS is considered an antigen target for the host antibody response and complement activation⁶⁰. While the conformation of the O-antigen can aid in evasion of the immune system, the endotoxicity of the LPS molecule is determined by the acylation state of the lipid A moiety^61-63. Hexa-acylated LPS is more potent in inducing an inflammatory response, compared to less-acylated LPS^61-63.

The HlyA toxin is expressed in approximately half of the clinical isolates from human UTI cases and expressed by an even greater fraction of the clinical pyelonephritis isolates^51,64. UPEC HlyA expression has been associated with more severe disease and complications such as kidney injury^65,66. HlyA is a member of the repeats-in-toxin (RTX) family of protein toxins, and is believed to be a pore-forming toxin⁶⁷. HlyA has been found to exhibit functions other than just lysis. While higher concentrations of HlyA cause lysis of various host cell types (including red blood cells, epithelial cells and leukocytes)^67-70, at lower (sublytic) concentrations it can alter host cell functions and signaling, modify inflammatory signaling, promote cellular detachment and induce cell death^70-72. Among the alterations of host cell functions is the induction of renal epithelial Ca²⁺ oscillations resulting in pro-inflammatory responses of these cells in vitro, but without lysis of the cells⁷³. In vivo, in the Tissue Microbiology UPEC kidney infection model, wt strains (HlyA+) and isogenic mutants lacking HlyA expression (HlyA-) have similar colonization kinetics⁶. However, the kinetics of the earliest host responses to strains lacking HlyA seem to be delayed⁶, indicating a role for HlyA in host immune signaling during in vivo kidney infection.

(18)

2.1.2 Host responses to UPEC kidney infection

Humans constantly encounter bacteria, both commensal and pathogenic. A well-developed immune response, able to differentiate both between self and non-self, as well as between harmful and non-harmful bacteria, is thus of utmost importance. A deficient immune response might reduce the ability to clear a bacterial infection, and lead to the development of more severe infections. At the same time overactivation of the immune response might lead to a dangerous cycle that can cause substantial tissue destruction⁷⁴. These deleterious events occur when the immunological homeostasis is disturbed, and it is thus important to continue mapping signaling systems responsible for the host-pathogen interactions during infections. There are two main types of immunity: the innate and the adaptive immune system⁷⁵. The innate immune system is the first line defense, while the adaptive immune response is activated at a later stage and culminates in antigen-specific responses as well as immunological memory in T and B lymphocytes⁷⁵. In this thesis we will focus on different aspects of the innate immune system.

During a UTI the first physical interaction between bacteria and host occurs at the epithelium of the urinary tract, where the epithelial cells form the first barrier. The epithelial linings have developed ways to prevent the adhesion of bacteria, detect invading microorganisms, and activate an immune response. Many of these responses are pathogen specific and depend on the bacterial virulence factors expressed. Thus, the host-pathogen interactions during infection are not a uniform response, but are highly specific to each individual infection.

The apical surface of the urothelium (epithelium of the proximal urethra, bladder, ureters, and renal pelvis) is covered with a mucin layer that acts as a first obstacle for bacterial adhesion^76,77. The epithelial cells also produce antibacterial peptides, which can kill the invading microbes⁷⁸, prevent their attachment⁷⁹, or inhibit biofilm formation⁵⁴. At the same time, urine flow eliminates unattached bacteria and acts as another physical barrier⁸⁰. Once infection is established there are other rapid host responses to clear the infection. Superficial bladder epithelial cells can slough off driven by a FimH dependent pilus-urothelial interaction⁴⁶. Despite these extensive host responses, bacteria can in certain cases still establish an infection, which calls for the involvement of larger host defense systems.

2.1.2.1 Pattern recognition receptors

The innate immune system provides initial discrimination between self and non-self^75,81. In order to clear attached bacteria and control the infection, the innate immune system signals the presence of bacteria⁷⁵. Pattern recognition receptors (PRRs) are specialized receptors of the innate immune system that recognize pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs)⁷⁵, as well as bacterial attachment to cell surfaces⁸². Mammals express many different PRRs, including Toll-like receptors (TLRs).

There are 13 members of the TLR family that have been identified in mammals (TLRs 1-13), where TLRs 1-10 are expressed in humans⁸³. TLRs are localized to either the cell surface or intracellularly, and they recognize different PAMPs such as lipids, lipoproteins, proteins, or

(19)

nucleic acids⁸³. This recognition initiates a rapid production of pro-inflammatory mediators that attract a variety of cell types including neutrophils, which are necessary for the clearance of bacteria during an infection⁸⁴.

