Erectile dysfunction resulting from pelvic surgery is associated with changes in cavernosal gene expression indicative of cavernous nerve injury

Abstract

Pelvic surgery, even without direct cavernous nerve injury, carries a high risk of post-operative erectile dysfunction. The present studies were aimed at identifying molecular mechanisms by which pelvic surgery results in erectile dysfunction. As a model of pelvic surgery, male Sprague–Dawley rats underwent pelvic laparotomy, avoiding direct cavernous nerve injury. A second group of animals, serving as a model of direct cavernous nerve injury, underwent bilateral transection of the cavernous nerve. Cavernosometry demonstrated, that even in the absence of direct nerve injury, the pelvic surgery model exhibited significant erectile dysfunction 3 days post-operatively. Gene expression profiling also demonstrated that even in this animal model of nerve-sparing pelvic surgery, the profile of differentially expressed genes in cavernosal tissue was indicative of cavernous nerve injury. In addition, although 6 hr after surgery there were significant changes in circulating cytokine/chemokine levels, an inflammatory response in the major pelvic ganglion, cavernous nerve and cavernosal tissue was only observed 3 days post-surgery. Our results validate a rat model of pelvic surgery exhibiting erectile dysfunction and suggest systemic release of cytokines/chemokines following surgical trauma might mediate a pathological inflammatory response in tissues distal to the site of surgical trauma, indirectly resulting in cavernous nerve injury and erectile dysfunction.

1 |. INTRODUCTION

Pelvic surgery is a major risk factor for the development of erectile dysfunction (ED) (Zippe et al., 2006). Although well-documented to occur after radical prostatectomy (RP), it is also associated with other pelvic surgeries, such as radical cystectomy and rectal surgery (Attaallah et al., 2014; Capogrosso et al., 2020). The role of nerve damage in the development of ED following RP has been recognised since 1982, when it was demonstrated that impotence following RP was secondary to injury of the pelvic plexus, the terminal branches of which form the cavernous nerve (CN) innervating the corpora cavernosa (Walsh & Donker, 1982). With the development and widespread availability of robot-assisted technology, nerve-sparing procedures were developed in an attempt to improve erectile function outcomes after RP. Although nerve-sparing techniques are now commonly employed in performing RP (Salonia et al., 2017) and have been shown to improve post-surgical recovery of urinary continence, the incidence of ED remains high (Coughlin et al., 2018). These observations suggest that in addition to direct CN injury (CNI), there are other mechanisms by which pelvic surgeries result in ED. This suggestion is strengthened by reports that patients undergoing pelvic surgeries that do not involve direct CN damage, such as surgical repair of abdominal aortic aneurysm and prostate biopsies, are also at increased risk for development of ED (Fainberg et al., 2020; Regnier et al., 2018).

Understanding the changes in global gene expression that occur in tissues distal to the site of injury has the potential to identify underlying mechanisms for the development and treatment for ED associated with nerve-sparing pelvic surgeries. Studies in other models of peripheral nerve injury have identified that injury impacts gene expression in distally located tissues in three phases (Bosse et al., 2006; Yao et al., 2012; Yi et al., 2017). In the first phase, changes in gene expression are primarily believed to be a response to the loss of neuronal excitatory signals. In this phase, which lasts approximately 6 hr post-injury, relatively few ontological groups of genes are affected, which commonly include ion transport, signal transduction and energy metabolism. In the second phase, occurring at time points 6–48 hr post-injury, there is modulated expression of genes involved in the inflammatory response. In the third phase, which lasts from 4 days to 4 weeks following nerve injury, differentially expressed genes are gradually normalised to pre-injury levels. Although these types of studies have been applied to models of direct CNI, the earliest time point investigated has been 48 hr post-surgery (Calenda et al., 2012; Hannan et al., 2015; Weyne et al., 2014). Therefore, the early changes in gene expression in cavernosal tissue that occur as a direct result of the loss of neural signals following CNI are unknown. At present, there have been no publications describing the changes in global gene expression in cavernosal tissue resulting from pelvic surgery where there is no direct CNI.

