Selim Cellek, MD, PhD; Trinity J. Bivalacqua, MD, PhD; Arthur L. Burnett, MD, MBA; Kanchan Chitaley, PhD; Ching-Shwun Lin, PhD
ONLINE: September 13, 2012 – The Journal of Sexual Medicine, DOI: 10.1111/j.1743-6109.2012.02916.x
Since the 1980s, the number of studies on the physiology of erectile function and the pathophysiology of erectile dysfunction (ED) has grown substantially. The risk of mistakes in research techniques has accompanied this growth. This review discusses common mistakes to help other scientists and researchers, especially those who are new to the field.
Measurement of Intracavernosal Pressure (ICP) to Cavernous Nerve Stimulation (CNS)
In vivo measurements of erectile function in experimental animal ED models may be done in a number of ways. The animals may be awake or anesthetized.
ICP may be measured in awake rats by using an implanted pressure transducer. However, this method is best for studying central control of erection.
Anesthetized rodents work well for measuring erectile responses in rats and mice and the authors provide extensive instructions on this procedure. Mean arterial pressure (MAP) and ICP can be continuously measured with a pressure transducer connected to a data acquisition system. This model is the most widely used approach for assessing erectile function and can be used at a variety of voltage or frequency responses, allowing the researcher to investigate suboptimal, physiological, and supraphysiological erectile responses.
The authors caution that some researchers report erectile responses to CNS only at high voltage or frequencies, incurring supraphysiological responses that may involve pelvic floor contractions. However, using only high voltage/frequencies “fails to assess the neurotransmission elicited by the CN and thus its influence on end-organ corporal function/dysfunction.”
CN Injury Models
These models are widely used today and provide a paradigm for ED associated with pelvic nerve trauma in humans. They also help researchers understand the neurobiology of penile erection and offer opportunities for developing therapeutic procedures for ED.
CN injury models involve physical damage to the CN that impairs its function. Such injuries may include a contusion, traction, or transection. The injury may come from a direct mechanical force or from intensive energy sources, such as heat, cold, radiation, or ultrasound.
In experiments, the researcher applies an injury that is similar to what one might find in a clinical context. Often, an injury from pelvic surgery, such as a radical prostatectomy, is mimicked.
These models help investigate therapeutic interventions for ED, including traumatic neurogenic ED and ED that results from cavernosal tissue degeneration. Study results may prompt the development of therapies that protect the cavernosal tissue.
These models also help scientists understand erection physiology.
The authors note the following “fundamental concepts associated with the scientific methodology of CN injury”:
- Different types of animals may be used, but small laboratory rodents, particularly rats, are used most often. Rats are can be dissected easily and may be resilient enough to have multiple surgical procedures.
- CN injury procedures involve anesthetizing the rodent and performing a low abdominal incision and pelvic dissection.
- Bilateral or unilateral injury may be performed, depending on the study design.
- Common types of injuries at the rodent level include crush, freezing, transection, and segmental excision or resection.
Best Practice Recommendations
The researcher should “establish a reliable and reproducible perturbation” appropriate for the animal species being used. If necessary, the injury may need to be repeated until the desired effect is realized. However, the injury should not be so excessive that it cannot be changed by therapy. It should also not so limited that the benefit of therapy cannot be determined.
In these experiments, proper positive and negative controls should be in place. The standard positive control is a sham treatment. The standard negative control is an ineffective intervention. Another control involves testing “the impact of the proposed agent independent of the perturbation effect.”
Results of experiments should be indexed (normalized) in order to analyze the entire study group and comparison group.
Researchers should consider the objective of their CN injury experiment when considering their strategy. Study outcomes may vary and both long-term and short-term study time frames may be appropriate.
The authors conclude their discussion of CN injury experiments with the following points:
- Scientists should carefully choose their models based on their research objectives.
- They should also take care not to choose an inappropriate model.
- They should consider the strengths and weaknesses of available models and use the one that will provide the most useful information.
- They should remember that animal models are “only a representation of the clinical condition and findings from their investigation should be accepted only as offering preclinical, ‘proof of concept’ evidence.”