2.1.2.2 Pathogen detection in the urinary tract

Bladder and renal epithelial cells have different ways of recognizing UPEC⁸². TLRs 2, 4 and 5 are the most important TLRs in the urinary tract⁸⁵. TLR2 can, in complexes with TLRs 1 and 6 detect triacylated lipoproteins, lipoteichoic acid, and peptidoglycans⁸⁵. TLR5 can detect flagellin protein monomers from the flagella of motile Gram-positive and Gram- negative bacteria amongst other things⁸⁵. The most important ligands of TLR4 is LPS, as well as Type I and P fimbriae⁸⁵. While bladder epithelial cells mainly recognize LPS through TLR4, renal cells do not upregulate their expression of TLR4 until after recognition of bacterial attachment or in presence of HlyA⁸². TLR4 signaling is activated when LPS-binding protein (LBP) binds to LPS on or shed from bacteria, and transfers it to either soluble or membrane bound cluster of differentiation-14 (CD14), which forms a receptor complex with TLR4 and myeloid differentiation factor 2 (MD-2)^86-88. Attachment of UPEC by binding of P-fimbriated bacteria to epithelial surface glycosphingolipids also results in an activation of TLR4⁸⁹.

2.1.2.3 Local pathophysiological changes during UPEC kidney infections

Once UPEC has attached to renal epithelial cells, they multiply within the renal tubule to establish colonization. This initial attachment of UPEC to the renal epithelial cells, as well as bacterial multiplication within the lumen has been visualized through intravital imaging in a Tissue Microbiology model of kidney infection in rodents (Figure 2)^6,90. Following bacterial colonization, several coordinated host responses occur, including immune signaling in renal epithelial cells, vascular alterations, and infiltration of neutrophils.

Figure 2. Tissue Microbiology model of a UPEC kidney infection. Live multiphoton imaging shows single UPEC bacteria (green, arrow) attaching to the renal proximal tubule epithelium (blue) at 1 h post infection, and subsequent multiplication of bacteria (green) within the tubule as the infection progress (5h post infection). Peritubular blood flow (red) is visualized surrounding the renal tubules (Månsson et. al., 2007).

In vitro studies have shown that bladder and renal epithelial cells produce signaling molecules upon recognition of UPEC^82,91. Cytokines released early by these epithelial cells include for example interleukin-6 (IL-6), interleukin-8 (IL-8), and tumor necrosis factor-α (TNFα)^82,91. Renal in vivo upregulation of several genes coding for pro-inflammatory cytokines has been found in an ascending rodent model of UPEC kidney infection⁹². Using

(20)

the Tissue Microbiology model of kidney infection such upregulation of pro-inflammatory cytokines has been detected within the first 5 h of infection⁷. These endothelial responses during early UPEC infection all contribute to a pro-inflammatory state at the site of infection, and result in the recruitment of neutrophils into the infected tissue that in turn mediate bacterial clearance^93,94. As neutrophils are attracted to the site of infection, they form neutrophil extracellular traps (NETs), which is another component of the host defense against invading pathogens⁹⁵. NETs are a network of extracellular fibers that serve to bind and kill pathogens⁹⁵. For immune cells to infiltrate the infection site, a number of vascular changes resulting in tissue remodeling have to occur.

Kidney infections cause four characteristic inflammatory vascular changes necessary for combating infection. Three of these are vasodilation, endothelial expression of cell-adhesion molecules, and increase in vascular permeability, which all contribute to a robust infiltration of immune cells⁷⁴, mainly neutrophils⁸⁴. This aids in combating the infection. The fourth vascular change that has been observed during kidney infection is an activation of coagulation in close proximity of the infection^7,90. This activation of the coagulation results in blood clotting in the local peritubular capillaries surrounding the infection site, and contributes to ischemia in the immediate vicinity of the infection (Figure 3)⁷. In parallel, renal epithelial cells change their morphology as the infection progress, and eventually slough off from the basement membrane. This epithelial remodeling has been found to be similar to how renal epithelial cells respond to sterile ischemic kidney injury⁷. While the sloughing of renal epithelial cells may result in clearance of some bacteria attached to these cells, it may also enable bacteria to access deeper structures. However, the infection-mediated clotting appear to be protective against bacterial dissemination, as heparin treatment prior to and during infection results in systemic spread of bacteria and urosepsis in rats⁷.