The role of local inflammation is widely accepted as a factor in the development of ED following RP, and there are several studies demonstrating that targeting the immune response in animal models of CNI can improve erectile function outcomes (Facio et al., 2016; Garcia et al., 2014; Matsui et al., 2017, 2021; Mulhall, 2009; Sezen et al., 2009; Yamashita et al., 2011). However, any surgical procedure represents a trigger for systemic inflammation through the release of proinflammatory factors from the site of surgical trauma, which may then impact distally located tissues (Margraf et al., 2020).

The present studies were performed in rat models to identify potential underlying molecular mechanisms that lead to the development of ED in patients that undergo nerve-sparing pelvic surgery. By performing global gene expression analysis of cavernosal tissue in a rat model of CNI, 6 hr post-injury, we generated a profile of gene expression changes associated with the loss of neuronal excitatory signals through direct CNI, with no evidence of an inflammatory response at this time point. This unique expression profile was compared with global cavernosal gene expression and inflammatory response markers in a rat animal model of pelvic surgery, which even in the absence of CNI resulted in ED 3 days post-surgically. Our results suggest that pelvic surgery (even in the absence of direct CNI) results in modulated cavernosal gene expression indicative of CNI, likely through an inflammatory response mediated by systemic release of cytokines/chemokines at the site of surgical trauma.

2 |. METHODS

Animal models.

The use of animals in the present study was carried out in accordance with the Guide for the Care and Use of Laboratory Animals, as adopted and promulgated by the U.S. National Institutes of Health, and was approved by the Animal Care and Use Committee of Albert Einstein College of Medicine. All animal studies complied with the ARRIVE guidelines. Male Sprague-Dawley rats, 8–10 weeks old (200–240 g), were obtained from Charles River, Kingston, NY. Rats were fed Purina laboratory rodent chow ad libitum and housed individually with a 07:00–19:00 light cycle. Rats were randomly divided into 3 groups (n = 6) and subjected to the following surgical procedures; i) naïve: (no surgery), ii) LAPA: (laparotomy surgery consisting of a 2–3-cm mid-abdominal incision, avoiding any injury to the CN or disturbance of the pelvic organs, followed by suturing of the incision approximately 5 min later) and iii) CNI: (CNI surgery consisting of similar procedures to LAPA, with the exception that the mid-abdominal incision was followed by bilateral CN transection, as previously described (Han et al., 2010; Tar et al., 2014)). The rationale for use of the CN transection model, rather than less severe injury models, was that it allowed identification of a profile of differentially expressed genes in cavernosal tissue that result from rapid and complete loss of neuronal signals due to CNI. Although CN crush in rats is often used in research as a model for the effect of nerve-sparing prostatectomy on erectile function, it is unclear from published studies if the effects involve a complete loss of the transmission of neuronal signals, or are secondary to an inflammatory response (Canguven & Burnett, 2008). In addition, there is a wide variation in how CN crush is performed, which complicates the comparison of molecular events between nerve crush studies performed using different procedures (Haney et al., 2018).

Erectile function assessment and collection of tissues to be analysed.

Erectile function in the naïve and LAPA groups was determined 3 days post-surgery by cavernosometry by electrostimulation of the CN and measuring the resulting intracorporal pressure/systemic blood pressure ratio (ICP/BP) response, as previously described (Draganski et al., 2018; Han et al., 2010; Tar et al., 2014). The ICP and the BP were recorded in real time using a bioinformation acquisition system (LabChart; ADInstruments, Colorado Springs, CO). The maximal ICP/BP ratios in the experimental animals were calculated for each level of stimulation. An ICP/BP >0.6 was commonly associated with a visible erectile response (engorgement of the cavernosal tissue). The results were analysed by comparing measures using two-tailed Student’s t test (alpha =0.05). The groups were considered to be significantly different at p < 0.05. Data were reported as the mean ±standard deviation. Following the determination of ICP/BP in the naïve and LAPA groups, animals were euthanised, and blood and tissues collected as described below. In the CNI group, determination of the erectile response by nerve stimulation is not possible (the nerve is transected distal to the site of electrostimulation). In this group, animals were euthanised, blood and tissues collected 6 hr following surgery, representing a time point where there was minimal inflammatory response.

