Paper II – Nerves can sense UPEC LPS and signal for local neuroimmune

the spleen is known to activate neutrophils²⁴⁵, the observed splenic activation may prime cells in the spleen that later move to the site of infection. Studies examining this would, however, need to go beyond the 4 h infection window, as the effects we have observed on Cxcl1 expression in the kidney occurs before massive neutrophil infiltration in the kidney.

Thus, in Paper I we show that both renal epithelial cells and sensory nerves can come in contact with UPEC during a kidney infection and respond to the infection. With regards to sensory nerves, we show that they have neural responses to both direct contact with UPEC expressing HlyA, and indirectly through eATP released from renal epithelial cells infected with HlyA-expressing UPEC. We found that this direct and indirect sensing of HlyA triggers inter-organ communication between the infected kidney and the spleen in vivo in an anti-inflammatory neuro-immune reflex.

5.2 PAPER II – NERVES CAN SENSE UPEC LPS AND SIGNAL FOR LOCAL

Figure 12. Higher concentrations of LPS induce immune IL-6 responses in DRG cells. IL-6 release from DRG cells stimulated with (A) 5 ng/ml or (B) 5 μg/ml purified LPSO6. Values are presented as means ± SD, n = 8-18. * = p<0.05 determined by two-sided student’s t-test (Paper II).

It has been reported that renal epithelial cells lack a robust immune response to LPS⁸². We confirmed that even though we stimulated renal epithelial cells with a higher concentration of LPS (5 μg/ml), they did not elicit an IL-6 response. Instead, the renal epithelial cells are more responsive to stimulation with live bacteria. However, the presence of HlyA during infection with LT004 resulted in lower IL-6 responses compared to infection with LT005 (HlyA-). Further probing showed that the presence of HlyA also results in increased cell death, indicative of HlyA levels rising to cytolytic levels in this static model of infection.

This is in contrast to the observed increases in IL-6 and IL-8 during a UPEC kidney infection in our biomimetic flow chamber model, but where HlyA remains at sublytic levels (Paper I)²²¹. This again highlights the biphasic action of HlyA, with sublytic concentrations of HlyA stimulating pro-inflammatory signaling^73,221, while cytolytic levels of HlyA result in increased cell death of renal epithelial cells and dampened cytokine production and release.

Having found such different trends in how renal epithelial cells and DRG cells respond to UPEC infection, we compared the IL-6 responses of these two cell types. We found that DRG cells respond more readily to LPS, while renal epithelial cells are more responsive to live bacteria, indicating differential inflammatory responses during UPEC infection. Bladder epithelial cells and renal epithelial cells have been shown to have differential responses to UPEC infection, where the bladder epithelial cells are more responsive to LPS, while renal epithelial cells are more responsive to attached live UPEC bacteria⁸². These differences in immune responses have been suggested to be of benefit of the host⁸². In the bladder it is considered beneficial for a robust immune response to be triggered as soon as bacteria and bacterial compounds are present in the bladder. However, in the kidney, which is sensitive to disturbances in homeostasis, an immune response is only desirable once UPEC has attached to the renal epithelium and established an infection, as the initiation of an immune response before an infection is established might lead to unnecessary tissue damage⁸². We suggest that sensory neurons add to this picture, where they might play a role in strengthening the immune responses to UPEC kidney infection once infection is established and LPS levels rise. The

sensory nerves can continue signaling for immune responses even when the renal epithelium slough off.

5.2.2 Nociceptor release of CGRP in response to UPEC LPS

Since we had found that sensory nerves can sense UPEC LPS and elicit an immune response, we next investigated if the same stimulation also gives rise to neural CGRP responses. Again, we found that DRG cells were responsive to the higher concentration of LPS (5 μg/ml), but not the lower concentration (5 ng/ml) (Figure 13A). Thus, is seems like DRG cells have dose-dependent immune (IL-6) and neural (CGRP) responses to UPEC LPS. Others have also shown that E. coli LPS can activate trigeminal ganglia sensory neurons in a dose-dependent manner¹³⁹. LPS from Porphyromonas gingivalis or E. coli lacking the myristoyl fatty acid moiety of the lipid A (i.e. penta-acylated lipid A) does not, however, seem to induce CGRP release from trigeminal ganglia sensory neurons, but rather enhance the CGRP responses to capsaicin^139,144. This indicates that both the concentration, and the structure of the LPS lipid A are important in the sensory neural responses to LPS.