2.1.2.4 Distant host responses to UPEC kidney infection

The immune response at the infection site during a UPEC kidney infection in rats is dynamic⁹⁶. As early as 8 h after onset of a localized infection in a single renal tubule, Interferon-γ (IFNγ) regulated genes are upregulated in the infected kidney tissue⁹⁶. IFNγ, a chemokine with both pro-inflammatory and anti-inflammatory effects⁹⁷, thus appears to modulate the immune response at the site of infection. Indeed, IFNγ has been found to be produced and released systemically within 8 h after the onset of a highly localized pyelonephritic infection⁹⁶. Further, the spleen has been identified as one of the sources of this systemic IFNγ⁹⁶. A similar splenic activation with increased Ifng expression in splenocytes has been observed in a model of ascending UPEC kidney infection in mice, as well as during

Figure 3. Infection-mediated clotting in peritubular capillaries. Live multiphoton imaging shows black silhouettes, indicative of a clot formation, within the blood vessel (red) surrounding a UPEC infected (green) renal tubule (blue) at 2.5 h post infection (Melican et.

al., 2008).

(21)

aseptic obstruction of the urethra for 6 hours⁹², implying an inter-organ communication between the affected kidney and the spleen.

Inter-organ communication networks have been described as networks that connect organs to coordinate responses under homeostasis and stress⁹⁸. The most commonly studied mediators of such inter-organ communication are migrating cells or proteins transported between different organs through the circulation⁹⁸. The spleen has gained increasing attention as a regulator of local and systemic inflammation through inter-organ communication⁹⁹. More recently the spleen has gained recognition for its major role in nerve reflexes to control immune responses to local and systemic inflammation, indicating a role for neural control of inflammation through inter-organ communication¹⁰⁰.

2.2 NEURAL CONTROL OF INFLAMMATION

The concept of neural control of inflammation is one of the most recent additions to the research of inter-organ communication networks. Accumulating evidence suggests that the nervous system receives information from the immune system or sites of inflammation via sensory neurons, and that it can modulate inflammation and alert distant organs to synchronize a response^13,101. The nervous system densely innervates nearly all tissues and is essential in regulating normal tissue function. Similar to how neural reflexes maintain homeostasis in the body by regulating vital functions such as heart rate, neural reflexes maintaining immune homeostasis can be activated when sensory neurons sense and respond to molecular products of inflammation^13,102. Signals arising from stimuli activating sensory neurons are orders of magnitude quicker than humoral signaling, indicating that such signals could mount a response quicker than traditional cytokine based immune signaling¹². Thus, it is not inconceivable that the body could take advantage of the innervation of barrier tissues for bacterial surveillance, which would result in activation of signals faster than those evoked by inflammatory mediators and cytokines.

Activation of sensory nerves in the periphery usually starts with stimulation of specialized receptors in sensory neurons, which results in rapid nerve firing that can give rise to both orthodromic inputs to the spinal cord and brain from the periphery, as well as antidromic inputs that relays a signal back to the periphery (axon reflex)¹³. With regards to the immune system, research has shown that orthodromic signals result in distant inter-organ communications, altering systemic immune responses to local and systemic inflammation, or rapid pain perception to alert the mammal to harmful stimuli¹⁰³. The antidromic signals result in the release of neurotransmitters in the periphery, which causes neurogenic inflammation^104-107. A few different neuro-immune reflexes have been described (Figure 4), including inter-organ reflex circuits, and reflex circuits that culminate in the release of neurotransmitters back in the affected organ. The gut-brain axis represents a neuro-immune reflex circuit that results in both local and distant responses.

(22)

Figure 4. Neural reflexes in immunity. The simple reflex originates with a stimulus that travels in a sensory (afferent) arc. The signals are relayed via interneurons and via a motor (efferent) arc to initiate a response in the tissue. Axon-axon reflexes have been described to be activated in response to bacterial infection. An afferent signal can split to send one signal to the central nervous system (CNS) and one efferent signal directly back to the infected tissue. The latter signal results in modulation of tissue/immune responses at the infection site. The inflammatory reflex originates with an afferent arc through sensory vagus nerve fibers. The signal continues through the brain stem, whereafter it descends through the motor vagus to activate the splenic nerve.