Determination of gene expression by RT-qPCR.

For determination of gene expression, corporal tissue, major pelvic ganglia (MPG) and CN were collected, placed in RNAlater (Qiagen) and stored at −80°C until RNA isolation. Relative gene expression for markers of inflammation was determined by RT-qPCR, as previously described (Draganski et al., 2018). Relative gene expression was determined by the ΔCt method with RPL19 used as reference gene. Student’s t test was performed on the results with a p < 0.05 deemed statistically significant.

Global gene expression analysis by RNA-seq.

Total corpora RNA was isolated from frozen tissue using RNeasy Fibrous Tissue Mini Kit (Qiagen), following the manufacturer’s instructions. RNA was sequenced by Genewiz using Illumina® NovaSeq (150 bp, >30 M coverage). Genewiz was blinded as to the nature of the samples being analysed. Sequence reads were trimmed to remove possible adapter sequences and nucleotides with poor quality using Trimmomatic v.0.36. The trimmed reads were mapped to the Rattus norvegicus Rnor6.0 reference genome available on ENSEMBL using the STAR aligner v.2.5.2b. The STAR aligner is a splice aligner that detects splice junctions and incorporates them to help align the entire read sequences after which BAM files were generated. Unique gene hit counts were calculated by using feature Counts from the Subread package v.1.5.2. Only unique reads that fell within exon regions were counted. As a strand-specific library preparation was performed, the reads were strand specifically counted. Following the extraction of gene hit counts, the gene hit counts table was used for downstream differential expression analysis (DEG). Using DESeq2, a comparison of gene expression between the naïve and the LAPA and CNI groups was performed. The Wald test was used to generate p-values and log2 fold changes in gene expression between the naïve and CNI and LAPA animal groups, with DEG defined as a log2 fold change >1, or <−1 and adjusted p-value < 0.01 compared to the naïve group (Table S1). Gene ontology analysis was performed on the statistically significant set of DEG’s using the online toolset of PANTHER (protein analysis through evolutionary relationships) (Thomas et al., 2003) to cluster DEG’s based on their biological processes and determine their statistical significance.

Determination of circulating chemokines/cytokines by Luminex.

Blood was collected from euthanised animals in EDTA-coated tubes (Vacuette, Greiner-Bio), and platelet free plasma stored at −80°C until assayed. The concentration of circulating cytokines was determined using a MILLIPLEX Rat Expanded Cytokine/Chemokine 27 Plex Kit (Millipore-Sigma, Burlington, MA) as per the manufacturer’s instructions. The assay was analysed in Luminex 200 (Millipore, Billerica, MA) using the xPONENT software (Millipore, Billerica, MA). Student’s t test was performed on the results with a p <.05 deemed statistically significant.

3 |. RESULTS

Erectile function was significantly decreased in the LAPA model compared to naive animals.

As shown in Figure 1, 3 days post-surgery at all levels of CN stimulation above 0.75 mA, there was a reduced erectile response in the LAPA compared to the naïve animal groups, indicated by a significantly lower ICP/BP. (Note: cavernosometry was not technically feasible in our animal model of direct CNI, because the CN is transected distal to the site of electrostimulation).

FIGURE 1.