Hypertension-Associated Erectile Dysfunction
The spontaneously hypertensive rat (SHR) is the most commonly-used animal for hypertension-associated ED studies. The authors describe a number of studies involving SHRs, including one by Behr-Roussel and colleagues, which supports the idea of ED as a biomarker for cardiovascular disease. Hypertension in SHRs mimic that in humans; however, hypertension in SHRs develops at a young age and “thus may present confounding compensatory or nonadaptive changes during development that may not be relevant to the vast majority of patients who develop hypertension in adulthood.”
The deoxycorticosterone-salt (DOCA-salt) animal is commonly used in hypertension studies, but few ED studies have used this model. The authors explain how these animals are used in scientific studies.
The one kidney-one clip (1K-1C) rat model is commonly used as a surgical model of non-renin-dependent renovascular hypertension and various studies are described.
Finally, the NOS inhibition, or Nw-nitro-L-arginine methyl ester (L-NAME) model may be used. However, the authors explain that this model’s limitation is “that by inducing hypertension with NOS inhibition, the system is biased to see ED and more specifically ED related to impaired NO signaling. . . As such, mechanistic extrapolation from this animal model to the human condition becomes potentially problematic.”
Cigarette Smoke-Associated Erectile Dysfunction
Many reports have associated cigarette smoke and passive smoking with ED. While researchers have investigated this connection, study results have not shown a causative relationship and have not given much information on “the pathway(s) by which smoke exposure results in an ED phenotype.”
The authors describe many animal studies investigating smoke exposure and ED. These types of studies may use dogs, rabbits, rats, and mice and usually involve one to ten weeks of artificially exhausted cigarette smoke inhalation. In some cases, the animals are injected with nicotine or the soluble components of cigarette smoke once nicotine and tar have been removed. Because the types of animals used and the smoke delivery methods vary, it may be difficult to compare the results of different studies.
While these studies have been worthwhile, some limitations “may lessen the translatability of any insight gained”:
- Lab results depend on the species – and the strain of species – used.
- Studies differ in terms of inhalation delivery methods, the time course of inhalation, and the components of cigarette smoke involved.
- When using animals, it is difficult to determine what concentration of smoke inhalation would apply to humans for direct cigarette smoking and passive smoke exposure. This also applies to the duration of smoke exposure.
- Reports on animal studies have focused on the NO signaling pathway, the main pathway for erections. However, smoking may affect other pathways as well, and these areas have not been fully studied.
Both hypertension and cigarette smoking are risk factors for ED and, taken with other comorbidities, may increase the risk. However, animal models have not been shown consistent results in these areas and no model is considered a “gold standard.” The authors suggest that developing such a gold standard approach and model for each area would benefit to ED research.
Measurement of NOSs, NO, and cGMP nNOS vs. eNOS vs. Inducible NOS (iNOS)
The authors explain that NO, an important component of penile erection, can be synthesized by three isoforms of NOS: eNOS, nNOS, and iNOS. nNOS is involved with starting an erection and eNOS is involved with maintaining it.
The activity and amount of both nNOS and eNOS must be assessed in preclinical models for ED, but they can be difficult to measure for various reasons, such as their labile and gaseous properties. In addition, iNOS is “not expressed under normal physiological conditions” and is not as involved in erectile function, although it may provide information about inflammatory conditions.
The authors cite the following as “common mistakes with quantification of eNOS and nNOS”:
- Quantification of one enzyme, not the other. Both enzymes should be measured.
- Absence of internal control. The protein load should be normalized using an internal control protein.
- Use of immunohistochemistry (IHC) alone. IHC should always be supported with Western blotting or enzyme-linked immunosorbent assay (ELISA).
- Isolation of the proteins. A well-validated method should be used to isolate eNOS and nNOS. Also, “the protein yield should be optimized for any of the protein isolation methods.”
- Using real time-polymerase chain reaction (RT-PCR) only. This method does not measure protein expression.