The activation of the splenic nerve results in stimulation of a subset of T cells to release acetylcholine.

Acetylcholine acts on splenic macrophages that express α7nAChR, and modulate their cytokine release. The enteric tissue protective reflex is stimulated by the presence of bacteria in the gut and result in the activation of efferent signals in enteric neurons and modulation of gut physiology (Tracey, 2016).

2.2.1 Inter-organ reflex circuits

A number of studies have mapped several inter-organ reflex circuits for neural control of inflammation. Cytokines, PAMPs and DAMPs have been shown to stimulate afferent arches through sensory nerves and activate action potentials that travel to the central nervous system, where the signals are either processed in the brain (for e.g. the perception of pain) or relayed to an efferent arch resulting in activation of organs involved in immune responses (including the spleen, lymph nodes and reticuloendothelial organs)¹⁰². Activation of nerve reflexes results in the release of neurotransmitters in these innervated tissues, and the neurotransmitters can thereafter stimulate cells of the innate and adaptive immune systems¹⁰². This neuro-immune signal results in modulation of the immune responses of these cells, and enables immune homeostasis¹⁰².

2.2.1.1 Pain, fatigue and fever responses during inflammation

Detection of tissue-damaging threats in the environment is essential for the homeostasis and survival of the organism. The body has its own ways of alerting the mind to a threat. In the case of inflammation or infection, the body can signal the threat through for example pain¹⁰⁸, fatigue¹⁰⁹, and fever¹¹⁰, all of which are common symptoms of inflammatory diseases and infections.

(23)

The pain response to tissue damage is mediated via the nervous system and includes the sensing of a noxious stimuli by sensory nerves in the periphery. This sensing results in the activation of signals (action potentials) that are transmitted to the central nervous system for perception of pain¹⁰⁸. Sensory nerves project to the dorsal horn of the spinal cord, where the signal is interpreted by second-order neurons that relay the information through the spinothalamic tract of the spinal cord to reach the thalamus in the brain¹⁰⁸. There third-order neurons send the signal to cortical areas (the somatosensory cortex), which results in perception of pain¹⁰⁸.

Pain and inflammation is often linked with fatigue¹⁰⁹. A fatigue state sends a signal to the body to rest¹⁰⁹, which may aid in alleviating pain. Another response to inflammation that is regulated by the central nervous system is fever, which is coordinated in the hypothalamus, and can result in nerve-mediated tissue responses that either reduce heat loss (e.g.

vasoconstriction) or generate heat (e.g. shivers)¹¹¹. Many theories regarding how the immune system signal to the central nervous system for fatigue and fever have focused on humoral signaling triggering these responses in the central nervous system. However, other signaling pathways that include the peripheral nervous system have been suggested. Watkins and colleagues have revealed that a fever response to intraperitoneal injection of IL-1β, and a pain response to the same stimuli is dependent on vagus nerve signaling^110,112.

2.2.1.2 The inflammatory reflex

Stemming from the vagus nerve studies performed by Watkins and colleagues, the inflammatory reflex was described in the early 2000s^113,114. The reflex has been studied in animal models of e.g. arthritis^115,116, endotoxemia113,117,118, sepsis^119,120, and recently also acute kidney injury^121-124. The inflammatory reflex is composed of an afferent sensory signal from the site of inflammation, via the vagus nerve, to the brainstem where the signals are relayed to the efferent vagus nerve, with the efferent signals travelling down to the celiac ganglion¹⁰². In the celiac ganglion the signal activates the splenic nerve leading to the secretion of norepinephrine in the spleen¹²⁵. There a subpopulation of splenic T cells that express choline acetyltransferase are stimulated by norepinephrine to produce and secrete acetylcholine¹²⁵. Secreted acetylcholine binds to α7 nicotinic acetylcholine receptors (α7nAChR) expressed on macrophages in the spleen, which respond by altering their cytokine expression¹¹⁷. This alteration results in a decrease in secretion of pro-inflammatory cytokines in the spleen^102,117. Stimulation of the efferent arch, termed the cholinergic anti- inflammatory pathway, has also been shown to inhibit neutrophil recruitment in infected tissue¹²⁶. While a lot of research has been done to study the cholinergic anti-inflammatory pathway¹⁰², the afferent arc is not as well understood.