Comparison of erectile function between naïve and laparotomized animals (LAPA) 3 days post-surgery validates LAPA as a model of ED associated with nerve-sparing pelvic surgery. Erectile function was determined by measuring the maximal intracorporal pressure/blood pressure (ICP/BP) ratio following different levels of CN stimulation (basal, 0.75, 1.0, 2.0, 4.0, 6 and 10 mA) in LAPA (n = 6) and naïve (n = 6) animal groups. Bars represent the mean ICP/BP ±SEM). *p < .05; **p < .01

In a rat model of direct CNI, there is no evidence of an inflammatory response in cavernosal tissue 6 hr after surgery.

We validated an animal model for determining the effects of direct CNI on cavernosal gene expression resulting from the direct loss of neuronal signalling, rather than associated with secondary effects mediated by an inflammatory response to surgical trauma. As shown in Figure 2, in the CNI group, 6 hr after surgery, there was no significant effect on cavernosal expression of three pro-inflammatory response markers (tnfα, il-1β, il-6). In comparison, in the LAPA group, 3 days after surgery, cavernosal expression of these pro-inflammatory markers was significantly upregulated.

FIGURE 2.

In a rat model of direct CN injury, there is no evidence of an inflammatory response in cavernosal tissue 6 hr after surgery. RT-qPCR was used to determine the expression of key markers of inflammation (tnfα, il-1β, il-6, vegf), transcriptional regulators of cytokine production and cell survival (nf-κb, nkap) and oxidative stress (ho-1) in cavernosal tissue at 6 hr following CN injury (CNI, n = 6) and 3 days post-surgery (LAPA, n = 6). Gene expression was normalised to RPL19. Bars represent mean mRNA expression relative to naïve animals (n = 6, dotted line). *p < 0.05; **p < 0.01; ***p < 0.001

Global gene expression analysis was used to profile differentially expressed genes in cavernosal tissue that are a response to direct CNI; this same profile was represented in cavernosal tissue from an animal model of pelvic surgery.

A total of 1,712 DEG were identified in the CNI compared to naïve group [519 (30%) with decreased and 1,193 (70%) with increased, gene expression], and 2,174 DEG were identified in the LAPA compared to naïve group [698 (32%) with decreased and 1,476 (68%) with increased, gene expression] (Figure 3). Surprisingly, there was considerable overlap of DEG (1,009) between the CNI and LAPA groups.

FIGURE 3.

Global gene expression analysis was used to profile differentially expressed genes in cavernosal tissue that are a response to direct CN injury; this profile was represented in cavernosal tissue from an animal model of nerve-sparing pelvic surgery. In the upper panel, a Venn diagram depicts the total numbers of significantly differentially expressed genes (DEG) in the LAPA and CNI animal groups compared to naïve animals (log2 fold change >1, or <−1 and adjusted p-value < 0.01). In the lower panel, the top 5 most significant ontologic groups are shown for the DEG that are unique or common to the LAPA and CNI groups (p-value < 0.001, with at least 20 DEG represented in an ontological group)

Gene ontology analysis was used to identify significant overrepresentation of DEG in functional ontological groups (p-value < 0.001, with at least 20 DEG represented in an ontological group, Table S2). The top 5 most significantly modulated ontological groups that are unique or shared by the CNI or LAPA animal groups are highlighted in Figure 3. This analysis confirmed the preliminary results shown in Figure 2 that in the CNI animal model (direct CNI at 6 hr post-surgery) there is no evidence of an inflammatory response in cavernosal tissue. The most significantly affected ontological groups were involved in transmembrane ion transport and energy metabolism, similar to changes in gene expression that occur at early times points in other models of peripheral nerve injury.