- Post-translational modification. The authors write, “It should be remembered that both nNOS and eNOS can go through post-translational modification through several mechanisms such as phosphorylation, myristoylation, and palmitoylation, as well as uncoupling and subcellular translocation.”
Measurement of NO and cGMP
NO is difficult to measure; however, several methods have been developed to do so. One common mistake in assessing NO concentrations is measuring nitrate and/or nitrate concentrations in the corpus cavernosum or in the systemic circulation. Several factors can affect nitrates and nitrate levels, so the results of such an assessment are not always reliable.
Using cGMP as an indirect measure of NOS activity can also be problematic. As the authors explain, “…cGMP is formed not only by other enzymes such as particulate guanylate cyclase but also rapidly metabolized by phosphodiesterases. Moreover, cGMP concentration in the penile tissue can be an indirect measure of the NOS activity within that tissue at that time point at its basal (inactivated); it would not be able to inform us about the maximum capacity of the enzyme activity.”
When measuring cGMP concentrations, the temperature of the microenvironment should be kept at a constant 4 degrees Celsius.
The authors explain two common mistakes when measuring ROS:
- Measuring ROS only, without measuring NOS content
- Measuring only one type of ROS among many other types
The authors add, “It should always be remembered that the measurement of ROS or antioxidant enzymes is mostly performed at a time point when the erectile tissue is not activated (basal, unstimulated conditions.)”
Stem Cells (SCs) in ED Field
The authors focus their review on using various types of SCs to treat various types of ED and discuss several related studies.
Intracavernous (IC) injection has been “the universally adopted” method of delivering stem cells, but using this method for SC transplantation “appears to lack scientific basis.” In cases of CN injury, the targeted tissue is not actually in the penis, but in the CN and/or their cell bodies in the major pelvic ganglia (MPG). However, research has shown that IC-injected SCs can travel to the MPG.
The authors note the following “key components” of a SC-for-ED preclinical trial:
- Stem cells
- ED animal model
- Erectile function assessment
- Histological assessment
The most preferable animal model for SC studies is the CN injury model, followed by diabetes mellitus models. Because problems with animal models and measuring ICP for erectile function assessment are discussed elsewhere in the article, the authors discuss histological assessments here.
Because SCs may differentiate into various cell lineages, most SC therapy studies use SCs that have unique traits or are prelabeled with markers or dyes so they can be easily identified. In SC-for-ED studies, choosing a suitable SC trait or label for accurate histological data can be a challenge.
The authors discuss the following traits and labels:
- LacZ is a bacterial gene that can convert a colorless substrate into a blue product. However, LacZ can still be difficult to detect.
- GFP is a green fluorescent protein from jellyfish. This may also be difficult to detect, unless the transfected GFP is “highly expressed or densely located.”
- DAPI is a fluorescent stain for the nuclei of fixed cells. However, since it does not penetrate intact cell membranes well, it is not suitable for living cells. “More importantly,” the authors state,”because it binds DNA noncovalently, it can disassociate from the DNA of labeled cells after transplantation and be adsorbed by host cells, resulting in false-positive detection.”
- DiI is a fluorescent stain for a cell membrane, but is not reliable for tracking transplanted cells.
- BrdU is a thymidine analog used to label cells by its incorporation into newly synthesized DNA. However, to detect BrdU, tissue samples must be pretreated in DNA denaturing conditions with strong acids and high temperatures. This process leads to distorted histological images.
- EdU is another thymidine analog developed to minimize the problems with BrdU. When using EdU, the tissue sample does not need to be treated. EdU is also easier to detect, but it can only be detected through fluorescence. Cell division may also be a problem when using EdU.
The authors point out that poorly prepared tissues and/or low resolution histological images may make assessment difficult to prove. Claims of cell differentiation should be supported by clear images.
In SC-for-ED studies, cell tracking methods, preparation of tissue samples, and the interpretation of data must be carefully attended to.
The authors note that the choice of topics in this article was based on their own experiences and that other topics, such as diabetes-induced ED and in vitro tissue bath experiments are equally important. They express hope that future articles will discuss these topics.