2.2.2 Local reflex circuits and neurogenic inflammation

While pain signaling occurs between a site of inflammation and the brain, nociceptor activation can also result in the release of neuropeptides, such as calcitonin gene-related peptide (CGRP), substance P and galanin, back into the inflamed tissue to cause vasodilation

(24)

and drive neurogenic inflammation^12,127. Some studies, however, also suggest that these neurotransmitters can attract and activate innate and adaptive immune cells and modulate immune responses^128-131. Such reflexes have been called ‘axon-axon reflexes’¹³. In certain cases, such as during defense against certain pathogens, this neural control of immunity through axon-axon reflexes may be of benefit for the host. In other cases, however, reflexes resulting in neurogenic inflammation appear to maintain inflammation^132-135. For example joint inflammation in rheumatoid arthritis has been found to be perpetuated by substance P, while denervation of the joint alleviates this inflammation¹³². Similar augmentation of inflammation by neurotransmitters has been found in models of airway inflammation, colitis, and psoriasis^133-135.

There is an increasing amount of research on how neural reflexes maintain homeostasis in the gut, which is highly colonized by different bacterial strains¹³⁶. This crosstalk between the nervous system and the gut is often referred to as the ‘gut-brain axis’¹³⁶. The reflex circuit functions to monitor and coordinate gut functions (such as immune responses, intestinal permeability, and enteric reflexes), thus providing a local response to stimuli in the intestines.

However, it has been shown that the gut-brain axis can also signal to the brain, which results in cognitive and emotional responses (indicating inter-organ communication)¹³⁶.

2.2.3 Neural sensing of bacterial infection

Given the ability of the nervous system to send signals within milliseconds, the sensory nervous system appears to be ideally placed to act as a first responder to pathogens and tissue injury¹³⁷. While the majority of research investigating neuro-immune interactions during infections has focused on endotoxemia and sepsis, recent studies have shed light on infection sensing and pain signaling during localized infection^101,137. Neural sensing of bacterial infections has mainly focused on the sensation of pain during infection. Traditionally it has been thought that the pain associated with bacterial infection arises from activation of sensory nerves by inflammatory mediators released from infected cells and immune cells. However, it is becoming more evident that the peripheral nerve system can sense and signal for the presence of bacterial compounds¹⁰¹. Studies have shown that sensory nerves express receptors for both PAMPs and DAMPs, and that sensory neurons can be excited through the activation of these receptors101,137,138.

2.2.3.1 Bacterial PAMPs stimulating nerve responses

Accumulating evidence has shown that sensory neurons can sense both Gram-positive and Gram-negative bacteria, and that these neuron-microbe and neuro-immune interactions play a critical role in host responses to bacteria^101,137. Sensory nerves have been found to be able to sense the presence of bacteria and signal for pain¹⁰¹. However, some bacteria have also been found to block pain signaling to evade detection¹⁰¹. Microbes can regulate the nervous system via PAMPs (Figure 5)¹⁰¹, as well as DAMPs and other products released during infection¹³⁷.

(25)

Figure 5. Neural sensing of pathogen-associated molecular patterns (PAMPs). Bacteria can directly activate nociceptors through neural sensing of PAMPs. (L-R) M. ulcerans dampens pain signaling through mycolactone binding to angiotensin II receptors (AT2R). M. tuberculosis produces sulfolipid-1 (SL-1), which induces calcium influx in nociceptors in lungs and mediates cough during infection. C. botulinum and C.

tetani secrete botulinum neurotoxin (BoNT) and tetanus neurotoxin (TeNT) that block nerve transmission.

The S. pyogenes toxin streptolysin S (SLS), as well as S. aureus toxins α-hemolysin, γ-hemolysin and phenol soluble modulin α3 (PSMα3) activate nociceptors and produce pain during infection by causing an influx of calcium. N-formyl peptides of both S. aureus and Gram-negative bacteria have been shown to signal via formyl peptide receptors (FPRs) on nociceptors and induce pain. Lipopolysaccharide (LPS) from Gram- negative bacteria can activate or sensitize transient receptor potential cation channels TRPV1 and TRPA1 either directly or through Toll-like receptor 4 (TLR4). Bacterial flagella can activate sensory nerves via TLR5 (Lagomarsino et. al., 2021).