Ontological analysis of DEG in the LAPA animals was also confirmatory of the results in shown in Figure 2; in this model of nerve-sparing pelvic surgery 3 days, post-operatively there was a pronounced inflammatory response in cavernosal tissue, indicated by significant overrepresentation of DEG involved in ontological groups related to immune function. Interestingly, in addition to DEG involved in the immune response, there were also changes in the same ontological groups that matched the profile associated with direct CNI (i.e. transmembrane ion transport and energy metabolism). In Table 1, we compare the expression levels of cavernosal DEG that are directly involved in ion transport that are common between the LAPA and CNI groups (the complete ontologic group of genes defined by PANTHER involved in membrane transport is compared in Table S3). Remarkably, there was not only an overlapping profile of DEG expression between the two groups, but there were also similar fold changes in expression levels of specific genes. These data demonstrate that pelvic surgery results in a profile of changed cavernosal gene expression indicative of direct CNI.

TABLE 1.

Comparison between expression levels of cavernosal DEG (directly involved in membrane ion transport) that are common between LAPA and CNI groups

		LAPA gene expression compared to naïve		CNI gene expression compared to naïve
Gene ID	Gene	log2FC	p-value	log2FC	p-value
Slc16a14	MONOCARBOXYLATE TRANSPORTER 14	−2.12a	2.56E–09a	−2.05a	2.57E–05a
Slc7a10	ASC-TYPE AMINO ACID TRANSPORTER 1	−1.86b	3.37E–03b	−1.82b	3.48E–03b
Slco1a4	SOLUTE CARRIER ORGANIC ANION TRANSPORTER FAMILY MEMBER 1A2	−1.81b	1.04E–11b	−1.61b	1.48E–04b
Kcna6	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY A MEMBER 6	−1.67b	3.17E–06b	−2.04a	7.09E–08a
Kcnk2	POTASSIUM CHANNEL SUBFAMILY K MEMBER 2	−1.36b	7.27E–05b	–1.11b	5.20E–03b
Slc22a1	SOLUTE CARRIER FAMILY 22 MEMBER 1	−1.28b	1.72E–03b	−1.3b	6.92E–03b
AC134224.3	POTASSIUM CHANNEL SUBFAMILY K MEMBER 7	−1b	4.65E–03b	−2.6a	1.67E–08a
Kcnma1	CALCIUM-ACTIVATED POTASSIUM CHANNEL SUBUNIT ALPHA-1	1.04c	4.86E–03c	1.86c	1.68E–07c
Slc38a1	SODIUM-COUPLED NEUTRAL AMINO ACID TRANSPORTER 1	1.09c	3.49E–03c	1.57c	1.75E–05c
Slc30a3	ZINC TRANSPORTER 3	1.28c	1.48E–04c	1.41c	3.55E–04c
Atp7b	COPPER-TRANSPORTING ATPASE 2	1.33c	4.26E–03c	1.59c	7.56E–03c