Having found that sensory neurons are responsive to another virulence factor (LPS), we wanted to continue by investigating if LPS could trigger a neuro-immune reflex in vivo. With our Tissue Microbiology UPEC kidney infection model it is hard to study the effect of LPS alone, as any LPS infused into the renal tubules would be directly flushed out due to the urine flow. However, as a UPEC kidney infection progresses, LPS levels are thought to rise in the infected tissue. This is due to pathophysiological changes such as the cessation of blood flow due to infection-mediated clotting in peritubular capillaries, as well as a slowdown of renal filtration rate due to bacteria clogging the renal tubule lumen^7,8. We therefore hypothesized that at later timepoints of infection neural sensing of LPS may be triggered as the LPS concentrations rise. We therefore utilized a 4 day retrograde model of UPEC kidney infection. To be able to distinguish our findings from the earlier reported neuro-immune reflex triggered by HlyA (Paper I)²²¹, we infected animals with both LT004 (HlyA+) and LT005 (HlyA-).

Since we had earlier observed nerve-driven inter-organ communication between the infected kidney and the spleen at an early phase of infection when LPS did not appear to play an important role in initiating a nervous response²²¹, this was the first reflex-circuit we wanted to investigate at this later stage of infection. We measured IFNγ levels in spleens of infected animals and compared those to sham-infected animals. After 4 days of infection no significant difference in splenic IFNγ was found between the groups (Figure 13B). This finding, together with the finding that we could not see any significant differences in bacterial loads in kidneys, bladder, or urine between LT004 (HlyA+) and LT005 (HlyA-) infected animals, highlights that HlyA appears to play its most predominant role in inter-organ communication during the earlier stages of infection, as we had previously reported²²¹. Our results also indicate that splenic IFNγ responses are dynamic. In an ascending UTI model in mice upregulation of Ifng could also be detected for up to 48 h, whereafter Ifng expression

returned to normal over the following 4 days⁹². This suggests that the splenic IFNγ response is involved in the host cytokine response during the earlier phases of kidney infection.

We continued by investigating if infection results in CGRP release in the kidney, indicative of an axon-axon reflex, since in our in vitro experiments we had found that LPS triggers the release of CGRP from DRG cells. We found elevated CGRP levels in kidneys of both LT004 (HlyA+) and LT005 (HlyA-) infected animals (Figure 13C). These results indicate a role for neuro-immune interactions also at this later time point of kidney infection. The elevated CGRP levels in infected tissue suggest that the infection might trigger an axon-axon reflex, and our in vitro data supports the concept of LPS as a potential driving factor in this response.

Figure 13. UPEC infection cause CGRP release. (A) CGRP release from primary DRG cells stimulated with LPSO6 (representative of LPS expressed by LT004 and LT005) at different concentrations. Capsaicin stimulated and unstimulated cells served as positive and negative controls respectively. Means ± SD are shown, n = 3. (B) Splenic IFNγ and (C) renal CGRP levels of LT004 (HlyA+) and LT005 (HlyA-) infected animals compared to sham-infected animals. Individual data points and median values (red bars) are shown in (B) and (C), n = 9. *=p<0.05 determined by one-way ANOVA with Turkey’s correction for multiple comparisons in (A) and Kruskal-Wallis analysis with Dunn’s correction in (B-C) (Paper II).

In the kidney, one known effect of CGRP is the modulation of vascular or tubular function¹⁷⁸. CGRP has also been found to increase the constriction frequency of lymphatic vessels²⁴⁷. CGRP treatment reduces cell count in draining lymph nodes during S. aureus infection¹⁴⁹, thus indirectly modulating host immune responses. CGRP receptors are also found on several hematopoietic cells, indicating that it may stimulate or modulate the responses of immune cells^248-250. Further, CGRP has been found to attract and activate both innate and adaptive immune cells129,131,251-253, and thus promote neurogenic inflammation. It also appears like CGRP can modulate the cytokine responses of immune cells, e.g. macrophages and dendritic cells131,149,249,250. In vivo, CGRP secreted from TRPV1+ nociceptors have been found to dampen TNF-α and CXCL1 production in the lung of mice with S. aureus pneumonia²⁵⁴. Further, CGRP may promote an immune shift from TH1 immunity toward TH2 immunity129,131,255,256. Thus, it is clear that CGRP has immune-modulating effects. The anatomical site, route of infection, types of neurons involved, target immune cell type, as well as time-point during an infection (early or late) could all impact whether CGRP release results in activation or inhibition of inflammatory responses. While we in this work strengthen the theory that nerves play a role in host defenses to kidney infection and show

that UPEC infection in the kidney triggers CGRP release, the downstream effect of this specific CGRP release remains to be investigated.

5.3 PAPER III – MECHANISMS BEHIND INFECTION-MEDIATED