Nociceptors have been found to express several TLRs, including TLRs 1-7, and 9, suggesting that nociceptors may directly detect pathogens^139-143. Bacterial flagella contain flagellin, which can activate A-fiber sensory neurons through TLR5, and cause a pain reaction¹⁴³. Bacterial products reported to stimulate TLR 1, 2, 3 and 7 have also been tested, but have not been found to affect neural excitation¹⁴². The role of LPS and TLR4-signaling has been confirmed in stimulation of nociceptors139,144-146. In vivo Calil et. al. showed that local intraplantar injection of LPS in mice causes TLR4-dependent pain responses¹⁴⁷. LPS stimulation has also been reported to induce a TLR4-mediated sensitization of the capsaicin receptor transient receptor potential channel V1 (TRPV1), but not a direct TLR4-mediated excitation^139,144. This means that LPS signaling through TLR4 may both directly activate sensory neurons, and sensitize sensory neurons to other stimuli.

While there are many reports on TLR4-dependent activation of sensory neurons by LPS, TLR4-independent pathways have also been suggested for neural activation via LPS^142,148. Ochoa-Cortez et al. showed that LPS can excite cultured colonic projecting DRG neurons, independent of TLR4¹⁴². Meseguer et. al. have shown that LPS can activate mouse trigeminal and vagal sensory neurons dependent on TRPA1, and this signaling through TRPA1 is independent of TRPV1 and TLR4¹⁴⁸. In contradiction to what Calil et. al. showed¹⁴⁷, the

(26)

researchers continue to demonstrate that intraplantar injection of purified E. coli LPS in mice cause acute pain responses and mechanical hyperalgesia dependent on TRPA1 but not TLR4 expression¹⁴⁸. The activating effect of LPS through TRPA1 is dependent on the lipid A moiety¹⁴⁸. Similar to how the structure and acylation state of the lipid A moiety alters immunogenicity, it also appears to play a role in activation of sensory neurons¹⁴⁸.

Different bacteria have also been found to stimulate nerve responses through signaling pathways other than TLR-signaling. Bacterial toxins that have been shown to stimulate neural responses are for example S. aureus pore-forming toxin α-hemolysin (Hla)^149,150, and S.

pyogenes streptolysin S (SLS)¹⁵¹. Further, both E. coli and S. aureus derived N-formyl peptides have been shown to signal via formyl peptide receptor 1 expressed by nociceptors, and cause pain reactions¹⁴⁹.

2.2.3.2 Neural sensing of tissue metabolic changes, DAMPs and cytokines released during bacterial infection

Inflammatory stimuli, including cytokines (e.g., IL-1β, TNFα), chemicals (e.g., histamine, bradykinin), lipid mediators (e.g., prostaglandins), DAMPs and microenvironmental metabolic changes have been shown to stimulate or sensitize sensory neurons¹³⁷. The stimulation of nerves by these factors could thus amplify sensory signals induced by bacterial stimulation during infection^12,152.

During early infection cytokines released by infected cells may stimulate and sensitize sensory neurons. Later during inflammation, immune cells (including neutrophils, macrophages and monocytes, mast cells and T-cells) can sustain the signal through the production of more cytokines as well as prostaglandins^153-156. Specifically, IL-1β, IL-6, IL- 17A, IFNγ, and TNFα have been shown to act directly on sensory neurons and sensitize their neural signaling^157-161. Some cytokines have also been demonstrated to have a direct neural excitation effect in sensory nerves¹⁶².

DAMPs are released from damaged or dying cells, and are able to signal to immune and nerve cells¹². Adenosine triphosphate (ATP) is one DAMP released during cellular injury^163-

165. DRG neurons are known to express purinergic receptors P2X^166,167, and P2Y¹⁶⁸, and have been found to sense ATP via these receptors^169,170. Further, ATP has been found to elicit the sensation of pain¹⁷¹. Some infections have been found to trigger ATP release from infected cells^172-174, and patients with UTIs have increased urinary levels of ATP¹⁷⁵, indicating that ATP may play a role in neural sensing of infection.

Tissue injury, inflammation, and ischemia cause local tissue acidosis^176,177. Increasing proton concentration can stimulate TRPV1 receptors or acid-sensing ion channels, and activate sensory neurons^176,177. While neural sensing of changes in pH during infection is not well studied, theoretically tissue acidosis could activate sensory neurons. Thus, a combination of PAMPs, DAMPs, cytokines and metabolic changes can all contribute to neural sensing of bacterial infection.