Kcnq4	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY KQT MEMBER 4	1.36c	3.73E–07c	1.23c	9.72E–05c
Slc38a4	SODIUM-COUPLED NEUTRAL AMINO ACID TRANSPORTER 4	1.47c	4.79E–05c	1.33c	3.74E–03c
Orai1	CALCIUM RELEASE-ACTIVATED CALCIUM CHANNEL PROTEIN 1	1.61c	4.29E–44c	1.21c	3.79E–13c
Slc43a1	LARGE NEUTRAL AMINO ACIDS TRANSPORTER SMALL SUBUNIT 3	1.67c	6.42E–20c	1.05c	3.20E–05c
Slc9a7	SODIUM/HYDROGEN EXCHANGER 7	1.69c	2.93E–05c	2.09d	1.19E–05d
Kcnq5	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY KQT MEMBER 5	1.91c	8.19E–07c	1.27c	1.69E–03c
Slco5a1	SOLUTE CARRIER ORGANIC ANION TRANSPORTER FAMILY MEMBER 5A1	2.04d	2.20E–04d	2.37d	5.89E–04d
Cacnb1	VOLTAGE-DEPENDENT L-TYPE CALCIUM CHANNEL SUBUNIT BETA-1	2.44d	5.65E–07d	2.15d	2.77E–06d
Slc12a2	SOLUTE CARRIER FAMILY 12 MEMBER 2	2.58d	3.95E–05d	3.25e	1.25E–08e
Atp1b2	SODIUM/POTASSIUM-TRANSPORTING ATPASE SUBUNIT BETA-2	2.67d	4.28E–14d	1.71c	5.07E–05c
Atp1a2	SODIUM/POTASSIUM-TRANSPORTING ATPASE SUBUNIT ALPHA-2	2.87d	4.67E–13d	2.26d	3.55E–07d
Kcnj12	ATP-SENSITIVE INWARD RECTIFIER POTASSIUM CHANNEL 12	2.87d	6.63E–09d	2.08d	3.48E–06d
Slc16a3	MONOCARBOXYLATE TRANSPORTER 4	2.99d	7.06E–16d	2.51d	7.80E–10d
Stac3	SH3 AND CYSTEINE-RICH DOMAIN-CONTAINING PROTEIN 3	3.07e	4.91E–05e	2.58d	1.23E–04d
Scn1b	SODIUM CHANNEL SUBUNIT BETA-1	3.13e	1.34E–07e	2.37d	5.14E–12d
Scn4a	SODIUM CHANNEL PROTEIN TYPE 4 SUBUNIT ALPHA	3.13e	1.34E–07e	3.09e	4.78E–08e
Kcnf1	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY F MEMBER 1	3.29e	4.09E–07e	1.59c	4.77E–03c
Kcnj11	ATP-SENSITIVE INWARD RECTIFIER POTASSIUM CHANNEL 11	3.29e	4.09E–07e	3.13e	1.77E–07e
Kcnc4	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY C MEMBER 4	3.45e	1.72E–06e	3.07e	1.47E–07e
Kcnn4	INTERMEDIATE CONDUCTANCE CALCIUM-ACTIVATED POTASSIUM CHANNEL PROTEIN 4	3.45E+00e	1.72E–06e	4.37f	1.40E–12f
Slc38a3	SODIUM-COUPLED NEUTRAL AMINO ACID TRANSPORTER 3	3.49e	8.15E–19e	1.76c	2.83E–06c
Kcna7	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY A MEMBER 7	3.62e	4.13E–07e	3.31e	5.10E–07e
Cacna1s	VOLTAGE-DEPENDENT L-TYPE CALCIUM CHANNEL SUBUNIT ALPHA-1S	3.76e	1.43E–06e	3.01e	2.33E–05e
Kcnc1	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY C MEMBER 1	4.35f	1.05E–20f	2.44d	1.09E–05d
Kcne1	POTASSIUM VOLTAGE-GATED CHANNEL SUBFAMILY E MEMBER 1	7.88f	6.37E–09f	9.69f	6.29E–16f

^a<−2 log2FC.

^b−1 to −2 log2FC.

^c1 to 2 log2FC.

^d2 to 3 log2FC.

^e3 to 4 log2FC.

^f> 4 log2FC.

In a rat model pelvic surgery, there is evidence of an inflammatory response in tissues distal to the site of injury 3 days post-surgery.

As a possible mechanism by which pelvic surgery effects tissue distal to the site of surgical trauma, we determined the inflammatory response in MPG and CN tissue. RT-qPCR demonstrated that, similar to cavernosal tissue, 3 days post-surgery (Figure 2), in the nerve-sparing animal model (LAPA), there was upregulation of pro-inflammatory genes (tnf-α, il-1b and il-6) in both MPG and CN tissues (Figure 4a and 4b, respectively). (Note: in Figure 4b, gene expression could only be determined in the LAPA group, because of the loss of CN tissue in generating the CNI model). An inflammatory response in MPG was observed 6 hr post-surgery in the animal model of CNI, likely because of the direct proximity of MPG and CN, such that the surgical procedures for CNI result in unavoidable MPG injury activating local repair and regeneration mechanisms (Cooke, 2019).

FIGURE 4.

In a rat model pelvic surgery, there is evidence of an inflammatory response in tissues distal to the site of injury 3 days post-surgery. RT-qPCR was used to determine the fold-change in gene expression compared to naïve animals in (a) MPG tissue and (b) CN tissue, of key markers of inflammation (tnfα, il-1β, il-6, vegf), transcriptional regulators of cytokine production and cell survival (nf-κb, nkap) and oxidative stress (ho-1) at 6 hr following CN injury (CNI, n = 6) and 3 days post-surgery (LAPA, n = 6). (Note: in Figure 4b gene expression could only be determined in the LAPA group, because of the loss of CN tissue in generating the CNI model). Gene expression was normalized to RPL19. Bars represent mean mRNA expression relative to naïve animals (N = 6, dotted line). *p < 0.05; **p < 0.01; ***p < 0.001

Levels of circulating chemokines/cytokines are differentially modulated 6 hr and 3 days post-surgery.

In order to explore a potential role for circulating chemokines/cytokines in mediating the inflammatory response seen in the cavernosal tissue 3 days following nerve-sparing procedures, a Luminex analysis was conducted on plasma from the CNI and LAPA groups. In Figure 5, immune factors with significant differences in concentration compared to naïve animals. Out of 17 cytokines/chemokines analysed, the concentration of acute-phase inflammatory factors IL-6, IL-1β and IL-5 was increased in the CNI group 6 hr post-surgery but unchanged in the LAPA group 3 days post-surgery compared to naïve animals. However, in the LAPA group we observed lower plasma concentrations of a different set of cytokines (IP-10, VEGF and CX3CL1) compared to naïve animals 3 days post-surgery.

FIGURE 5.

Levels of circulating chemokines/cytokines are differentially modulated at 6 hr and 3 days post-surgery. Luminex analysis was used for quantification of circulating chemokine/cytokines in naïve, CNI and LAPA groups. Bars represent the average plasma concentrations in pg/ml (n = 6 animals per group). *p < 0.05. (Note: data are only shown where there was a significant difference in chemokine/cytokine levels compared to naïve animals)

4 |. DISCUSSION

These studies are the first to investigate the underlying molecular mechanisms by which nerve-sparing pelvic surgery results in ED. We demonstrate that simple pelvic laparotomy of the rat results in ED 3 days after surgery, validating it as a model of pelvic surgery resulting in ED. This is also the first report using a rat model of direct CNI 6 hr post-surgery to investigate global changes in cavernosal gene expression. The earliest time point used to investigate the molecular events following direct CNI prior to these studies was 48 hr post-injury (representing the second phase of gene expression changes, where changes in gene expression may be secondary to activation of the inflammatory response) (Calenda et al., 2012; Hannan et al., 2015; Weyne et al., 2014). Our studies demonstrate that whilst at 6 hr post-surgery there is already significant modulation of circulating cytokine/chemokine levels, there is no evidence that the expression of key markers of the inflammatory response is changed in cavernosal tissue. Global gene expression analysis confirmed that 6 hr after CNI there is no evidence for modulated expression of genes involved in the inflammatory response in cavernosal tissue. Therefore, we were able to define a profile of DEG in cavernosal tissue that are a response to direct CNI, and do not involve secondary effects related to the inflammatory response. By comparing this profile of DEG, with those in our animal model of nerve-sparing pelvic surgery 3 days post-operatively, we identified there are changes in cavernosal gene expression indicative of both an inflammatory response and direct CNI.

Our global gene expression at the 6 hr time point demonstrated that direct CNI results in a rapid, early phase, of modulated gene expression in cavernosal tissue, likely in response to the loss of neuronal signalling. In this first phase, DEG is significantly over-represented in a small number of ontological groups, primarily involved in ion transport/signal transduction and energy metabolism. Modulated gene expression in these ontological groups has been reported in the initial response to injury in other peripheral nerves (Bosse et al., 2006; Yao et al., 2012; Yi et al., 2017). Ion channels have a well-documented role in smooth muscle function and erectile physiology, and changes in the activity of these pathways in cavernosal tissue are likely to contribute to the development of ED (Linton et al., 2012; Thornbury et al., 2019; Toktanis et al., 2018).