(27)

2.2.3.3 Neuro-immune signaling during urinary tract infections

UTIs are associated with several symptoms of nerve-mediated inter-organ communication including pelvic pain and frequent urination (bladder infection) or lower back pain and fever (kidney infection)³⁰. Further, the urinary tract is extensively innervated to maintain function and homeostasis^178-180. Thus, it is not a stretch to imagine that nerves also play a role in coordinating host responses to urinary tract infections.

Rudick and colleagues have shown that a UPEC strain associated with cystitis (NU14) causes pelvic pain when instilled into the bladders of mice¹⁸¹. Further probing showed that the pain response is not dependent on mast cells or bladder inflammation, nor is it dependent on bacterial colonization or bacterial expression of Type 1 pili¹⁸¹. Rather, LPS triggers the pain response¹⁸¹. This finding was further strengthened by showing the LPS mediated pain response is TLR4-dependent¹⁸¹. LPS from a bacterial strain causing asymptomatic bacteriuria does not cause acute pain, but rather its LPS can act as a TLR4 antagonist and attenuate the pain caused by NU14¹⁸¹. This indicates that the LPS structure has a role in pain responses to UTIs.

A genetically modified NU14 strain with a modified LPS O-antigen does not elicit the acute pain responses of NU14¹⁸². Instead this strain and its LPS can cause chronic pain responses that persist after clearance of bacteria, suggesting that the O-antigen of the LPS can modulate pain responses during cystitis¹⁸². Interestingly, another UPEC strain that lacks O-antigen can induce both acute and chronic pain responses¹⁸². Similarly to the acute pain responses to NU14 bladder infection, chronic pain responses were found to be TLR4-dependent, but independent of bladder inflammation, adaptive immune cells or hematopoietic cells, and bacterial colonization¹⁸². The chronic pain responses observed after infection with UPEC lacking O-antigen have also been found to be TRPV1-dependent and proposed to be sustained by CC chemokine receptor 2¹⁸³.

Both TRPV1 and TLR4 appear to be important for pain responses during UPEC bladder infection. However no co-localization between TLR4 and bladder sensory fibers has been found¹⁸⁴. This suggests that the LPS-TLR4 signaling may occur elsewhere rather than at the sensory nerve ending. The researchers suggest that the interaction between LPS and TLR4 occurs at the urothelium that in turn mediates the TLR4-dependent UTI pain signals¹⁸⁴. While this indicates a neuroepithelial signaling for pain, the exact signaling pathways that cause the nerve activation are not yet known.

Neuroepithelial signaling during bladder infection has also been found to modulate the inflammatory response to infection¹⁸⁵. In vitro, UPEC associated with cystitis has been found to induce the release of substance P from nerve cells, while a strain associated with asymptomatic bacteriuria did not induce a similar response¹⁸⁵. Similar trends have been observed in vivo in an experimental cystitis model, where a cystitis UPEC strain causes upregulation of the expression of substance P and its receptor neurokinin-1 receptor (NK1R) in bladder tissue¹⁸⁵. Infected animals also have elevated levels of Substance P in urine,

THE DYNAMICS OF INTRA- AND INTER-ORGAN COMMUNICATION DURING AN ACUTE KIDNEY INFECTION

From the Department of Neuroscience Karolinska Institutet, Stockholm, Sweden

THE DYNAMICS OF INTRA- AND INTER- ORGAN COMMUNICATION DURING AN

ACUTE KIDNEY INFECTION

Svava Steiner

Stockholm 2022

THE DYNAMICS OF INTRA- AND INTER-ORGAN COMMUNICATION DURING AN ACUTE KIDNEY INFECTION

THESIS FOR DOCTORAL DEGREE (Ph.D.)

Svava Steiner

“It is in moments of illness that we are compelled to recognize that we live not alone but chained to a creature of a different kingdom, whole worlds apart, who has no knowledge of us and by whom it is impossible to make ourselves understood: our body.”

Marcel Proust, 1920, The Guermantes Way

To my grandparents, who have inspired and still inspire me with their

creativity, curiosity, hard work and kindness (some in their own

special way).

POPULAR SCIENCE SUMMARY OF THE THESIS

ABSTRACT

LIST OF SCIENTIFIC PAPERS

SCIENTIFIC PAPERS NOT INCLUDED IN THIS THESIS

CONTENTS

LIST OF ABBREVIATIONS

1 INTRODUCTION

2 LITERATURE REVIEW