Although the immune response is aimed at correcting injury, systemic release of endogenous mediators, such as pro-inflammatory chemokines/cytokines, has a well-documented link to pathology in tissues distal to the site of surgical trauma (Huber-Lang et al., 2018; Lord et al., 2014; Margraf et al., 2020). Our analysis demonstrated that 6 hr post-surgery, the concentration of IL-1β in plasma was threefold higher than in naïve animals. IL-1β is the major pro-inflammatory cytokine responsible for mediating the induction of acute phase protein synthesis (Cahill & Rogers, 2008; Luo & Zheng, 2016; Stylianou & Saklatvala, 1998). Therefore, circulating IL-1β is potentially a factor in activating the inflammatory response we observed 3 days post-surgery in cavernosal tissue, MPG and CN. Circulating IL-6 was also increased 6 hr post-surgery compared to naïve animals. IL-6 has direct effects on endothelial nitric oxide synthase activity and expression as well as increasing vascular superoxide, which rapidly inactivates NO, limiting NO bioavailability (Didion, 2017) and therefore smooth muscle relaxation.

Our studies have focussed on the effect of pelvic surgery on erectile function, providing evidence that even with nerve-sparing procedures, there is nerve injury distal to the site of trauma, likely mediated through an inflammatory response. However, patients undergoing pelvic surgery also suffer dysfunction of other organ systems that are distal to the site of injury (for example, bladder and GI tract dysfunction) (Cheung & Sandhu, 2018; Sanda et al., 2008; Wilson, 1975). It remains to be determined if similar molecular mechanisms play a role in the development of these other iatrogenic pathologies. It has been suggested that veno-occlusive mechanisms may also play a role in development of ED following pelvic surgery (De Luca et al., 1996) (Calmasini et al., 2019; Rodrigues et al., 2015) and it remains to be determined if the inflammatory response to surgical trauma might also effect vascular function. Our studies were also limited to 3 days post-surgery; although other studies have shown the inflammatory response resolves 2–4 weeks after trauma, we have not determined erectile function in our animal model of nerve-sparing pelvic surgery at longer time points. This is of potential relevance in explaining why a large number of patients fail to recover erectile function; even when nerve-sparing techniques are employed, CNI may persist even after the inflammatory response has resolved.

Technical limitations in our studies stem from the feasibility of blinding the experimenters, at least at the level of collecting data, because the surgical techniques used to create the animal models are obvious to the experimenters performing the physiological determinations and collecting tissue. However, at the level of analysis, an outside contractor was used to perform RNAseq and generate the global gene expression comparisons between groups; and this contractor was not aware of the procedures used in groups of animals prior to isolation of RNA. To minimise variations between different groups of animals, as they were received they were randomly assigned to groups, and all animals experienced similar housing conditions and underwent similar procedures for anaesthesia and euthanasia.

Overall, our data from an animal model suggest that even with pelvic surgery that avoids direct CNI, there are changes in the cavernosal gene expression profile indicative of CNI. This effect is likely mediated by systemic release of cytokines/chemokines at the site of surgical trauma, resulting in a pathological inflammatory response in tissues distal to the site of surgery. Our studies are supportive of other pre-clinical studies which suggest a rationale by which modulating the inflammatory response may have positive outcomes on erectile function following pelvic surgery (Facio et al., 2016; Garcia et al., 2014; Matsui et al., 2017, 2021; Mulhall, 2009; Sezen et al., 2009; Yamashita et al., 2011). However, the success of such a strategy remains to be determined at the clinical level.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available in the supplementary material of this article.